INTERNATIONAL MONETARY FUND

GLOBAL
FINANCIAL
STABILITY
REPORT
Shifting Ground
beneath the Calm

2025
OCT
INTERNATIONAL MONETARY FUND

GLOBAL
FINANCIAL
STABILITY
REPORT
Shifting Ground
beneath the Calm

2025
OCT
 ©2025 International Monetary Fund

Cover and Design: IMF CSF Creative Solutions Division

Cataloging-in-Publication Data

IMF Library

Names: International Monetary Fund.

Title: Global financial stability report.
Other titles: GFSR | World economic and financial surveys, 0258-7440
Description: Washington, DC : International Monetary Fund, 2002- | Semiannual | Some issues also have thematic
titles. | Began with issue for March 2002.
Subjects: LCSH: Capital market—Statistics—Periodicals. | International finance—Forecasting—Periodicals. |
Economic stabilization—Periodicals.
Classification: LCC HG4523.G557

ISBNs: 979-8-22902-318-4 (paper)

979-8-22902-545-5 (ePub)
979-8-22902-543-0 (web PDF)

Disclaimer: The Global Financial Stability Report is a survey by the IMF staff published twice a year,
in the spring and fall. The report draws out the financial ramifications of economic issues highlighted
in the IMF’s World Economic Outlook. The report was prepared by IMF staff and has benefited
from comments and suggestions from Executive Directors following their discussion of the report
on April 11, 2025. The views expressed in this publication are those of the IMF staff and do not
necessarily represent the views of the IMF’s Executive Directors or their national authorities.

Recommended citation: International Monetary Fund. 2025. Global Financial Stability Report: Shifting
Ground beneath the Calm. Washington, DC, October.

Please send orders to:

International Monetary Fund, Publications Services
PO Box 92780, Washington, DC 20090, USA
Tel.: (202) 623-7430 Fax: (202) 623-7201
[email protected]
bookstore.IMF.org
eLibrary.IMF.org
CONTENTS

Assumptions and Conventions vii

Further Information viii

Preface ix

Foreword x

Executive Summary xii

IMF Executive Board Discussion of the Outlook, September 2025 xvii

Chapter 1. Shifting Ground beneath the Calm: Stability Challenges amid Changes in
Financial Markets 1
Chapter 1 at a Glance 1
Introduction 1
Financial Market Developments and Asset Valuations 2
Financial Conditions Ease, but the Growth-at-Risk Metric Remains Elevated 9
Emerging and Frontier Markets 10
Sovereign Bond Markets 14
Financial Intermediaries 21
Corporate Credit Risk 29
Policy Recommendations 34
Box 1.1. Global Real Estate Update 38
Box 1.2. Low Interest Rates in China Could Imperil Bank Profits and Lending 41
Box 1.3. Banks and Insurers Are Deepening Ties with the Private Credit Ecosystem 43
References 45
Chapter 2. Risk and Resilience in the Global Foreign Exchange Market 47
Chapter 2 at a Glance 47
Introduction 47
Macrofinancial Shocks and the Global Foreign Exchange Market: A Conceptual Framework 51
The Evolving Landscape of the Global Foreign Exchange Market 53
Macrofinancial Uncertainty and FX Trading Dynamics 56
Uncertainty Shocks and Stress in Foreign Exchange Markets 59
Spillovers of Foreign Exchange Market Stress to Other Asset Classes 64
Conclusion and Policy Recommendations 65
Box 2.1. Foreign Exchange Market Dynamics around the April US Tariff Announcement 68
Box 2.2. The Relevance of Settlement Risk in Foreign Exchange Markets 70
Box 2.3. Implications of Operational Disruptions in Foreign Exchange Markets 72
References 74
Chapter 3. Global Shocks, Local Markets: The Changing Landscape of
Emerging Market Sovereign Debt 77
Chapter 3 at a Glance 77
Introduction 77

International Monetary Fund | October 2025 iii

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Framework for Assessing EMDE Sovereign Debt Markets 79

Recent Trends in EMDE Sovereign Debt Markets 80
EMDE Bond Market Sensitivity to Global Shocks 85
Vulnerabilities: Limited Absorption Capacity and the Sovereign-Bank Nexus 87
Deepening Local Currency Bond Markets to Enhance Financial Stability 92
Policy Recommendations 97
Box 3.1. Local Currency Debt and Domestic Investors in the Corporate Sector in
Emerging Market and Developing Economies 100
Box 3.2. Case Studies of Local Currency Bond Market Reforms in Emerging Market and
Developing Economies Supported by IMF Technical Assistance 102
References 103
Figures
Figure ES.1. Economic Uncertainty and Financial Volatility xii
Figure ES.2. Near-Term Growth-at-Risk Framework Forecast xii
Figure ES.3. US Dollar versus Interest Rate Differentials xiii
Figure ES.4. Forced Sales of Bond Funds under the Waterfall Approach xiii
Figure ES.5. Share of Total Assets of Weak Banks, by Region xiii
Figure ES.6. Effects of Investor Composition on Emerging Market Local Bond
Market Sensitivity xiv
Figure ES.7. Estimated “r–g” (Five-Year Ahead) in Emerging Markets,
by Average Credit Rating Band xiv
Figure ES.8. Stablecoin Market Cap xv
Figure ES.9. Effect of an Increase in Microfinancial Uncertainty on Foreign
Exchange Market Conditions xv
Figure 1.1. Asset Prices amid Still-Elevated Uncertainty 3
Figure 1.2. Some Developing Divergences 5
Figure 1.3. Equity Valuation Pressures 7
Figure 1.4. Expected Profit Margins Have Been Revised Down in Most Cases 9
Figure 1.5. Financial Conditions Index 10
Figure 1.6. Global Growth-at-Risk 11
Figure 1.7. Pressures on Major Emerging Markets Have Eased, but Uncertainties Linger 12
Figure 1.8. Eurobond Issuance Has Remained Robust, but High Yields and the
Upcoming Maturity Wall Have Prompted Frontier Economies to
Explore Alternative Funding Strategies 13
Figure 1.9. Rising Bond Supply across Major Advanced Economies
Has Steepened Yield Curves 14
Figure 1.10. Widening Emerging Market Swap Spreads Are Increasing the Cost of Financing 15
Figure 1.11. Convenience Yields amid Rising Safe-Asset Supply 16
Figure 1.12. Anatomy of Two Stress Episodes from a Treasury Market Perspective,
March 2020 and April 2025 18
Figure 1.13. Structural Resilience in Repo Markets 19
Figure 1.14. Bond Mutual Fund Flows, Treasury Holdings, and
Forced Liquidations under Stress 20
Figure 1.15. Global Bank Stress Tests Results 22
Figure 1.16. Bank Exposure to Nonbank Financial Intermediaries 24
Figure 1.17. Banks’ Interest Margins Remain Resilient amid Concerns about
Potential Valuation Losses 26
Figure 1.19. Stablecoins Are Growing alongside Tokenized Assets, with
Notable Cross-Border Flows 27

iv International Monetary Fund | October 2025

 CONTENTS

Figure 1.20. The Rise of Stablecoins Comes with Potential Concerns over
Financial Stability 28
Figure 1.21. Corporate Fundamentals and Risks 30
Figure 1.22. Corporate Debt Sensitivity Analysis 31
Figure 1.23. Credit Risk and Fundamentals of Direct Lending 32
Figure 1.24. Cyclicality and Liquidity Risk for Increasing
Retail Participation in Private Credit 33
Figure 1.25. Vulnerabilities Posed by the Rising Ownership of the High-Yield
Bond Market, by Investment Funds and Exchange-Traded Funds 35
Figure 1.1.1. Commercial Real Estate Activity 38
Figure 1.1.2. Commercial Real Estate Headwinds 39
Figure 1.1.3. Residential Real Estate Activity 40
Exchange-Traded Funds
Figure 2.1. Key Developments in the Global Foreign Exchange Market 48
Figure 2.2. Macrofinancial Uncertainty and Foreign Exchange Market Conditions 50
Figure 2.3. Shock Transmission to Foreign Exchange Market Conditions and
Macrofinancial Feedback Effects 51
Figure 2.4. Structure and Trends in the Global Foreign Exchange Market 54
Figure 2.5. Bank and Nonbank Financial Institutions’ Presence in the Global
Foreign Exchange Market 55
Figure 2.6. Net Purchase of US Dollars and Macrofinancial Uncertainty 57
Figure 2.7. Effect of Macrofinancial Uncertainty on Foreign Exchange Spot Activity 58
Figure 2.8. Effect of Macrofinancial Uncertainty on Foreign Exchange Swap Flows 59
Figure 2.9. Effect of Global Macrofinancial Uncertainty Shocks on Foreign
Exchange Market Conditions 60
Figure 2.10. Foreign Exchange Market Fragilities as Amplifiers of Shock Transmission 62
Figure 2.11. Dealers’ Constraints and Foreign Exchange Market Conditions 63
Figure 2.12. Policy Mitigating Factors 63
Figure 2.13. Financial Spillovers of Foreign Exchange Market Stress 64
Figure 2.1.1. Broad US Dollar Index and VIX, January 2, 2025, to May 15, 2025 68
Figure 2.1.2. Net Spot US Dollar Flows before and after the US Tariff Announcement on
April 2, 2025 69
Figure 2.2.1. Effect of CLS Entry on Excess Foreign Exchange Returns and Volatility 70
Figure 2.3.1. Effect of Interdealer Platform Disruption on Market Liquidity 73
Figure 3.1. Financial Stability Framework for Local Sovereign Debt Markets in
Emerging Market and Developing Economies 79
Figure 3.2. Recent Trends in Emerging and Frontier Debt Markets 80
Figure 3.3. Composition of Marketable Domestic Public Debt in Selected Emerging Market and
Developing Economies 81
Figure 3.4. Portfolio Flows, Nonresident Holdings, and Investor Returns for
Selected Emerging Market and Developing Economies 83
Figure 3.5. Investor Base in Selected Emerging Market Local Currency
Government Bond Markets 84
Figure 3.6. Local Stress Index and the Investor Base for Local Currency
Bond Markets in Emerging Markets 85
Figure 3.7. Effect of Global Risk Factors on Local Currency Bond Markets in
Emerging Markets and the Role of Investor Composition 86
Figure 3.8. Investor Composition and the Long-Term Effect of a VIX Increase on
Local Currency Bond Markets in Emerging Markets 88

International Monetary Fund | October 2025 v

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 3.9. Financial Sector Assets and Absorption Challenges in

Selected Emerging Markets 89
Figure 3.10. Recent Trends in Sovereign-Bank Nexus Risks in Selected Emerging Market and
Developing Economies 91
Figure 3.11. Building Blocks of Local Currency Bond Market Resilience in
Emerging Market and Developing Economies 93
Figure 3.12. Local Currency Bond Market Development in Emerging Market and
Developing Economies 96
Figure 3.1.1. Corporate Debt in Emerging Markets 100
Online-Only Annexes
Online Annex 1.1. IMF Global Stress Test (GST) Scenarios
Online Annex 1.2. Definitions of Nonbank Financial Institutions (NBFIs)
Online Annex 1.3. Stress Test of Banks Exposures to NBFIs
Online Annex 1.4. Mutual Funds and Their Impact on Core Bond Markets
Online Annex 1.5. Corporate Debt-at-Risk Sensitivity Analysis
Online Annex 1.6. Estimating US Dollar Exposures
Online Annex 1.7. Sovereign Bond Markets
Online Annex 1.8. Short-Term Rate Sensitivities to Reserves and Dealer Positions
Online Annex 2.1. Glossary of Key Foreign Exchange (FX) Market Terms
Online Annex 2.2. Data Description and Sources
Online Annex 2.3. How Has the Global FX Market Evolved over Time?
Online Annex 2.4. How Do Macrofinancial Uncertainty Shocks Affect Trading Activity?
Online Annex 2.5. How Do Macro-Financial Uncertainty Shocks Affect FX Market Conditions?
Do Structural Vulnerabilities in FX Markets Amplify the Effects of These Shocks?
Online Annex 2.6. Does Stress in FX Markets Spillover to Other Financial Markets?
Online Annex 2.7. Foreign Exchange Market Dynamics around the US Liberation Day Tariff Announcement
Online Annex 2.8. The Relevance of Settlement Risk in Foreign Exchange Markets
Online Annex 2.9. Implications of Operational Disruptions in Foreign Exchange Markets
Online Annex 3.1. Measurement and Econometric Approach
Online Annex 3.2. Country Classification of Local Currency Bond Markets

vi International Monetary Fund | October 2025

ASSUMPTIONS AND CONVENTIONS

The following conventions are used throughout the Global Financial Stability Report (GFSR):
. . . to indicate that data are not available or not applicable;
— to indicate that the figure is zero or less than half the final digit shown or that the item does not exist;
– between years or months (for example, 2021–22 or January–June) to indicate the years or months covered,
including the beginning and ending years or months;
/ between years or months (for example, 2021/22) to indicate a fiscal or financial year.

“Billion” means a thousand million.

“Trillion” means a thousand billion.
“Basis points” refers to hundredths of 1 percentage point (for example, 25 basis points are equivalent to ¼ of
1 percentage point).

Minor discrepancies between sums of constituent figures and totals shown reflect rounding.

As used in this report, the terms “country” and “economy” do not in all cases refer to a territorial entity that is a
state as understood by international law and practice. As used here, the term also covers some territorial entities
that are not states but for which statistical data are maintained on a separate and independent basis.

The boundaries, colors, denominations, and any other information shown on the maps do not imply, on the part
of the International Monetary Fund, any judgment on the legal status of any territory or any endorsement or
acceptance of such boundaries.

International Monetary Fund | October 2025 vii

 FURTHER INFORMATION

Corrections and Revisions

The data and analysis appearing in the Global Financial Stability Report are compiled by IMF staff at the time
of publication. Every effort is made to ensure their timeliness, accuracy, and completeness. When errors are discov-
ered, corrections and revisions are incorporated into the digital editions available from the IMF website and on the
IMF eLibrary. All substantive changes are listed in the Contents page.

Print and Digital Editions

Print
Print copies of this Global Financial Stability Report can be ordered from the IMF Bookstore at imfbk.st/570343.

Digital
Multiple digital editions of the Global Financial Stability Report, including ePub, enhanced PDF, and HTML,
are available on the IMF eLibrary at eLibrary.IMF.org/OCT25GFSR.

Download a free PDF of the report and data sets for each of the figures therein from the IMF website at
IMF.org/publications/GFSR or scan the QR code below to access the Global Financial Stability Report web
page directly:

Copyright and Reuse

Information on the terms and conditions for reusing the contents of this publication are at IMF.org/external/
terms.htm.

viii International Monetary Fund | October 2025

PREFACE

The Global Financial Stability Report (GFSR) assesses key vulnerabilities the global financial system is exposed
to. In normal times, the report seeks to play a role in preventing crises by highlighting policies that may mitigate
systemic risks, thereby contributing to global financial stability and the sustained economic growth of the IMF’s
member countries.
The analysis in this report was coordinated by the Monetary and Capital Markets (MCM) Department under
the general direction of Tobias Adrian, Director. The project was directed by Charles Cohen, Advisor; Mahvash
Qureshi, Assistant Director; Thordur Jonasson, Assistant Director; Jason Wu, Assistant Director; Caio Ferreira,
Deputy Division Chief; Sheheryar Malik, Deputy Division Chief; Mario Catalán, Deputy Division Chief;
Andrea Deghi, Financial Sector Expert and Chapter 2 co-lead; Tomohiro Tsuruga, Senior Financial Sector Expert
and Chapter 2 co-lead; Arindam Roy, Senior Financial Sector Expert and Chapter 3 co-lead; Patrick Schneider,
Financial Sector Expert and Chapter 3 co-lead. It benefited from comments and suggestions from the senior staff
in the MCM Department.
Individual contributors to the report were John Caparusso, Yuhua Cai, Sally Chen, Yingyuan Chen,
Timothy Chu, Kay Chung, Fabio Cortes, Francesco de Rossi, Xiaodan Ding, Gonzalo Fernandez Dionis,
Andrew A. Ferrante, Deepali Gautam, Bryan Gurhy, Sanjay Hazarika, Zixuan Huang, Esti Kemp, Johannes
Kramer, Harrison Samuel Kraus, Seungduck Lee, Yiran Li, Xiang-Li Lim, Corrado Macchiarelli, Taneli Mäkinen,
Srobona Mitra, Benjamin Mosk, Kleopatra Nikolaou, Sonal Patel, Silvia Loyda Ramirez, Lawrence Tang,
Yuan Tian, Jeffrey Williams, Dmitry Yakovlev, Mustafa Yenice, Aki Yokoyama, Xuege Zhang, Yuchen Zhang, and
Jing Zhao. Chapter 2 also benefited from comments and suggestions from Angelo Ranaldo, External Advisor.
Javier Chang, Monica Devi, Srujana P. Tyler, and Yi Zhou were responsible for excellent coordination and
editorial support.
Rumit Pancholi from the Communications Department led the editorial team and managed the report’s
production, with editorial services from David Einhorn, Michael Harrup, Devlan O’Connor, and James Unwin.
This issue of the GFSR draws in part on a series of discussions with banks, securities firms, asset management
companies, hedge funds, standard setters, financial consultants, pension funds, trade associations, central banks,
national treasuries, and academic researchers.
This GFSR reflects information available as of September 12, 2025. The report benefited from comments
and suggestions from staff in other IMF departments, as well as from Executive Directors following their
discussions of the GFSR on September 29, 2025. However, the analysis and policy considerations are those
of the contributing staff and should not be attributed to the IMF, its Executive Directors, or their national
authorities. In the Executive Summary, data used in Figures ES.1, ES. 2, and ES. 3, and in Chapter 1, Figure 1.1
(all panels), Figure 1.2 (panels 1 and 2), Figure 1.3 (all panels), Figure 1.5 (all panels), Figure 1.6 (all panels),
reflect information through October 2, 2025.

International Monetary Fund | October 2025 ix

 FOREWORD

F
inancial conditions have eased since our are expanding. Therefore, stress in sovereign bond mar-
April 2025 Global Financial Stability Report, kets can transmit directly to banks or indirectly through
as policy uncertainty has receded somewhat, nonbanks. Stress tests for nonbanks in this report reveal
major central banks have become more accom- considerable scope for sell-offs in benchmark bonds.
modative, and the US dollar has weakened. Equity The growing size of nonbank financial intermedia-
markets have rebounded to record highs, corporate and tion could amplify these vulnerabilities. While helpful
sovereign funding spreads are at historically narrow in facilitating capital market activities and channeling
levels, and global funding liquidity remains abundant. credit to borrowers, their expansion raises the specter
However, the ground is shifting beneath this seem- of risk-taking and interconnectedness in the financial
ingly tranquil surface. Valuations of risk assets appear system. A key challenge revolves around the limited
stretched, especially as the global economy slows, and visibility into balance sheets and the interconnected-
concentration risks in certain segments have reached ness of nonbank financial institutions. Stronger data
historic highs. History reminds us that asset prices can and disclosures are critical to both diagnose vulnerabil-
abruptly correct following booms in the technology ities and guide policy responses during stress events.
sector. To the extent that stock market-driven wealth Growing macrofinancial uncertainty can strain
effects support strong consumption, a correction could even the highly liquid foreign exchange market, as
have broader implications for the real economy. Higher we discuss in Chapter 2. Such uncertainty may raise
long-term yields on major sovereign benchmarks— funding costs, impair liquidity, and heighten foreign
most notably for US Treasuries and euro area bonds— exchange volatility—effects that are notably pro-
could reverberate across the system, influencing nounced in emerging markets. These pressures can also
hedging strategies and reshaping correlations with risky spill over into other asset classes, triggering broader
assets. Structural improvements in market resilience, negative feedback loops, most evident in the presence
including central clearing and leverage requirements, of significant currency mismatches and fiscal vulner-
should help, although they remain a work in progress abilities. These considerations are especially relevant
in many jurisdictions. given US dollar softness and a substantial increase in
Concerningly, many advanced economies— foreign exchange hedging demand this year.
especially those with the most elevated debt levels— The evolving risk environment carries significance
have yet to present credible strategies to stabilize rising for emerging markets and developing economies
debt trajectories, even as new spending pressures (EMDEs). As discussed in Chapter 3 of this report,
emerge. With more fiscal risks, a higher level of term for emerging markets with strong fundamentals, a shift
premia could become a defining feature of global toward financing themselves in local currencies has
financial markets in years to come. In this environ- stabilized bond yields and bolstered market liquidity
ment, central bank independence is integral to ensure during periods of global stress. By contrast, emerg-
monetary policy continues to focus on maintaining ing markets and developing economies with weaker
price stability and anchoring inflation expectations. policy credibility and limited domestic savings remain
One of the most troubling shifts is the potential dependent on foreign currency borrowing, and may
erosion of the hedging role of longer-term bonds, also overly rely on domestic banks to buy government
exposing fragilities in the financial sector-sovereign bonds. Although funding costs remain contained for
nexus. Financial sector exposure to sovereign assets most EMDEs so far this year, new major shocks could
remains elevated across both banks and nonbanks. still test their resilience.
While banks globally are generally well capitalized, a At the same time, new financial market innovations,
vulnerable subset persists in most jurisdictions, and such as stablecoins backed by short-term government
banks’ exposures to nonbank financial intermediaries securities, have introduced new participants in

x International Monetary Fund | October 2025

 FOREWORD

sovereign debt markets and payment systems. In with central bank mandates, and strengthening finan-
weaker economies, these developments may lead to cial sector supervision. Reinforcing the independence
currency substitution and reduce the effectiveness of and credibility of central banks helps to anchor expec-
policies, like monetary policy. They could also alter the tations and bolster confidence in the policy framework.
bond market structure with potential implications for We also advocate for enhanced reporting and oversight
credit disintermediation. Possible runs on stablecoins of nonbank financial institutions, ongoing efforts to
may also generate forced sales of reserve assets, poten- improve the efficiency of local bond markets, and
tially disrupting market functioning. implementation of internationally agreed prudential
Navigating these challenges will require continued standards, including on cryptoassets.
vigilance by policy authorities. Our primary recom-
mendations emphasize fiscal discipline to ensure debt Tobias Adrian
sustainability, a close monetary policy focus in line Financial Counsellor

International Monetary Fund | October 2025 xi

 EXECUTIVE SUMMARY

Shifting Ground beneath the Calm Figure ES.1. Economic Uncertainty and Financial Volatility
(Percentile)
Global financial markets appear calm despite contin-
October 2024 GFSR
ued trade and geopolitical uncertainty (Figure ES.1). Economic uncertainty measures April 2025 GFSR
However, this issue of the Global Financial Stability Latest
100 Average post-pandemic
Report highlights several signs of shifting grounds in
90
the financial system that could raise vulnerabilities if
80
associated risks are not addressed. Accordingly, the 70
IMF’s growth-at-risk framework shows that risks to 60
global financial stability remain elevated (Figure ES.2). 50
Policymakers are urged to stay vigilant and respond 40
promptly to changing circumstances. 30
The first sign that the ground is shifting is the 20
10
continued appreciation of risk asset prices. Markets
0
appear to downplay the potential effects of tariffs on Trade Policy Economic Policy Geopolitical VIX Index MOVE Index
Uncertainty Uncertainty Risk Index
growth and inflation. IMF staff models show that
valuations of some risk assets have once again become Sources: Bloomberg Finance L.P.; Baker, Bloom, and Davis 2016; Caldara and others
stretched after the brief correction from the April 2 2020; Caldara and Iacoviello 2022; and IMF staff calculations.
Note: Percentiles are derived from monthly data since January 1997. “Average
tariff announcement by the United States. Meanwhile, postpandemic” is the average percentile since January 2022. The latest levels for the
the US dollar has depreciated by 10 percent so far VIX and MOVE indices are as of October 2, 2025. GFSR = Global Financial Stability
Report; MOVE = Merrill Lynch Option Volatility Estimate; VIX = Chicago Board
this year, having decoupled relative to wide US–G10 Options Exchange Volatility Index.
interest rate differentials in the months following the
announcement (Figure ES.3), amid concerns about
Figure ES.2. Near-Term Growth-at-Risk Framework Forecast
US policy uncertainty, and as investors reassessed (Historical percentile rank)
the dollar’s decade-long bull run. Any further abrupt
0
correction of asset prices could be exacerbated by
these changing asset correlations, straining financial
20
markets. For example, foreign exchange markets
have undergone structural shifts in recent years yet
40
have not experienced significant dollar weakness (see
Chapter 2).
60
Another crucial sign is that debt has continued to
shift toward the government sector as expanding global
80
fiscal deficits propel sovereign bond issuance. In major
advanced economies, sovereign bond markets increas-
100
ingly depend on price-sensitive investors, exerting 2012 13 14 15 16 17 18 19 20 21 22 23 24 25
upward pressure on term premiums and long-term
Sources: Bank for International Settlements; Bloomberg Finance L.P.; EUROPACE
yields. In emerging markets, governments have turned AG/Haver Analytics; IMF, International Finance Statistics database; and IMF staff
calculations.
to domestic investors for financing. Although this
Note: The black line traces the evolution of the fifth percentile threshold (the
has reduced reliance on foreign currency debt, it may growth-at-risk metric) of the near-term forecast densities, where the lower percentiles
yet create fragilities such as a stronger bank-sovereign represent a higher downside risk. The intensity of the shading depicts the percentile
rank for the growth-at-risk metric; the quintiles with the lowest percentile ranks are
nexus (see Chapter 3). shaded the brightest red and the highest are brightest green.

xii International Monetary Fund | October 2025

 Executive Summary

Figure ES.3. US Dollar versus Interest Rate Differentials Figure ES.4. Forced Sales of Bond Funds under the Waterfall
(Index, left scale; percent, right scale) Approach
(Billions of dollars)
110 2
April 2 tariff
announcement 1.9 400
Cash
1.8 350 Money market mutual fund shares
Treasury bills
1.7 300 Commercial paper
105
US Treasuries
1.6 250
1.5
200
1.4
100 150
1.3
100
DXY 1.2
US-G10 average (right scale) 1.1 50

95 1 0
Jan. Apr. Jul. Oct. Jan. Apr. Jul. Oct. April 2025 outflows March 2020 outflows 99th pct outflows
2024 2024 2024 2024 2025 2025 2025 2025 + 60 bps rate shock + 80 bps rate shock + 100 bps rate shock

Sources: Bloomberg Finance L.P.; and IMF staff calculations. Sources: Lipper; Securities and Exchange Commission N-PORT; and IMF staff
Note: The US-G10 average is the nominal 10-year interest rate differential between calculations. N-PORT data are taken from the second quarter 2025 batch, retaining
the United States and the average of the G10 countries. DXY is the US dollar index, only submissions for the first quarter of 2025.
which indicates the general international value of the dollar. Note: Under the “waterfall” approach, US Treasuries are sold after cash and other
liquid assets are depleted. See Online Annex 1.4 for more details. bps = basis points;
pct = percentile.

Finally, the growing size of nonbank financial inter-

mediaries (NBFIs) and deepened ties with banks have
Figure ES.5. Share of Total Assets of Weak Banks, by Region
heightened their sectoral interconnection. This Global (Percent of assets, vertical axis; number of banks, bars)
Financial Stability Report documents the expanding
50
role of NBFIs in core sovereign bond markets and cor- Adverse scenario: banks below CET1 of 7 percent (plus G-SIB buffer)
Adverse scenario with NBFI shock: banks below CET1 of 7 percent
porate debt markets, including through the participa- (plus G-SIB buffer)
40
tion of retail investors in private credit. Such links raise
the specter of excessive risk taking and interconnected-
30
ness in the financial system.
57
89
82 50
20
Vulnerabilities and Uncertainties
10
These shifts can cause vulnerabilities in the financial
system. Although sovereign bond markets in major
0
advanced economies have stabilized since the abrupt Global AE
sell-off prompted by the April 2 tariff announcement,
Sources: Call report data; European Banking Authority; Fitch Connect; Fitch
steepening yield curves, more negative swap spreads, Solutions; S&P Capital IQ Pro; and IMF staff calculations.
and the erosion of convenience yields indicate that Note: The figure shows the number of banks falling below the 7 percent CET1 ratio
plus a G-SIB buffer under the IMF Global Bank Stress Test adverse scenario, with an
bond market functioning is on shakier footing. Bond additional NBFI shock for euro area and US banks. The NBFI shock assumes that risk
market functioning could be tested if yields rise weights increase from 20 percent to 50 percent and all available commitments are
drawn. AE = advanced economy; CET1 = Common Equity Tier 1 capital; G-SIB =
abruptly—for instance, a scenario analysis for bond global systemically important bank; NBFI = nonbank financial intermediaries.
mutual funds shows that forced US Treasury liquida-
tions as a result of large fund outflows and an abrupt
increase in yields could reach almost $300 billion
(Figure ES.4). Given the crucial role of core sovereign global banks by assets would see their Common Equity
bonds as benchmarks and collateral, their deterioration Tier 1 capital ratio fall below the important threshold
has implications for the broader financial markets. of 7 percent plus a G-SIB buffer. However, additional
In the banking sector, capital ratios have improved. shocks to NBFIs could increase this share of weak
In an adverse macroeconomic scenario, the IMF’s banks by assets to 21 percent, highlighting the linkages
Global Stress Test reveals that about 18 percent of between banks and NBFIs (Figure ES.5).

International Monetary Fund | October 2025 xiii

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure ES.6. Effects of Investor Composition on Emerging Figure ES.7. Estimated “r–g” (Five-Year Ahead) in Emerging
Market Local Bond Market Sensitivity Markets, by Average Credit Rating Band
(Basis points) (Percent)

25

A/A−/BBB+
20

BBB/BBB−
15 (stable outlook)

10 BBB/BBB−
(negative outlook)

5
BB+/BB/BB−

0
All Stress All Stress All Stress All Stress −4 0 4 8
VIX +10 ppt Nonresident Resident banks Resident NBFIs Ex-ante r–g (five-year ahead)
average impacts (mean and +1SD) (mean and +1SD) (mean and +1SD)
Yield spreads (bps) Sources: Bank of England; Bank of Japan; Bloomberg Finance L.P.; Citi; Consensus
Economics; EUROPACE AG/Haver Analytics; European Central Bank; Federal Reserve
Source: IMF staff calculations. Bank; J.P.Morgan; London Stock Exchange Group; and IMF staff calculations.
Note: Bars indicate the estimated impact on yield spreads of a 10-percentage-point Note: The r–g estimates are computed from current 5-year, 10-year, and implied
increase in the VIX, along with the effects of a one standard deviation increase in 5-year forward yields, considering differences in the term premium. Inflation and
investor participation for nonresidents, resident banks, and resident NBFIs. Solid growth estimates are from Consensus Economics or, when unavailable, from World
bars indicate an amplification effect; hollow bars indicate attenuation. Shaded bars Economic Outlook database forecasts. The credit ratings are the median ratings from
indicate statistical insignificance. See Online Annex 3.1 for more information. three international rating agencies. Data include 14 major emerging markets: Brazil,
“Stress” refers to a subsample in which the VIX is above its 75th historical percentile. Chile, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Peru, the Philippines,
The sample is Brazil, China, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Poland, Romania, South Africa, and Thailand. r = long-term real interest rates;
Peru, Poland, Romania, South Africa, Thailand, and Türkiye. bps = basis points; g = long-term growth rates.
NBFIs = nonbank financial intermediaries; ppt = percentage point; VIX = Chicago
Board Options Exchange Volatility Index.

Emerging market government debt has grown margins in some sectors, adversely affect debt-servicing
significantly across most countries, but its structure has abilities, and make stretched corporate equity and
increasingly diverged. Emerging markets with stron- bond valuations vulnerable to corrections. In a scenario
ger economic fundamentals have been able to finance whereby additional tariffs are phased in and at the
debt largely from domestic resident investors in local same time firms face higher refinancing costs, the share
currency (see Chapter 3). The shift toward local gov- of corporate debt with an interest coverage ratio falling
ernment bond markets is empirically associated with below 1 would reach 55 percent in some countries. A
increased resilience to global shocks—an increase in weak tail of firms appears to already be struggling in
the resident investor share is associated with a decline the current environment. Despite the wave of restruc-
in the sensitivity of emerging market bonds to shocks turings, liquidity remains strained among the more
to the VIX, the Chicago Board Options Exchange’s vulnerable borrowers in the leveraged loan and private
Volatility Index (Figure ES.6). Nonetheless, increased credit markets. This has contributed to an increase in
local currency financing may create other fragilities, borrower downgrades.
such as a stronger bank-sovereign nexus. For weaker Stablecoins are growing rapidly and playing a larger
emerging market economies, on the other hand, role in financial intermediation, led by stablecoins
debt service burden is mounting, with long-term real pegged to the US dollar (Figure ES.8). The continued
interest rates (r) that are higher than long-term growth growth of stablecoins could have three main financial
rates (g) (Figure ES.7). That could expose emerg- stability implications: (1) weaker economies may face
ing markets to funding risks, as fiscal consolidation currency substitution and reduced effectiveness of
would be challenging for them (see the October 2025 policy tools, (2) bond market structure could change
Fiscal Monitor). with potential implications on credit disintermedi-
The corporate sector has been resilient so far, ation, and (3) investor runs out of stablecoins may
although tariffs could put pressure on corporate profit generate forced selling of reserve assets. Potential

xiv International Monetary Fund | October 2025

 Executive Summary

Figure ES.8. Stablecoin Market Cap Figure ES.9. Effect of an Increase in Macrofinancial
(Billions of dollars) Uncertainty on Foreign Exchange Market Conditions
(Percentage points, left scale; basis points, right scale)
300
USDC USDT Other
0.16 3.0
Exchange rate return volatility Bid-ask spreads (right scale)
250
2.5
0.12
200
2.0

150
0.08 1.5

100 1.0
0.04
50 0.5

0 0 0
2019 21 22 23 24 25 VIX index US EPU MOVE index

Sources: Bloomberg Finance L.P.; Reuter 2025; and IMF staff calculations. Sources: Baker, Bloom, and Davis 2016; Bloomberg Finance L.P.; LSEG Datastream;
Note: USDC = US Dollar Coin, issued by Circle Internet Group Inc.; USDT = US Dollar and IMF staff calculations.
Tether, issued by Tether Limited. Note: The figure depicts the response of weekly excess exchange rate return volatility
(bars) and bid-ask spreads (diamonds) against the US dollar after large macrofinancial
uncertainty shocks. Macrofinancial uncertainty is measured using the Chicago Board
Options Exchange Volatility Index (VIX), the US Economic Policy Uncertainty (EPU)
systemic effects would be conditional on stablecoins’ index, and the Merrill Lynch Option Volatility Estimate (MOVE) index. Large
uncertainty shocks are defined as dummy variables equal to 1 when the AR(1)
continued growth. residual error term of the respective indicator exceeds two standard deviations.
Despite deep liquidity, global foreign exchange
markets remain vulnerable to episodes of increased
macrofinancial uncertainty. As Chapter 2 shows, flight any monetary easing and maintain their commitment
to quality and increased demand for hedging during to price stability. This cautious approach would also
such periods can raise foreign currency funding costs help temper further valuation pressures on risk assets.
and impair foreign exchange market liquidity, reflected Central bank operational independence remains crucial
in wider bid-ask spreads and heightened exchange for anchoring inflation expectations and enabling
rate return volatility (Figure ES.9). These pressures central banks to achieve their mandates.
may be exacerbated by structural fragilities in the Urgent fiscal adjustments to reduce deficits are cru-
foreign exchange market, including large currency cial to protect the resilience of sovereign bond markets.
mismatches, concentrated dealer activity, and increased High debt and delayed fiscal adjustments in many
NBFI involvement. Strains in foreign exchange market countries could further raise borrowing costs for gov-
conditions can spill over into other asset classes, ernments, underscoring the need for more ambitious
tightening broader financial conditions and potentially fiscal measures to reduce sovereign risks. Improve-
posing risks to macrofinancial stability. Moreover, the ments in market structure—such as expanding central
expansion of foreign exchange trading has heightened clearing for cash bond and repo transactions to lower
settlement risk—the possibility that one party deliv- counterparty risks, improving balance sheet efficiency,
ers currency without receiving the countervalue. and boosting transparency—would also enhance bond
Operational risks to foreign exchange market infra- market resilience. Standing liquidity facilities are vital
structure, such as technical failures and cyberattacks, to backstop these markets.
further threaten market functioning. Even though a softer US dollar has tempered the
external headwinds faced by emerging markets in recent
months, these markets remain vulnerable to changes
Policy Recommendations in investor sentiment. When signs of fragility such as
Macroeconomic stability is crucial to financial rising inflation expectations and surges in exchange rate
stability. For tariffed jurisdictions facing weaker and capital flow volatility are observed, emerging mar-
demand, a gradual easing of the policy rate could be kets should use foreign exchange interventions, mac-
appropriate. For countries where inflation is still above roprudential measures, and capital flow management
target, central banks need to proceed carefully with consistent with the IMF’s Integrated Policy Framework,

International Monetary Fund | October 2025 xv

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

provided that these measures do not impair progress borders. To address liquidity mismatches in investment
on necessary fiscal and monetary adjustments. Further funds, it is key to further improve and expand the
developing local bond markets by enhancing macro- availability and usability of liquidity management
economic fundamentals—such as raising domestic tools. To address the risks that crypto assets such as
financial savings and strengthening fiscal and mone- stablecoins could pose to macroeconomic and financial
tary credibility—is essential to increase debt-carrying stability, policymakers should implement the Financial
capacity. Other policies to deepen emerging market Stability Board’s high-level recommendations, includ-
bond markets include enhancing the predictability ing establishing effective risk-management frameworks,
and transparency of debt issuance, developing efficient safeguarding anti-money laundering/combating the
repo and money markets, strengthening primary dealer financing of terrorism measures, and ensuring that
frameworks, and diversifying the investor base. relevant authorities have the powers they need and can
The IMF’s Global Stress Test underscores the impor- cooperate effectively.
tance of improving capitalization to address risks from To address financial stability risks arising from stress
weak banks. Implementation of internationally agreed- in the foreign exchange market, policymakers should
upon standards that ensure sufficient levels of capital enhance surveillance, including systematic foreign
and liquidity, notably Basel III, is paramount during exchange liquidity stress testing that captures inter-
times of high economic uncertainty. The efficiency of actions with underlying vulnerabilities. It is essential
regulations should be ensured by reviewing any undue to close foreign exchange data gaps and ensure that
complexity without undermining the overall resil- capital and liquidity buffers in financial institutions
ience of the banking sector or international minimum are adequate and supported by robust crisis manage-
standards. National authorities should strengthen the ment frameworks. Strengthening the global financial
financial sector safety net to protect the banking sector safety net—including through sufficient international
against potential financial stability risks from weak reserve buffers and an expanded network of central
banks. This includes establishing emergency liquidity bank swap lines—could help mitigate foreign exchange
assistance frameworks, ensuring that banks can quickly market volatility. This effort would also benefit from
access central bank funding, and advancing recovery a macroeconomic policy mix aligned with the IMF’s
and resolution frameworks to manage shocks without Integrated Policy Framework. Enhancing the opera-
systemic disruption or taxpayer losses. tional resilience of key foreign exchange market par-
Effective regulatory oversight of NBFIs and digital ticipants, including against cyber risks, and promoting
assets such as stablecoins calls for improved data broader use of payment-versus-payment arrangements
collection, coordination, and analysis, including across could further reduce settlement risks.

xvi International Monetary Fund | October 2025

IMF EXECUTIVE BOARD DISCUSSION OF THE OUTLOOK,
SEPTEMBER 2025

The following remarks were made by the Chair at the conclusion of the Executive Board’s discussion of the
Fiscal Monitor, Global Financial Stability Report, and World Economic Outlook on September 29, 2025.

E
xecutive Directors broadly agreed with staff ’s Directors broadly underscored the need to
assessment of the global economic outlook, reinvigorate multilateral cooperation to meaningfully
risks, and policy priorities. They welcomed reduce trade policy uncertainty by re-anchoring trade
the recent economic resilience despite repeated in an open, rules-based and transparent system. They
shocks, noting the importance of stronger economic acknowledged the need to modernize trade rules and
fundamentals and policy frameworks in EMDEs. lower barriers, including through regional agreements
Directors acknowledged, however, that major policy that remain open to and do not discriminate against
shifts are reshaping the global economic landscape third parties. There was general recognition that
and broadly concurred that the recent resilience, also trade diplomacy should work hand in hand with
supported by temporary factors, could be fragile as a coordinated approach to implement domestic
lingering vulnerabilities, elevated policy uncertainty, macroeconomic adjustments and address distortions
and fragmentation continue to weigh on growth behind internal and external imbalances. Attention
prospects. At the same time, a view was held that was also brought to the role of the global financial
staff ’s overall characterization of the global economic safety net in mitigating systemic risks and, in this
environment is overly pessimistic. Directors cautioned regard, the importance of continued progress on Fund
that protectionism and significant cuts to foreign aid concessional resources and a strong, quota-based, and
disproportionately affect the outlook for the world’s adequately resourced IMF at its center.
poorest economies, undermining their convergence Directors highlighted the need for the Fund to
prospects. provide tailored fiscal advice that takes country
Directors broadly concurred that risks to the specific circumstances into account. They stressed
outlook are tilted to the downside, including from the importance of rebuilding fiscal buffers and
prolonged policy uncertainty and any escalation creating space for new spending demands while
in trade tensions, as well as from rising fiscal safeguarding debt sustainability. Directors called for
vulnerabilities, increased fragilities in financial markets, fiscal consolidation with realistic and credible plans
and their potentially adverse interactions. With high that are anchored in robust medium term fiscal
debt service obligations and rollover needs, a continued frameworks and combine spending rationalization and
rise in government borrowing costs would further revenue generation, while protecting the vulnerable.
reduce fiscal space, challenging efforts to rebuild fiscal They emphasized the need to prioritize measures
buffers and making bond market functioning more that raise efficiency of public spending and support
fragile. Directors also acknowledged that stretched sustainable and inclusive private sector led growth,
risk asset valuations and higher interconnectedness while avoiding blanket spending cuts. Where new
between banks and nonbank financial institutions discretionary support is warranted, it should be
(NBFIs) has kept financial stability risks elevated. They transparent, targeted, and temporary. Directors
also recognized the risks stemming from eroding good noted the potential for reforms to pensions, health
governance and the independence of key economic care, wage bills, and tax expenditures to create fiscal
institutions. Labor supply shocks, regional conflicts, room for spending that promotes long run economic
including Russia’s war in Ukraine, and commodity growth. In countries where debt is unsustainable, they
price volatility are additional risks to the outlook. emphasized the importance of cooperation through the

International Monetary Fund | October 2025 xvii

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

G20 Common Framework and the Global Sovereign strengthen regulation of NBFIs and digital assets,
Debt Roundtable to seek timely and orderly debt including stablecoins.
restructuring. Directors acknowledged the importance of boosting
Directors emphasized the importance of central productivity and re-igniting growth over the medium
bank independence and their insulation from political term. They called for comprehensive and carefully
pressures for the anchoring of inflation expectations sequenced structural reform packages, taking into
and the pursuit of price stability in line with their account country-specific circumstances including social
respective mandates. Monetary policy should be data- and political economy considerations. Priority reforms
driven, calibrated to country-specific circumstances— include encouraging labor mobility and participation,
with careful assessment of the nature of shocks and increasing digitalization and AI readiness, and
the output gap—and clearly communicated. In improving the business climate and competition to
economies experiencing supply shocks, a gradual reallocate labor and capital to the most productive
easing of the policy stance should be considered firms. Directors generally welcomed the Fund’s analysis
provided that disinflation is clearly established. Where on industrial policies, with many calling for further
weaker demand dominates, cautious consideration work in this area, including expanding its scope to
can be given to a reduction in policy rates. A prudent include a discussion of spillover risks and related
approach to monetary policy easing can also help policy advice. Directors cautioned that the expanding
contain asset valuation pressures. For countries use of industrial policies involves opportunity costs
experiencing excessive exchange rate volatility and and tradeoffs, including fiscal costs, higher consumer
with shallow foreign exchange markets, the use of prices, and resource misallocation. Where pursued,
temporary foreign exchange interventions and capital industrial policies should be transparent and focus on
flow measures may be appropriate, consistent with the addressing market failures, targeting areas with the
advice of the Integrated Policy Framework, alongside highest potential for positive spillovers and impact on
further deepening local bond markets while managing supply-side capacity and job creation, supported by
risks from the bank-sovereign nexus. Directors also complementary structural reforms. Directors generally
called on the authorities to continue to use their noted that strong governance is key for their successful
macroprudential tools, as appropriate, and generally implementation and called on governments to stay
supported the consistent and timely implementation agile in monitoring their impact and scaling back or
of internationally-agreed regulatory frameworks, like discontinuing ineffective measures. A few Directors
Basel III, to mitigate macro-financial stability risks. also stressed the importance of leveraging historical
It will also be important to address data gaps and experiences in the conduct of industrial policies.

xviii International Monetary Fund | October 2025

 1
CHAPTER
SHIFTING GROUND BENEATH THE CALM: STABILITY CHALLENGES
AMID CHANGES IN FINANCIAL MARKETS

Chapter 1 at a Glance
• Recent months have seen continued appreciation in risk asset prices and a depreciation of the US dollar.
Meanwhile, government debt has continued to rise, and nonbank financial intermediaries (NBFIs) and
stablecoins have continued to grow.
• Markets appear complacent as the ground shifts. Despite trade tensions, geopolitical uncertainties, and
rising concerns about sovereign indebtedness, asset prices have returned to stretched valuations and
financial conditions have broadly eased.
• Although these shifts have been under way in recent years, new evidence points to increasing vulnerabili-
ties in the financial system:
◦ Valuation models indicate that risk asset prices are well above fundamentals, increasing the probability
of disorderly corrections when adverse shocks occur.
◦ Analysis of sovereign bond markets highlights growing pressure from widening fiscal deficits on the
functioning of markets.
◦ Stress tests for banks and NBFIs reveal increasing interconnectedness and persistent maturity
mismatches that could amplify shocks.
• These vulnerabilities reinforce each other, keeping global financial stability risks elevated. For example,
◦ An abrupt yield increase—triggered, for instance, by debt sustainability concerns—could strain banks’
balance sheets and pressure open-ended funds.
◦ Heightened interconnectedness between banks and NBFIs would exacerbate adverse shocks.
• This chapter urges policymakers to
◦ Remain attentive to potential risks to inflation, especially where inflation is still above target, and
preserve central bank operational independence;
◦ Curb government deficits;
◦ Implement internationally agreed-upon prudential standards;
◦ Strengthen financial sector safety nets and NBFI oversight, and
◦ Promote effective regulation and supervision of stablecoins.

Introduction of tariffs on growth and inflation (see the October

The world economy faces persistent trade and 2025 World Economic Outlook) as well as other
geopolitical uncertainties, while structural challenges potential adverse developments. This chapter shows
continue to weigh on medium-term growth. Yet, that beneath the calm surface, the ground is shifting
after a brief jolt from the United States’ April 2 tariff in several parts of the financial system, giving rise to
announcements, global financial markets have largely vulnerabilities. Global financial stability risks remain
brushed off subsequent shocks and uncertainties. Asset elevated, according to the IMF’s growth-at-risk metrics,
prices have rebounded strongly since the April 2025 having receded only modestly since the April 2025
Global Financial Stability Report, and, after an abrupt Global Financial Stability Report.
tightening in early April, financial conditions across The first sign that the ground is shifting appears
regions have eased back to accommodative levels. in asset price movements, covered in the next section,
The apparent calm masks a degree of complacency. “Financial Market Developments and Asset Valuations.”
Markets seem to have downplayed the potential effects The US dollar has depreciated by 10 percent to date

International Monetary Fund | October 2025 1

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

this year, despite April’s sell-off in risk assets and “Corporate Credit Risk,” appear to be struggling in
better-than-expected US economic data in the several an environment of higher tariffs and refinancing rates,
months that followed. This reflects a reassessment of and borrower downgrades and restructurings have
the dollar’s decade-long bull run and increased hedg- risen. Nonetheless, retail investors are increasingly
ing by non-US investors against further weakening. interested in private credit markets and high-yield
Meanwhile, IMF staff models ascertain that valuations bond funds, which could amplify credit downturns.
of risk assets have again become stretched. An abrupt
correction of asset prices could be exacerbated by these
unusual asset correlations and lead to an unwinding Financial Market Developments and
of leverage and straining financial markets. This strain Asset Valuations
could include foreign exchange markets, which have Asset Prices Rebound and Volatility Subsides
undergone structural shifts yet have not experienced amid Elevated Uncertainty
significant dollar weakness (see Chapter 2). Since the April 2025 Global Financial Stability
The second sign of shifting ground is that debt Report, financial markets have largely rebounded from
continues to move toward the government sector. As the broad-based sell-off that followed the April 2 tariff
detailed in the “Sovereign Bond Markets” section, announcement. In part this was because the 90-day
expanding fiscal deficits continue to propel sovereign pause was announced a week later, sequential trade
bond issuance. In advanced economies, sovereign bond agreements then resolved some uncertainty, and global
markets are increasingly dependent on price-sensitive economic data remained solid. Despite the intermittent
investors to buy new issuances. While bond market market gyrations since April from tariff-related news,
functioning has been stable to date, scenario analyses buyers have been ready to step in based on the belief
show that abrupt yield increases would strain bank bal- that any adverse tariff impacts would be temporary and
ance sheets and add liquidity pressures at open-ended eventually reversed. As a result, market volatility across
funds. Stress in core bond markets, although a tail asset classes has declined, on net, in contrast with the
risk, could have broad and disruptive ramifications for still-elevated economic, trade, and geopolitical uncer-
financial markets, given bonds’ role as key benchmarks tainty (Figure 1.1, panel 1). This decline in volatility
and collateral. In emerging markets, governments have has been supported by expectations of further easing of
turned more to domestic investors for financing in monetary policy across most major advanced econo-
recent years. Although this reduces reliance on foreign mies and emerging markets (Figure 1.1, panel 2).1
currency debt, it may also create fragilities such as a At present, the global economy has shown tenuous
stronger bank-sovereign nexus (see Chapter 3). resilience2 (see the October 2025 World Economic
Nonbank financial intermediaries (NBFIs) continue Outlook).3 Nonetheless, with tariffs settling at their
to grow and deepen their ties with banks. The section highest levels in almost a century, a slowdown in the
“Financial Intermediaries” documents the expanding global economy is beginning to emerge as front-loaded
role of NBFIs in core sovereign bond markets and cor- consumption and investments fade. In addition,
porate debt markets, including private credit. Although market expectations for near-term US inflation remain
the IMF’s Global Stress Test (GST) shows that the elevated amid high trade policy uncertainty, whereas
weak tail of global banks has diminished compared
with two years ago, a sizable group of weak banks
1Still, the lasting impact of tariffs on the global economy, particu-
remains, and banks have also become more exposed to
larly in the United States, remains a significant unknown, prompting
NBFIs—heightening interconnections and the fragili- caution in central bank communications.
ties across both sectors. In addition, the global growth 2Specifically, global growth in the first half of the year was larger

of stablecoins could offer investors alternatives to tradi- than predicted in the April 2025 World Economic Outlook, but
higher-frequency indicators for July and August point to drags on
tional safe assets and bank deposits and could influence
global economic activity. In addition, expectations for inflation have
cross-border capital flows. These trends raise the specter been revised upward in the United States but downward in many
of excessive risk taking, rising leverage, and maturity other jurisdictions, consistent with the expectation of a supply shock
mismatch vulnerabilities in the financial system. in the tariffing country and demand shocks in tariffed countries.
3Temporary factors include front-loading of consumption and
While not an imminent financial stability investment, inventory management strategies, implementation delay
risk, weaker firms, as documented in the section of tariffs, and strong profit margins.

2 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.1. Asset Prices amid Still-Elevated Uncertainty

Financial market volatility has declined since Monetary policy easing is expected to continue, Market expectations of inflation depend on oil
April, but measures of economic uncertainty although with divergence across countries. prices and trade policy uncertainty.
remain elevated.
1. Economic Uncertainty and 2. Advanced Economy and Emerging Market 3. Market-Based Inflation Expectations versus
Financial Volatility Interest Rate Paths Oil Price and Trade Policy Uncertainty
(Percentile) (Percent) (Percentile, left scale; percent, right scale)
October 2024 GFSR April 2025 GFSR United States Euro area Crude oil futures daily percentile (30D MA)
Latest Average postpandemic United Kingdom EM average (ex. BRL) Trade Policy Uncertainty daily percentile
Economic uncertainty measures BRL (right scale) (30D MA)
12 16 EUR two-year inflation swap (right scale)
100 USD two-year inflation swap (right scale)
10 15 100 3.25
50 8 14 80 2.25
60 1.25
6 13 40
0 0.25
4 12 20 −0.25
Trade Policy
Uncertainty
Economic
Policy
Uncertainty
Geopolitical
Risk Index

VIX Index

MOVE
Index

Jan. 2023
Apr. 2023
Jul. 2023
Oct. 2023
Jan. 2024
Apr. 2024
July 2024
Oct. 2024
Jan. 2025
Apr. 2025
Jul. 2025
Oct. 2025
2 11
0 10
2022 23 24 25 26

Despite economic surprises trending negative ... equity prices globally have risen, largely on Corporate spreads have narrowed.
for several months following the April sell-off, the back of outperformance in AI-related sectors,
investor sentiment has continued to improve ... particularly the M7.
4. Economic Surprise and Investment 5. Performance of Global Equities, by Selected 6. Investment-Grade and High-Yield Corporate
Sentiment Indices in the United States, 2025 Economies and Sector, 2025 Bond Spreads, 2025
(Z-score, left scale; percent, right scale) (Normalized, January 1, 2025 = 100; (Basis points)
percentage points)
Bloomberg Surprise Index US IG Europe IG UK IG
AAII US investor sentiment (right scale) United States Europe US HY Europe HY UK HY
0.3 45 Japan World 500
April EM ex. China China
40 April 2 tariff April 9 tariff pause 125
0.2
announcements
35 400
100
0.1
30
75 300
0 25 Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct.

−0.1 20 Jan. 1, 2025 Apr. 8, 2025 Net April 2 tariff April 9 tariff pause
to Apr. 8, 2025 to latest announcements 200
15 60
−0.2 30
10 10
0 100
−0.3 −10
5
−30
−0.4 0 0
M7
Info tech
Communication
services
Industrials
Materials
Consumer
discretionary
Financials
Utilities
Energy
Real estate
Health care
Consumer
staples

Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct.

Sources: American Association of Individual Investors; Bloomberg Finance L.P.; IBES DataStream; Baker, Bloom, and Davis 2016; Caldara and others 2020; Caldara and
Iacoviello 2022; and IMF staff calculations.
Note: In panel 1, percentiles are derived from monthly data starting January 1997. “Average postpandemic” is the average percentile since January 2022. “Economic Policy
Uncertainty” and “Trade Policy Uncertainty” are the indices of Baker, Bloom, and Davis (2016); the Geopolitical Risk Index is from Caldara and Iacoviello (2022). Economic
Uncertainty Measures are text-based. The latest level for economic uncertainty measures is the latest available for each corresponding monthly series. The latest levels for
the VIX and the MOVE indices are as of October 2, 2025. Solid lines in panel 2 are actual central bank policy rates. Dotted lines are forecast future policy rates derived from
swap curves. The average emerging market central bank (excluding Brazil) includes India, Hungary, Mexico, and South Africa. In panel 3, Trade Policy Uncertainty (30D MA)
is the 30-day moving average of percentiles calculated from the entire daily series in Caldara and others (2020). Crude oil futures (30D MA) correspond to the 30-day
moving average of percentiles calculated from the third generic crude oil futures contracts for West Texas Intermediate, due to expire in around three months from the date
of this publication. In panel 4, the Bloomberg Surprise Index is the Bloomberg ECO Surprise Index for the United States. The values are z-scores, representing the number
of standard deviations that analysts’ expectations, as surveyed by Bloomberg, are above or below normal surprise levels. The AAII index is compiled from the AAII weekly;
data indicate how bullish the surveyed members feel about equity markets in the next six months. Panel 5 uses the S&P 500 Index for the United States, Euro Stoxx 600 for
Europe, Topix Index for Japan, MSCI All Country World Index for World, MSCI Emerging Market excluding China for emerging markets excluding China, and Shanghai
Shenzhen CSI 300 Index for China. Global Industry Classification Standard level 1 sectors are used for the MSCI All Country World Index. Panel 6 uses option-adjusted
spreads. Data labels in the figure use International Organization for Standardization (ISO) country codes. AAII = American Association of Individual Investors; AI = artificial
intelligence; EM = emerging market; ex. = excluding; GFSR = Global Financial Stability Report; HY = high yield; IG = investment grade; M7 = Magnificent 7; MOVE =
Merrill Lynch Option Volatility Estimate; VIX = Chicago Board Options Exchange Volatility Index.

International Monetary Fund | October 2025 3

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

euro area inflation expectations have anchored as oil the section “Expanding Fiscal Deficits Exert Pressure on
prices have declined (Figure 1.1, panel 3). Although Bond Market Stability”).
inflation effects from trade policy are expected to be One noteworthy development has been the weak-
largely temporary, as indicated by inflation swaps ness in the US dollar against a basket of both G10
pricing, it is now more likely that tariffs may be set- and emerging market currencies. This has persisted
tling at high levels for an extended period.4 As a result, for several months after the April tariff announcement
exporters around the world who are most affected by despite a strong rally in risk assets as well as rising gold
tariffs could gradually be shifting some tariff-related prices, even with a wide differential between US and
costs onto consumers to mitigate pressures on their G10 interest rates that had supported the dollar in
profit margins (see the section “The Corporate Sector recent years (Figure 1.2, panel 2). Overall, the dollar
Is Resilient to Tariffs So Far”).5 has depreciated by about 10 percent so far this year
Despite recent economic data surprises trending neg- against major currencies. Analysts have put forth a
ative until more recently, equity market sentiments have number of possible drivers for dollar weakness, from a
continued to remain high (Figure 1.1, panel 4), buoyed revaluation of dollar strength amid concerns over the
by optimism about mega-cap stocks related to infor- US fiscal position to a shift in allocation away from
mation technology (IT) and artificial intelligence (AI) US-dollar-denominated assets driven by concerns
(Figure 1.1, panel 5), which are perceived to be less neg- about US policy uncertainty. Although cross-border
atively affected by tariffs. Corporate credit spreads have data do not support notions of a broad pullback in
tightened since April (Figure 1.1, panel 6). Given these non-US investor holdings,7 increased currency hedging
developments, equity and corporate credit valuations activity to mitigate losses on unhedged dollar exposure
have returned to being fairly stretched, and concen- appears to have emerged as a contributor to recent
tration of valuations at a handful of firms—especially dollar weakness.
the Magnificent 7 and AI-related stocks in the broad By way of context, non-US investors’ holdings of
benchmark equity index—is at historical highs (see the US dollar assets have risen steadily over time, with a
section “Equity Markets Exhibit High Valuations and large portion not matched by commensurate dollar
Concentration Risks”).6 liabilities. For example, total non-US investor hold-
ings of US securities increased from $16 trillion to
$31 trillion from 2015 to 2024. These holdings are
The Dollar, Bonds, and Risk Assets Diverge characterized by incomplete hedging, given the foreign
Since the April 2025 Global Financial Stability Report, exchange mismatches they present, and could be sub-
longer-term sovereign bond yields in most advanced ject to sudden, large-scale sell-offs (and therefore can
economies have risen, even as investors expect monetary be deemed “need to be hedged”). This rise in expo-
policy to continue to ease. Term premiums have been sure has been driven by macroeconomic factors, such
driven up by a rising bond supply, and there has been as current account surpluses, savings gluts, relatively
ongoing quantitative tightening by central banks as limited investment opportunities in non-US markets,
well as a slowdown in duration demand, including by and yield-seeking behavior. The mutually reinforcing
liability-driven investors (Figure 1.2, panel 1; see also dynamics between trade and finance are facilitated by
the unparalleled depth and liquidity of US financial
markets (see Chapter 2 of the July 2025 External Sector
4Pricing for two-year, three-year, and five-year inflation swaps for
Report).
the United States, along with the five-year inflation swap measure,
suggests that medium and longer-term inflation expectations, while
Although currency hedging can mitigate risks
more elevated since April, have not become unanchored so far. associated with incomplete hedging, the modest depth
5As noted in the October 2025 World Economic Outlook, tariffs
of foreign exchange markets in many jurisdictions
theoretically lead to currency appreciation for the tariff-imposing
relative to large dollar asset exposures and the dollar’s
country, mitigating the impact of tariffs on prices. However, US
dollar appreciation has not happened to date. Instead, dollar depre- strength over the past decade have made hedging
ciation may mean that exporters have less room to absorb tariffs
without a deterioration in their profits, thus leading to pass-through 7After a brief period of outflows in April, Treasury securities

to importing firms and consumers. experienced net inflows of about $105.5 billion. US equity net
6The Magnificent 7 companies are Alphabet, Amazon, Apple, inflows were $95.4 billion over April and May, according to Treasury
Meta, Microsoft, Nvidia, and Tesla. International Capital System data.

4 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.2. Some Developing Divergences

Longer-term bond yields have seen upward pressure from higher term The US dollar has weakened, while also having decoupled relative to
premiums. interest rate differentials for several months after the April tariff
announcement.
1. Decomposition of Changes in Longer-Term Bond Yields since Early 2. Dollar versus Interest Rate Differentials
April 2025 in Selected Advanced Economies (Index, left scale; percent, right scale)
(Basis points)
50 Average expected short term 110 2
April 2 tariff
Term premium 1.9
40 announcements
Change in fitted yield 1.8
30 1.7
105
20 1.6
1.5
10 1.4
100 1.3
0
DXY 1.2
−10 US-G10 average (right scale) 1.1
−20 95 1
Jan. Apr. Jul. Oct. Jan. Apr. Jul. Oct.
10-year
5-year, 5-year

10-year
5-year, 5-year

10-year
5-year, 5-year

10-year
5-year, 5-year

10-year
5-year, 5-year
2024 2024 2024 2024 2025 2025 2025 2025

AUS DEU GBR JPN USA

The optimal currency hedging ratio has US dollar asset holdings are large across Jurisdictions with large dollar asset exposures
increased recently, an indication that non-US jurisdictions in absolute terms as well as relative relative to the size of FX markets have wider CIP
investors ought to increase their hedging to FX market depth. deviations, resulting in tighter dollar funding
against further dollar depreciation. conditions.
3. Optimal Volatility-Minimizing FX Hedge 4. US Dollar Exposures and Ratio Relative to 5. US Dollar Asset Exposures Scaled by FX
Ratio for Non-US Investors, by Asset Class the Size of Local FX Markets Transaction Volume and CIP Deviations
(Percent) (Trillions of dollars, left scale; ratio, scaled by across Regions
local FX market monthly transaction volume, (Basis points, y-axis; ratio, x-axis)
right scale)
Range of 11 currencies
Median USD assets USD assets, scaled by
100 Equities Treasury 100 6 FX volumes (right scale) 50
45 20
80 80 5 40
4 35 0

CIP deviations
60 60 30
3 25 −20
40 40 20
2 −40
15
20 20 1 10 −60
5
0 0 0 0 −80
2022 23 24 25 2022 23 24 25 0 10 20 30 40 50 60
CYM
JPN
GBR
CAN
LUX
IRL
HKG
DEU
NLD
CHE
TWN
KOR
FRA
NOR
AUS
BMU
ITA
ESP
MEX
DNK

USD assets scaled by FX volume

Sources: Bloomberg Finance L.P., BNP Paribas; Bank for International Settlements; US Department of the Treasury; and IMF staff calculations.
Note: The 10-year yield can be split into different time horizons, as different factors may be at work over the near and medium versus the longer term. In panel 1, the
5-year-5-year forward conveys information contained within the latter half of a 10-year bond’s maturity—that is, spanning years 6 to 10—thus parsing out the influence of
cyclical factors that may be predominant drivers at shorter-term horizons. Term premium estimates follow Adrian, Crump, and Moench (2013). Early April refers specifically
to April 3. In panel 2, the US–G10 average is the nominal 10-year interest rate differential between the United States and the average of the G10 countries. In panel 3,
minimum return volatility hedge ratios by asset classes are estimated based on Wilcock and others (2025), with two-year rolling volatility and correlation of each local
currency exchange rate against the US dollar and S&P 500 index (equities) or J.P.Morgan Global Bond Index local currency US country index (Treasury). The shaded areas
indicate the range of the estimated hedge ratio for 11 currencies: British pound sterling, Japanese yen, Canadian dollar, euro, Chinese yuan, Hong Kong dollar, Swiss
franc, new Taiwan dollar, Australian dollar, Norwegian krone, and Republic of Korean won. In panel 4, US dollar exposures are estimated by focusing on cross-border
portfolio investments, loans, and deposits, using multiple databases, as of 2023 (see Online Annex 1.6 for details). Panel 5 plots the average CIP deviations versus US
dollar exposures scaled by the monthly FX transaction volume for 12 currencies, including advanced economies and emerging market economies. CIP deviations are
computed as the 12-year average of the difference between the three-month US interest rate and the foreign country’s interest rate, adjusted by the annualized forward
premium. Data labels in the figure use International Organization for Standardization (ISO) country codes. CIP = covered interest parity; DXY = US dollar index; FX =
foreign exchange; USD = US dollar.

International Monetary Fund | October 2025 5

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

expensive. Globally, hedge ratios for insurance com- on these exposures. Such hedging after dollar weakness
panies, pension funds, and mutual funds—which would involve selling US dollars forward or repatriat-
invest in dollar assets and have mostly local-currency-­ ing dollar deposits, so amplifying dollar weakening in a
denominated liabilities—are found to be considerably self-fulfilling manner. Consequently, “rush to currency
less than 100 percent, as evidenced in, for instance, hedge” behavior reportedly increased in the months
Du and Huber (2024) and Shin, Wooldridge, and after the April 2 tariff announcements (Parsons and
Xia (2025).8 Davis 2025).
More recently, the case to increase hedge ratios has According to IMF staff analysis, these hedging
strengthened. A measure of optimal hedge ratio from a assets include security portfolios, loans, and deposits,
non-US investor’s perspective—based on the mini- and they are especially large in international finan-
mization of asset return volatility—has significantly cial centers and jurisdictions with large NBFIs. In
increased recently for portfolios of both US equities some economies, dollar exposures are disproportion-
and Treasuries across currencies (Figure 1.2, panel 3).9 ately large relative to the depth of the local foreign
Many investors appear underhedged compared with exchange market (Figure 1.2, panel 4; see also Online
this optimal ratio (Shin, Wooldridge, and Xia 2025). Annex 1.6).
Amid dollar weakening, non-US investors with signif- The financial stability risk of a “rush to hedge”
icant unhedged dollar exposures could be prompted is that selling the US dollar forward could increase
to increase currency hedging to mitigate further losses dollar funding pressures, especially in shallower foreign
exchange markets with limited hedging instruments
8Based on evidence provided in Du and Huber (2024), hedge
and where absorption capacity for hedging flows is
ratios for insurers, pension funds, and mutual funds stood at around
44 percent, 35 percent, and 21 percent, respectively, as of June 2020, lower. With many foreign investors selling US dollars
suggesting incomplete currency hedging. More recent estimates of forward, the relative price of the US dollar forward
hedge ratios reveal that these could, in comparison, be even more versus spot would decline, resulting in larger devia-
conservative, as evidenced in Shin, Wooldridge, and Xia (2025).
Hedging practices also vary substantially across investor types and
tions from covered interest parity (CIP), an indicator
countries depending on investment objectives and risk tolerance. for dollar funding pressures. Indeed, jurisdictions
Japanese life insurers, for instance, own a sizable amount of US with larger US dollar asset exposure relative to foreign
dollar assets against yen-denominated liabilities, which in principle
exchange market depth currently have wider CIP
needs to be hedged to minimize currency mismatches. In practice,
they typically hedge 50 percent to 70 percent of their bond portfo- deviations (Figure 1.2, panel 5).10,11
lios (see McGuire and others 2021), as a 100 percent currency hedge
may not necessarily be optimal from a risk-management perspective.
Other long-term investors, and more specifically pension funds, Equity Markets Exhibit High Valuations and
may decide to not actively currency hedge their dollar exposure, but
may instead change allocation to dollar assets through multiyear
Concentration Risks
strategy asset allocation reviews, depending on cross-asset correla- The rebound in global equity prices since April
tions, liquidity conditions, and hedging costs; as well as discretionary
views about the market, particularly, the trajectory of the dollar.
has outpaced expected future earnings, reflecting
Some dollar exposures are not necessarily actively managed and thus buoyant investor sentiment (see the section “Asset
remain relatively insensitive to market developments. For instance, Prices Rebound and Volatility Subsides amid Elevated
dollar exposures associated with a non-US firm’s direct investment
Uncertainty”). In particular, the S&P 500 12-month
in the United States, where the firm has its operations, may not
be currency hedged. Another example is foreign reserve buffers forward price-to-earnings (P/E) ratio has climbed back
held by monetary authorities, which are not held for investment to about the 96th percentile since 1990 while con-
returns, but rather serve as a first line of defense against excessive tinuing to trade at a premium compared with other
exchange volatility and funding pressures. In a longer time horizon,
a decline in direct investment in the United States or a decrease in advanced and emerging markets (Figure 1.3, panel 1).
the dollar’s share of foreign exchange reserves could contribute to the
trend of dollar weakening. In fact, the US dollar share of interna-
tional reserves has declined since the turn of the century, reflecting 10Two related investigations in the literature are those conducted

portfolio diversification by central bank reserve managers, potentially by Du and Huber (2024) and Dao and Gourinchas (2025). Du and
exerting downward pressure on the dollar over time (Arslanalp, Huber (2024) document the strong correlation between hedging
Eichengreen, and Simpson-Bell 2022). US dollar exposures in Figure activity and cross-currency covered interest parity (CIP) deviations.
1.2, panels 1 and 4, are not aimed to include direct investments and Dao and Gourinchas (2025) also uncover the relationship between
foreign exchange reserves. the difference between external dollar assets and liabilities and CIP
9From 2021 to 2023, when the dollar strengthened and the deviations.
correlation between local currencies and dollar assets was higher, the 11In some cases (for example, the Taiwan dollar), offshore forward

hedge ratio needed to minimize asset returns from a non-US investor transactions are nondeliverable and so are not used for dollar
perspective was low relative to the current level. funding.

6 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.3. Equity Valuation Pressures

Global equity valuations rebounded with the US equity valuations appear stretched relative to fundamentals, on the basis of staff estimates, but ...
tariff pause.
1. Twelve-Month-Forward P/E Ratios 2. Twelve-Month-Forward P/E Ratios versus Model-Implied Estimates
(Percentile since 1990) (Percentiles since 1990)
United States AEs ex. US EMs Range 2 SD 1 SD Actual Fair value
100 100 100
90 90
80
80 80
70
70 60
60 60 50
50 Overvaluation 40
30
40 40
20
30 10
20 20 0
April 2 tariff April 9

Jan. 2025
Feb. 2025
Mar. 2025
Apr. 2025
May 2025
Jun. 2025
Jul. 2025
Aug. 2025
Sep. 2025
10 announcement tariff pause
0 0
Jan. 2024 Jul. 2024 Jan. 2025 Jul. 2025 1995 2000 05 10 15 20 25

... current overvaluation is still below historical Concentration risk, however, has risen to Concentration risk is far less pronounced in other
peaks, for example, during the dot-com bubble. historically high levels, with a narrow group of major global indices.
IT- and AI-related stocks predominantly driving
the S&P 500.

3. Kernal Density of Historical Over- (Under-) 4. Concentration Risk in the S&P 500 5. Comparing Normalized Concentration Risk
Valuations since 1995 (12-month-forward P/E ratio, percent, left scale; across Selected Indices
(Density, y-axis; percentile points, x-axis) inverse Herfindahl-Hirschman index (Percentage point change since 2015, left scale;
[concentration risk], right scale) current level, right scale)
Latest S&P 500 P/E ratio Change since 2015
Peak dot-com bubble S&P 500 tech P/E ratio 30 Latest (right scale, inverted) 0
Pre DeepSeek Concentration risk (right scale, inverted)
0.035 50 40 20
10
More concentration concentration
Higher

0.025 concentration 70 10
Less 20
concentration 0
0.015 30 100
30
Lower

−10
0.005 130
−20 40
Under valuation Over valuation S&P Euro Topix FTSE CSI
−0.005 10 160 500 Stoxx 500 100 300
1990 95 2000 05 10 15 20 25 ex. UK
−30
−25
−20
−15
−10
−5
0
5
10
15
20
25

Sources: Bloomberg Finance L.P.; LSEG DataStream; and IMF staff calculations.
Note: Panel 1 shows the percentiles of 12-month-forward (P/E) ratios since 1990 or the start of the data series. AEs (excluding the United States) and EMs are MSCI series,
while the United States is the S&P 500. Panel 2 compares the US P/E series from panel 1 with the central tendency of distribution of model-implied, fair-value estimates.
Estimates are based on weekly data. The model used here is based on an extended equity market asset valuation model discussed in Online Annex 1.1 of the October 2019
Global Financial Stability Report, which relates equity index on various proxies for earnings growth, equity risk premium, and term premium. For the analysis discussed
here, conditioning variables (proxies) are shuffled over multiple configurations used for parameter estimation through a bootstrapping methodology. The methodology
delivers a distribution of 3,600 fair-value estimates at each point in time, based on randomized sampling within the preceding five years of weekly data. The light gray
shaded region shows the range of estimates, and the dark gray shaded region shows estimates within two standard deviations around the mode. The black shaded region
shows estimates within one standard deviation around the mode. The fair-value series (green line) is the R2 weighted value of all estimates at each point in time. The
distribution’s mode generally closely tracks the estimated fair value. Panel 3 shows the model estimation errors, calculated as the percent difference between actual and
fair-value estimates, for the entire time series. Positive (negative) error (or deviation) indicates overvaluation (undervaluation). The vertical lines flag this error at each
corresponding point. In panel 4, the green dotted line (concentration risk) is the inverse of the Herfindahl-Hirschman Index (with the y-axis in reverse order) and measures
the effective number of equal weighted stocks driving the index. Normalized concentration risk in panel 5 is the inverse of the Herfindahl-Hirschman Index divided by the
total number of constituents within the corresponding index. AEs = advanced economies; AI = artificial intelligence; CSI = China Securities Index; EMs = emerging
markets; ex. = excluding; FTSE = Financial Times Stock Exchange; IT = information technology; P/E = price/earnings.

International Monetary Fund | October 2025 7

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

In view of the structure of the economy and equity sharp correction.14 Valuations would collapse as a
markets having evolved significantly over the past result, making the broader benchmark index vulnerable
few decades, a simple historical comparison of the to downturns.15
forward P/E ratios may not provide the most adequate Expected returns and valuations depend on expec-
assessment of valuations. IMF staff therefore estimate tations for corporate profitability. Investors typically
a large set of equity valuation models to ascertain a regard higher expected profit margins as a positive
possible range of fundamentals-based valuation for signal about the quality and sustainability of earnings,
the S&P 500. which tends to drive up equity prices. In general,
Model estimates suggest that the fair-value forward tariffs on imports should increase the cost of goods
P/E ratio should be about the 81st historical percen- sold, leading firms to either absorb the costs, thereby
tile (Figure 1.3, panel 2, green line). Comparing this directly impacting profit margins, or pass them on to
model-implied fair value with actual observed forward consumers as higher prices.
P/E suggests that the equity valuation is currently Over this year, stock analysts have meaningfully
stretched, with an estimated overvaluation of about revised down expected profit margins for most firms.
10 percentage points. However, in several past episodes By contrast, margins for the Magnificent 7 have been
the overvaluation was even higher; for instance, during revised up (Figure 1.4, panel 1), suggesting that tariffs
the dot-com bubble in the early 2000s (Figure 1.3, are not perceived to impact these companies as much
panel 3). as they may hurt other firms. A forward-looking risk
Of greater concern, concentration risk within the is for the effects of tariffs to eventually lead to margin
S&P 500 is at a historic high, with a narrow group compression across most S&P 500 sectors, including
of stocks spanning mega-cap IT and AI-related firms the Magnificent 7.
driving the broader index. The IT sector accounts Looking across regions, some analysts have revised
for a weight of 35 percent of the total S&P 500, down year-end profit margins on the assumption that
similar to during the dot-com bubble, but with the the full impact of tariffs has yet to percolate through the
Magnificent 7 alone accounting for 33 percent of global economy. Although current profit margins are
the index. Consequently, a measure of concentration high compared with median levels over the past decade
risk based on the Herfindahl-Hirschman Index is (Figure 1.4, panel 2), expectations of lower profit
now substantially higher than during the dot-com
bubble (Figure 1.3, panel 4, green line). Furthermore,
while concentration risk for the S&P 500 index has 14As discussed in the April 2025 Global Financial Stability Report,

in an environment of stretched valuations, materialization of an

witnessed an increase by about 20 percentage points
adverse shock could sour investor sentiment, triggering a sudden
over the past decade (when normalized by number stock market correction that reflects a sharp repricing of risk. This
index constituents) comparable benchmark indices in repricing may be amplified by abrupt deleveraging of balance sheets,
different jurisdictions have been characterized by far involving asset fire sales to meet margin calls or satisfy risk limits
amid a spike in market volatility (see Adrian, Malik, and Wu [2024]
less of an increase over the same period (Figure 1.3, for a recent example of such a mechanism playing out). As equity
panel 5).12 Against substantial AI-related investments valuations collapse, adverse spillovers to the wider market could also
(for example, information-processing equipment, data occur because of price correlations across different asset classes. In
this context, as the net worth of borrowers falls at an accelerated rate
centers)13 the possibility of mega-cap stocks failing and the risk management constraints of lenders become increasingly
to generate expected returns to justify current lofty binding, credit provision to the wider economy can be significantly
equity valuations could trigger deterioration in investor impaired, eventually weighing on output. Overall, a market sell-off
can be exacerbated by a negative feedback loop playing out between
sentiment and make the stocks susceptible to sudden,
pricing of risk and deleveraging, resulting in tightening financial
conditions, with sharp, possibly nonlinear, declines in economic
activity (Kiyotaki and Moore 1997; Bernanke, Gertler, and Gilchrist
1999; Brunnermeier, Eisenbach, and Sannikov 2013).
12Concentration risk is measured as the inverse of the 15This high concentration comes alongside historically elevated

Herfindahl-Hirschman Index. US household exposure to equities (as a share of total household

13Private fixed investment in information-processing equipment, assets), currently about 30 percent and on an upward trajectory since
which can be used a proxy for AI investments in data centers, has the global financial crisis (see the April 2025 Global Financial Sta-
contributed around 57 percent of US real GDP growth since the bility Report for a discussion). A major portion of rising household
fourth quarter of 2024 (evidence based on data sourced from Federal exposure is to benchmark indices, in particular the S&P 500 (largely
Reserve Economic Data and US Bureau of Economic Analysis; see in 401k retirement accounts and through passive investment vehicles
also Paul Krugman, “About That Stock Market,” August 6, 2025, and exchange-traded funds). This exposure makes household balance
and “What Happens if AI Hits an Energy Wall?,” August 19, 2025, sheets vulnerable to sharp corrections and prolonged declines in the
https://paulkrugman.substack.com/.). index, potentially more so currently, given high concentration.
8 International Monetary Fund | October 2025
 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.4. Expected Profit Margins Have Been Revised Down in Most Cases

Analysts have divergent profit margin Looking across regions, margins are now higher ... but declines in expected profit margins after
expectations for the M7 and the rest of the than in the past decade ... the April 2 tariff announcement may weigh on
S&P 500. prices and valuations.
1. Analysts’ Revisions of Expected Profit 2. Corporate Profit Margins for Selected 3. Analyst Revisions of Expected Corporate
Margins for the M7 and S&P 493 Regions Profit Margins for Selected Regions
(Expected year-end 2025, normalized; (Percent) (Percentage point change of expected year-end
100 = January 1, 2025) Range since 2015 Median Latest 2025 profit margins since April 2)
106 16 0.2
S&P 493 profit margins
M7 profit margins 14 0
104
12 −0.2
102 10
−0.4
100 8
−0.6
6
98 4 −0.8
96 2 −1
0 −1.2
94
United States
Europe
United Kingdom
Japan

Brazil
Canada
Brazil
Mexico
China
India
Hungary

United States

Canada
China
Japan

Euro area
United Kingdom

Mexico
Hungary
India
92
Jan. Feb. Mar. Apr. May Jun. Jul.
2025 2025 2025 2025 2025 2025 2025

Advanced economies Emerging markets Advanced economies Emerging markets

Sources: Bloomberg Finance L.P.; LSEG DataStream; and IMF staff calculations.
Note: Panel 1 shows the year-end 2025 earnings and profit margin estimates for M7 and for S&P 500 companies excluding the M7 (the S&P 493). The M7 is Alphabet,
Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla. Expected earnings are calculated as the sum of expected year-end 2025 net income for all companies in the M7 and
S&P 493, respectively. Profit margins are calculated as the sum of expected year-end 2025 revenue divided by net income. The series are normalized to equal 100 on
January 1, 2025. Panel 2 shows the range of quarterly 12-month trailing profit margins since 2015. Panel 3 bars depict the percentage point change in analysts’ estimates
for year-end 2025 profit margins since April 2. M7 = Magnificent 7.

margins may weigh on equity prices and valuations for Financial Conditions Ease, but the
these regions over the near term (Figure 1.4, panel 3).16 Growth-at-Risk Metric Remains Elevated
16While
The rebound of asset prices globally since the
tariffs on internationally sourced inputs can lead
to higher costs for product-based companies, it is argued that April 2025 Global Financial Stability Report, along-
service-based companies—for example, related to AI or encompassed side a weaker dollar, have eased financial conditions
within the Magnificent 7—can also be adversely affected by tariffs around the world (Figure 1.5, panel 1). The abrupt
weighing on their margins. A tariff-related increase in direct costs
necessary for service delivery could lead to higher cost of goods sold
tightening in financial conditions after the April 2
(that is, covering products and services), possibly including any raw tariff announcement proved short-lived, as financial
materials, labor, outsourced services, equipment, and technology, conditions in the euro area and the United States
among others. In addition, investments in capital expenditure may
have returned to levels immediately before the event.
be exposed to rising costs if any of the inputs are subject to tariffs.
Specifically, tariffs impacting major investment in AI-related infra- Conditions in other advanced economies and emerg-
structure, including semiconductors and data centers, could com- ing markets (including China) have become more
press margins for AI service providers. More specifically, data center accommodative. For the United States, although the
inputs include hardware such as server and storage arrays (chips),
networking equipment (switches, routers, fiber optics), and power softening of real estate prices has continued (Box 1.1,
systems (uninterruptible power supplies, generators, transformers); an improvement in corporate valuations (equity
and infrastructure such as real estate, cooling systems, software, prices, corporate bond spreads) amid falling volatility
racks, and cabling (as well as utilities and labor). To date, tariffs have
been initiated on semiconductors (100 percent), steel and aluminum
drove financial conditions back into easy territory by
(50 percent), and copper (50 percent), all key inputs of data center historical standards (Figure 1.5, panel 2). In emerging
infrastructure. Firms providing the infrastructure will likely raise markets (including China), external financing risks
costs to firms providing AI services or purchasing the infrastructure.
have lowered amid a weaker dollar, which has eased
Sector-specific restrictions, such as the Digital Service Tax in the
European Union, can decrease revenues for firms providing services. their financial conditions. That said, the Financial
Last, the latest earnings reports from some AI-related firms have Conditions Index for China does not capture the
highlighted that tariffs and new export controls have raised costs recent slowdown in bank lending (Box 1.2).
throughout their supply chains, further indicating that these could
undermine financial performance (see, for example, Nvidia Corpora- The IMF’s updated growth-at-risk (GaR) assess-
tion’s 10-Q filing for the period ending July 27, 2025). ment reveals that near-term downside risks to financial
International Monetary Fund | October 2025 9
 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.5. Financial Conditions Index

Financial conditions have eased significantly since early April... ... amid rising corporate valuations and falling volatility.
1. Financial Conditions Index 2. Key Drivers of the Financial Conditions Index
(Number of standard deviations over the long-term averages) (Contribution to standard deviations over the long-term averages)
United States Euro area Other AEs Interest rates Corporate valuations House prices
China EMs ex. China External financing risks Aggregate
1.0 1.0 0.6
Apr. 2025 GFSR United Euro area Other AEs China EMs ex.
States China 0.4
0.5 0.5
0.2

0 0 0
−0.2
−0.5 −0.5 −0.4
−0.6
−1.0 −1.0
−0.8
−1.5 −1.5 −1.0
25:Q1

25:Q1
25:Q2
25:Q3

2024:Q3
24:Q4
25:Q1
25:Q2
25:Q3
23:Q4
24:Q1
24:Q2
24:Q3
24:Q4

25:Q1
25:Q2
25:Q3

2024:Q3
24:Q4
25:Q1
25:Q2
25:Q3

2024:Q3
24:Q4
2025:Q1
May 4
Jun. 3
Jul. 3
Aug. 2
Sep. 1
22:Q4
23:Q1
23:Q2
23:Q3

Oct. 1

2024:Q3
24:Q4
25:Q1
25:Q2
25:Q3

2024:Q3
24:Q4
2022:Q1
22:Q2
22:Q3

Sources: Bloomberg Finance L.P.; Dealogic; EUROPACE AG/Haver Analytics; national data sources; and IMF staff calculations.
Note: The IMF FCI is designed to capture the pricing of risk. It incorporates various pricing indicators, including real house prices. Balance sheet or credit growth metrics are
not included. A lower FCI implies easier financial conditions and vice versa. For details, see Online Annex 1.1 in the October 2018 Global Financial Stability Report. In panel 1,
the shaded area on the right side shows the daily FCIs starting April 1, 2025. These daily FCIs are approximate values that are estimated using available high-frequency
market data, whereas the long-term standard deviations and averages are calculated over Q1 1990 and Q3 2025. In panel 2, the key drivers of the FCI show the contributions
of underlying components, which are the weighted average of the z-scores of these components. The series “aggregate” represents the sum of these contributions and is
similar to the FCI values shown in panel 1. The series “Since April GFSR” shows a simple average of aggregated z-scores and their drivers between April 12 and September 11,
2025. AEs = advanced economies; EMs = emerging markets; ex. = excluding; FCI = Financial Conditions Index; GFSR = Global Financial Stability Report.

stability have declined since the April 2025 Global cial markets eased. Although dollar depreciation
Financial Stability Report, albeit slightly. Easier global may reduce the value of emerging market residents’
financial conditions were partially offset by a slight holdings of dollar assets, it has also alleviated pres-
slowdown in already sluggish private sector credit sures on emerging markets’ asset and funding markets
growth, which has shifted just below the 10th per- (Figure 1.7, panel 1). Subdued energy prices have
centile of its historical distribution. The current GaR provided some relief by containing import costs and
metric suggests that one-year-ahead global growth is reducing external vulnerabilities, particularly for energy
forecast to fall below 0.5 percent, with a 5 percent importers. In addition, steady progress on disinflation
chance (Figure 1.6, panel 1, blue dot). Although this has allowed several emerging market central banks to
reflects a 0.1 percentage point improvement in the ease policy rates and so further support domestic finan-
GaR metric compared with April (red dot), it is still cial conditions. Nonetheless, several emerging market
around the 30th historical percentile, suggesting that central banks have been cautious in easing policy rates,
risks are still above historical standards (Figure 1.6, with rate cuts proceeding gradually as banks focus not
panel 2). Overall, the balance of risks to global growth only on current headline inflation, but also on the tra-
over the next year continues to be tilted to the down- jectory of inflation and the stickiness of core inflation.
side, with the probability of growth falling below 2 Although recent market developments are benign, the
percent remaining broadly unchanged compared with large debt burden alongside high real interest costs (r)
April. relative to long-term growth prospects (g) for some
emerging markets remains a lingering concern, posing
ongoing challenges to fiscal sustainability.
Emerging and Frontier Markets The more favorable external environment has helped
narrow hard currency bond spreads, although implied
Pressures on Major Emerging Markets Ease, but foreign exchange volatility has declined for most
Investors Remain Cautious markets. Domestic equity markets have rebounded but
As the dollar weakened and trade deals started corporate bond spreads declined. This more conducive
to be reached, pressures on emerging market finan- environment has catalyzed a rebound in capital flows,
10 International Monetary Fund | October 2025
 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.6. Global Growth-at-Risk

Global financial stability risks expected over the near term have declined only slightly since April and remain somewhat elevated by historical norms, with
the balance of risks to global growth still tilted to the downside.
1.Near-Term Growth Forecast Densities 2. Near-Term Growth-at-Risk Forecast
(Probability density) (Historical percentile rank)
0.45 0
0.40
0.35 Density for Density for 20
year 2026: at year 2025: at
0.30 2025:Q3 2025:Q1
40
0.25
0.20
5th percentiles 60
0.15
0.10 80
0.05
0 100
−3 −2 −1 0 1 2 3 4 5 6 2012 13 14 15 16 17 18 19 20 21 22 23 24 25

Sources: Bank for International Settlements; Bloomberg Finance L.P.; EUROPACE AG/Haver Analytics; IMF, International Finance Statistics database; and IMF staff
calculations.
Note: In panel 1, the mode (that is, the most likely outcome) of the latest estimate of growth forecast density accords with the IMF’s October 2025 World Economic Outlook
database forecast for 2026. Near-term corresponds to growth expected one year ahead. The global conditional forecast density model employed here augments
information on current quarter growth and financial conditions (see the April 2018 Global Financial Stability Report) with a proxy for global credit growth (Adrian and others
2022). This credit growth variable is constructed as a PPP-GDP weighted aggregate of country-specific quarterly growth rates in total credit to the private nonfinancial
sector provided by domestic banks and all other sectors of the economy. Credit data are sourced from the Bank for International Settlements. The sample of countries
accounts for 90 percent of total GDP of all systemically important jurisdictions, covering all major advanced and emerging market economies. Given lags in availability of
the Bank for International ‘Settlements’ credit data, credit growth for the current quarter conservatively reflects the latest available reading, available as of 2025:Q1. In
panel 2, the black line traces the evolution of the fifth percentile threshold (the growth-at-risk metric) of the near-term forecast densities, where the lower percentiles
represent higher downside risk. The intensity of shading depicts the percentile rank for the growth-at-risk metric. The quintiles with the lowest percentile ranks are shaded
the brightest red and the highest are shaded brightest green. PPP = purchasing power parity; Q = quarter.

with inflows primarily benefiting funds dedicated ones. The ratio of the implied volatilities of three-
to local currency bonds (Figure 1.7, panel 2). The month 10-delta butterfly options to three-month
contrast with lackluster flows into hard currency funds at-the-money options—a proxy for the expensive-
suggests that global investors have renewed interest in ness of protection against large currency moves—is
diversifying their asset holdings into emerging market currently much higher than historical average for
bonds to avoid being overly exposed to the dollar. sub-investment-grade emerging market currencies
Relatively tight spreads on hard currency issuances (Figure 1.7, panel 4, yellow line), while lower than his-
may also have limited their appeal to global investors, torical average for investment-grade emerging market
contributing to the subdued fund flows. currencies. This signals that some investors anticipate
Stretched valuations in some emerging market sharp currency moves in weaker emerging markets.
assets could increase the vulnerability of these assets to Part of this cautious positioning may stem from con-
adverse trade and geopolitical shocks. Hard currency cerns about domestic fiscal dynamics. In several emerg-
emerging market sovereign spreads have compressed ing markets, elevated debt burdens, high interest costs,
despite persistent macroeconomic uncertainty and softening growth momentum are raising questions
(Figure 1.7, panel 3).17 After rising sharply in April about future fiscal trajectories. Emerging markets with
2025, investment-grade emerging market spreads have weaker credit ratings also tend to have projected long-
since narrowed to levels last seen in 2007, while high- term real interest rates higher than their long-term real
yield spreads have fallen to post-pandemic lows, which growth prospects (Figure 1.7, panel 5), which could
raises concerns about whether valuations reflect the undermine long-term debt sustainability. This loop
underlying fragilities and potential external shocks. of high real interest costs and mounting debt burdens
Developments in emerging market currency option could exacerbate borrowing costs and fiscal pressures,
markets indicate that investors are cautious about making fiscal consolidation especially challenging for
emerging market currencies, especially lower-rated these sovereigns. Moreover, should global financial con-
ditions tighten again or growth underperform, pressure
17Benchmark spreads are from JPMorgan indices. on sovereign creditworthiness could swiftly resurface.
International Monetary Fund | October 2025 11
 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.7. Pressures on Major Emerging Markets Have Eased, but Uncertainties Linger
EM stress has declined since the April 2025 Global Financial Stability Report, but tight risk pricing masks ongoing uncertainties.
1. EMs’ Financial Market Stress Heatmap, September 2015 to September 2025
Short-Term Policy Rates
Ex-ante real policy rate
Ex-post real policy rate
FX Market (versus USD)
At-the-money option-implied volatility
FX-implied yield differential
External Funding Market
Sovereign credit default swap
Sovereign hard currency spread
Domestic Bonds
Yield spread (10-year minus 2-year)
Local corporate spread
Domestic Equities
Equity realized volatility
Forward price-to-earning ratio
Fiscal Sustainability
Ex-ante r–g
10-year bond-swap spread
Sep. Mar. Sep. Mar. Sep. Mar. Sep. Mar. Sep. Mar. Sep. Mar. Sep. Mar. Sep. Mar. Sep. Mar. Sep. Mar. Sep.
2015 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25

Fund flows have mostly benefited local currency funds, whereas hard Tight emerging market hard currency sovereign spreads could heighten
currency and blended funds have not had similar inflows. repricing risk should adverse events materialize.
2. Exchange-Traded Fund and Mutual Fund Cumulative Flows 3. EM Sovereign Spread, Uncertainty Index
(Percentage of assets under management, cumulative since the end of 2021) (Z-score, left scale; percentile, right scale)
Hard currency Local currency Blended EM investment-grade spread (left scale)
EM B-rated spread (left scale)
5 3 1.2
Trade Policy Uncertainty Index (right scale)
0 Global Economic Policy Uncertainty
−5 0.8
1 Index (right scale)
−10
0.4
−15
−20 −1
0
−25
−30 −3 −0.4
Jan. 2022 Oct. 2022 Jul. 2023 Apr. 2024 Jan. 2025 Apr. 2024 Aug. 2024 Dec. 2024 Apr. 2025 Aug. 2025

Foreign exchange forward markets for lower-rated emerging markets are Lower-rated emerging markets also have a larger interest burden, further
weary over potential two-way tail risks. complicating fiscal consolidation efforts.
4. 10 Delta Three-Month Butterfly over Three-Month At-The-Money 5. Estimated r–g (Five-Year Ahead), by Average Credit Rating Band
Option-Implied Volatility (Percent)
(Z-score)
2 A/A–/BBB+
Investment grade Sub–investment grade
1 BBB/BBB– (stable outlook)

0 BBB/BBB– (negative outlook)

−1 BB+/BB/BB–

−2 −4.0 0 4.0 8.0

Jan. 2021 Oct. 2021 Jul. 2022 Apr. 2023 Jan. 2024 Oct. 2024 Jul. 2025 Ex-ante r–g (5-year ahead)

Sources: Bloomberg Finance L.P.; Consensus Economics; EPFR; EUROPACE AG/Haver Analytics; IMF, World Economic Outlook database forecasts; and IMF staff
calculations.
Note: In panel 3, spreads are normalized using a z-score on weekly data points (average from daily observations) from January 2024 to September 2025. Percentiles for the
uncertainty indices are derived from weekly and monthly data points starting January 2024. Data for the Global Economic Policy Uncertainty Index are presented as a
monthly series, incorporating the most recent available month. In panel 4, the ratio of butterfly to at-the-money option-implied volatility is normalized using a z-score on
weekly data points (average from daily observations) from January 2021 to September 2025. In panel 5, the r–g estimates are computed from current 5-year, 10-year, and
implied 5-year forward yields, considering differences in the term premium. Inflation and growth estimates are from Consensus Economics or, when unavailable, from
World Economic Outlook database forecasts. Data include 14 major emerging markets: Brazil, Chile, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Peru, the
Philippines, Poland, Romania, South Africa, and Thailand. EM = emerging market; FX = foreign exchange; r = long-term real interest rates; g = long-term-growth rates;
USD = US dollar.

12 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.8. Eurobond Issuance Has Remained Robust, but High Yields and the Upcoming Maturity Wall Have Prompted
Frontier Economies to Explore Alternative Funding Strategies

International hard currency bond issuance has Sovereign spreads have tightened for frontiers, ... while maturing debt is accelerating.
remained robust in 2025, but market access has but yields remain high ...
been uneven.
1. Frontier Market Eurobond Issuance 2. Emerging Market Spreads and Yields 3. Upcoming International Maturing Debt of
(Millions of dollars, left scale; 12-month (Percent, left scale; basis points, right scale) Frontier Economies
issuance, right scale) 75th percentile (right scale) (Billions of dollars)
Europe and Central Asia MENA 25th percentile (right scale) Africa Latin America
SSA South and SE Asia Share with yields above 10 percent Asia Middle East
LAC Share with spread above 1,000 basis points Europe
6 80 (right scale) 1,500 16
14
5
60 12
4 1,000
10
3 40 8
6
2 500
20 4
1
2
0 0 0 0

2019:Q1
19:Q3
20:Q1
20:Q3
21:Q1
21:Q3
22:Q1
22:Q3
23:Q1
23:Q3
24:Q1
24:Q3
25:Q1
25:Q3
26:Q1
26:Q3
27:Q1
27:Q3
24
Feb. 2024

Jun. 2024
Apr. 2024

Aug. 2024
Oct. 2024
Dec. 2024

Jun. 2025
Feb. 2025
Apr. 2025

Aug. 2025

2002
04
06
08
10
12

20
22
14
16
18

Sources: Bloomberg Finance L.P.; and IMF staff calculations.

Note: Frontier economies are defined here as countries included in J.P.Morgan’s NEXGEM index. LAC = Latin America and the Caribbean; MENA = Middle East and North
Africa; SSA = sub-Saharan Africa.

Frontier Economies Explore Alternative Funding according to Panama’s Economy Ministry. Egypt issued
Strategies a $1 billion sovereign sukuk through a private placement
as part of its strategy to diversify funding sources. The
Primary market bond issuance by frontier economy
issuance was fully subscribed by Kuwait Finance House.
borrowers reached just over $13 billion by the end of
Angola entered into a $1 billion structured financ-
August 2025 (Figure 1.8, panel 1) but market access
ing arrangement linked to its own sovereign bonds.18
remained uneven. As highlighted in previous issues of
Although cost-effective relative to market rates, the deal
the Global Financial Stability Report, frontier borrowers
included contingent liabilities that triggered additional
have increasingly resorted to shorter tenors and smaller
payments amid market volatility earlier this year. Even
deal sizes to cater to cautious investors amid still-high
as there may be advantages to some of these alternative
global yields. Despite the easing of global financial
funding arrangements, such as when they allow the issuer
conditions and the weakening of the dollar, the median
to pay maturing debt without causing much market
sovereign eurobond yield among emerging market issu-
pressure, a broader shift toward private funding raises
ers now exceeds 6.5 percent, and with several frontier
transparency and debt sustainability concerns. This is
economy bonds trading above 10 percent, this raises
especially so when these obligations are not subject to the
concerns about refinancing costs (Figure 1.8, panel 2).
same market discipline or reporting standards as publicly
Rollover risks are amplified because large amounts of
traded bonds. These developments underscore a grow-
bonds need to be repaid in late 2025 and early 2026
ing divergence in financing conditions across frontier
(Figure 1.8, panel 3), especially for sub-Saharan issuers.
economies between those able to issue in public markets
Under challenging conditions in bond markets, some
at reasonable cost and those reliant on less-conventional
frontier economies are exploring alternative funding
and potentially more fragile forms of borrowing.
strategies, including private placements, bilateral loans,
and other financing instruments. For example, in early
2025, Panama secured a €1.2 billion bilateral loan from a 18See statements from respective authorities (Egypt Ministry of

subsidiary of Bank of America with a two-year maturity, Finance 2025; Panama Ministry of Economy and Finance 2025).

International Monetary Fund | October 2025 13

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.9. Rising Bond Supply across Major Advanced Economies Has Steepened Yield Curves
Longer-term bonds are under increased Investor concerns about larger fiscal deficits are ... with price-sensitive investors expected to
pressure amid greater supply. increasingly reflected in widening swap demand higher term premiums as
spreads ... compensation for absorbing rising bond supply,
exerting upward pressure on yields.
1. G4 and Emerging Market 10-Year 2. G4 GDP-Weighted 30-Year Swap Spreads 3. G4 30-Year Government and Duration Term
Aggregate Yields, G4 30-Year and versus Fiscal Deficits Premium Conditional on Projected
10-Year Yield Spreads (Basis points, left scale; percent, right scale) Free Float
(Basis points, left scale; percent, right scale) (Percent)
G4 30y swap spread
G4 30y—10y spread G4 10y G4 fiscal deficit, average over Government term premium
140 EM 10y (right scale) 10 80 the next 5 years (right scale) −2 Duration term premium 3
9 60 −2.5
120
8 40 2
100 −3
April 7 20
2025 6 −3.5 1
80 0
GFSR 5 −4
60 −20
4 −4.5 0
3 −40
40 −5
2 −60 Dec. Dec. −1
20 −80 −5.5 2025 2026
1
0 0 −100 −6 −2
2005 10 15 20 25 2005 10 15 20 25 67.5 70 72.5 75 77.5 80 82.5 85
Free float

Sources: Bank of England; Bank of Japan; Bloomberg Finance L.P.; European Central Bank; Federal Reserve Bank; JPMorgan; London Stock Exchange Group; US
Congressional Budget Office; and IMF staff calculations.
Note: In panel 1, the G4 composite reflects GDP-weighted average yields across the euro area, Japan, the United Kingdom, and the United States. The emerging market
composite reflects the GDP-weighted average across Brazil, Chile, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Peru, the Philippines, Poland, Romania, South
Africa, and Thailand. In panel 2, G4 fiscal balances are gauged by the GDP-weighted average net lending estimate over the next five years, as seen in the World Economic
Outlook, at a given point in time. The dashed line extends the historical series by splicing earlier net lending estimates with a linear projection of G4 net lending inferred
from Congressional Budget Office deficit data. Swap spreads are computed as the difference between overnight swap rates and sovereign yields of the same maturity.
Overnight rates are extended historically using interbank rates. In panel 3, the free float is the share of government bonds outstanding held by private investors, excluding
central bank holdings. The term premium is defined as compensation that investors require for bearing the risk that interest rates may change over the life of the bond.
These are estimated using Adrian, Crump, and Moench (2013) for rates for US treasuries, bunds, gilts, and Japanese government bonds, then aggregated using G4 GDP
weights. Duration term premiums are calculated based on market pricing of fully collateralized and centrally cleared interest rate swaps, which are a key intermediation
instrument to strip out pure interest rate risk and isolate credit and spread risk. They reflect compensation for taking on duration risk, which is exclusively driven by
conjunctural factors, rather than shifts in perceived creditworthiness. (See Online Annex 1.7 for the definition and methodology of term premium and duration term
premium calculations). A diamond indicates the latest observation and a line shows linear fit over the current yield regime. Vertical lines mark projected free-float levels,
based on World Economic Outlook database projections and central bank surveys. Ellipsoids show 95 percent confidence bands, obtained through bootstrapped
regressions in conjunction with sampling from central bank survey distributions. EM = emerging market; G4 = Group of Four; GFSR = Global Financial Stability Report.

Sovereign Bond Markets denced by a notable steepening of yield curves among

the G4—US Treasuries, European government bonds,
Bond market stability is fundamental to financial
UK gilts, and Japanese government bonds (Figure 1.9,
stability because key sovereign bonds serve as bench-
panel 1)19—alongside a widening in swap spreads (that
marks for asset prices and collateral in lending and
is, spreads becoming more negative). The widening of
derivative transactions. Although bond markets have
swap spreads, broadly capturing rising credit risk and
stabilized since the abrupt sell-off after the April 2
tariff announcement, steepening yield curves, more
19Across the G4, the fiscal challenge involves not only the size of
negative swap spreads, and the persistent erosion of
the deficit but also the level and the trajectory of public debt (high
convenience yields point to bond markets being on and rising), weak growth prospects, and high debt-service burdens.
shakier footing than they seem. For example, in the United States, the One Big Beautiful Bill Act
(Public Law 119–21; enacted July 4, 2025) is projected by the
Congressional Budget Office/Joint Committee on Taxation to raise
Expanding Fiscal Deficits Exert Pressure on Bond US federal deficits by about $3 to $3.5 trillion over the next decade.
While the act itself does not include a universal tariff, revenue
Market Stability prospects were linked to later “reciprocal” tariffs enacted under
Investor concerns about large fiscal deficits appear emergency authority through the International Emergency Economic
Powers Act. However, these were invalidated by federal courts (with
to have added more pressure on long-term bond yields. temporary stays in place), rendering any tariff-based fiscal offset
Across major advanced economies, the pressure is evi- legally tenuous and excluded from credible deficit scoring.

14 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.10. Widening Emerging Market Swap Spreads Are Increasing the Cost of Financing

The swap spread for median emerging markets has widened (become Higher sovereign spreads and covered interest parity deviations are also
more negative) as the debt-to-GDP ratio has climbed. associated with wider swap spreads.
1. Average Spread between 10-Year Emerging Market Interest Rate Swap 2. Average Annual Change in Swap Spread for High- versus
and 10-Year Government Bond, versus Debt-to-GDP Ratio Low-Vulnerability Emerging Markets
(Percentage points, left scale; percent, right scale) (Basis points)
10-year interest rate swap versus 10-year bond

0 High vulnerability Low vulnerability 5

Debt/GDP 44
−0.2
42 0
−0.4
40
−0.6 −5
38
−0.8 36 −10
−1.0 34
Average ASW −15
−1.2 32
−1.4 30 −20
2016 17 18 19 20 21 22 23 24 25 USD sovereign Covered interest Domestic Foreign
spread parity debt/GDP holdings

Sources: Bloomberg Finance L.P.; Citi; EUROPACE AG/Haver Analytics; J.P.Morgan; and IMF staff calculations.
Note: The sample of countries includes Brazil, Chile, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Poland, and South Africa. In panel 1, the swap spread is
labeled as ASW. In panel 2, countries are divided into two groups for each variable: those with the largest increase/smallest decline for that variable in a given year and
those with the smallest increase/largest decline. The bars show the annual change in swap spreads in each grouping, average over 2016 to 2025. USD = US dollar.

funding pressures in the financial system, has been panel 1). Some countries (for example, Colombia,
increasingly driven by fiscal considerations, exhibiting Mexico, and South Africa) have experienced a decline
strong co-movement with the projected average budget in the swap spread by more than 100 basis points. The
balance over the next five years (Figure 1.9, panel 2; relative underperformance of bonds can add to fiscal
see also the October 2025 World Economic Outlook).20 strains, as rising debt is compounded by a higher cost
In parallel, the continuation of quantitative tighten- of interest. Assuming an average stock of domestic
ing by major central banks has increased the amount debt at 40 percent of GDP, a −50 basis point swap
of free-floating bonds in the market to be absorbed spread equates to an increased annual fiscal cost of 0.2
by price-sensitive investors, exerting upward pressure percent of GDP. In addition, the negative swap spread
on term premiums, all else equal, and keeping yields is likely to drive up interest rates of private sector debt
elevated (Figure 1.9, panel 3; and Figure 1.2, panel 1). or lead to some crowding out of private sector debt. A
Meanwhile, regulatory requirements limiting dealer growing disconnect between bond yields and domestic
balance sheets and falling demand from liability-driven interest rate swaps could also lead to a lower pass-
investors have likely exacerbated the widening of through of monetary policy on the real economy, given
spreads. the swap market’s close link to policy rates.
Outside G4 bond markets, emerging markets have Widening emerging market swap spreads reflect a
also seen their domestic swap spreads widen. In a premium that investors require to absorb large sov-
panel of major emerging markets, the spread between ereign bond issuances, even though increased buying
10-year interest rate swaps and 10-year local currency by large domestic investors such as pension funds and
bonds has turned more negative over the past decade, insurance companies has kept sovereign bond markets
declining by almost 50 basis points, mirroring the rise resilient (see Chapter 3). Swap spreads tend to turn
in domestic debt as a percentage of GDP (Figure 1.10, more negative for emerging markets whose US dollar
sovereign spreads or CIP deviations rose the most
20More precisely, swap spreads capture the difference between
during the average year or have experienced the largest
same-tenor swap rates and government bond yields. The spreads cap- increase in the holdings of domestic debt by foreigners
ture the funding advantage of sovereign bond issuers compared with
maturity-matched swap rates. A positive value shows that sovereign during the average year (Figure 1.10, panel 2).
bond yields are lower than interest rate swap rates. Increased foreign buying could be because higher bond

International Monetary Fund | October 2025 15

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.11. Convenience Yields amid Rising Safe-Asset Supply

Safe-asset supply has risen amid fiscal expansion Domestic convenience yields have been mostly The cross-border convenience yield for US
across jurisdictions, putting upward pressure on stable, although they have seen periods of Treasuries has seen more structural erosion over
bond yields. transitory erosion, particularly in US Treasuries. the past decade but has remained broadly stable
since April.
1. Safe-Asset Supply (Excluding Central Bank 2. Domestic Convenience Yields for 3. Cross-Border Convenience Yields for
Holdings) in Primary Reserve Currencies Five–Year Bonds Five–Year Bonds
(Billions of dollars) (Basis points) (Basis points)
Private label MBS
EA sovereigns (ex. DEU/FRA, pre–debt reform) USA Euro area Min/max range Median
EA sovereigns (ex. DEU/FRA, post–debt reform) GBR USA (alt. est.) JPY GBP EUR
DEU/FRA sovereigns UK gilts JGBs US Treasuries more
Agency MBS US Treasuries “convenient” than
FX reserves (right scale; USD, EUR, JPY) Sovereigns more FX-hedged foreign
”convenient” government bonds
FX reserves + stablecoin (right scale) 0 0 150 25
Global financial Central bank QT
crisis
50,000 COVID-19 Declining FX −50 −20
Euro area Euro area reserves 100
crisis debt reform 0
9,800
40,000 −100 −40
50

30,000 −150 −60 −25

7,800
0
20,000 −200 −80
5,800 April −50
US Treasuries less −50
2025 “convenient” than
10,000 −250
GFSR
−100
Sovereigns less FX-hedged foreign
”convenient” government bonds
0 3,800 −300 −120 −100 −75
2005 10 15 20 25 2005 10 15 20 25
2008
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25 (YTD)

Apr. 2025
May 2025
Jun. 2025
Jul. 2025
Aug. 2025
Sep. 2025

Apr. 2025
May 2025
Jun. 2025
Jul. 2025
Sep. 2025
Aug. 2025
Sources: AG/Haver Analytics; Bloomberg Finance L.P.; Currency Composition of Official Foreign Exchange Reserves; Du, Im, and Schreger 2020; EUROPACE; London Stock
Exchange Group; national authorities; and IMF staff calculations.
Note: In panel 1, yearly data are the quarterly averages of free-floating securities, that is, securities not held by central banks for monetary policy purposes. Euro area sovereigns
include securities that fall under the definition of Maastricht debt, as defined by the European Union, and so include local government debt (for example, German Länder debt).
Agency MBS are MBS issued or guaranteed by the US mortgage agencies: Fannie Mae, Freddie Mac, and Ginnie Mae. Private-label MBS are nonagency MBS. In panel 2, following
Krishnamurthy and Vissing-Jorgensen (2012), the domestic convenience yield is measured as the yield spread between five-year AAA-rated corporate bonds and government
bonds, adjusted for differences in credit risk. As a robustness check, the dotted line shows an alternative estimate for the United States that obtains as the median over a broad
range of five-year domestic convenience yield estimates following Krishnamurthy and Vissing-Jorgensen (2012), Mota (2023), and Acharya and Laarits (2023). In panel 3, following
Du, Im, and Schreger (2020), the cross-border convenience yield captures the five-year yield gap between foreign-exchange-hedged foreign government bonds and US Treasuries,
bunds, gilts, or Japanese government bonds, whereby the currency and interest rate risk of the former is hedged back into dollars using maturity-matched cross-currency swaps.
alt. est. = alternative estimate; DEU = Germany; EUR = euro; ex. = excluding; FRA = France; FX = foreign exchange; GBP = Great Britain pound; GFSR = Global Financial Stability
Report; JPY = Japanese yen; MBS = mortgage-backed securities; QT = quantitative tightening; YTD = year to date.

yields relative to domestic funding rates are attractive. play a critical role in the global financial system.
Foreign investors who prefer interest rate swaps to Fiscal expansion by major economies has ramped
bonds because of the smaller balance sheet impact or up safe-asset supply (Figure 1.11, panel 1), whereas
ease of access amid capital controls may have helped demand for long-duration safe assets has declined,
compress swap spreads. amid a reduction in foreign exchange reserves denom-
inated in the largest reserve currencies.21 Life insurers
and pension funds have also become more cautious
Convenience Yields for Longer-Duration Bonds buyers amid expectations of elevated interest rate
Are Somewhat Stable
Government bonds form the main component of
21The largest reserve-currency issuers defined here include those
what are known as safe assets—highly liquid instru-
sovereign issuers with the largest share of global reserves, including
ments with minimal credit risk that are expected to the United States, issuers in the euro area (particularly Germany and
preserve their value even in market stress—and these France), Japan, and the United Kingdom.

16 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

volatility.22 Meanwhile, other investors, such as buyers spread) domestic investors forego over high-grade
of money market funds (including tokenized ones) and corporate bonds, after credit and liquidity adjust-
stablecoins, have increased demand primarily toward ments. The cross-border convenience yield (CCY)
short-dated safe assets like sovereign bills (see the refers to the yield discount investors are willing to
section “Stablecoins’ Growth Could Affect Financial accept to hold US Treasuries, for example, relative to a
Stability”).23 These trends may have increased the foreign-currency-hedged equivalent security issued by
demand and supply imbalance for safe assets with another sovereign. From a cross-border investor’s per-
longer duration. spective, higher currency-hedged yields for other G4
Convenience yields measure the premium investors bonds relative to Treasury yields imply that Treasuries
are willing to pay to hold safe assets. Of particular are the preferred safe asset.
interest are convenience yields for bonds with longer DCYs for European government bonds, gilts, and
duration, given the glut of supply. An erosion in con- US Treasuries have not had clear directional trends
venience yields can both signal and amplify funding over the past few years, although they have seen
market strains, raising concerns about the safe assets’ bouts of volatility. On the other hand, the CCY for
utility as high-quality collateral, especially during stress Treasuries has seen a secular erosion against other G4
periods. Lenders in short-term funding and repurchase government bonds over the past decade (Figure 1.11,
(repo) markets could demand higher haircuts on safe panels 2 and 3, respectively). This suggests that
assets pledged as collateral when convenience yields Treasuries’ status as the preeminent safe asset may
erode, in turn pushing up funding costs. Banks and have been reduced and aligns with market commen-
investors could diversify toward substitute assets, which tary that global investors may become more cautious
will likely encapsulate far fewer safe-asset properties about investing in US assets. The market volatility
and have shallower market depth, again leading to in April provides some insight into the behavior of
upward pressure on funding spreads.24 the two convenience yield measures during stress.
Convenience yields can be measured along DCYs first declined sharply for European government
domestic and cross-border dimensions. The domestic bonds and US Treasuries, indicating that investors’
convenience yield (DCY) reflects the premium (or preference for these government bonds over high-qual-
ity corporate bonds strengthened, before gradually
22The reduced structural participation of long-term investors, returning to their prior levels. CCYs have remained
such as life insurers and pension funds, reflects a combination of stable, indicating that Treasuries’ safe-asset status
factors. These include changes in regulatory capital requirements (for
compared with other G4 government bonds has been
example, Solvency II) and the structural shift from defined-benefit to
defined-contribution plans that has curtailed risk tolerance for rising broadly maintained.
interest rate volatility, particularly affecting long-term bonds. From a financial stability perspective, the relative
23Results of changes in composition of the demand side of the
stability of convenience yields likely helped keep fund-
sovereign bond market structure and the increase in free-float supply
is to pressure yields higher. Indeed, swaptions-implied odds imply ing markets orderly during the April episode. The lack
higher longer-term yields one year ahead, despite ongoing monetary of any substantive erosion in convenience yields across
easing in three of the G4. This corresponds with the observation the G4 since April may suggest, for now, that substi-
that long-term yields across advanced economies have become more
correlated across jurisdictions, increasing the potential for a rapid
tutes for safe assets—particularly for US Treasuries,
transmission of shocks across borders (see the October 2024 Fiscal the largest contingent of safe-asset supply—are limited
Monitor). among both domestic and cross-border investors. The
24The erosion of convenience yields could trigger a cascade of col-
more secular decline of CCYs, however, indicates that
lateral preference shifts, leading to amplification of funding market
stress. If investors lose confidence in US Treasury’s safe-asset status, cross-border diversification in safe-asset holdings could
for instance, then repo market haircuts could widen dramatically. be under way.25
Or, in extreme cases, collateral that once traded at zero or near-zero
haircuts (as in packaged bond/futures transactions) could suddenly
require significant buffers and so offer little protection against forced 25This type of risk could run both ways: disorder in funding

liquidation spirals. Stress could get magnified through a breakdown markets—whether domestic or cross-border—could quickly spill
of collateral chains. More specifically, the same Treasury bond into bond markets, driving abrupt term premium shifts and spikes
normally circulates through multiple financing operations, effectively in interest rate volatility. Evidence from the euro area sovereign crisis
multiplying the system’s liquidity. When confidence erodes, institu- shows that a disorderly repricing of sovereign risk can contaminate
tions become reluctant to pledge these assets along these rehypoth- swap-based signals. Credit stress may spill over into pricing of high-
ecation chains, causing the chains to snap. Consequently, the pool grade-rated corporate and financial issuers when broader doubts
of effective collateral in circulation would shrink, reducing market about sovereign repayment capacity trigger extensive degrees of
liquidity and widening funding spreads. market fragmentation (as explained in Société Générale 2012).

International Monetary Fund | October 2025 17

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.12. Anatomy of Two Stress Episodes from a Treasury Market Perspective, March 2020 and April 2025
During both the March 2020 and the April 2025 market turmoil ... but basis trades did not unwind in April 2025, and bond fund
episodes, US Treasuries started to sell off beyond a certain VIX outflows were much more limited.
level ...
1. Intraday 10-Year US Treasury Yield 2. Change in Leveraged Fund US Treasuries Futures Positions and
(Yield in percent) US Bond Fund Flows during Distressed Market Episodes
(Change, in billions of dollars)
4.8 April 2025 tariff turmoil March 2020 pandemic turmoil 80
Mar. 2020 Apr. 2025

4.4 60

4 40

3.6 20

1.6 0

1.2 −20

0.8 −40

0.4 −60
15 35 55 75 95 Change in leveraged funds’ US bond fund flows
Implied volatility (VIX index) Treasury futures position

Sources: Bloomberg Finance L.P.; EPFR; and IMF staff calculations.

Note: Panel 1 is based on intraday US Treasury yield data between March 26 and April 9, 2025, for the tariff turmoil, and between March 2 and March 20, 2020, for the
COVID-19 pandemic market turmoil. Panel 2 shows the change in leveraged fund Treasury futures positions for all maturities between March 25 and April 8, 2025, for the
tariff turmoil, and between March 3 and March 17, 2020, for the pandemic market turmoil. US bond fund flows reflect EPFR Fund Flow statistics for March 12 to March 18,
2020, and April 3 to April 9, 2025. Matching dates are not available because of data constraints. VIX = Chicago Board Options Exchange’s Volatility Index.

Sovereign Bond Market Functioning Depends on can contribute to such a reversal, as funds may be
Nonbank Financial Intermediaries forced to sell more liquid assets, including US Treasur-
US Treasury markets weathered the April 2025 ies. But despite fund outflows being relatively limited
tariff turmoil, stopping short of the severe dislocations in April 2025 (Figure 1.12, panel 2), this tipping point
witnessed during the March 2020 “dash for cash” was reached earlier (Figure 1.12, panel 1). This could
episode. This relative stability raises an important mean that investors had begun to question the liquid-
question: does the market’s resilience reflect structural ity value of US Treasuries, or it could reflect concerns
improvements, or was the April shock merely less about US fiscal policy (see the section “Expanding Fis-
severe or different in nature? Although some structural cal Deficits Exert Pressure on Bond Market Stability”).
improvements occurred in the resilience of funding NBFIs’ amplification of market stress from the
markets, the short-lived nature of the shock limited side of leveraged funds was limited in April 2025
the unwinding of leveraged hedge fund positions in compared with the March 2020 episode. Cash-futures
Treasury securities and outflow pressures from open- basis trades—a popular hedge fund arbitrage strategy
ended funds. These NBFIs nonetheless remain vulnera- exploiting price differences between Treasury bonds
ble to large and persistent bond market shocks. and futures—did not unwind as they did in March
Similar to the pandemic market turmoil of March 2020 (Figure 1.12, panel 2). Instead, a different
2020, in April 2025, Treasury yields initially declined leveraged fund strategy that involves taking a long (or
as the Chicago Board Options Exchange’s Volatility short) government bond position while taking the
Index (VIX) increased, reflecting “flight to safety” opposite position in a long-term interest rate swap—
dynamics (Figure 1.12, panel 1). However, during the “swap spread” trade—did reportedly unwind and
both episodes, Treasury yields reached a tipping point contribute to Treasury market volatility. Nonetheless,
as market stress continued to increase, with Treasury the absence of significant unwinding of the much-
yields rising beyond a certain VIX level. As market larger cash-futures basis trades seems to have been cru-
stress increases, redemptions by mutual fund investors cial to preventing a 2020-like crisis. Risks to financial

18 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.13. Structural Resilience in Repo Markets

Repo spreads were kept low by abundant ... even though dealers’ balance sheets in repos The increase in centrally cleared hedge fund
reserves in 2025, with dealer balance sheets and Treasuries became heavier. repo (through sponsoring) likely moderated
exerting mild upward pressure ... market functioning pressures.
1. Sensitivity of Repo Spreads to Changes in 2. Treasury Inventory and Repo (Gross and 3. Sponsored Repo Share of Hedge Fund
Bank Reserves and Dealer Balance Sheets Net) Positions by US Primary Dealers, Repo Positions, 2020–25
(Beta coefficients in basis points) 2019–25 (Percent)
(Percent of total marketable debt)
0.5 50 10 45
Dealer gross repo
3-period moving average 40
40 (dealer Treasury inventory) 8 35
0
3-period moving average
30
30 (dealer net repo) 6
25
−0.5
20
20 4
15
−1 10
10 2
5
−1.5 0 0 0
Reserves Dealer position 2019 20 21 22 23 24 25 2020 21 22 23 24 25

Sources: Bloomberg Finance L.P.; Federal Reserve; Office of Financial Research; and IMF staff calculations.
Note: Panel 1 presents the average beta coefficients for 2025 from regressing repo spreads on changes in reserves and dealer balance sheets in a rolling window of one
year. See Online Annex 1.8 for further information. In panel 2, marketable debt excludes Federal Reserve holdings. Panel 3 shows repo volume sponsored by the FICC
divided by the total repo exposure of hedge funds that qualify to report under the Securities and Exchange Commission’s Private Fund form. As noted by the Office of
Financial Research, hedge fund borrowing cash makes up most FICC-sponsored repo volumes. FICC = Fixed Income Clearing Corporation.

stability remain high as large hedge funds still hold (Figure 1.13, panel 3). Moreover, higher haircuts
near-record net interest rate derivatives and leveraged involved in central clearing likely curbed repo leverage
repo positions, indicating that they still have substan- and enhanced market stability.
tial amounts of basis and swap spread trades. Whereas Outflows from open-ended bond funds were
investors have focused mostly on US Treasury market moderate in April 2025 and did not appear to induce
basis trades, record levels are noted in the United significant forced selling. US-domiciled bond mutual
Kingdom (Bank of England 2025) and rising trends funds manage about $5 trillion in assets, with almost
are evident in Canada (IMF 2025a) and expected in one-quarter allocated to US Treasuries (Figure 1.14,
Europe (ECB 2024). panel 1), making them a major player in this market.
Both circumstantial and structural factors helped Large redemptions can force funds to liquidate Treasury
support the functioning of the US Treasuries market holdings once their liquidity buffers are depleted, put-
in April 2025. On the circumstantial side, the April 2 ting upward pressure on yields. The magnitude of out-
tariff announcement was followed by a policy reversal flows in the April 2025 scenario is much smaller than in
within a week, limiting the duration and severity of the March 2020 scenario (Figure 1.14, panel 2). Bond
market stress. On the structural side, repo markets— funds are highly heterogeneous: Some may have ample
critical for funding basis trade strategies—remained liquidity buffers or even experience inflows. Others may
relatively stable during the episode. Regression analysis face outflows or thin cash buffers. Margin calls on deriv-
indicates that repo spreads remained contained overall ative contracts can compound to liquidity pressures,
in 2025 in part because of increased banking sector but an analysis of funds’ interest rate exposures through
reserves, likely aided by slower quantitative tighten- swaps and Treasury futures suggests their impact is likely
ing and supportive standing facilities (Figure 1.13, smaller. Under the scenario of a 100 basis point curve
panel 1), and in part because dealer balance sheet shift, variation margin calls would amount to about
usage exerted only limited upward pressure on rates. $20 billion (Figure 1.14, panel 3). Although some funds
Although dealers expanded their Treasury and repo face margin calls, others receive variation margin credit,
positions (Figure 1.13, panel 2), increased volumes underpinning the heterogeneity across funds.
of repo central clearing (through “sponsored clear- Bond mutual funds and the Treasury market vul-
ing”) helped preserve dealer balance sheet capacity nerabilities are intertwined. Combining the outflow

International Monetary Fund | October 2025 19

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.14. Bond Mutual Fund Flows, Treasury Holdings, and Forced Liquidations under Stress
US-domiciled bond mutual funds hold almost $5 trillion in total assets, of Liquidity pressures stemming from fund outflows can be large.
which one quarter is in US Treasuries.
1. Holdings of US-Domiciled Bond Funds 2. US-Domiciled Bond Fund Flow Scenarios
(Trillions of dollars, left scale; percent, right scale) (Billions of dollars)
Cash MMF shares Treasury bills Fund with inflows Funds with outflows
Commercial paper US Treasuries Other assets Net flows for all US bond funds
6 30 100

5 25 0
−100
4 20
−200
3 15
−300
2 10
−400
1 5 −500

0 0 −600
Nominal Share of total assets April 2025 March 2020 99th percentile
outflow scenario outflow scenario outflow scenario

The liquidity pressures from variation margin calls on interest rate Under a “waterfall” liquidation approach, larger shocks lead to larger
derivatives is less significant, with heterogeneity across funds. Treasury sales by magnitude and by share of liquidated assets.
3. Bond Fund Variation Margin Calls on Interest Rate Derivatives 4. Forced Sales under the Waterfall Approach
(Billions of dollars) (Billions of dollars)
UST futures variation margin (funds with margin calls) 400
Cash
UST futures variation margin (funds with margin credit)
MMF shares 350
IRS variation margin (funds with margin calls)
Treasury bills
IRS variation margin (funds with margin credit) 300
CP
UST futures net margin calls for all bond funds (IRS and UST futures)
US Treasuries 250
15
10 200
5
0 150
−5
−10 100
−15
−20 50
−25
−30 0
60 bps rate shock 80 bps rate shock 100 bps rate shock April 2025 outflows March 2020 outflows 99th percentile outflows
+ 60 bps rate shock + 80 bps rate shock + 100 bps rate shock

Sources: Lipper; Securities and Exchange Commission N-PORT; and IMF staff calculations. N-PORT data are taken from the second quarter of the 2025 batch, retaining only
submissions for the first quarter of 2025.
Note: See Online Annex 1.4. Scenarios project flow percentages observed during historical episodes to current holdings. If individual fund flow data are missing, the
relevant (for example, March 2020) flow percentage assigned is based on the median of its peer group by US mutual fund classification category. Margin calls from
derivatives contracts only include Treasury futures contracts and interest rate swaps and are based on a linearized pricing model (estimated contract duration). The 99th
percentile outflow scenario considers historical fund flow data for each individual fund, taking its 99th percentile of outflows, that is, its first percentile of flows. All flow
data are based on monthly flows. Higher-frequency flow patterns can deviate, and monthly flow data may not be representative of shorter-term inflow and outflow
patterns. bps = basis points; CP = commercial paper; IRS = interest rate swap; MMF = money market mutual fund; pct = percentile; UST = US Treasury bonds and notes.
Treasury bills have a maturity of one year or less. Treasury bonds and notes have longer maturities.

scenarios and interest rate shocks, forced sales of the total volume of forced sales and raise the propor-
Treasuries by US bond funds can be estimated using a tion of Treasury holdings liquidated.
“waterfall” approach, whereby Treasuries are sold after Forced sales by bond mutual funds played a pivotal
cash and other liquid assets are depleted. Assuming role in making the March 2020 market turmoil
the outflow patterns seen in April 2025, in conjunc- disorderly. These risks remain and may have grown
tion with a 60 basis point increase in interest rates, as the sector has expanded. By contrast, the absence
bond funds’ forced sales are estimated at $66 billion, of large outflows in April 2025 may help explain why
with over half the liquidation being Treasury securities conditions remained relatively orderly. Large, rapid
(Figure 1.14, panel 4, left bar). Larger shocks increase forced sales of Treasuries are more likely to overwhelm

20 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

dealer intermediation capacity. In a severely adverse The GST adverse scenario assumes stagflation
scenario in which bond fund outflows reach their 99th with tight financial conditions arising from intense
historical percentile and interest rates rise by 100 basis geopolitical turmoil and supply chain disruptions in
points, forced Treasury sales would exceed current commodities and goods markets. As documented in
dealer Treasury inventories (Figure 1.14, panel 4, right Online Annex 1.1, the scenario assumes an across-
bar), potentially overwhelming dealer intermediation the-board 10 percent increase in tariffs over the
capacity and likely causing disorderly conditions in baseline for advanced economies and some emerging
Treasury markets. markets, higher inflation from supply chain rechannel-
ing in goods and commodities, and a corresponding
1 percentage point increase in policy rates globally in
Financial Intermediaries the first year. Higher government debt, including from
Higher Capital Ratios Strengthen Global Banks, additional fiscal and demand support, causes inves-
but the Weak Tail of Banks Remains Substantial tors to panic, raising term premia by 300 basis points
The GST shows the global banking system remain- to 500 basis points across advanced economies and
ing broadly resilient under the July 2025 World Eco- emerging markets, depending on the scale of govern-
nomic Outlook reference scenario.26 However, under a ment debt.29 The term spread rises in the first year
severe stagflationary scenario, banks representing about before quickly reversing as recession sets in. Corporate
18 percent of global bank assets can be considered bond spreads also rise sharply as investors sell bonds,
weak, as their Common Equity Tier 1 capital (CET1) whereas weak consumer confidence and recession lead
ratio falls below 7 percent. The share of weak banks to reduced sales and higher corporate losses.
has materially improved since the October 2023 Global Under this adverse scenario, the aggregate global
Financial Stability Report, which considered a similarly CET1 ratio declines by a modest 70 basis points,
severe shock to the global economy and found almost from 13 percent in 2024 to 12.3 percent at the end of
one-third of bank assets to be weak.27 This improve- the stress horizon (Figure 1.15, panel 1). This result
ment is mostly a result of improved capitalization is driven primarily by larger loan losses and operat-
across most regions, particularly in the United States ing expenses, partially offset by improved net inter-
and large Chinese banks, which increased the global est income from a steeper yield curve (Figure 1.15,
average CET1 ratio from about 12.5 percent in 2022 panel 2). However, these results vary across regions.
to 13 percent in 2024. The steepening yield curve Capital depletion is larger for banks in the euro area
assumed in the adverse scenario also contributed to the and other non-US advanced economies than in other
results by increasing banks’ net interest margins, which regions because of greater sensitivity to macrofinancial
overcompensated for rising loan and bond valuation shocks (output, unemployment, and higher long-term
losses.28 interest rates) that translate into larger loan losses.
There are also significant differences within emerging
markets, as emerging market banks outside China ben-
efit from higher net interest margins (Box 1.3).30
26The Global Stress Test (GST) examined 669 banks from Although most banks remain resilient, the CET1
29 countries, accounting for 74 percent of global sector assets. The
29 countries in the sample are Australia, Austria, Belgium, Brazil, capital ratios of 82 of 669 banks globally are pro-
Canada, China, Denmark, Finland, France, Germany, Greece, India, jected to fall below 7 percent (the CET1 plus capital
Indonesia, Ireland, Italy, Japan, Korea, Mexico, The Netherlands, conservation buffer plus relevant global systemically
Norway, Portugal, Saudi Arabia, South Africa, Spain, Sweden,
Switzerland, Türkiye, the United Kingdom, and the United States.
important bank buffers) in the adverse scenario.
The July 2025 baseline scenario assumes stable unemployment, a Under stricter criteria—either a CET1 ratio falling
slight decline in global GDP growth before recovering to about
3 percent in 2027, and falling short-term interest rates to support
GDP recovery and growth that contribute to improvement in the 29Central banks react based on Taylor-type rules (see Vitek 2018),
global CET1 ratio of 6 basis points. while fiscal authorities provide demand support in large jurisdictions,
27The severity was similar, but the shock is more protracted except in high-debt countries.
than in the 2023 stress test (see the October 2023 Global Financial 30The assumption of a constant net interest margin—and of net

Stability Report). interest margins being maintained in an adverse scenario—is based

28See IMF (2025a; 2025b) for more detailed stress testing results on econometric results that do not show economically meaningful
of banks and nonbank financial intermediaries using country specific pass-through coefficients from short-term interest rates to deposit
scenarios and supervisory data. and lending rates.

International Monetary Fund | October 2025 21

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.15. Global Bank Stress Tests Results

In aggregate, global banks are resilient to the adverse macro scenario, in Rising loan losses and expenses are the main force behind capital
part because of high starting capital ratios. depletion under the adverse scenario.
1. CET1 Ratio by Region, 2024–27 2. Global Contribution to CET1 Ratio Change
(Percent) (Percentage points)
25 25
Baseline Adverse 0.5
1.8
11.5
20 20

15 13.0 15
−6.8 −0.1 −0.7 12.3
−1.7 −0.5
10 10
−4.8
5 5
Increase Decrease Total
0 0
2024
25
26
27
2024
25
26
27
2024
25
26
27
2024
25
26
27

Starting CET1

Loan loss
Net interest
income
Fee

Valuation

Expense
Liquidity
feedback
Tax

Dividends
Risk-weighted
asset
Ending CET1
Global AEs EMs EMs ex. China

The weak banks with CET1 ratio below 7 percent account for 18 percent of A few banks would fall below the minimum 4.5 percent CET1 ratio
banking sector assets. threshold.
3. Share of Total Assets in Weak Banks, by Region 4. Distribution of Banks in Adverse Scenario, by CET1 Ratio
(Percent; number of banks) (Percentage of total assets)
Baseline: Below CET1 of 7 percent (plus G-SIB buffer) <4.5 [4.5, 7) [7, 9) [9, 11) [11, 13) ≥13
50 Baseline: Below CET1 of 7 percent or CET1 falling by 100
5 percent more (plus G-SIB buffer)
90
Adverse: banks Below CET1 of 7 percent (plus G-SIB buffer)
40 Adverse: Below CET1 of 7 percent or CET1 falling by 80
5 percent more (plus G-SIB buffer) 70
30 100 60
137
50
82 50
20 32 37 40
30
10 20
7 12 10
17 38 10 27 7 11 3 6
0 0
Global Advanced Emerging Emerging markets 2024 Trough 2024 Trough 2024 Trough
economies markets excluding China G-SIB Non–G-SIB Non–G-SIB
advanced economies emerging markets

Sources: Fitch Connect; Fitch Solutions; and IMF staff calculations.

Note: In panel 2, “loan loss” refers to loan loss provisions and “valuation” to valuation losses. “Expense” refers to operating costs, assumed to be constant over time and
across scenarios as a share of risk-weighted assets. Liquidity feedback is based on the extra funding cost of pledging securities at stressed prices and the extra cost of
accessing central bank facilities. In panel 3, “global” refers to all banks in the sample. Recent FSAPs to two systemically important jurisdictions have found similar results
for the largest banks, using supervisory data provided for the stress tests performed by IMF staff. CET1 = Common Equity Tier 1 capital; G-SIB = global systemically
important bank.

below 7 percent or a decline of 5 percentage points Although no country’s banking system would fail
or more—the number of weak banks increases to 137 to meet the minimum 4.5 percent CET1 ratio under
globally, accounting for 25 percent of global bank the adverse scenario, a few institutions would not meet
assets (Figure 1.15, panel 3). These weak banks have the requirement under the adverse scenario. These
persistent vulnerabilities: most were already consid- distressed cases account for about 1 percent of global
ered weak in the April and October 2023 issues of the assets and would require a $25 billion recapitalization
Global Financial Stability Report. Rising bank exposures to bring the CET1 ratio back to 4.5 percent. None
to NBFIs could further increase capital depletion (see of the banks that would fall below the minimum
the section “Stronger Bank-Nonbank Nexus Increases 4.5 percent threshold are global systemically important
Contagion and Liquidity Risks”). banks (Figure 1.15, panel 4).
22 International Monetary Fund | October 2025
 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

In the United States and the euro area, about large banks also have concentrated exposures. Exposure
10 percent of total loans go to the manufacturing, to private equity and credit funds alone is substantial
retail, and wholesale trade sectors, which are vulnerable ($497 billion) and growing rapidly, up 59 percent
to trade tensions. Persistent geoeconomic tensions between the fourth quarter of 2024 and the second
could lead to turmoil in financial markets. This is what quarter of 2025 (Figure 1.16, panel 2).33 Banks are
happened in the first quarter of 2025, when expected increasingly lending to private credit funds because
default frequencies, an indicator of the probability of these loans often deliver higher returns on equity than
default, increased by 100 basis points for trade sectors traditional commercial and industrial lending, thanks
in the United States before subsiding by July. How- to the lower capital requirements allowed by their col-
ever, even if such a rise in expected default frequencies lateral structure. Concentration among private equity
were to continue for a year, the additional effects on and private credit borrowers is also increasing: five
banks would be small: the average additional CET1 large fund managers account for about one-third of the
ratio would decline by 10 basis points in the United aggregate loan commitments of the entire private credit
State and by 20 basis points in the euro area. The most and equity industry (Levin and Malfroy-Camine 2025;
affected banks also have higher capital ratios. Pandolfo 2025). US banks with high NBFI exposure,
defined as exposure greater than 100 percent of Tier 1
capital, also have a more fragile funding structure than
Stronger Bank-Nonbank Nexus Increases their low-exposure peers, relying more on noncore and
Contagion and Liquidity Risks wholesale funding (Figure 1.16, panel 3).
As NBFIs increase their share and importance in the Banks’ growing exposures to NBFIs mean that
global financial system, they are becoming increasingly adverse developments at these institutions—such as
reliant on banks for funding.31 Banks lend to a variety downgrades or falling collateral values—could signifi-
of NBFIs, including mortgage companies, investment cantly affect banks’ capital ratios. IMF staff assessed
funds, broker-dealers, and securitization vehicles. In the potential impact on euro area and US banks under
turn, these NBFIs lend directly to businesses and a scenario in which the average risk weight for NBFI
consumers, and conduct activities in government bond exposures rises from 20 percent to 50 percent and
and other capital markets (see Box 1.3 and section borrowers draw down 100 percent of credit lines and
“Sovereign Bond Market Functioning Depends on undrawn commitments. The results suggest that the
Nonbank Financial Intermediaries”). Banks’ exposure impact on banks’ solvency ratios could be substan-
to NBFIs is large: in Europe and the United States, tial. CET1 ratios would decline by more than 100
NBFI loans represent, on average, 9 percent of banks’ basis points in about 10 percent of US banks and
loan portfolio, with exposures amounting to about 30 percent of European banks (Figure 1.16, panel 4).
$4.5 trillion, of which $2.6 trillion corresponds to Furthermore, the IMF GST adverse scenario, com-
loans and the rest to undrawn commitments.32 bined with an additional NBFI shock for euro area
The growing exposure to NBFIs is generating con- and US banks, projects an increase in the share of
centration risk among some banks in the United States weak banks—mostly in Europe, as the most affected
and Europe (Figure 1.16, panel 1). In the United US banks are already classified as weak (Figure 1.16,
States, banks representing almost 50 percent of the panel 5). The average additional CET1 ratio impact
sample assets have exposures to NBFIs exceeding their is 120 basis points for euro area banks and 65 basis
Tier 1 capital. While large banks serve as the primary
lenders to NBFIs—accounting for 90 percent of all 33NBFI exposure is defined as the sum of NBFI loans and

lending to these intermediaries—exposure concentra- NBFI unused commitments. US banks report NBFI loans and
unused commitments by type of intermediary (mortgage, business,
tion is more severe among large regional banks and
consumer, private equity funds, and other). Private equity funds
those with assets under $100 billion. In Europe, some include capital call commitments and other subscription-based
facilities to private equity and venture capital funds, or any other
31The growth of the global NBFI sector outpaced banking sector partnership funds that raise capital through limited partnership
growth, with its share of total global financial assets at 49.1 percent arrangements. Loans in this category include capital call subscription
(or $239 trillion) in 2023 (FSB 2024). facilities, which are loans to private equity and private credit funds
32NBFI loan amounts are based on aggregate data for European secured by their limited partners’ undrawn capital commitments to
banks for the fourth quarter of 2024 and bank-level data for US the fund, and net asset value loans that are secured by one or more
banks for the second quarter of 2025, as reported by the European of the fund’s existing equity or debt assets. Amounts are based on
Banking Authority and the Federal Financial Institutions Examina- 134 banks reporting this level of public disclosure in the second
tion Council’s Consolidated Reports of Condition and Income. quarter of 2025.

International Monetary Fund | October 2025 23

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.16. Bank Exposure to Nonbank Financial Intermediaries

Several banks in the US and the euro area have US banks have substantial undrawn US banks with high NBFI exposure
exposures to NBFIs exceeding their capital. commitments with private credit and equity concentration tend to rely more on wholesale
funds. funding.
1. Banks’ NBFI Exposure to Tier 1 Capital 2. NBFI Exposure, by Type 3. Characteristics of Banks with High versus
(Percent) (Percent of total exposure by bank group) Low NBFI Exposure Concentration
(Based on standardized z-scores for each
All other Private equity funds financial metric)
Total NBFI exposure to Tier 1 capital
Consumer Business
Total NBFI loans to Tier 1 capital
Mortgage High NBFI exposure Low NBFI exposure
600 Loan and lease

$335B
$739B
$939B

$291B

$170B
$551B
$1.2T

$97B
allowance/total
500 100 loans/leases
80 Equity to Net loan
400 0.6
60 assets growth
0.4
300 40 Return on
0.2 Net loans and
20 average
200 0 leases/assets
0 equity
−0.2

commitments

Undrawn
commitments

Loans
Undrawn
commitments
Loans
Undrawn
commitments

Loans
Undrawn

Loans
100 Net interest Investment
margin portfolio
0 depreciation
United States Euro area Net noncore
Reliance on
Banks rated from highest to lowest funding On-hand
All G-SIBs Large Other wholesale
NBFI exposure concentration dependence liquidity/
regional banks funding
liabilities
banks

A deterioration in NBFI credit risk, combined Adding NBFI stress to the Global Stress Test A few banks could face liquidity pressures to
with the withdrawal of all unused adverse scenario for EU and US banks would cover potential outflows from NBFI credit and
commitments, could materially affect banks’ increase the share of weak banks. liquidity lines.
capital ratios.
4. Decline in CET1 Ratio and Share of Banks’ 5. Share of Total Assets of Weak Banks, 6. Number of Banks with Negative Net
Total Assets by Region Available Liquidity
(Share of banks in sample, left scale; (Percentage of assets, vertical axis; number of (Share of banks in sample, left scale; percent
percentage of total assets, right scale) banks, bars) of total assets, right scale)
Share of banks with CET1 decrease of more than
Adverse: banks below CET1 of 7 percent
100 bps Share of banks, narrow liquidity metrics
(plus G-SIB buffer)
Share of banks with CET1 decrease of 50 to 100 bps Percentage of total assets, narrow liquidity
Adverse with NBFI shock: banks below
Percentage of total assets with CET1 decrease of metrics (right scale)
CET1 of 7 percent (plus G-SIB buffer)
more than 100 bps (right scale)
80 Europe United States 80 50 16 Europe United States 9
70 70 14 8
60 60 40 12 7
50 50 10 6
57 30 5
40 40 89 50 8
82 20 4
30 30 6 3
20 20 10 4 2
10 10 2 1
0 0 0 0 0
50 100 50 100 Global AEs 50 100 50 100
Stress risk weight Range of unused commitment outflow

Sources: Consolidated Reports of Condition and Income; European Banking Authority; Fitch Connect; Fitch Solutions; S&P Capital IQ Pro; and IMF staff calculations.
Note: Panel 1 shows total NBFI exposure, which includes loans and undrawn commitments for the United States as of June 2025, and exposures for the euro area as of
June 2024 that include an estimate of unused commitments among banks. Concentration is measured by the ratio of NBFI exposure to Tier 1 capital; each bar represents a
bank; and the sample includes banks reporting NBFI exposures of at least 10 percent of Tier 1 capital. Panel 2 shows the breakdown of NBFI loans and undrawn
commitments by NBFI loan type (business, consumer, and mortgage intermediaries, private equity funds, and all other NBFIs) for banks whose regulatory reports show
total assets exceeding $10 billion. The “All other” category includes exposures to insurance companies, hedge funds, investment funds, and pension funds. “Private equity
funds” includes private credit. “Large regional” banks refers to non–G-SIBs with total assets of at least $100 billion; and “Other banks” refers to banks with total assets of
less than $100 billion. Panel 3 shows standardized z-scores for each financial metric for banks with high and low NBFI exposure concentration in the United States (see
Online Annex 1.2 for definitions of each financial metric). Panel 4 is based on 2024 second quarter data for 109 banks in the euro area and 2025 second-quarter data for
362 banks in the United States. The shock assumes that the risk weights for NBFI exposures increase from 20 percent to 50 percent and 100 percent and that NBFIs draw all
unused commitments available. Panel 5 shows the number of banks falling below the 7 percent CET1 plus G-SIB buffer under the IMF GST adverse scenario, with an
additional NBFI shock for euro area and US banks. See the section “Higher Capital Ratios Strengthen Global Banks, but the Weak Tail of Banks Remains Substantial” for a
description of the IMF GST severe scenario. The NBFI stress assumes that risk weights increase from 20 percent to 50 percent and that all available commitments are drawn.
Panel 6 shows the number of banks where the net available liquidity becomes negative. The assessment considers a narrow liquidity metric that includes cash and
balances at banks. AEs = advanced economies; bps = basis points; CET1 = Common Equity Tier 1 capital; G-SIB = global systemically important bank; GST = Global Stress
Test; NBFI = nonbank financial intermediaries.

24 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

points for US banks, with the larger effect in Europe for financial stability, particularly in an environment
reflecting higher NBFI exposure relative to risk- marked by fluctuating interest rates. Despite significant
weighted assets.34 In a more conservative scenario monetary policy easing across major economies, banks’
in which commitments are fully drawn down and interest margins have shown remarkable resilience
risk weights reach 100 percent, CET1 ratios fall by (Figure 1.17, panel 1).36 Banks seem to be position-
100 basis points or more in 50 percent of banks ing themselves for additional interest rate declines.
(representing 39 percent of total assets) in Europe and European and North American banks have reduced the
12 percent of banks (representing 67 percent of total sensitivity of their net interest income to downward
assets) in the United States (Figure 1.16, panel 4).35 interest shocks (Figure 1.17, panel 2).
Furthermore, although most euro area and US By contrast, banks may be more vulnerable to
banks have sufficient liquidity buffers to honor their abrupt increases in bond yields and interest rates.
NBFI commitments, a few could face liquidity pres- Two years after heavy bond portfolio losses led to the
sures and may need to use less-liquid assets to cover demise of Silicon Valley Bank, stress test results show
potential outflows from NBFI credit and liquidity that global banks could incur valuation losses of about
lines. Sensitivity analysis shows that if NBFI borrowers 1 percentage point of the CET1 ratio in an adverse
were to fully draw these lines, 4 percent of US banks scenario in which longer-term government bond risk
(representing less than 1 percent of total assets) would premiums surge by 300 basis points to 500 basis points
lack enough liquid assets to meet the outflows, turning across advanced and emerging market economies—
their net available liquidity negative. The number perhaps driven by fiscal risks or eroding convenience
of banks under severe liquidity stress would rise to yields. However, losses are much more meaningful
5 percent of banks (representing 5 percent of sample for North American and European banks, reaching
assets) in the euro area and 14 percent of banks (repre- 2.5 percentage points and 1.5 percentage points of
senting 8 percent of sample assets) in the United States their respective CET1 ratios (Figure 1.18, panel 1). In
if a stricter definition of liquid assets is applied, includ- addition, in Europe, the sensitivity of banks’ economic
ing only cash and deposits at other banks (Figure 1.16, value of equity to upward shifts in interest rates has
panel 6). The impact of these outflows is concentrated increased, making them more vulnerable to a rise in
among smaller US banks and large euro area banks long-term bond yields as a result of more bond supply
that provide large liquidity and credit facilities relative or quantitative tightening (Figure 1.18, panel 2).
to their size. These banks also have lower liquidity European banks may therefore be sensitive to a steep-
ratios, higher asset encumbrance in the euro area, and, ening of the yield curve, with net interest income also
in the United States, a higher share of noncore deposits under pressure if policy rates are cut.
and a lower initial CET1 ratio compared with peers.
There could be additional impact of liquidity stress on
the solvency of these banks, which is not considered. Stablecoins’ Growth Could Affect Financial
Stability
Stablecoins—crypto assets issued by private
Banking Sector Stability Depends on Navigating institutions that promise a stable nominal value in a
Interest Rate Challenges given currency—have become a key component of
The ability of banks to maintain stable interest the digital asset ecosystem. The market grew rapidly
margins and keep bond portfolio losses at bay is crucial from about $3 billion in 2019 to almost $300 billion
at the end of September 2025 driven mainly by
34The methodology for the NBFI shock assumes that risk weights USDT (Tether), USDC (Circle), and other fiat-backed
for NBFI loans increase from 20 percent to 50 percent and that
unused commitments are fully withdrawn. The capital and liquidity stablecoins pegged to the US dollar (Figure 1.19,
impacts do not incorporate credit valuation adjustments related to panel 1). The rise of stablecoins could have three main
banks’ derivative links with NBFIs, which could affect banks’ risk- financial stability implications: (1) weaker economies
weighted assets and liquid asset needs.
35Liquidity shocks are based on second-quarter 2024 data for 109 may face currency substitution and reduced effective-
euro area banks and second-quarter 2025 data for 362 US banks. ness of policy tools, (2) the bond market structure
This sample is also used for the sensitivity analysis in Figure 1.16 could change with potential implications on credit
(panel 4), and it differs from the one used for the GST, which
includes a smaller set of banks with more complete data on a wider
set of variables. 36See Box 1.2 for a discussion about banks in China.

International Monetary Fund | October 2025 25

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.17. Banks’ Interest Margins Remain Resilient amid Concerns about Potential Valuation Losses
Banks’ interest margins have demonstrated remarkable resilience. European and North American banks have improved their sensitivity to
downward interest shocks.
1. Interest Rates and Banks' Net Interest Income 2. Net Interest Income Sensitivity to Parallel Shock down to
(Percent of Tier 1 capital, left scale; percent, right scale) Interest Rates
(Percentage of Tier 1 capital)
North America Europe Japan
SOFR Euribor 3M TIBOR 3M Europe North America Japan China
35 (right scale) (right scale) (right scale) 6 6
30 5 4

25 4 2
0
20 3
−2
15 2
−4
10 1 −6
5 0 −8
0 −1 −10
2021 22 23 24
2020:Q1

24:Q3

26:Q1
25:Q1

25:Q3
20:Q3

21:Q1

21Q3

22:Q1

22:Q3

23:Q1

23:Q3

24:Q1

Sources: Bloomberg Finance L.P.; Capital IQ; Visual Alpha; and banks’ Pillar III disclosures, including regulatory filings.
Note: Panel 1 shows banks with shares quoted on stock exchanges. Panel 2 illustrates banks’ net interest income sensitivity over the next 12 months to a parallel interest
rate shock across maturities as defined by the Basel Committee standard on the interest rate in the banking book (for example, generally −200 basis points for major
currencies). The sample of banks encompasses global systemically important banks and the largest regional banks in the United States (10 banks) and Europe (20 banks).
3M = three-month; Q = quarter; SOFR = secured overnight financing rate; TIBOR = Tokyo Interbank Offered Rate.

Figure 1.18. Potential Valuation Losses Remain a Concern

The IMF’s Global Stress Test identifies significant risks to valuations in case European banks are marginally more exposed to an upward shift of rates
of a substantial increase in rates. curve.
1. Valuation Losses in the First Year of the IMF Global Stress Test under 2. Sensitivity of Selected European Banks’ Economic Value of Equity
the Adverse Scenario to Upward Parallel Shock of Interest Rates
(Percentage points of the CET1 ratio) (Percentage of Tier 1 capital)
0 5

−0.5 0
−5
−1.0
−10
−1.5
−15
−2.0
−20
−2.5 −25
−3.0 −30
Global United Euro area Other Emerging 2021 22 23 24
States advanced markets
economies

Sources: Banks’ Pillar III disclosures; and the IMF‘s GST.

Note: Panel 1 includes a sample of banks in the IMF’s GST. Bars show valuation losses in an adverse scenario in which government bond term premiums surges by 300 to
500 basis points, according to the GST scenario. Panel 2 includes a sample of European global systemically important banks and the largest 20 European banks. EVE is
defined as the present value of assets less the present value of liabilities of the banking book (under a static balance sheet perspective), excluding own equity and other
instruments that do not generate interest. When assets or liabilities do not have a specific contractual maturity or embed implicit optionality, as is typically the case of
demand deposits, banks are allowed to make behavioral assumptions on the expected duration of those assets or liabilities. EVE sensitivity is calculated as the EVE change
after six interest rate shock scenarios as defined for currencies in the Basel interest rate in the banking book standard; the most impactful scenario is largely the parallel
upward shock presented here (that is, an interest rate shock of generally +200 basis points for major currencies). CET1 = Core Equity Tier 1 capital; EVE = economic value
of equity; GST = Global Stress Test.

26 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.19. Stablecoins Are Growing alongside Tokenized Assets, with Notable Cross-Border Flows
The stablecoin market has risen rapidly and The tokenization of Treasury funds is also Cross-border flows of stablecoins reflect strong
now stands at record heights. growing fast. dollar demand outside the United States.
1. Stablecoin Market Cap 2. Tokenized Treasury Funds 3. Net Stablecoin Flows
(Billions of dollars) (Billions of dollars) (Billions of dollars)
300 8 Africa
USDC USDT Other Other Treasury funds Middle and
OUSG in: Eas
7 t
250 USTB 8.61 2 out:
USDY .21
6
BENJI
200

Asi n:
BUIDL 5

out: rica

a an
20
i 0
.8
e
North Am

d Pacific
54.06
150 4

3
100

o u t:
8.3 0
2
50
1

0 0
out
: in: 9
2020 21 22 23 24 25 La 0.4 2 i n : 9
2 4. pe
ti n 1 0.5 9
Jan. 2023
Apr. 2023
July 2023
Oct. 2023
Jan. 2024
Apr. 2024
Jul. 2024
Oct. 2024
Jan. 2025
Apr. 2025
Jul. 2025
o
A
th e m e ric E ur
C a ri a a n d
b bean

Sources: Bloomberg Finance L.P.; Reuter 2025; RWA.xyz; and IMF staff calculations.
Note: Panel 1 appears in Reuter (2025). Panel 3 shows estimates by Reuter (2025) of bilateral net outflows of stablecoins in 2024 across regions. Orange areas represent
net flows from North America, blue arrows represent flows from Asia and the Pacific, red arrows represent flows from Africa and the Middle East, and purple areas
represent flows from Latin America and the Caribbean. BENJI = Franklin OnChain US Government Money Fund; BUIDL = BlackRock USD Institutional Digital Liquidity
Fund; OUSG = Ondo Short-Term US Government Bond Fund; USDC = US Dollar Coin, issued by Circle Internet Group Inc.; USDT = US Dollar Tether, issued by Tether
Limited; USTB = Superstate Short-Duration US Government Securities Fund.

disintermediation, and (3) runs faced by stablecoins amid a flurry of new fiat-backed stablecoins emerging
may generate forced selling of reserve assets. Potential in 2025.38
systemic effects would be conditional on stablecoins’ Nevertheless, the speed and volume of the adop-
continued growth. tion of stablecoin remains unclear. Projections by the
Recent global legal and regulatory initiatives could US Treasury Borrowing Advisory Committee of an
foster the issuance and integration of stablecoins into eightfold increase in stablecoin market capitalization
the financial system by providing clarity on issuance to about $2 trillion by 2028—roughly $500 billion
and oversight parameters (IMF, forthcoming).37 Major annually—are driven primarily by expectations of
US banks are preparing for a shift from cautious broader use in payments and cash management.
observation to active participation and adoption However, increased adoption faces significant chal-
lenges: different stablecoins often operate on separate
37The Guiding and Establishing National Innovation for US
blockchains, increasing transaction costs and frag-
Stablecoins Act, signed into law in July 2025, establishes a frame-
mentation in payments; they typically do not offer
work for stablecoins intended to be used for payments. Aiming to yields, making them less attractive than money market
reduce legal uncertainty and support broader crypto adoption, the funds (Nikolaou 2025); and ongoing improvements in
law establishes an oversight framework for stablecoins, reserve assets
traditional payment systems could reduce the need for
requirements, and compliance with anti-money laundering/combat-
ing the financing of terrorism (AML/CFT) legislation. The Digital blockchain-based alternatives.39
Asset Market Clarity Act complements these efforts by providing
legal and regulatory clarity for digital assets and reinforcing the 38JPMorgan has partnered with Coinbase to expand stablecoin

legitimacy of private sector innovations that accept stablecoins as access among clients starting in Fall 2025. Bank of America is
payment on blockchain platforms. These US initiatives are in step developing its own stablecoin, with Citigroup and JPMorgan also
with global trends: The European Union’s Markets in Crypto-Assets evaluating issuance of their own stablecoins. Meanwhile, new US
Regulation enforces a framework for crypto assets, including licens- dollar–backed stablecoins have emerged (for example, USDS), while
ing and transparency standards and anti-money laundering/combat- mainstream payment and e-commerce platforms have integrated
ing the financing of terrorism requirements for stablecoin issuers and stablecoins (PayPal with PYUSD).
providers of crypto services across Europe. Hong Kong SAR’s new 39Stablecoin transactions exhibit fragmentation, although, in

regime positions stablecoins and tokenized assets at the heart of its practice, major stablecoins partner with exchanges, which offer yield
fintech strategy. to incentivize users to hold the stablecoin.

International Monetary Fund | October 2025 27

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.20. The Rise of Stablecoins Comes with Potential Concerns over Financial Stability
Dollar-backed stablecoins are buying more Treasury bills. The 2016 money market mutual fund reform spurred a large increase in
demand for Treasury bills, affecting the relative prices of other assets.
1. Ownership of US Treasury Bills 2. Implications of the 2016 Money Market Mutual Fund Reform
(Billions of dollars, left scale; percent, right scale) (Percent, left scale; basis points, right scale)
2025:Q2 Growth YE23 to 2025:Q2 (right scale) T-Bill share privately held US MMF T-Bill share
2,500 100 50 T-Bill spreads (right scale) Commercial paper spread 20
45 (right scale) 15
2,000 80 40 10
60 35 5
1,500 30 0
40 25 −5
1,000 20 −10
20 15 −15
500 0 10 −20
5 −25
0 −20 0 −30
MMFs Rest of Foreign Insurance Fed Stablecoins Mutual Jan. Jul. Jan. Jul. Jan. Jul. Jan.
world official companies funds 2014 2014 2015 2015 2016 2016 2017

Sources: Bloomberg Finance L.P.; Crane data; US Z.1. Statistics; and IMF staff calculations.
Note: The Treasury bill growth rate is calculated on the basis of outstanding total US Treasury debt, after excluding Federal Reserve Treasury holdings. The Treasury bill
spread is calculated as the difference between the three-month Treasury bill rate and the three-month overnight index swap rate. The spread for commercial paper is the
difference between 90-day nonfinancial commercial paper and the three-month overnight index swap rate. Fed = US Federal Reserve; MMF = money market fund;
Q = quarter; YE = year end.

At the same time, traditional financial instruments Stablecoins are typically legally required to be
such as deposits and money market mutual fund shares backed by high-quality liquid assets such as short-term
have turned to tokenization, creating representations of government bonds, demand deposits, and govern-
them as digital tokens on a blockchain (Box 1.2 in the ment money market funds. The mainstreaming and
October 2024 Global Financial Stability Report). The continued growth of stablecoins could have substantial
tokenized market has grown substantially (Figure 1.19, implications for these assets. Stablecoin issuers already
panel 2), although it remains small compared with hold significant volumes of short-term government
stablecoins, which dominate blockchain-based pay- debt and are among the largest buyers (Figure 1.20,
ments and settlements. Tokenization may allow these panel 1), already putting downward pressure on US
instruments to compete with stablecoins, though both Treasury bill yields (Ahmed and Aldasoro 2025).
could grow in parallel. The 2016 US Securities and Exchange Commission
To date, net stablecoin flows are largely flowing money market mutual fund reform illustrates how reg-
outward from North America to the rest of the world, ulatory or structural shifts can abruptly reshape demand
reflecting dollar demand in those regions (Figure 1.19, across asset classes and affect market pricing. The reform
panel 3; see also Reuter 2025). Easy access to dollar-­ triggered a large reallocation from prime money market
denominated stablecoins raises concerns about currency mutual funds to government money market mutual
substitution and reduced monetary policy transmission, funds, doubling demand for Treasury bills by nearly
particularly in jurisdictions with weak macroeconomic $500 billion during a period when supply remained
fundamentals. In addition, a shift from physical cur- broadly stable, along with reducing demand for com-
rency to stablecoins could reduce seigniorage, affect- mercial paper and other short-term private sector debt.
ing central bank income and dividend distribution. This shift modestly lowered Treasury bill yields and
Stablecoins also pose risks to capital flow management, raised commercial paper yields (Figure 1.20, panel 2).
notably for emerging market economies, as they allow The expansion of stablecoins could have similar
US dollar liquidity to move outside regulated channels, effects, depending on whether it creates new demand
potentially weakening the effectiveness of capital flow for short-term sovereign bonds, as in the money
and foreign exchange measures and increasing risks for market mutual fund reform case, or simply reallocates
illicit uses of stablecoins (Cardozo and others 2024). demand. If stablecoins grow at the expense of money

28 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

market mutual funds, yield effects may be muted, as One factor contributing to stretched valuations
demand will be reallocated from the funds. However, is that instead of using cash flow for investments
if stablecoins displace bank deposits, which fund (Figure 1.21, panel 2), firms have engaged in financial
longer-term bonds and loans, demand could shift engineering to support valuations. Share buybacks
toward Treasury bills. Such a shift may steepen yield have kept on growing—for example, so far this year,
curves and raise concerns about credit disintermedia- US financial, technology, and communications services
tion as banks could face reduced funding capacity for firms have bought back near $1 trillion of stocks on an
lending to households and businesses. Altering yield annualized basis (Figure 1.21, panel 3). In Japan, the
curve dynamics can also complicate interest rate con- ratio of share buybacks to market capitalization is on
trol by central banks. These concerns would be ampli- pace to reach around 2.4 percent in 2025, in contrast
fied were stablecoins denominated in foreign currencies to 1.1 percent in 2024. The elevated valuations are,
to be widely adopted. The impact would depend on however, facilitated at the cost of investments in future
the geographic adoption patterns, asset allocation growth opportunities.
strategies, and supply of short-term government bills: High valuations in stock markets and buoyant
an increase in bill issuance can mitigate price pressures, risk sentiment also may have helped lower corporate
though at the cost of higher exposures to short-term funding costs. In reality, default rates, especially for
interest rate risk for the government. leveraged loans, have been climbing, even though some
Implications of a wider stablecoin adoption can of the defaults are voluntary liability management
stretch beyond their impact on the yield curve. Because exercises, including debt exchanges (Khoda and others
stablecoins may be subject to run risk, fire sales of 2025; Figure 1.21, panel 4). This suggests that some
stablecoins’ reserve assets—such as bank cash depos- weaker firms are struggling in the current environ-
its and government securities—could spill over into ment. Indeed, funding liquidity is strained among
bank deposits and government bond and repo mar- vulnerable borrowers (see the section “Some Private
kets. This could increase volatility and require central Credit Direct Lending Borrowers Remain Vulnerable”).
bank intervention. Moreover, in a scenario of broader Looking ahead, stretched valuations in stocks and
adoption, any loss of parity with the reference currency corporate bonds are vulnerable to correction, owing to
would also impose direct losses and heightened uncer- the likelihood that tariffs will dampen corporate prof-
tainty on a large user base. Financial fragmentation in itability (see the section “Equity Markets Exhibit High
payment systems resulting from limited interoperabil- Valuations and Concentration Risks”). Tariffs could
ity among stablecoins, and between stablecoins and also have larger-than-expected effects on inflation and
existing financial market infrastructure, may further growth as a result of firms passing rising input costs
accentuate these risks. to consumers, potentially leading to higher inflation
with stagnant demand. Empirical evidence shows that
inflation and growth shocks cause corporate spreads
Corporate Credit Risk to widen and stock prices to decline (Figure 1.21,
panel 5). Lower stock prices worsen the effect of tariffs
The Corporate Sector Is Resilient to Tariffs So Far on credit fundamentals, while higher term premiums
Even though corporate profit margins have been may put additional pressure on corporate debt issuers
revised downward since the April 2025 Global Finan- (see the section “Expanding Fiscal Deficits Exert
cial Stability Report, corporate balance sheets in many Pressure on Bond Market Stability”).
countries are still healthy in aggregate, keeping corpo- IMF staff have developed a more comprehen-
rate credit risks at bay, although vulnerabilities remain sive assessment of the cross-country costs for firms
unresolved. In the United States, interest income resulting from higher effective tariff rates on their
on assets has increased more than liabilities during US exports. Tariff-related costs also depend on the
high-interest-rate years, lowering firms’ net interest share of exports to the United States in a country’s
payments (Figure 1.21, panel 1) and propping up total exports, the proportion of exporting firms in
their cash buffers. Net interest payments have recently the country, and country corporate-level factors (see
started to increase, as maturing corporate debts need Online Annex 1.5 for more details). For the average
to be refinanced at higher fixed rates (see the October country, additional tariffs would reduce firms’ profit
2023 Global Financial Stability Report). margin by 1 percentage point (Figure 1.22, panel 1,

International Monetary Fund | October 2025 29

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.21. Corporate Fundamentals and Risks

In the United States, ample interest income on Saving surpluses (cash flow as a percentage of ... enabling firms in many cases to engage in
financial assets had largely offset interest GDP) have reflected steady earnings amid a lack financial engineering to support equity
payments until recently. of attractive investment opportunities ... valuations.
1. US Nonfinancial Firms’ Net Interest 2. Global Nonfinancial Corporate Capital 3. Share Buybacks
Payments Expenditure and Cash Flow (Billions of dollars; 2025 bars are year-to-date
(Billions of dollars, left scale; percent, right (Percent of GDP) annualized)
scale)
Net interest payments Capex as a percentage of GDP 14 Topix 1,200
400 Fed Funds rates (right scale) 6 Cash flow as a percentage of GDP S&P 500
13 Stoxx Europ 600 1,000
350 5 12
800
4 11
300
10 600
3
250 9
2 400
8
200 1 200
7
150 - 6 -
2015 16 17 18 19 20 21 22 23 24 25
1999:Q1
2001:Q1
03:Q1
05:Q1
07:Q1
09:Q1
11:Q1
13:Q1
15:Q1
17:Q1
19:Q1
21:Q1
23:Q1
25:Q1

2012
13
14
15
16
17
18
19
20
21
22
23
24
25
Elevated corporate valuations have enabled That said, valuations are vulnerable to “stagflation”—high-inflation and low-growth surprises.
firms to restructure loans, leading to higher
loan default rates than for bonds.
4. US Corporate Debt Default Rates 5. Sensitivity of Global Risk Assets to the Stagflation Surprise Index
(Percent) (Basis points, percent, y-axis; points, x-axis)
High-yield corporate bond spread change Stock return
8 High-yield bonds 200 30
Broadly syndicated
loans
6 100 10

4 0 −10

2 −100 −30

0 −200 −50
2011 13 15 17 19 21 23 −1.5 −0.5 0.5 1.5 −1.5 −0.5 0.5 1.5

Sources: Bank of America; Bloomberg Finance L.P.; Bureau of Economic Analysis; European Central Bank; EUROPACE AG/Haver Analytics; Federal Reserve Bank of St. Louis;
IMF, World Economic Outlook database; Japanese Ministry of Finance; Refinitiv DataStream; US Department of the Treasury; and IMF staff calculations.
Note: Panel 2 shows the GDP-weighted average of the euro area, Japan, and the United States. In panel 5, the Stagflation Surprise Index is defined as the GDP-weighted
average of the spread of the Inflation Surprise Index and the Growth Surprise Index between the United States, the euro area, the United Kingdom, and Japan. A higher
index value means a larger stagflation surprise, indicating higher inflation, lower growth, or combined surprise relative to market expectations. High-yield corporate bond
spread changes are based on the Bloomberg Global High-Yield Corporate Bond Index, and stock returns are based on the MSCI All Country World Index.

green line). Some countries have firms with much capabilities (Figure 1.22, panel 2). Two extreme
higher sensitivity to additional tariffs (Figure 1.22, scenarios can provide an estimated range of deterio-
panel 1, red bubbles) and could experience a steeper ration in earnings. First, a 100 percent cost pass-
erosion of margins. through from a country’s exporters to US consumers
These estimated tariff costs can be translated into and second, a 0 percent pass-through with an equal
effects on corporate earnings and debt servicing distribution of tariff-related costs between import-

30 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.22. Corporate Debt Sensitivity Analysis

Tariff costs faced by firms are more pronounced for some countries, ... resulting in large shares of firms with an interest coverage ratio below 1
eroding their profit margin by 1 percent on average ... for a broad set of countries, while also accounting for higher debt
refinancing costs.
1. Effective Tariff Rates versus Impact of US Tariffs on Corporate 2. Sensitivity of Debt-at-Risk to Tariffs and Refinancing Costs
Profitability across Countries (Percent of total corporate debt)
(Percentage of revenue)
Debt whose ICR could slip below 1
5 Heigher sensitivity Others (range for 0–100 percent 70
passthrough of tariff costs)
12 60
US tariff impact on corporate

4 Debt whose ICR was below 1 in 2025:Q1

Subsample of countries with higher 50
sensitivity to US tariffs
profit margins

3 40
5
2 2 Cross-country average 30
impact of US tariffs on
4 20
corporate margins = 1%
1
10
0 0
0 10 20 30 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Additional effective tariffs Country

Sources: Dealogic; IMF, World Economic Outlook database; S&P Capital IQ; US Tariff Tracker, Center for Global Development; and IMF staff calculations.
Note: The sample includes Bangladesh, Brazil, Canada, Chile, China, Colombia, France, Germany, India, Japan, Korea, Malaysia, Mexico, the Philippines, South Africa,
Spain, Türkiye, the United Kingdom, the United States, and Vietnam. Countries are assigned numbers for both the panels. The additional tariffs are calculated relative to
January 20, 2025, and as of July 12, 2025. For panel 1, the impact of US tariffs on corporate profitability for a country is estimated by the interaction of the additional
tariffs, the share of export revenue exposed to US firms engaged in exports, and the proportion of exporting firms in total, which is proxied by goods exports as a
percentage of GDP in 2024 (see Online Annex 1.5 for details). Countries whose companies face a larger-than-average increase in implied tariff costs (that is, greater than
1 percent of revenue) are identified as having higher sensitivity (red bubbles). The corresponding labels for the red bubbles show the assigned country number. The
horizontal dashed green line is the simple average across countries in the sample, excluding the United States. Panel 2 shows the possible range of increases in
debt-at-risk under varying degrees of cost pass-through scenarios for each country in the sample and higher refinancing costs. Debt-at-risk is defined as the share of debt
with an interest coverage ratio below 1 in total. The countries are sorted by the size of debt-at-risk as of the first quarter of 2025 (see Online Annex 1.5 for details on
scenario construction and calculations). ICR = interest coverage ratio.

ing and exporting firms. A sensitivity analysis of and ensure sustainable operations, potentially
both scenarios shows that since debt refinancing raising inflation.
costs are rising in coming years as large volumes of
debt mature (see the October 2024 Global Financial
Stability Report and the April 2025 Global Finan- Some Private Credit Direct Lending Borrowers
cial Stability Report), a sizable share of firms could Remain Vulnerable
end up with an interest coverage ratio (the ratio of Elevated policy rates and uncertainty continue to
earnings over interest expenses) below 1, especially exert pressure on direct lending borrowers, although the
in countries where tariff costs are high (Figure 1.22, industry has demonstrated flexibility in managing short-
panel 2, red bars). Some countries already operating term pressures. Continued earnings growth, declining
with a low percentage of risky corporate debt (debt policy rates, the use of payment-in-kind features of
with an interest coverage ratio below 1) could expe- payment-in-kind features—that is, paying interest with
rience a large increase in their share, and this would additional debt—and recent restructurings (Figure 1.23
heighten credit risk. In addition to corporate debt panel 1, red bars) have helped lift some cash-flow
serviceability, higher tariff costs would be a drag on pressure from borrowers. As a result, although the
macroeconomic fundamentals. Firms sensitive to overall interest coverage ratio remains low (Figure 1.23,
tariffs would have limited scope to absorb the addi- panel 1, blue line), the share of borrowers with a cash-
tional tariff costs through improvement in opera- only interest coverage ratio below 1 has declined con-
tional efficiency. At the same time these firms would siderably, returning to levels observed before the interest
be reeling under higher refinancing costs. Hence, rate hiking cycle (Figure 1.23, panel 2).
they would be cornered into passing the additional Despite the wave of restructurings, liquidity remains
costs to consumers to manage their profit margins strained among the more vulnerable borrowers,

International Monetary Fund | October 2025 31

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 1.23. Credit Risk and Fundamentals of Direct Lending

Elevated policy rates keep the interest coverage ratio low, and selective The series of adjustments helped better position direct lending borrowers
defaults, or restructurings, have declined. to service their debt in the short term.
1. Average Interest Coverage Ratio for Direct Lending Borrowers and 2. Credit Rating Assessments with Cash Interest Coverage below 1
the Debt Restructuring (Selective-Default) Rate (Percent of total number of S&P Global credit estimates)
(Ratio, left scale; percent, right scale)
Selective defaults (right scale)
Apr. 2022–Feb. 2023 20
Average interest coverage ratio (left scale)
2.7 6.0 Fed rate hikes
5.5 18
2.5
5.0
16
2.3 4.5
4.0 14
2.1
3.5
1.9 3.0 12
2.5
1.7 10
2.0
1.5 1.5 8
2021 22 23 24 25:H1
2021:Q3
21:Q4
22:Q1
22:Q2
22:Q3
22:Q4
23:Q1
23:Q2
23:Q3
23:Q4
24:Q1
24:Q2
24:Q3
24:Q4
25:Q1

A vulnerable tail of direct lending borrowers is under pressure from weak ... with vintages before rate hikes having a higher share of weaker direct
liquidity ... lending borrowers.
3. Number of Ratings Upgrades and Downgrades of Direct Lending 4. Default Rates of Direct Lending Borrowers across Vintages
Borrowers (Percent in Morningstar DBRS sample)
(Number of actions in S&P Global Credit estimates)
Percentage of vintage defaulted (left scale)
Credit estimates: lowered Credit estimates: raised Percentage of vintage still active (right scale)
120 8 100
7 90
100 80
6
80 70
5 60
60 4 50
3 40
40 30
2
20 20
1 10
0 0 0
2019 20 21 22 23 24
2021:Q1
21:Q2
21:Q3
21:Q4
22:Q1
22:Q2
22:Q3
22:Q4
23:Q1
23:Q2
23:Q3
23:Q4
24:Q1
24:Q2
24:Q3
24:Q4
25:Q1
25:Q2

Vintage year

Sources: Morningstar DBRS 2025; and S&P Global Ratings 2025.

Note: In all four panels, data are based on the sets of direct lending borrowers reviewed by Morningstar DBRS and S&P Global (Morningstar DBRS 2025; S&P Global
Ratings 2025). In panel 1, selective defaults are defined in S&P Global Ratings (2025) and include amend-and-extend transactions and payment-in-kind switches allowing
borrowers to choose to pay interest expenses by issuing more debt. In panel 2, the interest coverage ratio is based on the cash-only portion of interest expense, excluding
the accrued payment-in-kind portion of the coupon’s interest rate. The interest coverage ratio is defined as earnings before interest, taxes, depreciation, and amortization
divided by interest expense.Fed = US Federal Reserve.

contributing to a rise in borrower downgrades continuing, and on the pace at which policy rates
(Figure 1.23, panel 3). Overall, defaults remain more normalize.
common among firms that borrowed from private
credit before monetary policy began to tighten in
2022 (Figure 1.23, panel 4). These older vintages of Broader Retail Participation in Private Credit
borrowers also include more firms constrained by Can Generate New Risks
liquidity (Morningstar DBRS 2024). Because direct Retail investors have become—and are expected
lending involves high leverage and is structured with to remain—major contributors of new funds for the
variable rates, borrowers depend on economic growth expansion of private credit (Figure 1.24, panel 1).

32 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.24. Cyclicality and Liquidity Risk for Increasing Retail Participation in Private Credit
Retail funds have become a major part of private credit assets under Perpetual nontraded business development companies’ funds closely
management. followed market sentiment in 2022–23
1. Assets under Management for Private Credit Funds 2. Perpetual Nontraded Business Development Companies’ Quarterly
(Billions of dollars) New Issuance and Redemptions
(Billions of dollars, left scale; percent, right scale)
Institutional Retail Recessionary sentiment Net issuance (left scale)
Redemptions (right scale)
2,000 12 4
3.5
1,600 10
3
8
1,200 2.5
6 2
800 1.5
4
1
400 2
0.5
0 0 0
2014 15 16 17 18 19 20 21 22 23 24

2022:Q1

22:Q2

22:Q3

22:Q4

23:Q1

23:Q2

23:Q3

23:Q4

24:Q1

24:Q2

24:Q3

24:Q4

25:Q1
Sources: Capital IQ; and PitchBook.
Note: In panel 2, the recessionary sentiment is Bloomberg’s consensus forecast of the recession probability in the United States within one year.

Private credit asset managers are developing new These vulnerabilities underscore the need for robust
products to attract retail investors, including retire- asset-liability management and sufficient sources
ment savings accounts, countering the slowdown in of liquidity to cover crowded redemptions during
institutional fundraising. The increasing share of retail a shock. Perpetual nontraded BDCs—vehicles that
investors in private credit can change the industry in allow periodic redemption windows—usually maintain
two important ways: by introducing higher liquidity leverage meaningfully below the regulatory maximum.
risks and by making investments more procyclical. This gives them room for additional borrowing if
Most private credit funds currently pose little matu- unexpected redemptions occur. However, a key risk to
rity transformation risk because traditional structures this liquidity management strategy is asset devaluation
like private credit collateralized loan obligations and during an economic shock, which would mechanically
closed-end funds do not typically allow redemptions increase the actual leverage, lower collateral value, and
during the fund’s lifespan. The expansion to retail inves- potentially reduce borrowing capacity.
tors is associated with the growth of semiliquid invest- Another liquidity management tool used by per-
ment vehicles that offer periodic windows of liquidity, petual nontraded BDCs is holding larger portfolios of
ranging from quarterly redemptions (regular or discre- marketable assets (mostly traded leveraged loans). These
tionary) to exchange-traded funds with daily liquidity. portfolios amount to 10 percent to 40 percent of total
Broader retail participation in private credit may assets, compared with the 1 percent to 3 percent range
also add procyclicality to fund inflows and outflows. typically observed in publicly traded BDCs that do
Some evidence suggests that private credit lending not permit redemptions. While part of this marketable
could be more stable than similar leveraged loans and portfolio may temporarily hold newly raised funds
less responsive to shocks than high-yield bond markets before deployment into private credit loans, it can
(see Chapter 2 of the April 2024 Global Financial also serve as a liquidity cushion against idiosyncratic
Stability Report). However, products with stronger redemption shocks and, to a lesser extent, against
retail participation, such as perpetual nontraded economywide downturns. That said, the effectiveness of
business development. companies (BDCs), seem to marketable securities as a buffer has its limits, consid-
track the cyclicality of market sentiment more closely ering that liquidity in semiliquid assets may evaporate
(Figure 1.24, panel 2). during times of stress.

International Monetary Fund | October 2025 33

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Investment Funds Dominate the High-Yield bonds issued by certain borrowers, particularly those
Bond Market rated CCC or lower (Figure 1.25, panel 5). Although
this concentration may be less concerning for the
Open-ended investment funds and exchange-
funds themselves, because they typically manage large
traded funds own a large share of high-yield bonds
volumes of relatively diverse assets, it is a risk for issu-
outstanding. Since 2015, these funds’ share of the US
ers, for whom prices could fall were a dominant debt
high-yield bond market has risen from 37 percent
holder to exit the market. This situation could become
to 45 percent. By comparison, their shares in other
especially problematic were it to coincide with a period
fixed-income markets, such as investment-grade prod-
when the company needed to refinance debt.
ucts and US Treasuries, have also risen but are signifi-
Exchange-traded funds have also grown their share
cantly smaller (Figure 1.25, panel 1).
of the US high-yield bond market to 7 percent in
This increased presence makes high-yield bonds
2024, from 3 percent in 2015. The sensitivity of
more vulnerable to the behavior of open-ended funds
high-yield bond exchange-traded funds to S&P 500
and exchange-traded funds, particularly because these
returns is higher than the sensitivity of their underly-
funds might face sudden runs. Monthly flow data for
ing index to S&P 500 returns (Figure 1.25, panel 6).
high-yield bond funds and exchange-traded funds over
This suggests that the rise in exchange-traded funds
the past decade show that in eight episodes globally,
may increase contagion risk and possibly amplify price
outflows exceeded 2 percent of assets under manage-
moves across asset markets during periods of stress.42
ment (over $10 billion of outflows). This is much
worse than other fixed-income sectors, which have
suffered outflows over 2 percent of assets in only one Policy Recommendations
or, at most, two instances during the same period
The ground is shifting in the financial system. Some
(Figure 1.25, panel 2). Furthermore, considering the
shifts have already been under way, but their growing
high-yield market is relatively illiquid, the impact of
intensity requires policymakers to remain vigilant and
these outflows on bond yields can be substantial. In
respond promptly to changing circumstances as they
the United States, which makes up about 60 percent
unfold.
of the global high-yield bond market, average monthly
To ensure macroeconomic stability, central banks
trading volume is about $200 billion.40 This implies a
should stay attentive to the risks to inflation associated
high outflow-to-trading-volume ratio compared with
with tariffs. So far, central banks that have started
other fixed-income markets (Figure 1.25, panel 3). The
easing cycles have cut interest rates gradually, in part
liquidity mismatch means that funds may face more
as insurance against the possible impact of tariffs on
significant and faster losses during market stress, as
the economy, including potentially weaker demand in
they are compelled to sell assets to meet redemption
tariffed jurisdictions, has yet to fully materialize. In
requests. A clear example occurred in March 2020,
jurisdictions where inflation is still well above target
when US high-yield bonds experienced a mark-
and where tariffs might constitute a supply shock,
to-market loss of 12 percent, considerably larger than
central banks need to proceed carefully with any easing
the 7 percent loss in US investment-grade bonds.41
and maintain their commitment to price stability man-
Increasing ownership of high-yield bonds by funds
dates. This cautious approach should also help temper
and exchange-traded funds can also heighten the
further valuation pressures in risk assets. Central bank
concentration risks of bond issuers. Bond funds and
operational independence remains critical for anchor-
exchange-traded funds that are not dedicated to the
ing inflation expectations and enabling central banks
high-yield asset class or indexed to high-yield bond
to achieve their mandates (see Chapter 1 of the Octo-
benchmarks have increased their holdings (Figure 1.25,
ber 2025 World Economic Outlook for key institutional
panel 4). Funds that are not indexed to benchmarks
features that can help preserve this independence).
can overinvest in certain issuers. For example, a single
investment fund can hold a substantial portion of the 42Greater investment in passive investment strategies, such as

exchange-traded funds, may be related to the rise in cross-asset

correlations during periods of stress, which is one of the main
40Average monthly trading volume on the high-yield bond market attributes of contagion. Benchmark-focused investors are more likely
is according to Tradeweb TRACE data. to be driven by common shocks than by idiosyncratic fundamentals
41Total return percentages in March 2020 are from the ICE Bank of the assets they invest in (see Chapter 1 of the April 2018 Global
of America US High-Yield and US Corporate Bond Indices. Financial Stability Report).

34 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Figure 1.25. Vulnerabilities Posed by the Rising Ownership of the High-Yield Bond Market, by Investment Funds and
Exchange-Traded Funds
Investment and exchange-traded funds own a large and rising share of the ... and their investor base is more flight prone.
high-yield bond market ...
1. Ownership Shares of Open-Ended Investment Funds and 2. Monthly Outflows of Open-Ended Investment Funds and
Exchange-Traded Funds Exchange-Traded Funds
(Percent) (Percentage of assets)
Global high-yield bonds
50 2015 −6
Global intermediate-term
45 2024
corporate bonds
40 −5
Global intermediate-term
35 government bonds
−4
30 Global intermediate term
(mixed bonds)
25 −3
20
15 −2
10 −1
5
0 0
US high-yield US investment-grade US Treasury market 2015 16 17 18 19 20 21 22 23 24 25
bond market bond market

The negative effects of outflows from the high-yield bond market can be The growing share of investment and exchange-traded funds has been
worsened by its low liquidity. driven by other bond investment funds and high-yield exchange-traded
funds.
3. Maximum Flows and Trading Volumes 4. Ownership Shares of US High-Yield Bond Market
(Billions of dollars) (Percent)
Trading volume of underlying asset class (average monthly amount) High-yield bond dedicated investment funds
3,500 Liquidity mismatches: ratio of maximum 0.1 Other bond investment funds 50
historical outflow to trading volume (right scale) High-yield dedicated exchange-traded funds
3,000
40
2,500
2,000 30
0.05
1,500 20
1,000
10
500
0 0 0
US high-yield US investment-grade US Treasuries 2015 16 17 18 19 20 21 22 23 24
bonds bonds

Other bond investment funds can own a large share of the debt of some ... whereas the greater sensitivity of exchange-traded funds to major liquid
issuers, increasing their concentration risk ... markets increases contagion risks.
5. Share of the Debt of Selected High-Yield Bond Issuers 6. Average Correlation to S&P 500 Index
(Percent) (12-month moving average)
30 High-yield dedicated fund ownership 1
Additional other bond funds ownership 0.9
25 Largest single fund ownership of other bond funds 0.8
20 0.7
0.6
15 0.5
0.4
10 0.3
US high-yield exchange-traded fund 0.2
5
US high-yield benchmark 0.1
0 0
CCC BB CCC CCC CCC B CCC CCC B CCC CCC CCC BB CCC 2011 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Ratings of individual high-yield firm borrowers

Sources: Bloomberg Finance L.P.; EPFR Global; EUROPACE AG/Haver Analytics; J.P. Morgan; TRACE; Tradeweb; and IMF staff calculations.
Note: The total debt of the issuing firms in panel 5 amounts to $90 billion, about 4 percent of the ICE Bank of America Global High-Yield Index. Panel 6 uses the iShares
iBoxx $ High-Yield Corporate Bond Exchange-Traded Fund as the proxy.

International Monetary Fund | October 2025 35

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Strengthening global financial safety nets and for- Standing liquidity facilities that backstop core govern-
eign exchange market transparency and resilience can ment bond markets are equally crucial.
mitigate the impact of abrupt asset price corrections— Emerging markets should deploy policies consistent
especially in light of recent dollar depreciation—when with the IMF Integrated Policy Framework to mitigate
market volatility spikes. The capacity and operational external pressures while further deepening local finan-
readiness of global financial safety nets,43 including cial markets, especially bond markets. A softer dollar
bilateral and regional currency swap lines, are crucial to has tempered the external headwinds for emerging
preserve stability in funding and foreign exchange mar- markets in recent months, and rate cuts could induce
kets amid unforeseen ramifications of dollar weakening more global flows into emerging market assets. That
during periods of market stress.44 Over the medium said, emerging markets with weaker fiscal positions—
term, the growing role of NBFIs and other new foreign for example, real financing costs outpacing real
exchange market participants underscores the need growth—are vulnerable to abrupt changes in investor
for better data reporting and transparency, stronger sentiment. To counteract the effects of capital inflow
systemic risk monitoring and stress testing (especially or outflow pressures, the use of foreign exchange inter-
on foreign exchange mismatches), and greater opera- ventions, macroprudential measures, and capital flow
tional resilience among key intermediaries to contain management measures may be appropriate under the
financial stability risks (see Chapter 2). Integrated Policy Framework for economies, especially
Urgent fiscal adjustments to curb government defi- if indicators of fragility such as rising inflation expec-
cits and improvements in market structure are crucial tations and surges in exchange rate and capital flow
to the resilience and functioning of core sovereign bond volatility are observed. For example, provided buffers
markets.45 High debt and delayed fiscal adjustments in are available, countries can deploy reserves through
many countries could further raise borrowing costs for foreign exchange interventions or temporarily relax
governments, underscoring the need for more ambi- macroprudential constraints to mitigate risks to mac-
tious fiscal measures to reduce sovereign risks. In addi- roeconomic and financial stability from capital outflow
tion, sustained trust in the institutional foundations in pressures. Such measures, however, should not impair
G4 economies has underpinned their sovereign bonds’ progress on necessary fiscal and monetary adjustments
safe-asset status for decades and needs to be preserved.46 or on the further development of local bond markets
These foundational elements can be complemented (see Chapter 3). Frontier economies should exercise
by improvements in market structure, particularly a caution against excessive reliance on less-conventional
continued migration toward the central clearing of and potentially more fragile forms of borrowing, such
cash bond and repo transactions to reduce counter- as private placements and bespoke instruments.
party risk, strengthen intermediaries’ capacity through The growing importance of NBFIs in financial
balance sheet netting, and increase transparency. intermediation highlights the need for sound oversight
of this segment. Regulators should improve data col-
lection, coordination, and analysis—particularly across
43Following IMF (2023), the global financial safety net refers borders—to ensure consistent oversight.
to the network of bilateral, regional, and multilateral liquidity To address liquidity mismatches in investment
arrangements that provide countries with access to foreign exchange
liquidity during periods of financial stress. The arrangements include funds, it is key to further improve and expand the
central bank swap lines, regional financing arrangements, and IMF availability and usability of liquidity management
lending instruments. tools.47 Timely and consistent implementation of
44Both unipolar and multipolar international monetary system

configurations can serve as a stable backstop for the global economy.

revised recommendations and guidance from the
However, risks of volatility and potential instability would rise Financial Stability Board (FSB) and International
during the transition between configurations (Chapter 2 of the July Organization of Securities Commissions is crucial to
2025 External Sector Report).
45“Core sovereign bond markets” includes markets covered in the
address structural vulnerabilities in open-ended funds.
section “Expanding Fiscal Deficits Exert Pressure on Bond Market The use of swing pricing and other antidilution mech-
Stability.” anisms can also be effective in mitigating liquidity
46See Gourinchas and Jeanne (2012), who argue that a global

safe asset must be backed not only by sufficient fiscal capacity and
liquidity support, but also by central bank independence and institu- 47Evidence from recent IMF Financial Sector Assessment Pro-

tional safeguards that prevent the monetization of debt and preserve grams indicates that there is room to further improve and expand
monetary credibility to prevent a fiscal-monetary nexus. the availability of liquidity management tools to fund managers.

36 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

mismatches by reducing incentives for investors to NBFIs from a systemwide perspective. In countries
redeem shares ahead of others, especially during with insufficient buffers, policymakers should consider
periods of market stress (see Chapter 2 of the October whether macroprudential buffers can still be built at
2022 Global Financial Stability Report). More definitive the current juncture to increase resilience against a
guidance to lengthen redemption frequency for funds range of shocks while avoiding a broad tightening of
investing in illiquid assets—including high-yield financial conditions. Were a downturn in activity to
bonds—could more fundamentally address liquidity lead to substantial financial stresses, such buffers could
mismatches, although they may require amendments be released to help banks absorb losses and support the
to the legal frameworks in some jurisdictions (IMF provision of credit to the economy, thereby reducing
2021). financial amplification of the downturn.
Broader retail participation in private credit could In light of risks to financial stability from weak
translate into herd behavior to redeem investments banks, continued efforts to strengthen the financial
during stress episodes. In line with FSB recommenda- sector safety net are critical. Central banks should
tions, private credit funds should create and redeem establish frameworks for emergency liquidity assistance
shares at a low frequency or require long notice or and stand ready to provide support to solvent and via-
settlement periods. Regulators should implement ble banks facing temporary liquidity shortfalls, subject
stringent requirements to ensure that private credit to strong safeguards (for example, forward-looking
firms use liquidity management tools and conduct solvency and viability assessments, appropriate interest
stress testing to assess the sufficiency of these tools rates, collateralization, and appropriate haircuts).
during economic downturns or episodes of procyclical Furthermore, all banks should periodically assess their
redemptions. Securities market regulators should also access to central bank lending, including their ability
ensure funds that permit retail participation clearly and to mobilize collateral quickly. Further progress on
comprehensively disclose potential risks and redemp- enhancing recovery and resolution frameworks is essen-
tion limitations to their investors. Increasing retail par- tial to ensure that authorities are well positioned to
ticipation requires close supervision of conduct risks, manage potential shocks without systemic disruption
as more frequent redemptions may exacerbate concerns or exposure of taxpayers to losses.
about valuations. Furthermore, the potential use of Potential increasing adoption of stablecoins could
continuation funds would require stricter oversight. impact safe-asset markets, financial intermediation, and
Global banking stress tests have found that improved monetary sovereignty. Effective regulation, supervision,
capitalization is key to addressing weak banks and and oversight of stablecoin arrangements is crucial to
enhancing banking sector resilience. To preserve mitigate financial stability and integrity risks, including
financial stability amid high economic uncertainty, it is those associated with stablecoin runs. A comprehensive
vital to implement Basel III and other internationally policy, legal, and regulatory response for crypto assets
agreed-upon standards that ensure sufficient capital is necessary to address the risks they pose to macro-
and liquidity in the banking sector. The efficiency economic and financial stability. Policymakers should
of regulations should be ensured by reviewing any implement the FSB’s high-level recommendations for
undue complexity without undermining the over- crypto assets and the broader IMF-FSB policy rec-
all resilience of the banking sector or international ommendations, ensuring that market and prudential
minimum standards. The increased interconnected- authorities possess adequate powers, effective risk
ness between banks and NBFIs means that strains at management frameworks are in place, anti-money
weaker, lightly regulated financial institutions can have laundering and combatting the financing of terrorism
significant consequences for banks and the broader measures in line with international standards are effec-
financial system. Supervisors should carefully monitor tively implemented, and relevant authorities cooperate
banks’ exposures to NBFIs by assessing the solvency with one another. It is also necessary to guard against
and liquidity implications of these exposures under excessive capital flow volatility and adopt unambiguous
adverse scenarios. Supervisors from all financial sectors tax treatment of crypto assets. Sound macroeconomic
and macroprudential authorities need to coordinate policies and credible institutional frameworks can
more closely to establish sound governance structures, ensure monetary sovereignty is preserved, even as the
mechanisms, and processes to monitor banks and stablecoin market continues to develop.

International Monetary Fund | October 2025 37

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 1.1. Global Real Estate Update

Global commercial real estate (CRE) prices across institutional demand, and increased cross-border
all regions have continued their tenuous recovery activity.
since the April 2025 Global Financial Stability Report Developments in the office market are heteroge-
(Figure 1.1.1, panel 1). The recent price recov- neous around the world. Some markets, such as in
ery may not have captured ongoing challenges in London, Paris, Sydney, and Tokyo, are experiencing
some sectors, such as offices. Indeed, in the United strong rental and leasing growth. Others, especially
States, delinquency rates, as indicated by CRE that major cities in the United States, continue to experi-
backs commercial mortgage-backed securities, rose ence elevated vacancy rates, reflecting differences across
to 7.29 percent for August, driven by continued cities in tenant preferences and adaptability to new
stress in the office sector (Figure 1.1.1, panel 1). workplace standards (Figure 1.1.1, panel 3).
Rising CRE prices coincide with positive investment Looking ahead, private CRE markets are still facing
momentum: after bottoming out in 2023, the direct headwinds. Real estate sentiment surveys indicate
CRE investment growth rate recovered to 34 percent that the share of investors expecting improvements
year-over-year in the latest quarter (Figure 1.1.1, in market conditions has declined in recent months,
panel 2), reaching $185 billion. Investment has been reportedly over concerns about market volatility,
buoyed by sectors experiencing strong structural construction cost pressures, and uncertainty around
demand, such as logistics, data centers, and multi- funding spreads (Figure 1.1.2, panel 1). An index of
family housing. Market analyses suggest that growth CRE market liquidity, which tends to lead to changes
remains driven by liquid debt markets, stronger in CRE valuations, deteriorated during the brief
global market turmoil in April (Figure 1.1.2, panel 2),
This box was prepared by Corrado Macchiarelli. suggesting that CRE sentiment is sensitive to broader

Figure 1.1.1. Commercial Real Estate Activity

Recent commercial real estate activity suggests Capital flows are driving growth in ... with some cities witnessing a
an uneven recovery. direct real estate transactions ... rebound in leasing and occupancy
more than others.
1. Global Property Fund Index and 2. Quarterly Year-over-Year 3. Commercial Real Estate
Commercial Mortgage-Backed Security Change in Direct Investment, Vacancy Rates and Rental
Delinquencies by Region Growth Rates in Major Cities
(Index growth, annualized; percent) (Percent) (Percent)
Global Global Vacancy rate Rental growth
Asia Pacific Americas
Continental Europe US Delinquency Rates Europe, Middle East, and Africa San Francisco
North America (Percent 30+ days) Asia Pacific Atlanta
200 Dallas
25 Washington, D.C.
Overall 150 Toronto
20
Sydney
15 Multifamily 100 New York
10 Stockholm
Lodging 50 Beijing
5 Paris
Retail 0 London
0 Amsterdam
−5 Industrial −50 Singapore
Berlin
−10 Office Seoul
−100
−15 Tokyo
2020:Q1
20:Q3
21:Q1
21:Q3
22:Q1
22:Q3
23:Q1
23:Q3
24:Q1
24:Q3
25:Q1

0 5 10 15
2019
20
21
22
23
24
25

−20
−10
0
10
20
30
40

Sources: MSCI; Trepp; JLL Research; and IMF staff calculations.

Note: In panel 1, the change in the MSCI Global Quarterly Property Fund Index is annualized. The last observation is for the second quarter
of 2025. “US delinquency rates” refers to August 2025. In panels 2 and 3, the last observation is for the first quarter of 2025. In panel 3,
transactions larger than $5 million exclude land/development and entity-level deals.

38 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Box 1.1 (continued)

Figure 1.1.2. Commercial Real Estate Headwinds
Forward-looking commercial real estate ... as reflected in liquid real estate Capitalization rates are widening,
sentiment has become somewhat more indices. indicating that investors are
downbeat ... demanding lower prices.
1. Real Estate Sentiment Survey 2. Liquid Real Estate Indexes 3. US Capitalization Rates, by Sector
(left scale) and the US 10-Year (Percent; December 2019 = 100) (Percent; two-year moving average)
Treasury versus Moody’s BAA
Corporate Yield (right scale) MSCI USA IMI liquid real estate Apartment
(Percent) MSCI UK IMI liquid real estate Retail
MSCI Europe ex. UK IMI liquid Office
Percentage expecting market real estate 130 Industrial 9
conditions to improve
Percentage expecting market 8.5
120
conditions to stay the same
8
Percentage expecting market
conditions to worsen 110 7.5
Moody’s BAA corporate bond yield
minus 10-year Treasury yield (reverse) 7
100
100 0 6.5
90
50 1 6
2 80 5.5
0
3 5
−50 70
4 4.5
−100 5 60 4
Nov. 2019
May 2020
Nov. 2020
May 2021
Nov. 2021
May 2022
Nov. 2022
May 2023
Nov. 2023
May 2024
Nov. 2024
May 2025

Dec. 2019
Jun. 2020
Dec. 2020
Jun. 2021
Dec. 2021
Jun. 2022
Dec. 2022
Jun. 2023
Dec. 2023
Jun. 2024
Dec. 2024
Jun. 2025

2009
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Sources: JLL Research; Federal Reserve Bank of St. Louis, Federal Reserve Economic Data; Bloomberg Finance L.P.; MSCI; and IMF staff
calculations.
Note: Panel 1 shows the response to the following question in JLL Research’s Real Estate Sentiment Survey: “Over the next six months, do
you think market conditions will improve, stay the same, or worsen?” Panel 2 illustrates the indexed performance of MSCI IMI Liquid Real
Estate indices for Europe (excluding the United Kingdom), the United Kingdom, and the United States. These indices track the performance
of publicly traded real estate securities, offering a liquid proxy for regional real estate dynamics. Index values are normalized to
December 2019. The last observation is for September 2025. Panel 3 displays a two-year moving average of quarterly capitalization rates for
four major US commercial real estate sectors. Capitalization rates are calculated as the ratio of net operating income to property value,
proxying real estate yields. In panels 1 and 3, the last observation is for the second quarter of 2025. ex. = excluding; IMI = Investable
Market Index.

market sentiment. Finally, CRE capitalization rates in 2023, residential real estate markets are entering a
the United States have increased in office and retail phase of uneven recovery. In some advanced econo-
segments, suggesting that new CRE investors will mies, price growth has resumed modestly, supported
likely demand lower property prices before they invest by falling interest rates (Figure 1.1.3, panels 1 and
(Figure 1.1.2, panel 3). This repricing process makes 2). Where household leverage and debt-servicing
refinancing of CRE debt more challenging at a time capabilities have eased, real home prices have in some
when a substantial volume of US CRE debt is due to cases also demonstrated stronger growth (Figure 1.1.3,
mature in a higher interest rate environment (see the panel 3). This indicates that less-constrained borrow-
April 2025 Global Financial Stability Report). ers are also more likely to support housing demand
Similar to CRE, after rising strongly immediately through increased credit uptake, thereby reinforcing
after the COVID-19 pandemic and then being price momentum, although the relationship varies
weighed down by higher interest rates in 2022 and considerably across countries.

International Monetary Fund | October 2025 39

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 1.1 (continued)

Figure 1.1.3. Residential Real Estate Activity
Residential real estate markets globally Some emerging market economies are Lower debt-service burdens are
are entering a phase of uneven facing extended declines. associated with stronger real house
recovery. price growth.
1. Year-over-Year Changes in 2. Distribution of Residential 3. Yearly Difference in Debt-Service
Residential Property Prices Property Prices over the Latest and Ratios versus House Price Growth
(Percent) Previous Quarter (Percentage points, x-axis;
Interquartile range (Percentage of countries) percent, y-axis)
Median Advanced economies (latest) 0.6
Norway

Debt-service ratio difference (year-on-year)

Emerging market economies Emerging market economies (latest)
15 Australia 0.4
Advanced economies Advanced economies (previous quarter)
Emerging market economies 0.2
10 60 Japan
(previous quarter) Belgium Netherlands
50 0
5 Germany United Kingdom
France Italy −0.2
40
0 United States Denmark
30 Canada Sweden −0.4
−5 Finland Spain
20 −0.6
Korea Portugal
−10 10 Households spending a smaller −0.8
share of income on debt repayments
−15 0 −1
Less or −5 to 0 to 5 to Greater −10 −5 0 5 10 15
2010
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

equal 0% 5% 10% or equal Real house price change

than −5% than 10% (year-on-year)

Sources: Bank of International Settlements; and IMF staff calculations.

Note: In panels 1 and 2, advanced and emerging market economies are calculated by the Bank for International Settlements, using rolling
GDP and purchasing power parity exchange rate weights. In panels 1, 2, and 3, the last observation for real house prices is for the first
quarter of 2025. In panel 3, the last observation for debt-service ratios is for the fourth quarter of 2024.

40 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Box 1.2. Low Interest Rates in China Could Imperil Bank Profits and Lending
Since 2022, China’s economy has endured relatively partially offset the erosion in asset yields by main-
weak growth compared with historical rates and sus- taining loan rates at a relatively high level to preserve
tained low inflation (IMF 2024, 2025b). In response, profitability (Figure 1.2.1, panel 2).
the People’s Bank of China has lowered key interest If this erosion in asset yields were to persist, banks’
rates to stimulate growth, bringing the benchmark equity base might weaken, hindering the sector’s
policy rate down to 1.4 percent from 2.2 percent ability to withstand negative shocks. Weighed by net
three years ago. Meanwhile, bond yields have fallen interest margins, both return on equity and return on
to near historical lows. This decline in interest rates assets for the banking sector have declined, falling to
has weighed on banks’ margins and profitability. Such 8.2 percent and 0.63 percent in the second quarter
erosion, combined with banks’ ongoing challenges of 2025, near their lowest in a decade, compared
to generate capital organically (for example, through with 8.9 percent and 0.69 percent a year earlier
retained earnings) could imperil bank balance sheets (Figure 1.2.1, panel 3). This decline in nominal rates
and stifle credit supply, raising financial stability con- could constrain lending—a scenario known as “rever-
cerns as well as imperiling China’s economic growth. sal interest rates,” whereby persistently low rates cut
Banks’ profitability pressures have intensified amid into banks’ profits and capital base, curbing lending,
continued margin compression. Average net inter- despite accommodative monetary policy (Abadi,
est margins across the banking system declined to a Brunnermeier, and Koby 2022; Wang 2025). Cur-
historic low of 1.42 percent in the second quarter of rently, capital buffers at the largest banks are adequate
2025, as the benchmark seven-day reverse repo rate (IMF 2025b). Still, despite such buffers and low rates,
also reached a historical low (Figure 1.2.1, panel 1). loan growth at these banks has slowed below the
The deposit spread—proxied by the gap between five-year average on subdued demand (Figure 1.2.1,
the one-year China government bond yield and the panel 4). Earlier this year, the authorities injected
one-year time deposit rate—compressed sharply in 500 billion yuan (or about $69 billion) of capital
late 2024 as bond yields fell faster than deposit rates, into large state-owned banks to help expand lending
underscoring pressure on asset yields. The loan spread capacity, reflecting concerns that declining profit-
(China’s one-year loan prime rate minus one-year gov- ability could constrain credit supply. Such concerns
ernment bond yield) has remained elevated (around underscore the difficult trade-offs policymakers face as
150 basis points), suggesting that banks have sought to they ease rates to low levels: whereas low rates support
growth in the short term, sustained low rates could
This box was prepared by Sally Chen, Lawrence Tang, and weaken bank profitability and reduce lending capacity
Jing Zhao. over time.

International Monetary Fund | October 2025 41

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 1.2 (continued)

Figure 1.2.1. Low Interest Rates Have Weighed on Bank Profitability in China
The net interest margin is at a historical low ... ... while deposit income has fallen and loan spreads are
stable.
1. Banks’ Net Interest Margin and the Seven-Day Repo Rate, 2. Deposit Income versus Loan Spreads
First Quarter of 2021 to the Second Quarter of 2025 (Basis points)
(Percent)
250 60
2
200 40
Net interest margin

1.8
150 20

1.6
100 0
2025:Q2 Loan spread (1y LPR—1y CGB)
1.4 Deposit spread
50 (1y CGB—1y time deposit, −20
right scale)
1.2 0 −40
1.3 1.5 1.7 1.9 2.1 Oct. Dec. Feb. Apr. Jun. Aug.
7D reverse repo rate 2024 2024 2025 2025 2025 2025

Bank profitability is at a 10-year low ... ... and loan growth has slowed.
3. Return on Assets and Equity 4. Loan Growth at Large Banks
(Percent) (Percent; year-over-year)
25 1.6 2025:Q2 2024:Q2 Five-year average 14
1.4 12
20
1.2 10
15 1
8
0.8
10 6
0.6
0.4 4
Return on equity
5
Return on asset (right scale) 0.2 2
0 0 0
2014 15 16 17 18 19 20 21 22 23 24 25 PSBC ABC BOC BoCom ICBC CCB

Sources: Bloomberg Finance L.P.; CEIC; J.P. Morgan; and WIND.

Note: In panels 1 and 3, net interest margin, return on equity, and return on assets reflect data for the entire banking sector in China. In
panel 2, the one-year time deposit rate is estimated by J.P. Morgan, currently at 0.95 percent, while the PBOC’s official benchmark one-year
deposit rate stands at 1.5 percent, which is generally viewed as the ceiling, and the one-year lending prime rate, which is a weighted
average of rates submitted by 20 quoting banks, stands at 3 percent currently. In panel 4, the loan growth analysis focuses on the six largest
state-owned banks: Postal Savings Bank of China (PSBC), Agricultural Bank of China (ABC), Bank of China (BOC), Bank of Communications
(BoCom), Industrial and Commercial Bank of China (ICBC), and China Construction Bank (CCB). CGB = China government bond; LPR = loan
prime rate; Q = quarter.

42 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

Box 1.3. Banks and Insurers Are Deepening Ties with the Private Credit Ecosystem
The exponential growth of private credit has raised rowers: banks earn fees for originating and servicing
concerns that credit provision is migrating from corporate loans and asset-based finance, which are
strictly regulated banks and relatively transparent pub- consequently booked in private credit funds (for
lic markets to the comparatively lightly regulated and example, forward-flow origination). Often, such part-
opaque private credit industry. The emerging financial nerships are complemented by an agreement that banks
system, however, is marked by intertwined operations provide leverage to engaged private credit funds and
whereby traditional institutions like banks and insur- additional banking services to private credit borrowers,
ers, as well as alternative nonbanks like private credit including revolving lines of credit. Although such part-
funds, are not substitutive entities but instead part of nerships in principle are beneficial for banks and private
an increasingly integrated system. Recent partnerships credit managers, they have not yet been tested over time.
among the private credit industry, banks, and insurers Some market participants raise concerns that the partner-
highlight that cooperation can generate significant ships may lead to looser underwriting standards and
economic benefits for the parties involved. To realize weaker loan monitoring.
these benefits for the broader economy, adjustments
to supervisory and regulatory approaches are needed Insurance Companies
to address the buildup of risks across sectors and Private credit has long been an important com-
borders. ponent of insurers’ portfolios, especially in North
America, where it represents about one-third of total
Banks investments (Figure 1.1.3, panel 1). Private credit
In the past decade, the private credit industry has instruments offer insurers additional spread for illi-
built a sizable channel for raising long-term capital quidity and supply long-duration assets to match their
from institutional investors. The “patient” nature of long-term liabilities. However, increasing exposure to
capital in most private credit balance sheets gave it private credit requires advanced asset-liability manage-
a competitive advantage in originating and retain- ment to account for higher asset illiquidity, policy sur-
ing credit in the riskiest areas, like leveraged finance render risk, and single-name concentrations. Whereas
to middle-market borrowers or subordinated debt some private credit investments represent simple credit
to commercial real estate transactions—areas often originated by nonbank lenders, a significant and
avoided by strictly regulated banks. To tap into other growing portion of insurers’ private credit portfolios
types of clients and credit products, several private is in structured instruments providing leverage to the
credit managers have entered more than 20 part- high-yielding part of the private credit ecosystem: for
nerships with banks in various countries in the last example, securitized products (such as middle-market
three years. Larger private credit managers have been collateralized loan obligations and commercial real
partnering with global banks with a wide network estate collateralized loan obligations), fund financing
of clients (in particular, global systemically import- through feeder notes, collateralized fund obligations,
ant banks) or smaller banks with deep expertise in and private placements of private credit funds’ debt.
a particular lending niche (for example, asset-based A growing share of insurers’ private credit exposure
finance). Such partnerships often aim to distribute is sourced through either affiliated private credit
private credit products to banks’ wealth management managers or partnerships with private credit managers,
clients or create channels for banks to offload capi- which requires special attention because of potential
tal-intensive assets to private credit funds, in line with conflicts of interest and the lack of transparency
sales of banks’ loan portfolios or the growing trend of (Cortes, Diaby, and Windsor 2023).
synthetic risk transfers (see the October 2024 Global Most insurers’ exposure to private credit is classified
Financial Stability Report). Smaller private credit as investment grade, and many private credit instru-
managers look for anchor bank partners to back their ments would be much less appealing if classified as
growth by providing leverage to their private credit below investment grade. The investment-grade status
funds and strengthening their pipeline of lending allows favorable risk-capital treatment and considers
deals. the instruments’ cash flows sufficiently reliable to
Many partnerships assume the “originate-to-distribute” qualify for asset-liability matching. Insurers’ search
model that relies on banks’ network of potential bor- for private credit exposures classified as investment

International Monetary Fund | October 2025 43

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 1.3 (continued)

Figure 1.3.1. Insurers’ Exposure to Private Credit
Insurers have diverse exposure to private credit instruments, Private ratings have been growing in number and importance
especially in North America and the United Kingdom. in the United States.
1. Moody’s Survey of Insurers’ Average Positions in 2. Privately Rated Securities Held by US Insurers
Private Credit Instruments (Number)
(Percent of total investments)
40 Commercial real estate Big Three rating agencies 9,000
lending Specialized rating agencies 8,000
35
Private placements
30 Residential mortgages 7,000
Infrastructure 6,000
25
Money market lending 5,000
20 Asset-based finance,
fund finance, and other 4,000
15
3,000
10 2,000
5 1,000
0 0
United States United Europe Asia Pacific 2019 20 21 22 23
and Canada Kingdom

Sources: Moody’s; NAIC; and IMF staff calculations.

Note: Panel 1 refers to Moody’s 2024. In panel 2, the “Big Three rating agencies” are Moody’s Investors Service, Standard & Poor’s, and Fitch
Ratings.

grade has changed the rating landscape in the United insurers’ capital and potentially causing liquidity gaps
States, with an increasing share of the assessment being because of insufficient cash flow from the defaulted
conducted by smaller rating agencies specializing in entities. Because reliable private ratings are key for
the private credit ecosystem (Figure 1.3.1, panel 2). insurers’ prudential regulation, it is imperative to
Misclassification of below-investment-grade instru- keep the risk of inflated ratings minimal by ensuring
ments into the investment-grade bucket may result in the soundness of private rating assessments and
default losses significantly exceeding those expected requiting adequate transparency of methodologies
during an economic shock, leading to the erosion of and reports.

44 International Monetary Fund | October 2025

 CHAPTER 1 Shifting Ground Beneath the Calm: Stability Challenges amid Changes in Financial Markets

References Du, Wenxin, Joanne Im, and Jesse Schreger. 2020. “The US
Abadi, Joseph, Markus Brunnermeier, and Yann Koby. 2023. Treasury Term Premium.” Journal of International Economics 112.
“The Reversal Interest Rate.” American Economic Review 113 European Central Bank (ECB). 2024. “Liquidity Risks in the
(8): 2084–120. Non-bank Financial Sector: Systemic Implications and Policy
Acharya, Viral V., and Toomas Laarits. 2023. “When Do Trea- Options.” In Financial Stability Review, May 2024, Box 3.
suries Earn the Convenience Yield? A Hedging Perspective.” Frankfurt am Main, Germany: European Central Bank.
NBER Working Paper 31863, National Bureau of Economic Egypt Ministry of Finance. 2025. “Egypt Resumes Plan of Inter-
Research, Cambridge, MA. national Issuance of Sovereign Sukuk.” News release, June 25.
Adrian, Tobias, Richard K. Crump, and Emanuel Moench. Financial Stability Board (FSB). 2024. Global Monitoring
2013. “Pricing the Term Structure with Linear Regressions.” Report on Non-Bank Financial Intermediation 2024. Basel,
Journal of Financial Economics 110 (1): 110–38. Switzerland: FSB.
Adrian, Tobias, Sheheryar Malik, and Jason Wu. 2024. “Global Gourinchas, Pierre-Olivier, and Olivier Jeanne. 2012. “Global
Fragilities Mount Despite Rate Cuts and Buoyant Market.” Safe Assets.” BIS Working Paper 399, Bank for International
IMF Blog, October 22. Settlements, Basel, Switzerland.
Ahmed, Rashad, and Iñaki Aldasoro. 2025. “Stablecoin and Safe International Monetary Fund (IMF). 2021. “Investment Funds and
Asset Prices.” BIS Working Paper 1270, Bank for Interna- Financial Stability: Policy Considerations.” Departmental Paper
tional Settlements, Basel, Switzerland. No 2021/018, International Monetary Fund, Washington, DC.
Arslanalp, Serkan, Barry Eichengreen, and Chima Simpson-Bell. International Monetary Fund (IMF). 2023. “The Global Financial
2022. “The Stealth Erosion of Dollar Dominance: Active Safety Net: Preparing for the Next Shock.” Finance & Develop-
Diversifiers and the Rise of Nontraditional Reserve Curren- ment, December: 18–19. https://www.imf.org/en/Publications/
cies.” IMF Working Paper 22/58, International Monetary fandd/issues/2023/12/PT-global-financial-safety-net
Fund, Washington, DC. International Monetary Fund (IMF). 2024. “People’s Republic
Baker, R. Scott, Nicholas Bloom, and Steven J. Davis. 2016. of China: 2024 Article IV Consultation.” IMF Country
“Measuring Economic Policy Uncertainty.” Quarterly Journal Report 2024/258, Washington, DC.
of Economics 131 (4): 1593–636. International Monetary Fund (IMF). 2025a. “Canada: Financial
Bank of England. 2025. “Financial Stability Report—July 2025.” Sector Assessment Program—Technical Note on Systemic
Bank of England, London. https://www.bankofengland.co.uk/ Liquidity Assessment.” IMF Country Report 25/233,
financial-stability-report/2025/july-2025 Washington DC.
Bernanke, Ben, Mark Gertler, and Simon Gilchrist. 1999. International Monetary Fund (IMF). 2025b. “People’s Republic
“The Financial Accelerator in a Quantitative Business Cycle of China: Financial Sector Assessment Program-Financial
Framework.” In Handbook of Macroeconomics, vol. 1, edited System Stability Assessment.” IMF Country Report
by J. B. Taylor and M. Woodford, 1341–93. Amsterdam, The 2025/100, Washington, DC.
Netherlands: Elsevier. International Monetary Fund (IMF). Forthcoming. “A Primer
Brunnermeier, Markus K., Thomas Eisenbach, and Yuliy on Stablecoins.” Washington, DC.
Sannikov. 2013. “Macroeconomics with Financial Frictions: Khoda, Neha, Dong Ba, Yvonne Tong, and Adam Vogel. 2025.
A Survey.” In Advances in Economics and Econometrics: Tenth “The Era of Pre-emptive LMEs.” Bank of America Global
World Congress, edited by Daron Acemoglu, Manuel Arellano, Research High Yield and Loan Strategy, July 24.
and Eddie Dekel. New York: Cambridge University Press. Kiyotaki, Nobuhiro, and Moore, John. 1997. “Credit Cycles.”
Caldara, Dario, and Matteo Iacoviello. 2022. “Measuring Geopo- Journal of Political Economy 105 (2): 211–48.
litical Risk.” American Economic Review 112 (4): 1194–225. Krishnamurthy, Arvind, and Annette Vissing-Jorgensen. 2012.
Caldara, Dario, Matteo Iacoviello, Patrick Molligo, Andrea “The Aggregate Demand for Treasury Debt.” Journal of
Prestipino, and Andrea Raffo. 2020. “The Economic Effects Political Economy 120 (2): 233–67.
of Trade Policy Uncertainty.” Journal of Monetary Economics Levin, John D., and Antoine Malfroy-Camine. 2025. “Bank
109 (C): 38–59. Lending to Private Equity and Private Credit Funds:
Cortes, Fabio, Mohamed Diaby, and Peter Windsor. 2023. “Private Insights from Regulatory Data.” Federal Reserve Bank of
Equity and Life Insurers.” Global Financial Stability Note Boston, Supervisory Research and Analysis Note 2025–02,
2023/001, International Monetary Fund, Washington, DC. February 5.
Dao, Mai, and Pierre-Olivier Gourinchas. 2025. “Covered McGuire, Patrick, Ilhyock Shim, Hyun Song Shin, and Vladyslav
Interest Parity in Emerging Markets.” IMF Working Paper Sushko. 2021. “Outward Portfolio Investment and Dollar
2025/057, International Monetary Fund, Washington, DC. Funding in Emerging Asia.” BIS Quarterly Review, December.
Du, Wenxin, and Amy W. Huber. 2024. “Dollar Asset Holdings Moody’s. 2024. “Insurers’ Private Credit Holdings Will Grow,
and Hedging around the Globe. NBER Working Paper 32453, with Benefits Outweighing Risks.” Moody’s Investor Service,
National Bureau of Economic Research, Cambridge, MA. June 2024.

International Monetary Fund | October 2025 45

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Morningstar DBRS. 2024. “Liquidity Support Remains a Key Working Paper 2025/41, International Monetary Fund,
Factor in Mitigating Defaults.” Commentary, November 11. Washington, DC.
Morningstar DBRS. 2025. “Leveraged Finance Spotlight: Private S&P Global Ratings. 2025. “Private Credit and Middle-Market
Credit Still Under Pressure Entering H2 2025.” Commentary, CLO Quarterly: Predictably Unpredictable Q3 2025.” S&P
July 29. Global Ratings, New York, July 25.
Mota, Lira. 2023. “The Corporate Supply of (Quasi) Safe Shin, Hyun Song, Philip Wooldridge, and Dora Xia. 2025.
Assets.” MIT Working Paper, Massachusetts Institute of Tech- “US Dollar’s Slide in April 2025: The Role of FX Hedging.”
nology, Cambridge, MA, September. BIS Bulletin 105, Bank for International Settlements, Basel,
Nikolaou, Kleopatra. 2025. “Money Market Fund Growth Switzerland.
during Hiking Cycles: A Global Analysis.” IMF Work- Société Générale. 2012. “Identifying the Sovereign Risk Compo-
ing Paper 2025/150, International Monetary Fund, nent in Credit Spreads.” February 20.
Washington, DC. Vitek, Francis. “The Global Macrofinancial Model.” IMF
Panama Ministry of Economy and Finance. 2025. “Panama Working Papers 2018/081. International Monetary Fund,
Asegura Financiamiento Estratégico Con Ahorros en Tasas de Washington, DC. doi.org/10.5089/9781484350706.001.
Interés.” Press release, March 5. Wang, Olivier. 2025. “Banks, Low Interest Rates, and
Pandolfo, Jordan. 2025. “Banks and Private Credit: Competitors Monetary Policy Transmission.” Journal of Finance 80 (3):
or Partners?” Federal Reserve Board of Kansas City. 1379–416.
Parsons, Joe, and Chris Davis. 2025. “Taiwan Turmoil: What Wilcock, Joshua, Alex Jekov, Oliver Brennan, Parisha Saimbi,
Drove the TWD Surge?” Risk.net, May 8. and Sam Lynton-Brown. 2025. “What Drives FX Hedge
Reuter, Marco. 2025. “Decrypting Crypto: How to Ratios?” BNP Paribas Markets 360 Strategy and Economics,
Estimate International Stablecoin Flows.” IMF July 3.

46 International Monetary Fund | October 2025

 2
CHAPTER
RISK AND RESILIENCE IN THE GLOBAL
FOREIGN EXCHANGE MARKET

Chapter 2 at a Glance
• The global foreign exchange (FX) market plays a key role in the international monetary and financial
system, and its smooth functioning is essential for maintaining global financial stability.
• Structural shifts, such as the increasing involvement of nonbank financial institutions (NBFIs) and growing
trade in derivatives, offer benefits but may also raise the global FX market’s vulnerability to adverse shocks.
• This chapter finds that increased macrofinancial uncertainty can strain FX market conditions by signifi-
cantly raising funding costs, impairing liquidity, and amplifying excess exchange rate return volatility.
• The effect of shocks is more pronounced for emerging markets and for currencies with high NBFI partici-
pation, concentrated dealer networks, and elevated hedging activity.
• FX market stress can spill over to other asset classes, tightening financial conditions and posing risks
to macrofinancial stability—especially in countries with significant currency mismatches and fiscal
vulnerabilities.
• Outages in critical payment systems and risk of settlement failure significantly impair market liquidity and
increase excess exchange rate return and its volatility, raising the cost of FX transactions.
• Amid elevated uncertainty and a shifting global economic landscape, investor strategies are evolving.
Following the US tariff announcements in early April 2025, investors in some countries reduced their US
dollar holdings, whereas others maintained exposures, highlighting diverging cross-country responses.

Key Policy Recommendations

• Enhance FX market surveillance through systemic risk monitoring, stress testing, and scenario analysis to
capture liquidity shocks and spillovers.
• Close critical FX data gaps by improving reporting and transparency, especially in regard to NBFIs and
bilateral exposures outside centralized infrastructures.
• Ensure robust liquidity and capital buffers, backed by effective safeguards, such as access to central bank
liquidity with proper oversight, sufficient international reserves, and expanded central bank swap lines.
• Strengthen operational resilience of financial market infrastructures and financial institutions through
cyber risk frameworks, contingency planning, and coordinated oversight.
• Reduce settlement risks and market inefficiencies in over-the-counter FX markets by encouraging
payment-versus-payment adoption and exploring digital innovations to develop interoperable financial
platforms.

Introduction the world (BIS 2025a; Figure 2.1, panel 1). It facili-
tates cross-border trade and financial transactions by
The global foreign exchange (FX) market is a cor-
enabling currency conversion and influencing exchange
nerstone of the international monetary and financial
rates.1 Cross-border transactions account for about
system. With an average daily turnover exceeding
two-thirds of global FX market turnover, with the
$9.6 trillion, the FX market has grown over the years
into the largest and most liquid financial market in 1The global FX market is broadly defined as a decentralized global

marketplace for trading currencies. It encompasses spot transactions

The authors of this chapter are Yuhua Cai, Andrea Deghi (co–team (that is, immediate exchanges of currencies at current rates) and a
lead), Seungduck Lee, Taneli Mäkinen, Yuan Tian, Tomohiro Tsuruga range of derivative instruments, including outright forwards, FX
(co–team lead), and Mustafa Yenice, under the supervision of Mario swaps, currency swaps, and currency options, commonly used for
Catalán and Mahvash Qureshi. Angelo Ranaldo served as an expert hedging, funding, and speculative purposes. See Online Annex 2.1
advisor. for definitions of technical terms.

International Monetary Fund | October 2025 47

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 2.1. Key Developments in the Global Foreign Exchange Market

The FX market has grown substantially, driven The US dollar continues to be the dominant The shares of participants have shifted, with
mainly by an increase in swap trades. currency in global FX trading. nonbank financial institutions taking a greater
role.
1. FX Market Turnover, by Instrument 2. FX Market Turnover, by Currency 3. FX Market Turnover, by Counterparty
(Daily average trillions of dollars, net-net basis) (Daily average trillions of dollars, net-net basis) (Daily average percent of total market
turnover, net-net basis)
10 Currency swaps USD EUR JPY 20 70
1998 2025
Options and other products GBP CNY Others
Outright forwards 60
8 16
FX swaps
50
Spot
6 12 40

4 8 30
20
2 4
10
0 0 0
1998 2001 04 07 10 13 16 19 22 25 2001 04 07 10 13 16 19 22 25 Reporting Other financial Nonfinancial
dealers institutions customers

Sources: BIS 2025a; and IMF staff calculations.

Note: Daily averages correspond to April. The BIS Triennial Central Bank Survey (2025a) adjusts figures reported on a net-net basis to correct for double counting in local
and cross-border interdealer transactions. In panel 2, because each foreign exchange transaction involves two currencies, the sum of the individual currencies is twice that
of the total turnover involved. CNY = Chinese yuan; EUR = euro; FX = foreign exchange; GBP = British pound; JPY = Japanese yen; USD = US dollar.

US dollar being the dominant trading currency (BIS diversity in FX markets, potentially increasing liquid-
2022; Figure 2.1, panel 2). ity, reducing transaction costs, and strengthening price
The structure of the global FX market has evolved discovery and risk sharing. However, many NBFIs are
across multiple dimensions, including the diversifica- subject to less regulatory oversight than traditional
tion of market participants, the expansion of traded banks and may lack access to central bank facilities.
instruments, and broader changes in the FX trading NBFI trading strategies, often driven by leverage,
ecosystem (Schrimpf and Sushko 2019; Chaboud, short-term arbitrage, and high-frequency trading, can
Rime, and Sushko 2023; Chaboud and others 2024). amplify market swings and shift inventory risk onto
For example, whereas trading in the 1990s took place market-making dealers.3 Moreover, many NBFIs (such
mostly between large dealers, such as commercial as mutual funds) exhibit liquidity mismatches, funding
and investment banks, nonbank financial institutions longer term or less liquid assets with short-term liabili-
(NBFIs) have since become increasingly important ties. This structural fragility can heighten systemic risk
players (Figure 2.1, panel 3).2 The share of spot during market volatility, as tighter funding conditions
trading has declined, whereas the use of derivatives, and rapid unwinding of positions may amplify liquid-
especially FX swaps, mostly for funding and hedging ity pressures in FX markets (see the October 2022
currency risk, has grown notably (Figure 2.1, panel 1). and April 2023 issues of the Global Financial Stability
The number of execution methods and trading plat- Report; FSB 2025).
forms has also expanded with the increasing electroni- The growing use of FX derivatives has enhanced
fication of the market. liquidity and risk management by enabling institu-
These shifts in the FX market have enhanced tions to hedge currency exposures and access foreign
competition and efficiency but have also introduced currency funding. However, it has also facilitated lev-
challenges for macrofinancial stability. For example, eraged investments and increased interconnectedness.
increasing NBFI participation contributes to more During stress, margin calls and forced deleveraging
can amplify volatility and liquidity strains (Borio,
2The decline in FX activity by nonfinancial firms suggests that

trade plays a more limited role than financial flows in driving FX 3Under stress, algorithmic and high-frequency traders often with-

transactions. draw, reducing liquidity and amplifying volatility (BIS 2020).

48 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

McCauley, and McGuire 2022; Nenova, Schrimpf, and shocks can interact with underlying vulnerabilities,
Shin 2025). Moreover, the opacity of over-the-counter such as dealer concentration and currency mismatches,
derivatives markets, which dominate FX trading, magnifying stress and propagating instability across
complicates risk monitoring for regulators and central financial markets. Historical episodes of elevated global
banks, potentially obscuring the buildup of systemic macrofinancial uncertainty show that FX funding
risks.4 and market liquidity pressures, reflected in wider
Compounding these challenges are two structural cross-currency bases (a measure of deviation from the
vulnerabilities: high dealer concentration and currency covered interest parity, or CIP), bid-ask spreads, and
mismatches. Nearly half of global FX turnover is excess exchange rate return volatility, tend to rise with
intermediated by a small group of dominant dealers— uncertainty (Figure 2.2).8
mostly large, regulated banks—leaving the market These FX market dynamics are especially relevant in
exposed should these institutions scale back activity today’s conditions of elevated policy uncertainty and
during stress (BIS 2022).5 Meanwhile, persistent cur- a shifting global macrofinancial landscape. Structural
rency mismatches, in which liabilities and assets are in changes in global trade and financial flows, driven by
different currencies, drive sustained demand for short- evolving trade policies, supply chain realignments, and
term FX derivatives, increasing rollover and funding geopolitical considerations, may be reshaping currency
risks when conditions tighten (FSB 2022). Disruptions demand and FX market behavior (External Sector
in these markets can sharply raise hedging costs, Report 2025; Box 2.1). At the same time, heightened
prompting the unwinding of positions that reinforce uncertainty around these developments raises the risk
volatility and further elevate costs.6 of abrupt shifts in investor sentiment and expectations,
In addition to structural fragilities, the global FX affecting capital flows and FX market conditions, par-
market is exposed to a range of external and opera- ticularly in economies with less liquid FX markets.
tional risks. The market’s central role in global finance Moreover, the expansion of FX trading has height-
makes it highly sensitive to macroeconomic develop- ened exposure to settlement risk: the possibility that
ments and policy shifts that influence cross-border one party will deliver its currency without receiving
trade and financial flows and affect currency valuation. what the counterparty owes. This risk is particularly
For example, an increase in macroeconomic uncer- acute in cross-border transactions, in which time zone
tainty can change investor risk sentiment and interest differences and operational delays can lead to failed
rate expectations, triggering rapid portfolio adjust- settlements and trigger liquidity shortfalls and systemic
ments, liquidity strains, and volatility (see, for exam- stress. Although certain risk mitigation arrangements,
ple, Berger, Chaboud, and Hjalmarsson 2009).7 These such as payment-versus-payment (PvP), have been
adopted over the years, most emerging market curren-
cies remain outside these frameworks.9 With the grow-
4Other changes in the FX ecosystem, such as trading platform
ing prominence of emerging market currencies in the
proliferation and increased electronification, have improved market
access, speed, and transparency but have also added complexity, frag- global FX market (Figure 2.1, panel 2), their exclusion
mentation, and operational risk. For example, algorithms help align from such mechanisms leaves a substantial portion of
prices in stable market conditions, but they can break down during
volatility, generating price discrepancies and a flight to more liquid
platforms (“liquidity mirage”; BIS 2020). Electronification can also 8The cross-currency basis measures the CIP deviation, reflecting
deepen informational asymmetry, giving technologically advanced the cost of swapping one currency for another; a widening basis
traders an edge and distorting price discovery (Ranaldo and Somogyi signals stress in FX funding markets. The bid-ask spread captures
2021). the difference between the prices at which a dealer is willing to buy
5During stress episodes, regulatory constraints can limit dealer (bid) and sell (ask) a currency; wider spreads suggest reduced market
capacity, further reducing liquidity, raising transaction costs, and liquidity and higher transaction costs. Excess exchange rate return
impairing trade execution (Aldasoro, Huang, and Tarashev 2021; volatility reflects the movement in returns unexplained by macroeco-
Huang and others 2025). nomic fundamentals; it captures the influence of factors like shifts in
6See, for example, Borio and others (2016); Barajas, Deghi, investor sentiment, liquidity conditions, or risk aversion.
Raddatz, and others (2020); Du and Schreger (2022); and Kloks, 9The CLS foreign exchange settlement system reduces settlement

Mattille, and Ranaldo (2023). risk for 18 currencies through its PvP mechanism, which settles
7Geopolitical disruptions, such as armed conflicts or sanctions, both sides of an FX transaction simultaneously. Some regional PvP
can also have an impact on FX markets by affecting cross-border systems offer similar functionality for select currencies traded against
trade and financial flows (April 2023 and April 2025 issues of the the US dollar but lack global integration, limiting efficiencies. Many
Global Financial Stability Report; Hui 2021; Hossain, Masum, and currencies remain outside PvP systems because of technical, regula-
Saadi 2024). tory, and economic constraints (Glowka and Nilsson 2022).

International Monetary Fund | October 2025 49

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 2.2. Macrofinancial Uncertainty and Foreign Exchange Market Conditions

Foreign exchange funding and market liquidity conditions deteriorated significantly during the global financial crisis and the COVID-19 market turmoil.
1. CIP Deviation Relative to US Dollar 2. Bid-Ask Spread Relative to US Dollar 3. Excess Exchange Rate Return Volatility
(Basis points) (Percent) Relative to US Dollar
(Percentage points)
50 Interquartile range 0.3 Interquartile range 3
Median Median
0 Global
Global financial
−50 0.2 crisis 2
financial Tariff Tariff
Global COVID-19 Tariff crisis announcement announcement
−100 announcement
financial COVID-19 COVID-19
crisis
−150 0.1 1
Interquartile range
−200 Median
−250 0 0
2005 08 11 15 18 21 25 2005 08 11 15 18 21 25 2005 08 11 15 18 21 25

Sources: Bloomberg Finance L.P.; LSEG Datastream; and IMF staff calculations.
Note: In panel 1, CIP deviation is calculated using three-month overnight index swap rates for 12 currencies against the US dollar. A negative widening basis signals stress
in dollar funding markets. Panel 2 shows the bid-ask spread calculated as [(ask rate − bid rate) ∕mid rate] × 100. Wider spreads suggest reduced market liquidity. In
panel 3, excess exchange rate return = log(exchange rate at time t ∕exchange rate at time t − 1) − log(forward rate at time t − 1 ∕exchange rate at time t − 1). In all panels,
“tariff announcement” refers to the April 2, 2025, US declaration of new import tariff rates. See Online Annex Figure 2.3.1 for details on the construction of the measures
used in this figure and illustrations by country group (that is, advanced economies and emerging market economies). CIP = coverage interest parity.

FX transactions exposed to settlement risk (Box 2.2). Against this backdrop, this chapter explores recent
This exposure is exacerbated by the risk of operational developments, vulnerabilities, and risks in the global
disruptions to FX market infrastructure: technical FX market and discusses policy options for mitigat-
failures, cyberattacks, or power outages could impair ing the risks. This chapter begins with a conceptual
FX market functioning, generating liquidity strains framework for how shocks stemming from macro­
and volatility and making delayed or failed settlements financial uncertainty or operational disruptions can
more likely.10 affect FX market conditions and financial stability.
Stress in FX markets can spill over to markets for It then presents stylized facts on the evolution of FX
other asset classes, posing risks to macrofinancial stabil- markets, covering key currencies, participants, and
ity. Elevated FX volatility and hedging costs, reflected interconnectedness. Next, it empirically analyzes three
in wider cross-currency bases, raise uncertainty and the key questions: (1) How do different macrofinancial
cost of managing currency exposure, potentially affect- uncertainty shocks affect FX trading across market
ing yields and risk premiums. Higher funding costs participants? (2) How do these shocks influence FX
can also erode the intermediation capacity of financial market conditions, as measured by cross-currency
institutions, tighten financial conditions, and amplify basis, excess exchange rate return volatility, and bid-ask
systemic stress, triggering an adverse macrofinancial spreads, and are the effects amplified by structural
feedback loop (Adrian and Shin 2014).11 market fragilities? (3) Does FX market stress spill over
into other financial markets, such as those for sover-
eign bonds and equities, with broader implications for
10Recent examples of operational disruptions to critical payment

infrastructure include the 2018 Fedwire cyber incident and the 2000
financial stability?
and 2025 TARGET2 outages. Even as contingency measures within To address these three key questions, this chapter
FX settlement infrastructure helped prevent systemic fallout, these draws on a unique data set covering FX spot and swap
events revealed operational fragilities and the importance of resilient
transactions across major advanced economies and
and coordinated backup systems (Khiaonarong, Leinonen, and
Rizaldy 2021). emerging market economies. The data, sourced from
11See, for example, Bruno and Shin (2015); Du, Tepper, and CLS Group, provide daily and weekly information from
Verdelhan (2018); Hofmann, Shim, and Shin (2020); Greenwood January 1, 2015, to May 31, 2025, on FX flows and
and others (2023); and Liao and Zhang (2025) for discussions of the
different channels through which tighter FX market conditions may pricing for 18 major currencies, disaggregated by four
be transmitted to the broader financial system. institutional sectors: banks, investment funds, other

50 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Figure 2.3. Shock Transmission to Foreign Exchange Market Conditions and Macrofinancial Feedback Effects

Amplified by
Change in FX flows FX market conditions
vulnerabilities
Financial flows Currency mismatch Higher FX volatility (excess return volatility)
Adverse macro- (funding; hedging; speculation)
financial or Trade flows Higher FX transaction costs (bid/ask)
operational shock Dealer concentration
Remittance flows Higher hedging and funding costs
(CIP premium)
NBFI share in customer flows
Higher settlement risk for non-PvP currencies
Dealer balance sheet
constraints • Lower risk adjusted return
of investment/lending in
foreign currency
• Lower risk appetite

Financial and nonfinancial Financial and liquidity

Other asset prices
sector balance sheets conditions

Macrofinancial stability

Source: IMF staff.

Note: Black arrows in the figure indicate the direct transmission channels. CIP = covered interest parity; FX = foreign exchange; NBFI = nonbank financial institution;
PvP = payment versus payment.

NBFIs, and nonfinancial firms.12 Macrofinancial uncer- and market conditions (Figure 2.3). For example,
tainty is captured using three commonly used indicators: a surge in financial uncertainty, often captured by the
the Chicago Board Options Exchange Volatility Index VIX, which measures option-implied volatility in the
(VIX), the Merrill Lynch Option Volatility Estimate US stock market, can dampen investor risk appetite
(MOVE) index, and the economic policy uncertainty and prompt a flight to quality, with investors real-
(EPU) index of Baker, Bloom, and Davis (2016). locating portfolios toward safer assets, such as those
denominated in US dollars (Figure 2.3; Caballero and
Krishnamurthy 2008).13
Macrofinancial Shocks and the Global Rebalancing of this type increases demand for
Foreign Exchange Market: A Conceptual dollar-denominated assets while leading to the unwind-
Framework ing of positions in other currencies. Concurrently,
The global FX market, as a decentralized, mostly financial institutions outside the United States, such as
over-the-counter arena, enables continuous and flexible nondealer banks, investment funds, and insurers, may
trading across time zones. It draws on a wide range of seek to hedge their increased dollar exposures using
participants, including dealers (banks and NBFIs that FX swap contracts, which involve buying US dollars
act as market makers), nondealer banks and NBFIs, in the spot leg of a contract while agreeing to sell
nonfinancial firms, central banks, and retail investors. them back in the forward leg of the contract. Dealer
It is broadly segmented into spot and derivatives mar- banks that intermediate the transactions involved in
kets, each catering to different investment horizons, these contracts must expand their balance sheets to
risk profiles, and participant needs. meet the increased demand, often borrowing dollars
An increase in global macrofinancial uncertainty
13US-dollar-denominated safe assets have long anchored the
can trigger significant shifts in cross-border trade and
international monetary system, reflecting the dollar’s central role in
financial activity, with direct implications for FX flows global trade and finance. Although geopolitical and economic shifts
could affect the use of international reserve currencies and the dol-
lar’s status as a global safe asset, evidence of any significant structural
12Online Annex 2.2 provides further details on CLSMarketData change in recent months remains limited (Chapter 1; External Sector
and the sample of economies included in the analysis. Report 2025, Chapter 2).

International Monetary Fund | October 2025 51

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

in the process. This can be costly when regulatory and restoring market functioning (Barajas, Deghi,
constraints limit dealer banks’ ability to supply Fendoglu, and Xu 2020; Aizenman and others 2021).
liquidity.14 As a result, the cost of these swaps rises In addition to raising the cost of FX swaps, an
and the cross-currency bases between the dollar and increase in uncertainty can make the global FX mar-
other currencies widen, effectively placing a limit on ket less liquid and volatile. As financial uncertainty
the extent of portfolio rebalancing by tightening FX increases, it can become more costly for dealers to hold
market conditions.15 foreign currencies, raising the cost of facilitating trades
Portfolio rebalancing triggered by a shock can create in the global FX market. This is partly because the
a vicious cycle. In addition to their use for hedging, rising risk of large losses from adverse exchange rate
FX swaps are used for speculation, such as carry trades, movements requires banks to hold more capital as a
and for short-term dollar funding of longer-term safety buffer.17 The expense of holding this additional
asset positions, a form of maturity transformation capital is passed on to customers through wider bid-ask
(October 2019 Global Financial Stability Report). When spreads: the gap between the prices at which dealers
the US dollar–foreign currency basis widens, it signals can buy and sell foreign currencies. Wider spreads
rising costs and reduced availability of dollar funding, make trading more expensive, discouraging participa-
particularly for institutions outside the United States. tion and making markets less liquid. With fewer trades
The funding strains imposed by these changes can taking place, prices can swing more sharply in response
force unhedged exposures or asset sales, increasing mar- to even small orders. This can lead to another vicious
ket volatility. Heightened volatility, in turn, may fuel cycle: higher volatility leads to wider spreads, which
investor risk aversion and flight to safety, intensifying reduces trading and amplifies market volatility.
demand for safe assets and FX swaps, further widening Structural market fragilities can amplify the trans-
the cross-currency basis and deepening funding stress mission of global shocks to FX market conditions
(Huang and others 2025).16 The resulting feedback (Figure 2.3). For example, when market making is con-
loop pressures dealer balance sheets and constrains deal- centrated among a small number of dealers, regulatory
ers’ ability to absorb risk, amplifying systemic risk and constraints arising from sharp declines in asset prices are
disrupting global financial intermediation. Such trans- more likely to lead to funding and liquidity stress, with
mission effects can be mitigated by requiring financial broader market implications. Similarly, when market
institutions to hold adequate foreign currency liquidity participants have smaller liquidity buffers or greater
buffers and stable dollar funding such as customer leverage, as some types of NBFIs often do, they are more
deposits. In addition, historical episodes of high macro- likely to engage in procyclical market behavior, which
financial uncertainty, such as the global financial crisis can amplify the impact of a shock on FX market con-
and the COVID-19 market turmoil, have highlighted ditions. In addition, larger underlying FX mismatches
the importance of central bank interventions, including across institutions increase the latter’s reliance on FX
those through dollar liquidity swap lines, for breaking swaps for hedging, resulting in more pronounced portfo-
cycles generated by shock-induced portfolio rebalancing lio adjustments when cross-currency bases widen.18
Stress in FX markets can be transmitted to the
14During periods of broad market volatility, dealer banks may
broader financial system and the real economy through
face tighter balance sheet constraints because of a rise in demand for various channels. Elevated FX funding costs, reflected
intermediation and a decline in asset values, making it harder for in a widening of cross-currency bases, can reduce the
them to meet regulatory requirements for leverage and limiting their profitability of financial institutions facing capital
ability to support FX swap markets.
15In addition, informational asymmetries in FX markets, in which

some participants know more than others, mean that large customer 17Specifically, the increased risk of sharp exchange rate losses

trades can signal valuable information. Dealers may interpret the can raise dealer banks’ value-at-risk estimates—measures of how
flows involved in these trades as reflecting private insights or strategic much a portfolio could lose over a set time period with a given
intent, prompting them to adjust prices or positions, which can trig- level of confidence—which are used to calculate regulatory capital
ger broader market reactions, increasing volatility, widening spreads, requirements.
and reducing liquidity. 18Eguren-Martin, Busch, and Reinhardt (2024) find that UK
16When volatility rises, arbitrage mechanisms that normally banks with greater currency mismatches—that is, those more reliant
keep prices aligned across venues and instruments can break down. on FX swaps for US dollar funding and hedging of exchange rate
Tighter balance sheet constraints increase the cost of providing risk—respond to a widening cross-currency basis by cutting back
liquidity and deter arbitrage, allowing price gaps to persist, distorting cross-border foreign currency lending more aggressively than banks
pricing, and amplifying market pressure. with matched exposures.

52 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

constraints such as leverage ratio requirements. As cap- Beyond macrofinancial uncertainty, operational dis-
ital constraints tighten, such institutions may respond ruptions can significantly impair FX market functioning
by deleveraging, contracting their balance sheets and and amplify volatility. Outages in trading platforms,
selling risky assets, including stocks and corporate messaging systems, or payment and settlement infra-
bonds in local currencies.19 A wider cross-currency structure can delay trade execution and settlement,
basis also implies a higher cost of hedging FX risk increasing market illiquidity and counterparty risk.
embedded in long positions of US-dollar-denominated Although the global FX market’s decentralized struc-
assets, prompting institutions to reduce their hedge ture and high substitutability across platforms have
ratios—the proportion of their foreign currency risk implied that disruptions to individual trading platforms,
exposure covered by hedging instruments—and take including interdealer platforms, have not been systemic
on greater FX risk. Market participants may mitigate to date, simultaneous outages across multiple platforms,
this risk through self-insurance strategies, increasing such as those from cyber incidents or power outages,
their holdings of safer and more liquid assets. As a could trigger systemic stress by cutting off access to
result, demand for sovereign bonds, particularly those liquidity and risk management tools (Box 2.3). Simi-
of short duration, can increase, especially in coun- larly, prolonged disruptions to payment systems and set-
tries with stronger fiscal fundamentals. This increased tlement infrastructures (for example, CLS; TARGET2;
demand can exert downward pressure on bond yields.20 Fedwire; and the Clearing House Automated Payment
Market liquidity and exchange rate volatility are System) are inherently more disruptive and require
also important in the transmission of FX market stress. robust safeguards and backup arrangements to contain
As noted earlier, higher transaction costs, reflected in systemic risks.
wider bid-ask spreads, can discourage market participa-
tion and impair the functioning of FX and related asset
markets. For financial institutions and firms engaged The Evolving Landscape of the Global
in cross-border trade and investment, this can translate Foreign Exchange Market
into reduced access to hedging instruments and trade The global FX market has expanded, with average
finance, dampening economic activity. Similarly, excess daily trading volumes increasing fivefold since the late
exchange rate return volatility can increase uncertainty 1990s. Growth in FX swap and spot transactions has
around asset valuations and macroeconomic outcomes, driven this expansion and accounts for the bulk of trad-
undermining investor confidence and prompting ing activity (Figure 2.1, panel 1). FX swap activity has
portfolio rebalancing away from riskier assets.21 In risen notably in recent years, reflecting increased NBFI
addition, excess volatility can tighten banks’ balance participation (Figure 2.4, panels 1 and 2). A majority
sheet constraints, potentially reducing domestic credit of these swaps are of short duration, typically with ten-
provision. Such dynamics would reinforce the feedback ors up to three months (Figure 2.4, panel 3), highlight-
loop between FX market stress and broader financial ing the continuous hedging needs of institutions and
instability. the market’s role in short-term liquidity management.
The US dollar remains the dominant trading
19When risk-weighted capital requirements become more bind-
currency in spot and swap markets. About one-fourth
ing, financial institutions may shift toward safer assets, increasing
demand for sovereign bonds, which typically carry zero risk weights. of transactions involve the euro against the US dollar,
In contrast, when leverage constraints tighten, FX market stress is and one-fifth involve the Japanese yen, underscoring
more likely to push local currency sovereign bond yields higher, as the importance of these two currencies among major
institutions face broader funding pressures rather than incentives to
rebalance toward low-risk assets.
currency pairs (Figure 2.4, panel 4). However, the
20These dynamics can also spill over into stock and other asset euro–US dollar share of total transactions has declined
markets through the actions of leveraged investors and financial from about one-third in 2015, whereas the share of
intermediaries, such as hedge funds, pension funds, and insurers. For
other currencies relative to the dollar has increased,
example, pension funds with internationally diversified portfolios
and local currency defined-benefit liabilities typically hedge FX reflecting a gradual diversification in trading activity.22
risk to manage asset-liability mismatches. However, as hedging cost
rises, these entities may choose to unwind their foreign positions, 22Figure 2.4, panel 4, is based on CLS data and covers FX trans-

transmitting stress from FX markets to other markets, amplifying actions settled by CLS. As shown in Figure 2.1, panel 2, the share
volatility across asset classes. of other currencies, particularly the Chinese yuan, has increased over
21Excess exchange rate volatility may also complicate monetary time. The share of the yuan, however, remains well below China’s
policy transmission, affecting macrofinancial stability. share in global output and trade.

International Monetary Fund | October 2025 53

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 2.4. Structure and Trends in the Global Foreign Exchange Market
Trading activity in FX swaps and forwards has surged in recent years ... ... with the increase driven largely by NBFIs.
1. FX Derivatives, by Instrument Type, 1998–2024 2. FX Derivatives, by Counterparty, 1998–2024
(Trillions of dollars) (Trillions of dollars)
150 FX swaps and forwards Reporting dealers 150
Currency swaps Other financial institutions
Derivatives Nonfinancial customers
Total OTC FX derivatives Total OTC FX derivatives
100 100

50 50

0 0
1998 2004 11 17 24 1998 2004 11 17 24

A large share of FX swaps have short maturities: three months or less. The euro and yen dominate FX trading against the US dollar, but the share
of other currencies has been rising.
3. Composition of FX Swaps, by Tenor, 2015:M1–2025:M5 4. FX Spot and Swap Turnover, by Currency Pair, 2015:M1–2025:M5
(Percent) (Trillions of dollars)
0 days 1–3 days 4–95 days 96–360 days ≥361 days EURUSD USDJPY GBPUSD USDOTH 700
100 EURJPY JPYOTH EUROTH OTHOTH
600
500
400
50 300
200
100
0 0
All Bank Fund Other NBFIs NFCs 2015 16 17 18 19 20 21 22 23 24 25

Sources: Bank for International Settlements, OTC Derivatives Statistics; CLS Group; and IMF staff calculations.
Note: Panels 1 and 2 are based on the Bank for International Settlements’ OTC Derivatives Statistics and reflect the FX derivatives outstanding by instrument type and
counterparty, respectively. Panels 3 and 4 are based on CLSMarketData by CLS Group. Panel 3 indicates the composition of FX swaps by tenor for all swap transactions and
transactions disaggregated by market participant, aggregating all swap flow data from January 2015 to May 2025. Panel 4 shows the composition of FX spot and swap
transactions aggregated annually by currency pair. EUR = euro; FX = foreign exchange; GBP = British pound; JPY = Japanese yen; M = month; NBFIs = nonbank financial
institutions; NFCs = nonfinancial corporation; OTC = over-the-counter; OTH = other currencies; USD = US dollar.

The dollar’s relative importance seems to have of the global FX network, whereas their financial and
remained stable through May 2025, with no major nonfinancial clients typically operate at the periphery.23
shift after the US tariff announcements in early Bank participation in FX trading varies considerably
April 2025 (Box 2.1). across currency pairs. For example, euro–US dollar
A large share of FX transactions takes place between transactions are intermediated by US banks as well as
banks (Figure 2.5, panel 1). Among NBFIs, investment by banks in some other major markets, such as France,
funds dominate FX trading, reflecting the funds’ grow- Germany, and the United Kingdom (Figure 2.5,
ing use of FX instruments for portfolio diversification panel 2). However, yen–US dollar transactions are
and risk management (BIS 2025b). Banks remain
central to the FX ecosystem, as evident from their high 23Because CLS membership largely comprises major banks,

each transaction in the data set is recorded relative to the banking

degree of interconnectedness with other market par-
sector. The data thus do not capture the interconnectedness
ticipants (Figure 2.5, panels 2 and 3). Banks in major between other institutional sectors, such as that between NBFIs and
economies, especially the United States, form the core nonfinancial firms.

54 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Figure 2.5. Bank and Nonbank Financial Institutions’ Presence in the Global Foreign Exchange Market

Interbank transactions account for most FX market activity. Banks in the United States and in major European economies are dominant
in euro–US dollar trades ...
1. FX Spot and Swap Turnover, by Counterparty, 2024 2. Network of Spot Euro–US Dollar Sectoral Trading Activity, 2024
(Trillions of dollars) (Share)
600 60 Banks
KR
500 50 EZ Investment funds
JP Other NBFIs
400 40 NFCs
ZA KR
DK KR
IT US
300 30
RW CH US
FR GB
200 20 US SG
RW GB CH
DK EZ ES FR
100 10 NL CH CA
EZ JP CA DE
HK GB
0 0 RW AU
SG
US
Bank–Bank Bank–Fund Bank–Other NBFIs Bank–NFCs GB NL AU
JP
(right scale) (right scale) (right scale)

... whereas Japanese and US banks dominate yen–US dollar trades. Dealer bank concentration thus varies across currency pairs and
instruments.
3. Network of Spot Yen–US Dollar Sectoral Trading Activity, 2024 4. Dealer Concentration in FX Swaps, 2015:M1–2025:M5
(Share) (Index)
RW Banks Interquartile range Median: all tenors
DK Investment funds Median: tenor 96–360 days Moderate concentration 3,000
Other NBFIs High concentration
KR
NFCs
CA
DK CN 2,400
US HK JP
US
RW JP CA DE
US 1,800
JP GB FR
US
IT
EZ HK
JP CH AU SG 1,200
ZA CH
SG
KR
NL KR 600
2015 17 19 21 23 25

The share of NBFI trading activity also varies by currency pairs ... ... as does NBFI hedging activity.
5. Share of NBFIs in FX Swaps, by Currency Pair, 2015:M1–2025:M5 6. NBFIs’ Hedging Pressure, by Currency
(Percent) (Percent)
30 Interquartile range NBFI swap share, median EUR JPY CHF GBP 100
80
60
20
40
20
10
0
−20
0 −40
2015 17 19 21 23 25 2015 17 19 21 23 25

Sources: CLS Group; and IMF staff calculations.

Note: Panel 1 shows the total spot and swap flow turnover by counterparty sector for 2024 across all currencies. Panels 2 and 3 show the structure of euro–US dollar and
yen–US dollar networks, respectively, depicting average net flow from each country-sector in the pair to the other country-sector in 2024. The size of the node represents
the market share of gross flows. The arrows represent the direction of net inflows. Data labels in panels 2 and 3 use International Organization for Standardization (ISO)
two-digit country codes. Panel 4 measures bank dealer concentration across currency pairs by computing the Herfindahl-Hirschman Index. The index scale ranges from 0
to 10,000. The horizontal lines refer to the Bank for International Settlements’ benchmarks for concentration in FX markets. Panel 5 shows the share of NBFIs by currency
pair and indicates the median and interquartile range of the distribution. In panel 6, hedging pressure is computed as the difference between short and long swap
outstanding positions of various currencies relative to the US dollar, normalized by the total outstanding FX swap positions. A positive hedging pressure reflects an
imbalance in NBFIs’ demand for hedging in the FX swap market. The measure is shown as a three-month rolling average (see Online Annex 2.3 for details). CHF = Swiss
franc; EUR = euro; EZ = euro area countries excluding France, Germany, Italy, The Netherlands, and Spain; FX = foreign exchange; GBP = British pound; JPY = Japanese
yen; M = month; NBFI = nonbank financial institution; NFC = nonfinancial corporation; RW = rest of the world.

International Monetary Fund | October 2025 55

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

predominantly facilitated by banks in Japan and the Nonresident NBFIs typically increase their holdings
United States (Figure 2.5, panel 3). There are thus of safe haven assets during periods of elevated macro­
different degrees of dealer bank concentration, defined financial uncertainty. For example, net purchases of US
here as the extent to which banks in specific jurisdic- dollars in both spot and swap markets tend to rise with
tions dominate trading activity in a particular currency spikes in the VIX or the US EPU index (Figure 2.6,
pair, with some currency pairs showing higher degrees panels 1 and 2).27 Similarly, net spot purchases of
of concentration, as reflected by the upper range of other safe haven currencies, such as the euro and the
the interquartile distribution in Figure 2.5, panel 4.24 Swiss franc, by nonresident NBFIs also react strongly
The degree of dealer concentration is also high in to these shocks (Online Annex Figure 2.3.5).
swap markets with longer tenors, suggesting that even In the recent episode of uncertainty triggered by US
though the global FX market is broadly diversified, tariff announcements in early April 2025, nonresident
certain segments remain reliant on a concentrated NBFIs increased their purchases of safe haven assets.
group of dealers. This reliance can amplify systemic However, overall net spot purchases of US dollars by
risk in the event of financial or operational disruptions both non-US banks and non-US NBFIs were relatively
affecting these key institutions. subdued compared with those in previous episodes,
The trading activity of NBFIs also varies signifi- such as the 2020 COVID-19 shock (see Box 2.1).28
cantly across currency pairs. Over the past decade, Demand for US dollar swaps by non-US NBFIs
the median share of NBFIs in FX trading activity rose sharply, suggesting a shift to hedge previously
has averaged about 8 percent (Figure 2.5, panel 5). unhedged exposures. Despite the magnitude of the
However, some currency pairs, such as euro–US dollar, shock, stress in the FX market remained limited, with
have had a significantly higher participation share, no major disruption.
exceeding 15 percent in recent years. This variation
reflects differences in market structure, liquidity, and
the importance of specific currency pairs to institu- Macrofinancial Uncertainty and FX
tional investors. Moreover, hedging of currency expo- Trading Dynamics
sures among NBFIs, measured by their net FX swap This section formally examines how an increase in
positions against the US dollar in various currencies, global macrofinancial uncertainty affects FX trading
has generally been on an increasing trend (Figure 2.5, activity, focusing on cross-border transactions involving
panel 6).25 Notably, this measure appears to be posi- major currencies against the US dollar. The analysis
tively correlated across major currencies, suggesting a considers several uncertainty measures, including
synchronized need for dollar hedging (“hedging pres- financial market volatility, monetary policy uncertainty,
sure”) that can strain liquidity and amplify volatility, and broader economic policy uncertainty.29 These
particularly in times of market stress.26 measures capture different dimensions of risk and
may influence market behavior and demand for the
dollar through distinct but related channels: financial
24Dealer bank concentration is measured using the volatility, proxied by the VIX, often triggers immedi-
Herfindahl-Hirschman Index, which reflects the share of trading ate liquidity needs and safe haven flows, as investors
activity in specific currency pairs in key financial jurisdictions—Can-
ada, France, Germany, Japan, Switzerland, the United Kingdom, and reallocate their investments into assets with cash-like
the United States—conducted by banks classified as dealer banks.
Banks in all other jurisdictions are treated as nondealer banks (see 27In general, both US and non-US NBFIs are net buyers of US
Online Annex 2.3 for details). dollars. Among banks, US institutions typically supply US dollars to
25Specifically, hedging pressure is defined as the net short swap their non-US counterparts in the spot market (Online Annex 2.3).
position of NBFIs in specific currencies with respect to the US dollar In the FX swap market, behavior varies by tenor: US banks rely on
relative to the total outstanding market swap position (Bräuer and short-tenor swaps for their FX funding, which non-US banks pro-
Hau 2023). This measure reflects both demand-side factors (for vide. For longer tenors, US banks provide hedges to non-US institu-
example, rollover needs of NBFIs) and supply-side constraints (for tions (Kloks and others 2023; Kloks, Mattille, and Ranaldo 2024).
example, dealer balance sheet limitations). Subsequent empirical 28These observations are supported by data on portfolio flows into

analysis attempts to disentangle these two channels. the United States among nonresident investment funds, which show a
26Hedging pressure from NBFIs across countries tends to be slowdown during April and May (Online Annex Figure 2.7.1). Bond
strongly correlated with these institutions’ net bond investment fund flows especially declined, reflecting in part weaker demand for
positions with respect to the United States, suggesting that greater US Treasuries (Grothe and others 2025; Jiang and others 2025).
exposure to US fixed-income assets is associated with higher demand 29See Online Annex 2.4 for further details on empirical methodol-

for FX hedging (see Online Annex 2.3; BIS 2025b). ogy and results for the analysis presented in this section.

56 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Figure 2.6. Net Purchase of US Dollars and Macrofinancial Uncertainty

Non-US nonbank financial institutions increase their purchase of dollars as ... whereas US banks act as net lenders.
uncertainty rises ...
1. Net US Dollar Spot Flows versus VIX and EPU, 2015:M1–2025:M5 2. Net US Dollar Swap Flows versus VIX and EPU, 2015:M1–2025:M5
(Weekly correlation in four-week moving averages) (Weekly correlation in four-week moving averages)
0.4 0.1
VIX US EPU VIX US EPU
0.3
0
0.2 Higher USD demand

−0.1
0.1

0
−0.2

−0.1 Higher USD demand

−0.3
−0.2

−0.3 −0.4
US banks US nonbank Non-US banks Non-US nonbank US banks US nonbank Non-US banks Non-US nonbank
institutions institutions institutions institutions

Sources: CLS Group; Haver Analytics; and IMF staff calculations.

Note: The figure displays the unconditional correlations between net US dollar spot and swap flows and various uncertainty measures across different country sectors.
EPU = economic policy uncertainty index; M = month; USD = US dollar; VIX = Chicago Board Options Exchange Volatility Index.

properties; monetary policy uncertainty (measured levels of economic policy uncertainty (Online Annex
by the MOVE index’s signal of volatility in the bond Figure 2.4.5).31
market) may affect expectations and funding costs; NBFIs, particularly investment funds, respond more
and broader economic policy uncertainty, captured by strongly to global uncertainty shocks than do banks
the EPU index, can weigh on longer-term investment or nonfinancial firms. After a spike in the VIX or the
decisions and influence demand for the dollar. MOVE index, weekly growth rates in trading volumes
An increase in uncertainty tends to raise nonresi- of nonresident NBFIs rise by about 40 percentage
dent demand for US dollars. The effects are particu- points, on average, compared with a 15 percentage point
larly pronounced after large shocks to the VIX or the increase in the rates for dealer and nondealer banks
MOVE index, defined here as unexpected changes (Figure 2.7, panels 2 and 3). This sectoral asymmetry
in the values of those indices exceeding two standard could reflect that NBFIs are more exposed to market-­
deviations (Figure 2.7, panel 1).30 The estimated driven risks and operate with tighter liquidity and mar-
effects are economically meaningful: uncertainty shocks gin constraints. Unlike banks, NBFIs rely more heavily
of the magnitude observed during episodes like the on market funding and collateralized borrowing, making
2020 COVID-19 turmoil can raise weekly spot trading them more vulnerable to asset price volatility and margin
growth by up to 24 percentage points. Additional anal- calls (Aramonte and Avalos 2021; FSB 2022).
ysis for flows into other safe haven currencies—such as Similar dynamics are evident in the FX swap mar-
the euro, Japanese yen, and the Swiss franc—confirms ket, in which NBFIs account for a growing share of
that these currencies also attract inflows during periods
of heightened global uncertainty (Online Annex Figure 31Although these findings highlight strong demand for safe haven

2.4.2). These results remain robust when the sample is currencies during major uncertainty shocks, caution is warranted in
restricted to the period preceding the 2025 US tariff extrapolating the patterns observed to future episodes. Deepening
global fragmentation that triggers major shifts in the use of inter-
announcements, a time marked by unusually high national currencies may alter traditional flight-to-safety dynamics.
Moreover, the nature of a particular shock matters: more localized
episodes, such as a rise in US-specific macroeconomic uncertainty
30Using an alternative measure of monetary policy uncertainty without a broader surge in global macrofinancial uncertainty, may
derived from the EPU index yields similar results (see Online Annex reduce nonresident demand for US dollars (see Online Annex 2.4;
Figure 2.4.1). Grothe and others 2025).

International Monetary Fund | October 2025 57

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 2.7. Effect of Macrofinancial Uncertainty on Foreign Exchange Spot Activity

After a spike in macrofinancial uncertainty, demand for US dollars among nonresident institutions, especially NBFIs, tends to rise.
1. Change in Spot US Dollar Inflows after an 2. Change in Spot US Dollar Inflows after a 3. Change in Spot US Dollar Inflows after a
Uncertainty Shock VIX Shock, by Counterparty MOVE Shock, by Counterparty
(Percentage points) (Percentage points) (Percentage points)
40 Baseline Large uncertainty shock 70 70
60 60
30
50 50
20 40
40
30
10 30
20
10 20
0
0 10
−10 −10 0
VIX MOVE US EPU Bank Dealer Nondealer NBFIs Bank Dealer Nondealer NBFIs
bank bank bank bank

Sources: Baker, Bloom, and Davis 2016; Bloomberg Finance L.P.; CLS Group; LSEG Datastream; and IMF staff calculations.
Note: “US dollar inflows” refers to spot transactions by non-US financial and nonfinancial institutions in 15 jurisdictions. The figure displays the impacts of various
uncertainty shocks on weekly changes in dollar inflows using a panel model. The model controls for a range of global and domestic macrofinancial factors, including the
Federal Reserve Bank of Chicago’s National Financial Conditions Index, a commodity price index, the US term spread, domestic term spreads in other countries, and the
spot exchange rate. The specification also includes country-sector fixed effects and country-time fixed effects. In panel 1, “Baseline” refers to the effects of a
one-standard-deviation increase in uncertainty measures on the outcome variable. “Uncertainty shocks” are defined as dummy variables equal to 1 when the first-order
autoregression residual of the underlying indicator exceeds two standard deviations. Whiskers show the 90 percent confidence intervals. EPU = economic policy
uncertainty index; MOVE = Merrill Lynch Option Volatility Estimate index; NBFIs = nonbank financial institutions; VIX = Chicago Board Options Exchange Volatility Index.

short-term US dollar funding and hedging. During FX exposures notably drive the hedging behavior of
episodes of elevated uncertainty, NBFIs increase FX participants, making currency mismatches a key ampli-
swap usage to secure liquidity or hedge currency fier of global shocks. Institutions holding substantial
exposures. The response is especially pronounced for open positions in foreign currencies, particularly the
longer-dated transactions (those of more than seven US dollar, face heightened amounts of valuation and
days) (Figure 2.8, panel 1), which are more indicative rollover risk during periods of volatility, prompting
of hedging, than for short-term trading or arbitrage a surge in hedging activity. The analysis shows that
flows (Du, Tepper, and Verdelhan 2018; Bräuer and countries whose banks have larger dollar funding
Hau 2023).32 Quantitatively, a VIX shock raises FX gaps tend to exhibit stronger responses to uncer-
swap activity by about 5 percentage points among tainty shocks (Figure 2.8, panel 3). Net international
banks but nearly twice as much among NBFIs, which investment positions in the dollar also help explain
is consistent with patterns observed in spot market cross-country differences in hedging behavior, as US
flows (Figure 2.8, panel 2). A similar response follows and non-US investors often face opposing hedging
shocks to the MOVE index, which captures expected needs (Figure 2.8, panel 4).33 These findings align with
volatility in US interest rates. However, the response studies showing that net hedging activity is propor-
appears somewhat weaker in this case, presumably tional to a country’s net investment position in the
because the regression model includes the term-spread corresponding currency (Gabaix and Maggiori 2015;
differential as a control variable, which may absorb Devereux and Yu 2020; Liao and Zhang 2025).34
part of the impact of the MOVE index on US dollar
33For instance, a country with a large long position in US dollar
swap inflows (Kumar and others 2023).
assets—such as holdings of US Treasuries—would face a greater
incentive to hedge FX risk when volatility rises, typically by selling
32Most sub-seven-day trades are used by banks and money dollars forward.
market desks to square books, fund inventories, or arbitrage rate 34The banking sector’s US dollar FX mismatch is measured as

differentials. These flows roll over daily and serve intraday liquidity the ratio of dollar-denominated assets minus dollar-denominated
needs rather than strategic hedging purposes. Short-dated swaps liabilities, normalized by dollar-denominated assets. Similarly, the net
are concentrated in the interdealer market for intraday liquidity investment position is defined as the difference between dollar long-
management or maturity transformation, in which banks use short- term debt held by foreign investors and foreign long-term debt held
dated interbank swaps to fund longer-dated dollar lending (Kloks by US institutions, normalized by the total amount of outstanding
and others 2023). long-term debt among US and foreign investors.

58 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Figure 2.8. Effect of Macrofinancial Uncertainty on Foreign Exchange Swap Flows

Heightened uncertainty raises FX swap activity in longer maturities, ... especially among NBFIs, likely reflecting tighter balance sheet
indicating stronger hedging rather than interdealer activity ... constraints and greater sensitivity to global shocks.
1. Change in US Dollar Swap Inflows after an Uncertainty Shock 2. Change in US Dollar Swap Inflows after a VIX Shock, by Counterparty
(Percentage points) (Percentage points)
15 All At or above 7 days 20

10 15

5 10

0 5

−5 0
All Below At or above All Below At or above Banks Dealer bank Nondealer bank NBFIs
7 days 7 days 7 days 7 days
VIX MOVE

Hedging demand tends to rise with FX mismatches, as institutions with larger currency gaps—measured by the bank funding gap or net foreign investment
positions—face greater exposure to exchange rate fluctuations.
3. Change in US Dollar Swap Inflows after a VIX Shock, by CCFR 4. Change in US Dollar Swap Inflows after a VIX Shock, by NIP
(Percentage points) (Percentage points)
30 Low CCFR High CCFR Low NIP High NIP 20
15
20
10
10 5

0 0
−5
−10
−10
−20 −15
All Below 7 days At or above 7 days All Below 7 days At or above 7 days

Sources: Baker, Bloom, and Davis 2016; Bloomberg Finance L.P.; CLS Group; LSEG Datastream; and IMF staff calculations.
Note: “US dollar swap flows” refers to swap transactions by non-US financial and nonfinancial institutions. FX swap transaction volumes are smoothed, with the four-week
average of weekly flows taken. The figure displays the impacts of different uncertainty shocks on weekly changes in dollar inflows using panel models with fixed effects.
The models control for global and domestic macrofinancial factors, including the Federal Reserve Bank of Chicago’s National Financial Conditions Index, a commodity price
index, the US term spread, domestic term spreads, the three-month overnight index swap covered interest parity deviation, and the spot exchange rate. The specification
also includes country-sector fixed effects and country-time fixed effects. An “uncertainty shock” is defined as a dummy variable equal to 1 if the first-order autoregression
residual of the underlying indicator exceeds two standard deviations. The banking sector’s US dollar FX mismatch is the ratio of its net to its total dollar-denominated
assets. The dollar net investment position is the difference between foreign holdings of US long-term debt and US holdings of foreign long-term debt, scaled by total
outstanding holdings of long-term debt. High (low) CCFR or NIP refers to economies with vulnerability levels above (below) the quarterly sample median. Whiskers show
the 90 percent confidence intervals. CCFR = cross-currency funding gap ratio; FX = foreign exchange; MOVE = Merrill Lynch Option Volatility Estimate index;
NBFIs = nonbank financial institutions; NIP = net (foreign) investment position; VIX = Chicago Board Options Exchange Volatility Index.

Uncertainty Shocks and Stress in Foreign The baseline analysis uses a panel of 11 major pairs of
Exchange Markets the US dollar with other currencies for which disag-
gregated data are available from CLS. In addition,
The effects of global uncertainty shocks on FX trad-
an alternative sample that includes a broader set of
ing dynamics can translate into FX market stress. To
emerging market currencies is used to assess the impact
evaluate these effects, three key measures of FX market
of uncertainty shocks on FX markets in that group of
conditions are considered here: (1) the cross-­currency
countries.36
basis (through CIP deviation), (2) annualized excess
spot-return volatility, and (3) quoted bid-ask spreads.35
36The emerging market sample includes 16 currencies, with cov-

erage varying across regressions due to data availability (see Online

35See Online Annex 2.5 for methodological details. Annex 2.5 for details).

International Monetary Fund | October 2025 59

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 2.9. Effect of Global Macrofinancial Uncertainty Shocks on Foreign Exchange Market Conditions

Uncertainty shocks widen CIP deviations, ... while also increasing excess exchange rate ... and widening bid-ask spreads.
signaling increased US dollar funding stress ... return volatility ...
1. Effect of an Increase in Uncertainty 2. Effect of an Increase in Uncertainty 3. Effect of an Increase in Uncertainty
Measures on CIP Deviation Relative to Measures on Excess Exchange Rate Return Measures on Bid-Ask Spreads Relative to
the US Dollar Volatility Relative to the US Dollar the US Dollar
(Basis points) (Percentage points) (Basis points)
10 0.20 4
Baseline Baseline
Large uncertainty shock Large uncertainty shock
0 0.16
3

−10 0.12
2
−20 0.08

Baseline 1
−30 0.04
Large uncertainty shock
−40 0 0
VIX US EPU MOVE VIX US EPU MOVE VIX US EPU MOVE

Sources: Baker, Bloom, and Davis 2016; Bloomberg Finance L.P.; CLS Group; LSEG Datastream; and IMF staff calculations.
Note: The figure shows the effects of one-standard-deviation increases in each uncertainty indicator and high uncertainty shocks on the three-month overnight index swap
CIP deviation, excess exchange rate return volatility, and bid-ask spreads over a one-week horizon. Large uncertainty (VIX, US EPU, or MOVE) shocks are represented as
dummy variables equal to 1 when the first-order autoregression residuals of the underlying indicators exceed two standard deviations. The effects are economically
significant: The standard deviation is about 40 basis points for CIP deviations, 0.3 percentage point for excess exchange rate return volatility, and 0.06 percent for bid-ask
spreads (normalized by the mid-rate). Whiskers show the 90 percent confidence intervals. CIP = covered interest parity; EPU = economic policy uncertainty index; MOVE =
Merrill Lynch Option Volatility Estimate index; VIX = Chicago Board Options Exchange Volatility Index.

Uncertainty shocks are a key driver of cross-currency ple, the results show that these tend to experience
bases relative to the dollar. A one-standard-deviation somewhat stronger and more persistent effects after
increase in global macrofinancial uncertainty indica- uncertainty shocks, across all measures of FX market
tors, such as the VIX or the EPU index, widens the conditions (Online Annex Figure 2.5.3). Cross-cur-
three-month basis by up to 13 basis points over a rency bases and bid-ask spreads widen, on average,
week (Figure 2.9, panel 1). The effect of changes in more than twice the amounts estimated for advanced
the MOVE index is not statistically significant once economies, and the estimated effects on excess
the effect of term-spread differentials is controlled for, exchange rate return volatility also increase notably.
consistent with the view that US interest rate volatility These results are aligned with those of earlier studies
transmits mainly through the yield curve (Kumar and (for example, Du and Schreger 2016, 2022; Dao and
others 2023). Large uncertainty shocks—those exceed- Gourinchas 2025; and Dao, Gourinchas, and Itskhoki
ing twice the standard deviation—produce dispropor- 2025) that find greater sensitivity of emerging market
tionately larger effects, indicating a nonlinear response. currencies to global shocks, possibly as a result of
Elevated uncertainty also impairs FX market liquid- structural factors like shallower markets, greater reli-
ity and increases volatility. In response to shocks to the ance on foreign currency financing, and more limited
VIX, EPU, and MOVE indices, weekly excess exchange access to dollar liquidity backstops.37
rate return volatility increases by about 5–10 basis
points, while bid-ask spreads widen by 1–3 basis 37Because data on the CIP deviation of overnight index swaps

points—equivalent to about half a standard deviation, are not available for many emerging markets, the analysis compar-
ing advanced economies and emerging market economies uses the
on average. These effects persist for up to three months,
Treasury CIP deviation, which is the difference between the yield on
peaking about four weeks after the shock (Figure 2.9, a US Treasury bond and the synthetic yield obtained by swapping a
panels 2 and 3; Online Annex Figure 2.5.2). foreign government bond into dollars through the FX swap market.
Notably, the effects of shocks are larger for emerg- Because much of the variation in Treasury CIP deviations could
reflect credit risk, the regressions include measures of expected
ing market currencies. When the analysis is extended default frequency in the banking sector or sovereign credit default
to include emerging market currencies in the sam- swap spreads. See Online Annex 2.3 for details.

60 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Effect of Foreign Exchange Market Fragilities Dealer balance sheet constraints represent another
key friction shaping the transmission of uncertainty
The impact of uncertainty shocks on FX markets
shocks to FX markets. Beyond demand-driven forces,
is shaped by underlying market fragilities. In the
FX market dysfunctions can arise as systematic
banking sector, US dollar funding needs amplify the
responses to market frictions that intensify under
effect of uncertainty on CIP deviations when the
elevated uncertainty. As outlined in the conceptual
cross-currency funding ratio (CCFR) is large, that
framework (Figure 2.3), increased market volatility
is, above the sample median, reflecting a significant
can constrain dealers’ balance sheet capacity, limiting
shortfall of dollar-denominated liabilities relative to
their ability to intermediate FX swaps and contributing
dollar-denominated assets that must be covered using
to CIP deviations (see, for example, Du, Tepper, and
FX swaps (Figure 2.10, panel 1). Elevated CCFR
Verdelhan 2018; Dao, Gourinchas, and Itskhoki 2025;
levels also amplify the effect on FX market volatility,
and Kubitza, Sigaux, and Vandeweyer 2025).39 To
with excess exchange rate return volatility rising in
formally assess this mechanism, the analysis interacts
response to uncertainty shocks when the CCFR is large
uncertainty shocks with a proxy for dealer balance
(Figure 2.10, panel 2). A second amplification chan-
sheet strength: specifically, the capital ratio of primary
nel arises from hedging pressure linked to currency
dealer banks, following He, Kelly, and Manela (2017).
mismatches on the balance sheets of NBFIs. As noted
A higher capital ratio, reflecting stronger equity
earlier, during periods of heightened macrofinancial
buffers, is associated with greater capacity to supply
uncertainty, NBFIs increase their dollar hedging activ-
derivatives and absorb risk. The findings suggest that
ity, typically by selling dollars forward in FX swaps.
stronger capital positions help mitigate the effects
As a result, hedging pressure tightens synthetic dollar
of uncertainty shocks, improving overall FX market
funding conditions and amplifies deviations from CIP
functioning by reducing CIP deviations and excess
(Figure 2.10, panel 3).
exchange rate return volatility (Figure 2.11, panels 1
Excess exchange rate return volatility is sensitive to
and 2).40
dealer concentration and the share of NBFI activity
(Figure 2.10, panel 4). In concentrated markets, as
reflected in a high Herfindahl-Hirschman Index, fewer Role of Policy Factors
dealers dominate, potentially reducing competition and
Policy backstops are critical for stabilizing the
market depth. This raises transaction costs and limits
global FX market during adverse shocks. Among the
the market’s capacity to absorb shocks. Similarly, a siz-
most effective tools are the Federal Reserve’s US dollar
able presence of price-taking NBFIs (for example, real-
liquidity swap lines, which provide selected foreign
money investors or macro funds) may increase order
central banks with direct access to dollar fund-
flow imbalances during stress, raising market volatility.
ing. These arrangements ease dollar funding stress,
These institutions typically demand liquidity without
limiting CIP deviations and helping to stabilize FX
providing it, contributing to wider spreads and higher
swap markets. During the 2020 COVID-19 tur-
levels of execution risk. Unlike hedging pressure, which
moil, the Federal Reserve expanded its swap lines to
affects valuations, dealer concentration and the NBFI
additional central banks and introduced the Foreign
share in a currency’s trade appear to affect market
and International Monetary Authorities Repo Facil-
outcomes more through intermediation and liquidity
ity, offering temporary liquidity against US Treasury
provision channels.38
collateral. Analysis shows that newly activated swap
38The distinction lies in the transmission channels: hedging pres- lines reduced CIP deviations by up to 30 basis points,
sure affects valuation—the pricing of forward rates relative to interest nearly offsetting the entire impact of the initial VIX
differentials—whereas dealer and investor structure affects liquidity
shock, and significantly lowered excess exchange rate
and intermediation capacity. CIP reflects a valuation arbitrage condi-
tion among spot, forward, and interest rates. When hedging demand
outpaces dealers’ ability to supply synthetic US dollars, forward 39Dealer balance sheets have not kept pace with the expansion

prices deviate from arbitrage-consistent levels. This distorts relative of the US Treasury supply since the global financial crisis (Online
pricing without necessarily affecting transactional liquidity. As a Annex Figure 2.5.1).
result, CIP deviations are more responsive to FX mismatches and 40The results support the idea that tighter dealer constraints

hedging imbalances, whereas spreads and volatility reflect frictions in reduce dealers’ ability to intermediate (He, Kelly, and Manela 2017;
liquidity provision. Duffie 2023).

International Monetary Fund | October 2025 61

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 2.10. Foreign Exchange Market Fragilities as Amplifiers of Shock Transmission

Banks’ FX mismatches increase the cost of US dollar funding through ... and amplify the impact of uncertainty shocks on excess exchange rate
swaps ... return volatility.
1. Effect of an Increase in Uncertainty on CIP Deviation Relative to 2. Effect of an Increase in Uncertainty on Excess Exchange Rate Return
the US Dollar Conditional on Banks’ FX Mismatch Volatility Conditional on Banks’ FX Mismatch
(Basis points) (Percentage points)
10 0.20
Baseline CCFR interaction effect Baseline CCFR interaction effect

0 0.15

−10 0.10

−20 0.05

−30 0

−40 −0.05
VIX Large VIX US EPU Large US EPU VIX Large VIX US EPU Large US EPU
shock shock shock shock

Increased NBFI hedging activity amplifies the effect of financial uncertainty ... whereas dealer concentration and greater participation of NBFIs in a
on CIP deviation ... currency’s trading amplify excess exchange rate return volatility.
3. Effect of an Increase in the VIX on CIP Deviation Relative to 4. Effect of an Increase in the VIX on Excess Exchange Rate Return
the US Dollar Conditional on Market Fragilities Volatility Conditional on Market Fragilities
(Basis points) (Percentage points)
10 0.10
Baseline Additional effect Baseline Additional effect
5 0.08
0.06
0
0.04
−5
0.02
−10
0
−15 −0.02
−20 −0.04
Hedging pressure HHI Share of NBFIs Hedging pressure HHI Share of NBFIs

Sources: Bloomberg Finance L.P.; LSEG Datastream; and IMF staff calculations.
Note: The figure shows the effects of one-standard-deviation increases in each uncertainty indicator and high-uncertainty shocks on three-month overnight index swap
covered interest parity deviations and excess exchange rate return volatility, along with the amplification effects due to various FX market vulnerability measures over a
one-week horizon. Large uncertainty (VIX or US EPU index) shocks are represented as dummy variables equal to 1 when the first-order autoregression residuals of the
underlying indices are two standard deviations above average. The CCFR measures a country’s banking sector’s US dollar mismatch as the difference between its US dollar
assets and US dollar liabilities, divided by US dollar assets, using quarterly Bank for International Settlements data. “Additional effect” refers to the additional impact of
uncertainty shocks when vulnerabilities are one standard deviation above their averages. “Hedging pressure” measures the net hedging activity of NBFIs and is calculated
as the difference between their aggregate short and long FX swap (forward) positions, scaled by the global average of outstanding US dollar contracts. The HHI is the sum
of the squared market shares of all bank dealers. “Share of NBFIs” captures the proportion of non-interdealer swap market activity accounted for by NBFIs. The three
vulnerability measures are computed for each currency area. Whiskers show the 90 percent confidence intervals. See Online Annex 2.3 for details on variable construction.
CCFR = cross-currency funding ratio; CIP = covered interest parity; EPU = economic policy uncertainty index; FX = foreign exchange; HHI = Herfindahl-Hirschman Index;
NBFI = nonbank financial institution; VIX = Chicago Board Options Exchange Volatility Index.

return volatility (Figure 2.12, panels 1 and 2). These funding dries up, whereas larger reserve buffers may
outcomes underscore the importance of swap lines in also enhance a sovereign’s perceived creditworthiness
mitigating market dysfunction.41 and help mitigate flight-to-quality pressures, thereby
International reserves are a stabilizing force during mitigating FX market stress. The analysis here shows
stress episodes. Central banks can use reserves to that economies with stronger reserve buffers—about
provide domestic dollar liquidity when private one standard deviation above the average—experience
notably smaller CIP deviations and lower excess
41The stabilizing effect of these new swap lines is consistent with
exchange rate return volatility following macrofinan-
Barajas, Deghi, Fendoglu, and Xu (2020); Barajas, Deghi, Raddatz,
and others (2020); Aizenman and others (2021); and Bahaj and Reis cial uncertainty shocks (Online Annex Figure 2.5.4,
(2022). panels 1 and 2).
62 International Monetary Fund | October 2025
 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Figure 2.11. Dealers’ Constraints and Foreign Exchange Market Conditions

A higher capital ratio boosts dealers’ capacity to supply FX liquidity, limiting ... whereas a lower ratio limits dealers’ willingness to intermediate,
CIP deviation ... worsening FX market conditions.
1. Effect of an Increase in Uncertainty on CIP Deviation Conditional on 2. Effect of an Increase in Uncertainty on Excess Exchange Rate Return
Dealers’ Capital Ratios Volatility Conditional on Dealers’ Capital Ratios
(Basis points) (Percentage points)
40 0.1
Baseline Baseline
Dealer capital interaction effect Dealer capital interaction effect
30
0
20

10
−0.1
0

−10
−0.2
−20

−30 −0.3
VIX Large VIX US EPU Large US EPU VIX Large VIX US EPU Large US EPU
shock shock shock shock

Sources: Bloomberg Finance L.P.; LSEG Datastream; and IMF staff calculations.
Note: The figure shows the effects of one-standard-deviation increases in the VIX and the US EPU index and their associated uncertainty shocks on FX market conditions,
along with the mitigating effects of a proxy of dealer balance sheet strength. Large uncertainty (VIX or US EPU) shocks are represented as dummy variables equal to 1 when
the first-order autoregression residuals of the underlying indices are two standard deviations above the average. Dealer balance sheet strength is proxied by the capital
ratio of primary dealer banks from He, Kelly, and Manela (2017). Whiskers show the 90 percent confidence intervals. CIP = covered interest parity; EPU = economic policy
uncertainty index; FX = foreign exchange; VIX = Chicago Board Options Exchange Volatility Index.

Figure 2.12. Policy Mitigating Factors

FX swap lines help reduce CIP deviations ... ... and lower excess exchange rate return volatility.
1. Mitigating Effect of FX Swap Lines on CIP Deviations after 2. Mitigating Effect of FX Swap Lines on Excess Exchange Rate Return
External Shocks Volatility after External Shocks
(Basis points) (Percentage points)
50 0.3
40 Baseline Baseline
Swap line interaction effect Swap line interaction effect
0.2
30
20
0.1
10
0 0
−10
−0.1
−20
−30
−0.2
−40
−50 −0.3
VIX Large VIX US EPU Large US EPU VIX Large VIX US EPU Large US EPU
shock shock shock shock

Sources: Bloomberg Finance L.P.; LSEG Datastream; and IMF staff calculations.
Note: The figure shows the effects of a one-standard-deviation increase in the VIX and the US EPU index and their associated uncertainty shocks on FX market conditions,
along with the mitigating effects of policy backstops like new central bank swap lines. Large uncertainty (VIX or US EPU) shocks are represented as dummy variables equal
to 1 when the first-order autoregression residuals of the underlying indices are two standard deviations above the average. Currencies in the sample with new swap lines
are the Danish krone, the Norwegian krone, the Singapore dollar, and the Swedish krona. Whiskers show the 90 percent confidence intervals. CIP = covered interest parity;
EPU = economic policy uncertainty index; FX = foreign exchange; VIX = Chicago Board Options Exchange Volatility Index.

International Monetary Fund | October 2025 63

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 2.13. Financial Spillovers of Foreign Exchange Market Stress

FX funding stress can tighten financial ... particularly when currency mismatches are Fiscal vulnerabilities can amplify the spillover of
conditions ... large. FX funding stress onto sovereign bond yields.
1. Effect of Cross-Currency Bases on 2. Effect of Cross-Currency Bases on Financial 3. Public Debt and the Effect of Cross-Currency
Financial Conditions Conditions with FX Mismatches Bases on Five-Year Sovereign Bond Yields
(Standard deviations) (Standard deviations) (Basis points)

1.0 Confidence interval High FX imbalances 4 High debt to GDP 75

Estimate Low FX imbalances Low debt to GDP
0.8 3 50

25
0.6 2
0
0.4 1
−25
0.2 0 −50

0 −1 −75
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
Weeks Weeks Weeks

Sources: Bloomberg Finance L.P.; CLS Group; LSEG Datastream; and IMF staff calculations.
Note: Panel 1 shows estimates from regressions of financial conditions on the cross-currency bases of local currencies against the US dollar. Panel 2 shows estimates from
regressions of financial conditions on the cross-currency bases and on the cross-currency bases interacted with a dummy variable equal to 1 when FX mismatches in a
country are above the sample median for the period. Panel 3 shows estimates from regressions of five-year local currency sovereign bond yields on the cross-currency
bases and on the cross-currency bases interacted with a dummy variable that takes the value 1 when the debt-to-GDP ratio of a country is above the sample median for the
period. In all panel regressions, a granular instrumental variables approach is used in which the cross-currency bases are instrumented with variables that capture
idiosyncratic shocks to demand for dollar funding in the FX swap market for three different tenors (less than 7 days, between 7 and 35 days, and more than 35 days). The
cross-currency basis and the idiosyncratic demand shocks are standardized for each currency and tenor. The shaded areas represent 90 percent confidence intervals,
obtained using Driscoll-Kraay standard errors, with the number of lags equal to 4√T , in which T denotes the number of time periods in the sample. The specifications include
time and currency effects. The currencies in the sample are the euro, the Japanese yen, the British pound, the Swiss franc, the Canadian dollar, the Australian dollar, the
New Zealand dollar, the Swedish krona, and the Norwegian krone, trading against the US dollar. FX = foreign exchange. See Online Annex 2.6 for further details.

Spillovers of Foreign Exchange Market FX market stress tightens overall financial condi-
Stress to Other Asset Classes tions. Shocks that widen cross-currency bases drive
down risky asset prices, tightening aggregate financial
Stress in the global FX market, given its size and
conditions. A one-standard-deviation widening tight-
deep linkages with other financial markets, can
ens financial conditions by 0.4 to 0.7 standard devi-
trigger cross-market spillover effects. To identify these
ations over the following year. This is a sizable effect,
spillover effects, the analysis employs panel regres-
about half the tightening observed during the dash-for-
sions and a granular instrumental variables approach.
cash episode induced by COVID-19 in March 2020,
The results show that a widening of cross-currency
and highlights the potential systemic implications of
bases triggers a flight-to-quality, which compresses
FX market dislocations for credit spreads, equity prices,
local currency sovereign bond yields and reduces
and funding costs (Figure 2.13, panel 1).
stock prices (Online Annex Figure 2.6.1). Specif-
The transmission of FX market shocks to broader
ically, a one-standard-deviation widening (about
financial conditions is amplified by vulnerabilities,
25 basis points) reduces longer-term sovereign bond
such as currency mismatches on the balance sheets
yields by about 25 basis points, with effects lasting
of financial institutions or elevated public debt
up to three months.42 Shorter-term yields fall even
(Figure 2.13, panel 2). In economies with a low level
more sharply, reflecting increased demand for less
of FX mismatches, the effect of a cross-currency basis
interest-rate-sensitive assets.
widening is negligible because the spillover effects
of an increased cost of funding or hedging using FX
swaps are likely to be smaller. In contrast, countries
42This is an economically meaningful effect, considering, for
with a high level of mismatches experience a tightening
example, that the five-year US Treasury yield fell by about 50 basis
points in the week following the collapse of Silicon Valley Bank in of financial conditions by as much as two standard
March 2023. deviations following a one-standard-deviation shock

64 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

to cross-currency bases. Fiscal vulnerabilities also play address vulnerabilities and mitigate associated risks,
a role: the effect of cross-currency basis widening on policy actions could focus on three key areas.
five-year sovereign bond yields is greater for economies
with high public debt relative to GDP (Figure 2.13,
panel 3), consistent with flight to quality favoring Strengthening Surveillance to Monitor Systemic
fiscally sound economies.43 Risk Arising from FX Market Stress
These findings underscore the systemic importance Although stress testing and systemic risk monitoring
of FX markets, in which stress can transmit into have advanced, the role of FX markets as a conduit for
tighter financial conditions, exacerbating downside tail risk transmission and cross-border spillovers remains
risks to real GDP growth and threatening macrofi- underappreciated. A more structured surveillance
nancial stability (Adrian, Boyarchenko, and Giannone approach is needed to better capture FX market vulner-
2019; October 2024 Global Financial Stability Report). abilities and their potential to disrupt macrofinancial
stability. Enhancing FX liquidity stress tests is essential
Conclusion and Policy Recommendations to assess the sectoral resilience to funding shocks and
sudden tightening in spot and swap market conditions.
The global FX market has expanded significantly Systemwide stress tests should incorporate scenarios
over time. This growth has been accompanied by nota- involving heightened volatility, as well as wider bid-ask
ble structural changes, including a rising presence of spreads and cross-currency bases, while factoring in
NBFIs and increased reliance on FX swaps for liquidity FX market vulnerabilities, to assess how FX market
management and currency risk hedging. These devel- disruptions could transmit across the financial system.
opments present opportunities and challenges for Monitoring and mitigating rollover and liquidity risks
market resilience and policy frameworks. This chap- from short-tenor FX swap positions, which are widely
ter’s analysis shows that the global FX market reacts used for funding and hedging, is also essential, as these
strongly to macrofinancial shocks despite its deep can amplify stress during market disruptions.
liquidity. Heightened risk aversion tends to increase Scenario analysis is crucial to evaluate the impact
demand for safe assets, straining FX market and of operational disruptions on FX market functioning
funding liquidity conditions, particularly in emerging and broader financial stability. The scenarios employed
markets. Structural vulnerabilities, such as high dealer in such analysis should include severe and persistent
concentration, the growing role of NBFIs, and inten- technical failures in primary trading venues and critical
sified FX hedging and funding pressures, can amplify payment systems, cyberattacks, physical disasters,
these effects. Moreover, operational disruptions in FX and defaults by major FX dealers, and consider the
market infrastructure can restrict trading and impair availability of contingency measures. Cyberattacks, in
liquidity, which exacerbates market stress. Given the particular, pose growing risks, with the potential to
central role of FX markets in the financial system, impair liquidity, delay settlements, and trigger systemic
such stress can spill over into debt and equity markets, stress across markets (April 2024 Global Financial
tightening overall financial conditions and posing risks Stability Report).
to macrofinancial stability. Closing data gaps is essential to strengthen the mon-
The findings also highlight that FX settlement risk itoring of FX market risks. The decentralized nature
remains a material concern, particularly for economies of FX trading makes comprehensive and timely data
that cannot access robust risk mitigation infrastructure. collection a persistent challenge. Key gaps in data avail-
The adoption of simultaneous settlement systems, ability include limited visibility into bilateral exposures,
such as PvP platforms, significantly reduces excess FX settlement practices, intraday trading, and counterparty
returns and volatility and can thereby reduce settle- concentrations, particularly for transactions carried on
ment uncertainty and currency risk premiums. outside centralized infrastructures like PvP systems.
A shifting global macrofinancial landscape under- Much of this information is held privately by NBFIs or
scores the need to strengthen FX market resilience. To embedded in bilateral dealer relationships not sub-
ject to reporting requirements. Addressing these gaps
43Similar results for sovereign bond yields are obtained when fiscal
through enhanced regulatory reporting and improved
vulnerability is proxied by sovereign credit default swap spreads data sharing is essential to strengthen surveillance and
(Online Annex 2.6). support resilient FX markets.

International Monetary Fund | October 2025 65

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Ensuring Adequate Capital and Liquidity Buffers Financial Market Infrastructures (BIS-CPSS-IOSCO
at Financial Institutions, Supported by a Robust 2012), financial market infrastructures should
Crisis Management Framework identify plausible sources of operational risk and
implement robust systems, policies, and procedures
Regulatory and supervisory agencies should ensure
to ensure high reliability. This implementation
that financial institutions with a dominant and sys-
should include comprehensive business continuity
temic role in FX markets maintain adequate hedges
planning, cyber resilience frameworks, and regular
and capital and liquidity buffers. Strengthening access
testing of contingency arrangements. Given the
to intraday central bank liquidity and credit facilities,
substantial netting efficiency provided by settlement
including for NBFIs, alongside stronger regulatory and
and clearing systems, a prolonged failure could
supervisory oversight to limit moral hazard (April 2023
have serious effects, such as preventing participants
Global Financial Stability Report), can help prevent
from accessing additional liquidity they may require
payment gridlocks during market stress.44 Moreover,
to fulfill payment obligations and capital needed
supervisors and banks should effectively monitor and
to cover potential counterparty losses.45 Financial
manage liquidity risks in significant currencies.
institutions, likewise, should adopt comprehensive
Economies relying heavily on external financing
operational risk management practices that address
should maintain sufficient international reserve buffers
vulnerabilities in technology, processes, and third-
to safeguard against external shocks (IMF 2016).
party dependencies (April 2024 Global Financial
Strengthening and expanding the network of cen-
Stability Report). Given the interconnected nature
tral bank swap lines can enhance global FX liquidity
of FX markets, disruptions in one jurisdiction can
backstops and help reduce contagion risks. Notably,
affect counterparties globally, underscoring the need
the IMF’s lending toolkit plays a vital role within the
for coordinated responses among central banks and
broader global financial safety net, offering tailored
agencies with oversight responsibility for financial
instruments to support countries facing FX liquidity
market infrastructures.
pressures during adverse shocks.
Reducing FX settlement risks requires wider
Managing systemic risk arising from stress in FX
adoption of PvP arrangements. In the interim,
markets may also require a policy action mix in line
dealer banks can strengthen risk controls through
with the IMF’s Integrated Policy Framework. This
alternative arrangements, including “pre-settlement
is particularly important when external shocks lead
netting,” which reduces settlement risk by bilaterally
to undesirable macroeconomic fluctuations, particu-
offsetting gross obligations, and “on-us” settlement,
larly in the presence of significant FX mismatches on
in which both legs of FX trades are settled within
private sector balance sheets or shallow FX markets,
the same institution, thereby mitigating counter-
as reflected in excess exchange rate return volatility,
party exposure (BIS 2025a). Strong anti-money
or wider bid-ask spreads and CIP premiums. The
laundering/combating the financing of terrorism
response may include FX intervention and macro-
measures should also be implemented to reduce
prudential and capital flow management measures
uncertainty in settlement. Policy initiatives that
calibrated to country-specific conditions (IMF 2023).
leverage digital technologies, if properly designed,
offer promising avenues for helping to address set-
Adequate Management of Operational and tlement risk and increase the safety and efficiency of
Settlement Risk cross-border payments. These include linking faster
payment systems or developing cross-border central
Strengthening the operational resilience of finan-
bank digital currency, as explored in recent Bank for
cial market infrastructures is critical to safeguarding
International Settlements initiatives (IMF 2024a).46
FX market stability. In line with the Principles of

44Effective oversight of NBFIs is essential to limit moral hazard 45For example, CLS’s netting process typically reduces funding

and reduce central banks’ exposure to collateral and credit risks. requirements by approximately 96 percent (CLS Group 2025).
Liquidity support facilities should be temporary, collateralized with 46The cross-border use of FX-denominated central bank digital

appropriate haircuts, and priced at a penalty rate, to safeguard finan- currency, if such currency is not properly designed and regulated,
cial stability. These measures should align with the Financial Stability could displace domestic currencies, weaken monetary policy control,
Board’s recommendations to address vulnerabilities in NBFIs, and heighten vulnerability to shocks (IMF 2023; October 2024
including liquidity mismatches and leverage (FSB 2025). Global Financial Stability Report, Chapter 3).

66 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Finally, FX markets could lower transaction costs and dealer constraints in over-the-counter markets
and volatility by migrating toward well-­designed (Adrian and Mancini-Griffoli 2023; IMF 2024b).47
financial platforms, which have the potential to
reduce counterparty and settlement risks, as well 47“Financial platforms” refers to infrastructures that facilitate the

as costs associated with information asymmetry interoperability of financial assets.

International Monetary Fund | October 2025 67

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 2.1. Foreign Exchange Market Dynamics around the April US Tariff Announcement
On April 2, 2025, the United States announced Figure 2.1.1. Broad US Dollar Index and VIX,
increased tariff rates on imports, marking a major January 2, 2025, to May 15, 2025
policy shift with potential implications for global (Index)
trade and investment. The announcement triggered
a sharp reaction in financial markets: measures of VIX Nominal Broad USD Index (right scale)
financial uncertainty like the Chicago Board Options 60 April 2 tariff 132
Exchange Volatility Index spiked, and the broad US announcement
dollar index depreciated by about 2 percent on impact 128
(Figure 2.1.1). Although market volatility eased after 40
the suspension of some tariffs on April 9, the dol- 124
lar continued to weaken. Overall, despite the large
magnitude of the shock, foreign exchange (FX) market 20
120
conditions remained broadly orderly, with no major
disruptions observed. Using data from the CLS FX
settlement system, this box examines how FX market 0 116
Jan. Feb. Mar. Apr. May
dynamics evolved around the announcement, focusing 2025 2025 2025 2025 2025
on changes in spot and swap dollar trading volumes
across countries and sectors. Sources: Federal Reserve Bank of St. Louis, Federal Reserve
Spot dollar purchases by nonresident investors Economic Data; and IMF staff calculations.
rose notably ahead of April 2, increasing by about Note: USD = US dollar; VIX = Chicago Board Options Exchange
Volatility Index.
$265 billion on a net basis between January 1 and
April 1. Following the tariff announcement, purchases
continued to rise through mid-April but have since April 2 tariff announcement (Figure 2.1.2, panel 4).2
declined (Figure 2.1.2, panel 1). As of the end of May, Compared with what took place after the COVID-19
cumulative net spot purchases remained broadly stable. shock, hedging demand from these investors—which
Cross-country differences are evident in trading involves selling US dollar forward contracts—has been
patterns. For example, Canada was a net buyer of spot stronger and more persistent.3 Although the overall
dollars from November 2024 through mid-April 2025 cumulative change in swap positions has been only
but shifted to net selling thereafter. Similarly, spot slightly larger than that of the COVID-19 episode,
dollar sales on a net basis by major euro area countries combined with the muted net spot dollar purchases,
increased after April 2. Across sectors, most of non-US this may have contributed to US dollar depreciation
institutions’ activity has been driven by nonbank pressure during April and May.4
financial institutions around the tariff episode, which
2Cumulative swap flows in panel 4 of Figure 2.1.2 should not
contrasts with what took place in previous macrofi-
be interpreted as net mark-to-market positions. This is because
nancial shocks, such as the COVID-19 market turmoil
they do not account for maturity and refinancing activities. In
in March 2020, when banks dominated FX trading addition, each flow is recorded using the forward rate fixed at the
(Figure 2.1.2, panels 2 and 3).1 time of the contract, without incorporating after-the-fact valua-
The FX swap activity of non-US nonbank inves- tion changes that are due to the shifts in market exchange rates.
3Swap dollar flows are positive for non-US banks, indicating
tors against the US dollar increased notably after the
that they sell dollar hedges, that is, they buy dollar forward
contracts. A similar pattern holds for US banks, which are net
hedge sellers, whereas US nonbank institutions are net buyers of
1Net purchases of other safe haven currencies, such as the FX hedges.
euro and Japanese yen, rose notably after April 2, exceeding 4Note that a shift in trading activity in other sectors, such as

levels observed during the COVID-19 turmoil (Online Annex the official sector, or by other institutions not captured by the
Figure 2.7.2). Moreover, as Canada became net sellers of US CLS settlement data may also have contributed to US dollar
dollars during this period, the country appeared to shift toward depreciation pressures following April 2 (see, for example, Jiang
the euro, and major euro area countries moved toward the yen. and others 2025).

68 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Box 2.1 (continued)

Figure 2.1.2. Net Spot US Dollar Flows before and after the US Tariff Announcement on April 2, 2025
(Billions of dollars)

Net purchases of US dollars increased in the first quarter of Non-US nonbank institutions have been active buyers of US
2025 but have stabilized since then. dollars in the spot market.
1. Cumulative Net US Dollar Spot Flows, by Purchaser 2. Cumulative Net US Dollar Spot Flows, by Sector,
Nationality, 2024:M11–2025:M5 2024:M11–2025:M5
600 USA CAN Non-US bank 400
Tariff Tariff
EA (DE, FR, IT) EA other Non-US nonbank
400 JPN GBR US bank
ROW Total non-US US nonbank 200
Total non-US
200

0 0

−200
−200
−400

−600 −400
Nov. 4 Dec. 3 Jan. 3 Feb. 3 Mar. 4 Apr. 2 May 1 Nov. 4 Dec. 3 Jan. 3 Feb. 3 Mar. 4 Apr. 2 May 1
2024 2025 2024 2025

Spot US dollar purchases by non-US investors decreased more ... although hedging demand has been higher.
after the April US tariff announcement than they did during
the COVID-19 episode ...
3. Cumulative Net US Dollar Spot Flows of 4. Cumulative Net US Dollar Swap Flows of
Non-US Institutions Non-US Institutions
400 t − 6w t − 6w 2,000
t − 2w t − 2w
t+0 t+0
t + 2w t + 2w 1,000
300
t + 6w t + 6w

200 0

100 −1,000
Buying USD hedge

0 −2,000
Bank Non- Non-US Bank Non- Non-US Bank Non- Non-US Bank Non- Non-US
bank total bank total bank total bank total
COVID-19 shock April 2025 US tariff COVID-19 shock April 2025 US tariff
(t = Mar. 9, 2020) (t = Apr. 3, 2025) (t = Mar. 9, 2020) (t = Mar. 31, 2025)

Sources: CLS Group; and IMF staff calculations.

Note: In panel 1, residency is based on the institutions participating in CLS Settlement System. Panels 3 and 4 present net US dollar spot
and swap flows around two major shocks—COVID-19 and the April US tariff announcement—denoted as time t, with t − h and t + h
representing h weeks before and after the event, respectively (h = 2 and 6). For COVID-19, t is set as March 9, the date of a sharp stock
market decline amid broader financial turmoil triggered by pandemic fears. In panel 3, for the US tariff announcement shock, t is defined as
April 3, reflecting the timing of the US tariff announcement after market close in several major economies. In panel 4, t is set as March 31 to
align with the weekly frequency of FX swap flow data; a positive bar indicates that the institutional sector is a net buyer of US dollars in the
forward leg of swap contracts and a net seller of US dollars in the near leg of swap contracts; US dollar swap flows with tenors of less than
35 days are excluded from panel 4. Country names are represented by International Organization for Standardization (ISO) three-digit
country codes except EA (DE, FR, IT) = euro area (Germany, France, and Italy); EA other = euro area except Germany, France, and Italy;
ROW = rest of the world.

International Monetary Fund | October 2025 69

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 2.2. The Relevance of Settlement Risk in Foreign Exchange Markets

Foreign exchange (FX) settlement risk, often Figure 2.2.1. Effect of CLS Entry on Excess
referred to as Herstatt risk, arises when one party Foreign Exchange Returns and Volatility
in a currency trade delivers the currency it sold but (Basis points)
fails to receive the currency it bought. Such risk is
significant because such failures can trigger liquidity 0
pressures, credit losses, and even systemic disruptions,
especially during periods of market stress. Whereas −10
payment infrastructure and risk mitigation tools have
improved over the years, settlement risk remains a −20
concern, particularly in emerging market and devel-
oping economies. These jurisdictions often lack access
−30
to robust simultaneous settlement mechanisms like
payment-versus-payment (PvP) systems, rely more
on correspondent banking relationships that intro- −40
Excess Volatility Excess Volatility
duce additional layers of counterparty exposure and returns returns
operational complexity, and operate within payment Hungary’s access to CLS Panel analysis
and legal frameworks that may not align with global (2002–25)
FX settlement arrangements. This box reviews key
Sources: Bloomberg Finance L.P.; Haver Analytics; LSEG
historical episodes, outlines the evolution of mitigation Datastream; and IMF staff calculations.
strategies, and assesses the relevance of settlement risk Note: The figure shows the difference-in-difference impact of
for countries without access to global PvP platforms Hungary’s entry into the CLS system on excess foreign exchange
like the CLS system.1 rate returns and the associated volatility, using daily data for a
one-month window before and after the event. The analysis
Key Episodes and Risk Mitigation Frameworks includes relevant control variables, including country and time
fixed effects. Standard errors are clustered at the country level. The
The term “Herstatt risk” originates from the estimated effects are statistically significant at the 10 percent level
or lower.
failure of Bankhaus Herstatt on June 26, 1974,
when German regulators closed the bank after it had
received Deutsche mark payments from US counter-
parties but before it delivered the corresponding US
dollars later that day. The default triggered widespread in March 2020, Barclays suffered a $129 million FX
disruption in the multilateral net settlement system loss when its counterparty, UAE Exchange, failed to
and prompted the establishment of the Basel Commit- deliver the currency owed amid COVID-19-related
tee on Banking Supervision later that year. Settlement market stress. These incidents underscore enduring
risk has persisted well beyond this episode. In 2008, vulnerabilities in FX settlement, particularly during
Kreditanstalt für Wiederaufbau, a German state- periods of heightened market volatility or institutional
owned investment and development bank, transferred fragility.
€300 million (equivalent to $426 million at the time) In response to these vulnerabilities, the global finan-
to Lehman Brothers on the morning of Lehman’s cial community has progressively developed strategies
bankruptcy filing but never received the corresponding to mitigate FX settlement risk. A pivotal advancement
payment, resulting in a unilateral loss. More recently, was the establishment in 2002 of CLS, a multicur-
rency PvP system designed to ensure simultaneous
1FX settlement risk can be mitigated through two key
settlement of both legs of an FX transaction. CLS
significantly reduces counterparty risk by linking set-
approaches: presettlement netting, which reduces the amount to
be exchanged by offsetting obligations between counterparties, tlements across time zones and currencies and covers
and simultaneous settlement mechanisms, such as PvP or on-us a substantial share of global FX turnover (see Online
settlement. PvP ensures that each currency leg settles only if Annex 2.2 for further details). In parallel, international
the other does, eliminating settlement risk. In on-us settlement, standard-setting bodies, such as the Basel Committee
both legs are processed within the same institution. However,
on Banking Supervision, the Committee on Payments
protection against loss is assured only if settlement occurs simul-
taneously or within preauthorized credit lines; this is known as and Market Infrastructures–International Organization
“on-us with loss protection.” of Securities Commissions, and the Global Foreign

70 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Box 2.2 (continued)

Exchange Committee, have issued guidelines to regional currencies that did not join CLS at the time,
strengthen legal certainty, promote settlement finality, are used as benchmarks for comparison.2
and encourage broader adoption of PvP mechanisms. The analysis shows that CLS entry led to a signifi-
Notably, the FX Global Code, published in 2017 cant decline in the average daily excess exchange rate
under the auspices of the Bank for International Set- returns and volatility of the forint—the former by
tlements, sets out principles for good market conduct, about 11 basis points (bps) and the latter by 3 bps—
including robust risk management and settlement when the one-month periods before and after CLS
practices to enhance transparency and integrity in FX participation are compared (Figure 2.2.1). Notably,
markets. Many jurisdictions have also implemented the average excess returns over the month before CLS
real-time gross settlement systems and improved their accession were about 28 bps, suggesting that CLS
legal frameworks to support netting arrangements and participation eliminated this excess return, bringing
cross-border enforceability. Despite these advance- the currency closer to the value expected for it in
ments, about 25 percent of the deliverable turnover the forward market. These results suggest a mean-
of currencies is without risk mitigation mechanisms ingful reduction in market uncertainty and support
(Glowka and Nilsson 2022). the hypothesis that PvP systems help mitigate credit
risk in FX markets, thereby lowering settlement risk
PvP Systems and Currency Risk Premiums premiums.
In typical FX transactions, mismatches in global
time zones and payment system operating hours 2. Panel Analysis
expose participants to settlement risk, including the A number of currencies are settled through PvP
possibility of counterparty default. PvP systems like systems, such as CLS, the B3 Foreign Exchange
CLS eliminate such risk. By reducing settlement Clearinghouse in Brazil, the Clearing Corporation
uncertainty and counterparty exposure, PvP mech- of India Limited’s Forex Settlement, and the Hong
anisms could lower the settlement risk premiums Kong Special Administrative Region–based Clearing
investors may require for bearing settlement risk and House Automated Transfer System. To estimate the
reduce the volatility of such premiums. To exam- impact of PvP adoption, a panel regression analysis
ine these effects empirically, two complementary is conducted over January 2000 to May 2025. The
approaches are employed here: a natural experiment results support earlier findings, showing that CLS
based on Hungary’s accession to CLS in 2015 and participation is associated with a significant decline
a broader panel analysis covering 26 currencies over of 34 bps in excess FX returns and 3 bps in volatility,
2000–25, including 16 currencies settled through CLS on average (Figure 2.2.1). This reinforces the view
and 4 currencies with other PvP arrangements (those that PvP systems can contribute to global FX market
of Brazil, India, Malaysia, and Thailand). stability by lowering settlement risk and the associated
risk premiums.
1. Hungary’s Accession to CLS
2The koruna and zloty are selected as control group currencies
Hungary’s currency, the forint, joined the CLS
system on November 16, 2015. To assess the impact based on Czechia’s and Poland’s comparable macroeconomic
characteristics to those in Hungary, with the validity of this
of this change on the forint’s excess FX rate returns
selection supported by exchange rate correlation analysis and a
and volatility, a difference-in-difference analysis is parallel trends test conducted using one year of pre-CLS data.
conducted. The Czech koruna and Polish zloty, similar See Online Annex 2.8 for details on the empirical methodology.

International Monetary Fund | October 2025 71

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 2.3. Implications of Operational Disruptions in Foreign Exchange Markets

The resilience of foreign exchange (FX) markets is market, they provide a natural setting for isolating the
critical to global financial stability, given their central effects of operational disruptions, reducing the risk of
role in facilitating cross-border trade, investment, and confounding from broader financial market impair-
exchange rate determination. Although FX markets are ments (see Online Annex 2.9 for methodological
typically deep and liquid, they also depend on a con- details).
centrated set of trading venues and dealer banks for The analysis focusing on the currencies directly
execution and price formation.1 This reliance creates affected by the outages suggests a deterioration in
potential vulnerabilities: operational disruptions, such market liquidity across multiple dimensions. During
as technological failures, natural disasters, and cyber the outages,2 the cost of FX transactions, as measured
incidents, affecting these core infrastructures and inter- by bid-ask spreads, increased in the spot and swap
mediaries can impair market functioning and liquidity, markets (Figure 2.3.1, panel 1).3 An examination of
with spillovers to other financial markets, as discussed the volumes traded in the spot market on the days of
in the conceptual framework in this chapter’s text. the outages reveals a decline of $2.8 billion, on aver-
This box analyzes how operational disruptions can age, across the affected currencies.4 Further analysis
affect FX market conditions by estimating the effects suggests that a $1 billion decrease in trading volume is
of outages at two primary trading venues, Electronic associated with a 0.3 basis point widening of bid-ask
Broking Services (EBS) and London Stock Exchange spreads in the spot market, which could be considered
Group’s FX Matching, as a case study. an estimate of the causal effect of volumes on bid-ask
The core of the global FX network overlaps with spreads.5 Larger drops in trading volumes because
the interdealer market (Figure 2.5, panels 2 and 3), in of more severe incidents, such as prolonged outages
which transactions between dealer banks are facili- or simultaneous disruptions across multiple trading
tated by primary market venues, most notably EBS venues, could lead to more pronounced widenings of
and FX Matching. These platforms play a central role bid-ask spreads.
in price formation and liquidity provision. In 2023, The market for the affected currencies also
EBS experienced an outage; in 2015, trading on the becomes less liquid, as measured by the price impact
FX Matching platform was disrupted. Both outages of trading volume (Figure 2.3.1, panel 2; Ranaldo
occurred when the London and New York trading and Santucci de Magistris 2022). Economically,
sessions in the FX market overlapped, a period of high however, the effect is moderate: the trading volume
market liquidity (King, Osler, and Rime 2012). There required to move daily FX returns by one stan-
are no detailed accounts of the exact nature of the dard deviation (about 3.4 percent) declines from
disruptions, but publicly available information indi- $19.2 billion to $18.6 billion. The deterioration of
cates that both outages lasted for some time and likely market liquidity is also reflected in an increase of
had a meaningful impact on trading in the interdealer about 0.2 standard deviations (about 11 percent) in
market.
To examine the effect of these outages on FX mar- 2The effects of the outages on bid-ask spreads are quantified

ket conditions, key market liquidity indicators, such as over their assumed durations, deduced from publicly available
spot and forward bid-ask spreads, are analyzed across information (Mackenzie Smith 2015; Lambert 2023).
3The increase in bid-ask spreads for currencies directly affected
two dimensions—currencies that are primarily traded
by the outages suggests that disruptions to the interdealer market
on the affected platform versus those that are not, and can raise inventory holding costs for dealer banks, thereby
over time for all currencies—to assess the aggregate contributing to wider bid-ask spreads (Amihud and Mendelson
effects. Because the outages at EBS and FX Matching 1980).
4All dollar volumes in this box are adjusted for inflation using
directly affected the interdealer segment of the FX
the US Bureau of Labor Statistics Consumer Price Index and
expressed in December 2024 US dollars.
1Over the past two decades, the number and types of FX 5The relation between bid-ask spreads and trading volume

trading venues (for example, multidealer platforms, single-dealer is estimated using the outages as an instrumental variable for
platforms, electronic communication networks, and retail trading volume. The estimates are comparable with those
platforms) have increased notably. However, core liquidity and in Bessembinder (1994), which translate to an increase of
price discovery in the interdealer segment, which underpins 0.2–1.1 basis points for a decrease of $1 billion, in December
the broader FX market, remains concentrated in a few venues, 2024 US dollars, in forecastable futures trading volume for the
notably Electronic Broking Services and London Stock Exchange German mark, Japanese yen, British pound, and Swiss franc
Group’s FX Matching. between January 1979 and December 1992.

72 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Box 2.3 (continued)

the dispersion of transaction prices across counterpar- conditions. Nonetheless, the increase in transaction
ties, suggesting that some must accept less favorable costs is economically moderate, with bid-ask spreads
terms of trade. across all major currency pairs against the US dollar
Further analysis of the aggregate effects of the widening, on average, from 3 to 4 basis points in the
outages indicates that transaction costs in the spot spot market and from 4 to 5 basis points in the swap
and swap markets rose significantly across all curren- market during the outages.
cies, including those mainly traded on venues that Together, the results suggest that even relatively
remained operational (Figure 2.3.1, panel 2). This short-lived outages of trading platforms can materi-
suggests the presence of liquidity spillovers across ally affect FX market liquidity. This underscores the
market venues, which is possibly due to the migra- importance of resilient infrastructures and intermedi-
tion of trading from the affected platforms to other aries in safeguarding against more severe operational
venues, leading to congestion and strained liquidity disruptions that could pose risks to financial stability.

Figure 2.3.1. Effect of Interdealer Platform Disruption on Market Liquidity

(Standard deviations)

Outages on interdealer platforms reduce liquidity for affected ... and raise transaction costs for those traded on other
currencies ... platforms.
1. Effect of FX Interdealer Platform Disruption for Currencies 2. Effect of FX Interdealer Platform Disruption for
Primarily Traded on the Platform All Currencies
0.4 1.0

0.9

0.8
0.3
0.7

0.6

0.2 0.5

0.4

0.3
0.1
0.2

0.1

0 0
Spot bid-ask Forward bid- Realized Price Spot bid-ask Forward bid-ask
spread ask spread illiquidity dispersion spread spread

Sources: Bloomberg Finance L.P.; CLS Group; and IMF staff calculations.
Note: The bars in the figure represent estimated coefficients from panel regressions of the outcome variables on an indicator variable for
platform outage. In panel 1, the indicator variable is equal to 1 during a platform outage only for the currencies traded primarily on the
platform and 0 otherwise. In panel 2, the indicator variable is equal to 1 during a platform outage for all currencies. “Realized illiquidity,”
defined as in Ranaldo and Santucci de Magistris (2022), refers to the ratio of the realized absolute variation of intraday returns to the volume
of transactions in billions of US dollars and measures the price impact of trading volume. Price dispersion is the coefficient of variation of
transaction prices for each pair of currencies traded across different counterparty sectors. Bid-ask spreads are sampled at 30-minute
intervals, whereas the other measures are constructed at a daily frequency. The sample period covers the day of each outage as well as
90 days before and after. All the measures are standardized separately in each of the two 181-day windows and for each currency traded on
the platform. The currencies in the sample are the euro, the Japanese yen, the British pound, the Swiss franc, the Canadian dollar, the
Australian dollar, the New Zealand dollar, the Swedish krona, and the Norwegian krone, trading against the US dollar. The specifications in
panel 1 include time and currency-year effects, and those in panel 2 include currency-year effects. The specifications for the bid-ask spreads
also include currency–time of day–year effects. The error bars represent 90 percent confidence intervals, obtained using Driscoll-Kraay
4
standard errors, with the number of lags equal to √ T , in which T denotes the number of time periods in the sample. FX = foreign exchange.

International Monetary Fund | October 2025 73

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

References Barajas, Adolfo, Andrea Deghi, Claudio Raddatz, Dulani

Adrian, Tobias, Nina Boyarchenko, and Domenico Giannone. Seneviratne, Peichu Xie, and Yizhi Xu. 2020. “Global Banks’
2019. “Vulnerable Growth.” American Economic Review 109 Dollar Funding: A Source of Financial Vulnerability.” IMF
(4): 1263–89. Working Paper 20/113, International Monetary Fund,
Adrian, Tobias, and Tommaso Mancini-Griffoli. 2023. “The Rise Washington, DC.
of Payment and Contracting Platforms.” IMF Fintech Note Berger, David, Alain Chaboud, and Erik Hjalmarsson. 2009.
23/005, International Monetary Fund, Washington, DC. “What Drives Volatility Persistence in the Foreign Exchange
Adrian, Tobias, and Hyun Song Shin. 2014. “Procyclical Lever- Market?” Journal of Financial Economics 94 (2): 192–213.
age and Value-at-Risk.” Review of Financial Studies 27 (2): Bessembinder, Hendrik. 1994. “Bid-Ask Spreads in the
373–403. Interbank Foreign Exchange Markets.” Journal of Financial
Aizenman, Joshua, Yothin Jinjarak, Donghyun Park, and Economics 35 (3): 317–48.
Huanhuan Zheng. 2021. “Good-Bye Original Sin, Hello Risk Borio, Claudio, Robert McCauley, and Patrick McGuire. 2022.
On-Off, Financial Fragility, and Crises?” Journal of Interna- “Dollar Debt in FX Swaps and Forwards: Huge, Missing, and
tional Money and Finance 117: 102442. Growing.” BIS Quarterly Review (December): 67–73.
Aldasoro, Iñaki, Wenqian Huang, and Nikola Tarashev. 2021. Borio, Claudio, Robert McCauley, Patrick McGuire, and
“Asset Managers, Market Liquidity and Bank Regulation.” Vladyslav Sushko. 2016. “Covered Interest Parity Lost:
BIS Working Paper 933, Bank for International Settlements, Understanding the Cross-Currency Basis.” BIS Quarterly
Basel, Switzerland. Review (September): 45–64.
Amihud, Yakov, and Haim Mendelson. 1980. “Dealership Bräuer, Leonie, and Harald Hau. 2023. “Can Time-Varying
Market: Market-Making with Inventory.” Journal of Financial Currency Risk Hedging Explain Exchange Rates?” CEPR
Economics 8 (1): 31–53. Discussion Paper No. 18516, CEPR Press, Paris and
Aramonte, Sirio, and Fernando Avalos. 2021. “The Rise of Pri- London.
vate Markets.” BIS Quarterly Review 37 (December): 69–82. Bruno, Valentina, and Hyun Song Shin. 2015. “Capital Flows
Bahaj, Saleem, and Ricardo Reis. 2022. “Central Bank Swap and the Risk-Taking Channel of Monetary Policy.” Journal of
Lines: Evidence on the Effects of the Lender of Last Resort.” Monetary Economics 71: 119–32.
Review of Economic Studies 89 (4): 1654–93. Caballero, Ricardo, and Arvind Krishnamurthy. 2008. “Col-
Baker, R. Scott, Nicholas Bloom, and Steven J. Davis. 2016. lective Risk Management in a Flight to Quality Episode.”
“Measuring Economic Policy Uncertainty.” Quarterly Journal NBER Working Paper 12896, National Bureau of Economic
of Economics 131 (4): 1593–636. Research, Cambridge, MA.
Bank for International Settlements and International Chaboud, Alain, Lisa Chung, Linda S. Goldberg, and Anna
Organization of Securities Commissions. 2012. “Principles Nordstrom. 2024. “Towards Increasing Complexity: The
for Financial Market Infrastructures.” Committee on Evolution of the FX Market.” Liberty Street Economics (blog),
Payment and Settlement Systems and Technical Committee January 11, 2024.
of IOSCO, Basel, Switzerland. Chaboud, Alain, Dagfinn Rime, and Vladyslav Sushko. 2023.
Bank for International Settlements (BIS). 2020. “FX Execution “The Foreign Exchange Market.” In Research Handbook
Algorithms and Market Functioning.” Markets Committee of Financial Markets, edited by Refet S. Gurkaynak and
Papers, BIS, Basel, Switzerland. Jonathan H. Wright, 253–75. Cheltenham, UK: Elgar.
Bank for International Settlements (BIS). 2022. BIS Quarterly CLS Group. 2025. “Settlement.” CLS Group, https://www.cls-
Review (December). https://www.bis.org/publ/qtrpdf/r_ group.com/products/settlement/.
qt2212.htm. Dao, Mai, and Pierre-Olivier Gourinchas. 2025. “Covered
Bank for International Settlements (BIS). 2025a. “Triennial Interest Parity Deviations in Emerging Markets: Measurement
Central Bank Survey: Preliminary Results of OTC Foreign and Drivers.” IMF Working Paper 25/057, International
Exchange Turnover in April 2025.” Monetary and Economic Monetary Fund, Washington, DC.
Departments, BIS, Basel, Switzerland. Dao, Mai, Pierre-Olivier Gourinchas, and Oleg Itskhoki. 2025.
Bank for International Settlements (BIS). 2025b. Financial “Breaking Parity: Equilibrium Exchange Rates and Currency
Conditions in a Changing Global Financial System. BIS Annual Premia.” IMF Working Paper 25/153, International Monetary
Economic Report, Basel, Switzerland. Fund, Washington, DC.
Barajas, Adolfo, Andrea Deghi, Salih Fendoglu, and Yizhi Xu. Devereux, Michael B., and Changhua Yu. 2020. “International
2020. “Strains in Offshore US Dollar Funding during the Financial Integration and Crisis Contagion.” Review of Eco-
COVID-19 Crisis: Some Observations.” Global Finan- nomic Studies 87 (3): 1174–212.
cial Stability Note 20/001, International Monetary Fund, Du, Wenxin, and Jesse Schreger. 2016. “Local Currency Sover-
Washington, DC. eign Risk.” Journal of Finance 71 (3): 1027–70.

74 International Monetary Fund | October 2025

 CHAPTER 2 Risk and Resilience in the Global Foreign Exchange Market

Du, Wenxin, and Jesse Schreger. 2022. “CIP Deviations, the International Monetary Fund (IMF). 2016. “Guidance Note
Dollar, and Frictions in International Capital Markets.” In on the Assessment of Reserve Adequacy and Related
Handbook of International Economics, vol. 6: 147–97. Considerations.” IMF Policy Paper, Washington, DC,
Du, Wenxin, Alexander Tepper, and Adrien Verdelhan. 2018. June 3, 2016. https://www.imf.org/en/Publications/
“Deviations from Covered Interest Rate Parity.” Journal of Policy-Papers/Issues/2016/12/31/Guidance-Note-on-the
Finance 73 (3): 915–57. -Assessment-of-Reserve-Adequacy-and-Related-Consider-
Duffie, Darrell. 2023. “Resilience Redux in the US Treasury ations-PP5046.
Market.” Paper presented at the Federal Reserve Bank of International Monetary Fund (IMF). 2023. “Integrated
Kansas City’s Jackson Hole Symposium, Jackson Hole, Policy Framework—Principles for the Use of Foreign
Wyoming, August 25. Exchange Intervention.” IMF Policy Paper 23/061,
Eguren-Martin, Fernando, Mattias O. Busch, and Dennis Washington, DC.
Reinhardt. 2024. “Global Banks and Synthetic Funding: The International Monetary Fund (IMF). 2024a. “Central Bank
Benefits of Foreign Relatives.” Journal of Money, Credit, and Digital Currency: Progress and Further Considerations.” IMF
Banking 56 (1): 115–52. Policy Paper 24/052, Washington, DC.
Financial Stability Board (FSB). 2022. “US Dollar Funding and International Monetary Fund (IMF). 2024b. “G-20 Note on
Emerging Market Economy Vulnerabilities.” Report to the Financial Platforms: What Are They and What Are Their
Group of Twenty, FSB, Basel, Switzerland. https://www.fsb Macro-financial Implications?” IMF, Washington DC.
.org/2022/04/us-dollar-funding-and-emerging-market Jiang, Zhengyang, Arvind Krishnamurthy, Hanno N. Lustig,
-economy-vulnerabilities. Robert Richmond, and Chenzi Xu. 2025. “Dollar Upheaval:
Financial Stability Board (FSB). 2025. “Leverage in Nonbank This Time Is Different.” Working Paper, Northwestern
Financial Intermediation: Final Report.” Policy document, University.
FSB, Basel, Switzerland. https://www.fsb.org/2025/07/ Khiaonarong, Tanai, Harry Leinonen, and Ryan Rizaldy. 2021.
leverage-in-nonbank-financial-intermediation-final-report. “Operational Resilience in Digital Payments: Experiences and
Gabaix, Xavier, and Matteo Maggiori. 2015. “International Issues.” IMF Working Paper 21/288, International Monetary
Liquidity and Exchange Rate Dynamics.” Quarterly Journal of Fund, Washington, DC.
Economics 130 (3): 1369–420. King, Michael R., Carol Osler, and Dagfinn Rime. 2012.
Glowka, Marc, and Thomas Nilsson. 2022. “FX Settlement Risk: “Foreign Exchange Market Structure, Players, and
An Unsettled Issue.” BIS Quarterly Review (December): 75–81. Evolution.” In Handbook of Exchange Rates, edited by
Greenwood, Robin, Samuel Hanson, Jeremy C. Stein, and Adi Jessica James, Ian W. Marsh, and Lucio Sarno, 3–44.
Sunderam. 2023. “A Quantity-Driven Theory of Term Premi- Hoboken, NJ: Wiley.
ums and Exchange Rates.” Quarterly Journal of Economics 138 Kloks, Peteris, Edouard Mattille, and Angelo Ranaldo. 2023.
(4): 2327–89. “Foreign Exchange Swap Liquidity.” Research Paper 23-22,
Grothe, Magdalena, Peter McQuade, Martino Ricci, and Luca Swiss Finance Institute, Zurich, Switzerland.
Tondo. 2025. “Recent Patterns in Global Risk Behaviour in Kloks, Peteris, Edouard Mattille, and Angelo Ranaldo. 2024.
Financial Markets.” Column, voxEU, Center for Economic “Hunting for Dollars.” Research Paper 24-52, Swiss Finance
Policy Research, August 12, 2025. Institute, Zurich, Switzerland.
He, Zhiguo, Bryan Kelly, and Asaf Manela. 2017. “Intermedi- Kloks, Peteris, Patrick McGuire, Angelo Ranaldo, and Vladyslav
ary Asset Pricing: New Evidence from Many Asset Classes.” Sushko. 2023. “Bank Positions in FX Swaps: Insights from
Journal of Financial Economics 126 (1): 1–35. CLS.” BIS Quarterly Review (September): 17–31.
Hofmann, Boris, Ilhyock Shim, and Hyun Song Shin. 2020. Kubitza, Christian, Jean-David Sigaux, and Quentin
“Bond Risk Premiums and the Exchange Rate.” Journal of Vandeweyer. 2025. “The Implications of CIP
Money, Credit and Banking 52 (S2): 497–520. Deviations for International Capital Flows.” Working
Hossain, Ashrafee T., Abdullah-Al Masum, and Samir Saadi. Paper 3017, European Central Bank, Frankfurt am Main,
2024. “The Impact of Geopolitical Risks on Foreign Germany.
Exchange Markets: Evidence from the Russia–Ukraine War.” Kumar, Abhishek, Sushanta Mallick, Madhusudan Mohanty, and
Finance Research Letters 59 (January): 104750. Fabrizio Zampolli. 2023. “Market Volatility, Monetary Policy
Huang, Wenqian, Angelo Ranaldo, Andreas Schrimpf, and Fabri- and the Term Premium.” Oxford Bulletin of Economics and
cius Somogyi. 2025. “Constrained Liquidity Provision in Cur- Statistics 85 (1): 208–37.
rency Markets.” Journal of Financial Economics 167: 104028. Lambert, Colin. 2023. “What Happens When a Primary FX
Hui, Hon Chung. 2021. “The Long-Run Effects of Geopolitical Venue Goes Offline?” The Full FX, October 4, 2023.
Risk on Foreign Exchange Markets: Evidence from Some Liao, Gordon Y., and Tony Zhang. 2025. “The Hedging
ASEAN Countries.” International Journal of Emerging Markets Channel of Exchange Rate Determination.” Review of Finan-
17 (6): 1543–64. cial Studies 38 (1): 1–38.

International Monetary Fund | October 2025 75

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Mackenzie Smith, Robert. 2015. “Thomson Ranaldo, Angelo, and Paolo Santucci de Magistris. 2022.
Reuters Matching Suffers Outage.” FX Markets, “Liquidity in the Global Currency Market.” Journal of Finan-
July 1, 2015. https://www.fx-markets.com/technology/ cial Economics 146 (3): 859–83.
trading-systems/2415705/thomson-reuters-matching Ranaldo, Angelo, and Fabricius Somogyi. 2021. “Asymmetric
-suffers-outage Information Risk in FX Markets.” Journal of Financial
Nenova, Tsvetelina, Andreas Schrimpf, and Hyun Song Shin. Economics 140 (2): 391–411.
2025. “Global Portfolio Investments and FX Derivatives.” Schrimpf, Andreas, and Vladyslav Sushko. 2019. “FX Trade
Working Paper 1273, Bank for International Settlements, Execution: Complex and Highly Fragmented.” BIS Quarterly
Basel, Switzerland. Review (December): 39–51.

76 International Monetary Fund | October 2025

 3
CHAPTER
GLOBAL SHOCKS, LOCAL MARKETS: THE CHANGING LANDSCAPE
OF EMERGING MARKET SOVEREIGN DEBT

Chapter 3 at a Glance
• Amid rising global sovereign debt levels and heightened vulnerabilities to global shocks, this chapter exam-
ines the changes in emerging market and developing economies’ (EMDEs’) domestic debt markets.
• The structure of government debt has increasingly diverged in emerging markets with stronger economic
fundamentals from others that continue to face significant financing and debt challenges.
• Many emerging markets with strong fundamentals have been able to issue domestically in local currency
and, given the subdued interest from international buyers, have found new resident buyers.
• This shift toward local currency issuance has supported resilience, as EMDEs with higher shares of local
currency debt and more diverse investor bases have exhibited more stable bond yields and market liquidity
during periods of global stress.
• In contrast, EMDEs with weaker policy credibility and shallower pools of domestic financial savings
remain reliant on foreign currency borrowing, short-term local currency debt, or less stable funding
sources.
• The growing sovereign-bank nexus in some EMDEs warrants attention, as it may mask underlying weak-
ness in debt absorption capacity and amplify financial stability risks.

Policies to Address Financial Vulnerabilities

• Enhancing macroeconomic fundamentals—such as raising domestic financial savings and strengthening
fiscal and monetary credibility—remains essential to increase debt-carrying capacity and attract stable
sources of long-term funding.
• Proven positive steps can also be taken to deepen EMDEs’ local currency bond markets and enhance their
functioning, with benefits for financial stability. These include enhancing the predictability and trans-
parency of debt issuances, developing efficient repo and money markets, strengthening primary dealer
frameworks, and diversifying the investor base.

Introduction $12 trillion excluding China), and nonresident portfolio

inflows have slowed. Although many EMDEs and their
This chapter examines the evolution of emerging
local bond markets have demonstrated resilience to a
market and developing economies’ (EMDEs)1 domestic
variety of global shocks (see Chapter 2 of the October
debt markets against a backdrop of rising debt levels
2025 World Economic Outlook), the contrast with those
and heightened vulnerabilities to global shocks. Over
that have faced significant distress and macroeconomic
the past decade, total government debt among EMDEs
instability highlights the structural and market chal-
has more than doubled to nearly $30 trillion (close to
lenges related to domestic bond market development.
EMDEs have experienced significant outflows from
The authors of this chapter are Sally Chen, Andrew Ferrante,
Bryan Gurhy, Zixuan Huang, Kay Chung, Esti Kemp, Xiang-Li Lim, their domestic local currency bond markets (LCBMs)2
Shijia Luo, Sonal Patel, Arindam Roy (co-lead), Patrick Schneider
(co-lead), Xuege Zhang, Zhongxia Zhang, and Jing Zhao, under the 2The term “local currency bond market” (LCBM) used through-

oversight of Charles Cohen, Thor Jonasson, and Jason Wu. out this chapter refers to marketable securities issued by the gov-
1In this chapter, “emerging market and developing economies” ernment in local currency in the domestic market. Countries most
(EMDEs) is used as a general term covering a full economy sample commonly issue in local currency in their domestic markets and in
of 56 economies, which are classified into 12 major emerging mar- hard currency (most often US dollars or euros) in international mar-
kets, 7 other emerging markets, and 37 frontier markets (see Online kets. While exceptions to this pattern exist, data limitations preclude
Annex 3.2 for the full list). more detailed analyses of this issue at present.

International Monetary Fund | October 2025 77

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

and financial stress during global shocks like the 2013 emerging and frontier markets, with emerging markets
“taper tantrum.” As a result, EMDEs have sought to further classified into major and other emerging markets
increase the role of resident buyers in their financing on the basis of market size and fragmentation.4
strategies. In addition, weak returns in LCBMs over This chapter estimates the effects of global shocks
the past decade—driven largely by continuing dollar on LCBMs and how these effects are associated with
strength—have made them a less appealing asset class the degree of participation by nonresident versus
for global investors benchmarked to US dollar assets. domestic investors, as well as the split between banks
Considering these developments, EMDEs have had and nonbank financial institutions (NBFIs) within
two main options for funding increased debt issuance: domestic investors. Empirical results confirm that
find more resident buyers for local currency debt or the presence of more nonresident investors is indeed
continue to rely on foreign-currency-denominated associated with greater sensitivity of domestic markets
sovereign bond issuance or external loans. Compared to global shocks, while the presence of more domestic
to many advanced economies, financial markets in investors—notably banks—is associated with lower
EMDEs tend to be less developed, and their domestic sensitivities (see the “EMDE Bond Market Sensitivity
debt markets are more exposed to market stress and to Global Shocks” section).
spillovers from global shocks. Although these results suggest that more resident
A select group of major emerging markets has buyers of local currency debt tend to improve resil-
largely been able to rely on local currency issuance that ience to global shocks, this does not mean that more
has been increasingly absorbed by domestic inves- domestic buyers are always better. This chapter also
tors amid higher domestic financial savings. This has explores the drawbacks that may be associated with
helped reduce the risks stemming from both “original an overreliance on domestic issuance and demand (see
sin” (currency mismatch) and “original sin redux” the “Vulnerabilities: Limited Absorption Capacity and
(nonresident outflows). Other EMDEs have expanded the Sovereign-Bank Nexus” section). To this end, this
borrowings largely through relatively shorter maturity chapter highlights the risk of overborrowing and the
financing from domestic banks and the central bank adverse feedback loops that could ensue if domestic
and often continue to rely on expensive foreign cur- banks were to absorb excessive amounts of sovereign
rency debt. Last, several EMDEs have had to resort to debt (that is, the sovereign-bank nexus), which could
domestic debt restructuring because of unsustainable lead to large financial stability downsides in cases of
public debt burdens.3 debt distress or restructuring. Resilience in EMDEs,
Although all government debt is considered, this therefore, depends on macroeconomic factors such
chapter focuses on LCBMs and investigates how the as monetary and fiscal credibility (see Chapter 2 of
changes in composition of debt issuance, investor the October 2025 World Economic Outlook), as well
absorption, and market structure have influenced as sufficiently deep and liquid sovereign debt markets
resilience to external shocks (see the “Recent Trends that feature a diverse domestic buyer base with high
in EMDE Sovereign Debt Markets” section). It has absorption capacity.
been well established that LCBMs play a critical role Rapid expansion of LCBMs without adequate
in enhancing macrofinancial stability and deepening absorption capacity and strong monetary and fiscal
domestic financial systems. By reducing currency anchors can lead to overreliance on captive investors
mismatch and rollover risks, they insulate public like banks and central banks, raising financial stability
finances from external shocks and support coun- risks and resulting in the crowding out of private
tercyclical responses. To help overcome limitations
from wide variations in data coverage, this chapter 4See Online Annex 3.2 for economy classification of local

focuses on a newly compiled data set of government currency bond markets. “Major EMs” (12) are those that have
local currency marketable bonds above 25 percent of GDP, with a
debt issued in domestic markets in 56 EMDEs, minimum of 50 percent of bonds exceeding $1 billion. “Other EMs”
broken down by investor type, which constitutes (7) are non–frontier markets that have local currency marketable
over 90 percent of local currency government debt bonds of more than 10 percent of GDP and at least 15 percent of
bonds above $1 billion. Economies classified as frontier markets
outstanding in EMDEs. EMDEs are classified into
(37) are a sample of economies that are either part of the JPMorgan
Next Generation Markets Index, are lower-income countries with
3Emerging and frontier markets who restructured their sovereign outstanding Eurobonds, or that have local currency marketable
domestic debt since 2010 include economies such as Argentina, bonds-to-GDP >10 percent; and 15 percent of outstanding bonds
Ghana, Jamaica, and Sri Lanka. with size >$250 million equivalent.

78 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Figure 3.1. Financial Stability Framework for Local Sovereign Debt Markets in
Emerging Market and Developing Economies

(Domestic savings and market

• Underdeveloped local markets • Deep local markets

Debt absorption capacity

• Excess domestic savings • Insulation from global shocks

High
• Bubbles in other local assets • Strong asset-liability matching

infrastructure)
• Capital flow and asset price • Crowding out
volatility • Financial repression
• Sudden stop and debt • Sovereign/bank nexus risks

Low
sustainability risks

Low High
Resident share of sovereign debt

Source: Authors.

credit. In extreme cases, unlike sovereign external Framework for Assessing EMDE
debt restructuring, domestic debt restructuring can Sovereign Debt Markets
impose disproportionate losses on domestic banks and
The framework in Figure 3.1 highlights the inter-
financial institutions, threatening systemic stability and
action of domestic absorption capacity and the role of
transmitting sovereign stress across the economy (IMF
resident investors, as well as the consequences of this
2021).
interaction for financial stability. Absorption capacity
Against this backdrop, LCBM development has
requires both strong macroeconomic fundamentals
two aims: (1) to reduce currency mismatch and
to generate sufficient domestic financial savings and
sudden stop risks by anchoring financing in local
sound financial market systems to channel these
currency and (2) to limit losses and spillovers to
savings into the LCBM. The framework assumes two
domestic investors should a domestic debt restruc-
core financial stability objectives by sovereign issu-
turing be required. To conclude, this chapter pro-
ers: (1) expand local currency issuance to domestic
vides policy advice on developing a resilient LCBM,
investors to reduce both currency mismatch and the
drawing on findings from the IMF and the World
risk of capital outflows and (2) minimize the risks to
Bank’s LCBM diagnostic framework and on broader
domestic financial institutions by building an investor
technical assistance for LCBM development (see
base with a larger and more diverse share of resident
the “Deepening Local Currency Bond Markets to
buyers willing and able to hold more local currency
Enhance Financial Stability” section). While improv-
government bonds.
ing macroeconomic fundamentals—such as raising
Broadly speaking, this interaction leads to four pos-
domestic financial savings and ensuring a stable
sible outcomes. When high debt absorption capacity
macrofinancial environment—remains essential for
is successfully used to increase the share of domes-
LCBM development, a strong policy framework
tic buyers, EMDEs are more insulated from global
and robust financial market systems are critical for
shocks because assets and liabilities in the economy are
channeling financial savings into a well-functioning
matched in local currencies. Even in this case, however,
local market. Foundational market infrastructure
there is the trade-off that bond markets and resident
(including money markets, primary markets, and
investors might be more exposed to local shocks.5
secondary markets) must be developed, legal certainty
When absorption capacity is low but domestic buyers
provided, and sustained efforts to deepen the investor
are nonetheless forced to buy sovereign debt, financial
base through sound debt management practices and
repression and sovereign-bank nexus risks may ensue.
market communication undertaken. In the absence of
The more unusual case of an economy with ample
these elements, efforts to deepen sovereign debt mar-
kets often stall, raising financial stability risks from
5For example, this can include inflation shocks that lead
poor price discovery, shallow liquidity, and excessive
to valuation losses in bonds held by resident investors despite
reliance on banks and public institutions to absorb well-functioning markets. This chapter does not analyze this domes-
government debt. tic trade-off.

International Monetary Fund | October 2025 79

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 3.2. Recent Trends in Emerging and Frontier Debt Markets

Government debt has increased sharply over the Gross financing needs have remained above Major EMs primarily issue in local currency,
past 15 years, with the median debt-to-GDP ratio prepandemic levels for many economies. while many other EMs and FMs still rely more on
reaching close to 60 percent. foreign currency issuance.
1. Government Debt 2. Gross Financing Needs 3. Foreign Currency versus Local
(Trillions of dollars, left scale; percent of GDP, (Share of economies, left scale; median percent of Currency Debt, 2024
right scale) GDP, right scale) (Percent share and ratio)
China, foreign currency >0.15 >0.10 >0.05
Major EMs Other EMs FMs
China, local currency <0.05 Median
EM ex. China, foreign currency
30 EM ex. China, local currency 70 100 14 1
Median debt-to-GDP
90 0.9
25 ratio (right scale) 60 12
80 0.8

Local currency share of debt

50 70 10 0.7
20
40 60 8 0.6
15 50 0.5
30 40 6 0.4
10 30 0.3
20 4
5 20 0.2
10 2
10 0.1
0 0 0 0 0
2010 12 14 16 18 20 22 24 0 0.5 1.0 1.5 2.0
2013
14
15
16
17
18
19
20
21
22
23
24
25
26
Ratio of increase in local currency debt
to total debt increase since 2010

Sources: IMF, World Economic Outlook database; IMF staff calculations; and sovereign investor base estimates by Arslanalp and Tsuda (2014).
Note: Panel 1 includes the maximum sample of 56 countries (see Online Annex 3.2). Panel 2 includes a subset of 45 countries based on data availability. Panel 3 includes
only countries where the general government debt as a percentage of GDP increased between 2010 and 2024. “Major EMs” are Brazil, China, Colombia, Hungary, India,
Indonesia, Malaysia, Mexico, the Philippines, Poland, South Africa, and Thailand. Frontier markets are classified according to the methodology in the October 2025 Global
Financial Stability Report, Chapter 3, footnote 5. “Other EMs” are the balance of the sample. EM = emerging market; ex. = excluding; FM = frontier market.

potential absorption capacity (that is, high domestic (nearly $12 trillion excluding China), with the median
financial savings) but without a well-developed debt debt-to-GDP ratio reaching close to 60 percent of
market to absorb these savings could lead to asset bub- GDP (Figure 3.2, panel 1). Gross financing needs are
bles in other local markets such as real estate or public forecast to ease slightly but remain above the levels
equities. In the worst case, EMDEs with both low seen immediately before the pandemic in many econ-
shares of resident buyers and low absorption capacity omies, leaving them more vulnerable to future shocks
are forced to rely on foreign borrowing and are more (Figure 3.2, panel 2).
vulnerable to sudden stops of capital flows and debt Foreign currency borrowing has become less
sustainability risks. With many EMDEs starting in this prominent in some EMDEs, but progress has been
low/low corner, the challenge has been to move to the uneven, and the currency composition of govern-
high/high quadrant without getting stuck in the bad ment borrowing still varies considerably across
equilibrium of overreliance on a high share of domestic economies. Major emerging markets, a minority
bank investors while still lacking adequate absorption of our broad sample, have more than two-thirds of
capacity. total government debt in local currency and have
avoided large net foreign currency issuance since
Recent Trends in EMDE Sovereign Debt 2010. In contrast, other emerging and frontier mar-
Markets kets still rely significantly on foreign currency debt
amid less developed LCBMs (Figure 3.2, panel 3).
Financing Needs Are Growing as Public Debt Expansion of LCBMs in EMDEs has taken place
Rises amid widely varying macroeconomic and institu-
Government debt in EMDEs has been rising tional conditions, shaping the depth and resilience
rapidly since 2010, reaching close to $30 trillion of LCBMs to different degrees (see the “Deepening

80 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Figure 3.3. Composition of Marketable Domestic Public Debt in Selected Emerging Market and Developing Economies
Major EMs issue mostly on local currency markets, while other EMs and Regarding marketable debt, major EMs are able to issue a greater amount
FMs rely more on international bonds and external loans, respectively. of longer-term local currency bonds, while other EMs and FMs issue more
short-term securities and foreign currency bonds.
1. Composition of Central Government Debt 2010 versus 2024 2. Composition of Marketable Debt, by Instrument Type, 2010–25
(Percent) (Percent of total)
Domestic marketable Domestic non-marketable Foreign currency bonds Treasury bills
External marketable External non-marketable Local currency Treasury bonds
100 6 100
10 9
10
10 26 31
80 10 41 80
8 56
13 38 60 61 63
11 71 64
60 78 72 60
5 80
21 9 11
8
40 2 2 13 40
71 74 13
7 15 27
20 45 45 48 38 20
40 9 8 35 31
21 25
13 13 14 12
0 0
2010 2024 2010 2024 2010 2024 2010 2018 2024 2010 2018 2024 2010 2018 2024
Major EMs Other EMs FMs Major EMs Other EMs FMs

Most EMs and FMs extended their local currency bond maturities, but Carrying cost of domestic debt portfolio in real terms exceeds projected real
maturity declined for Other EMs, reflecting cost and absorption constraints. growth in several countries.
3. Average Time to Maturity for Domestic Debt in Emerging Market and 4. Net Interest Cost of Domestic Debt Portfolio in 2024 versus
Developing Economies, 2010 versus 2024 Projected Inflation and Real Growth Rate, 2025–29
(Number of years) (Percentage points)
8 20
2010 2024
15
6
10

4 5

0
2
−5

0 −10
Major EMs Other EMs FMs
ZAF
BRA
HUN
MEX
COL
POL
MYS
CHN
IND
PHL
ARG
PER
DOM
CHL
ROU
SAU
TUR
EGY
PAK
LKA
KEN
UGA
ARM
JAM
CRI
NAM
ZMB
TUN
BWA
GHA
DZA
VNM
NGA
Major EMs Other EMs FMs

Sources: Bank for International Settlements; Bloomberg Finance L.P.; national authorities; national sources; and IMF staff calculations.
Note: Panel 1 shows the annual weighted average share of total marketable debt issuance for economy groupings since 2010 split between the domestic and foreign
jurisdiction of debt instruments. Panel 2 shows the simple average composition of marketable debt for a group of 36 EMDEs. Panels 3 and 4 refer to domestic debt only.
Local currency bond market classification is based on two thresholds: (1) domestic debt-to-GDP ratio and (2) the share of bonds exceeding a minimum outstanding size.
“Major EMs” are defined by domestic debt-to-GDP ratios above 25 percent with at least 50 percent of bonds exceeding $1 billion outstanding. “Other EMs” meet a
10 percent domestic debt-to-GDP threshold and have at least 20 percent of bonds above $1 billion. “FMs” are those within the broader frontier sample that meet a
10 percent domestic debt-to-GDP threshold with at least 15 percent of bonds above $250 million. See Online Annex 3.2 for details of this classification. EM = emerging
market; FM = frontier market. Data labels in the figure use International Organization for Standardization (ISO) country codes.

Local Currency Bond Markets to Enhance Financial Major emerging markets issue primarily in local cur-
Stability” section). Although not within the scope of rency in their domestic debt markets.7 However, other
this chapter, corporate debt in more developed large emerging and frontier markets also rely significantly
emerging markets has also migrated toward local on foreign currency denominated international bonds
currencies (Box 3.1).6 and external loans, respectively (Figure 3.3, panel 1).

7At the end of 2024, a few emerging markets had a modest share
6A well-functioning LCBM is foundational for development of of foreign-currency-denominated bonds in their domestic bonds
domestic corporate bond market by, for example, providing a reliable outstanding, notably Argentina and Türkiye, alongside some recent
local currency yield curve benchmark (IFC 2025). restructuring cases such as Ghana and Sri Lanka.

International Monetary Fund | October 2025 81

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

In terms of maturity, many major emerging markets recent months. Inflows to local currency debt averaged
have been able to rely on long-term local currency over 1 percent of GDP in aggregate (excluding China)
bonds to meet financing needs (Figure 3.3, panel 2), from 2010 to 2014 but under 0.5 percent of GDP
thereby mitigating rollover risks, and have extended from 2015 to 2024,10 with inflow cycles becoming
their maturity profiles over the past 20 years as a smaller and shorter (Figure 3.4, panel 1). Staff analysis
result of improved macroeconomic stability and a finds that a strong dollar and higher US Treasury
larger institutional investor base. For major emerging yields have played significant roles in curbing flows to
markets, the average time to maturity of debt reached LCBMs,11 yet other related recent work suggests that
seven years in 2024 (Figure 3.3, panel 3), and the the role of the global financial cycle in total portfo-
average cost on the domestic debt portfolio declined lio debt flows is overstated.12 Nonresident holdings
marginally. Nevertheless, some major emerging mar- have stagnated in many countries as a share of GDP,
kets have had to compensate investors for additional although they remain significant and continue to play
risk through the use of inflation-linked or floating-rate an important role in some local markets (Figure 3.4,
instruments, whereas others also use marketable sukuk panel 2).
to meet investor preferences.8 Many frontier markets For global investors, total returns on the emerging
have also significantly extended maturities since 2010, market local currency bond index have been per-
although some have seen the average interest cost on sistently weak over the past decade, primarily under-
domestic debt portfolios rise significantly. mined by poor currency returns amid a strong dollar
Extending debt maturities in countries with less cycle (Figure 3.4, panel 3). Risk-adjusted returns have
stable macroeconomic environments and fiscal anchors lagged comparable asset classes such as US high-yield
can lead to rising term premiums. This highlights corporate bonds, likely denting risk appetite for the
the trade-offs faced by debt managers in balancing asset class (Figure 3.4, panel 4). Returns on emerging
funding costs and refinancing risks for local currency market hard currency bonds have performed somewhat
borrowings.9 In several economies, the real interest better. Net international sovereign bond issuance has
rate on outstanding domestic bonds exceeds projected continued at a robust pace, with total outstanding debt
real GDP growth over the next five years, suggesting reaching over $1.4 trillion in 2025 despite outflows
that the net real carrying cost of domestic debt may of around 20 percent of assets under management
impose fiscal burdens in the years ahead (Figure 3.3, from dedicated emerging market hard currency funds
panel 4). since 2022, suggesting an increased role for crossover
investors.13
Over the past decade, the structure of the invest-
Weak Returns Have Weighed on Nonresident ment base for domestic local currency debt has
Investor Risk Appetite changed materially. For many emerging markets,
Portfolio flows to LCBMs have broadly deceler- the nonresident share of local currency debt peaked
ated over the past 10 years despite a modest uptick in nearly a decade ago (Figure 3.5, panel 1), although
the decline accelerated after the pandemic. The
8Asian emerging markets tend to rely on a high share of fixed-rate decline generally reflects a significant increase in net
domestic bonds. In Latin America and Central and Eastern Europe, issuance alongside tepid inflows, rather than large
issuance has also included a significant amount of floating, or outflows, outside select cases (Figure 3.5, panel 2).
inflation-linked, bonds. Emerging markets like Indonesia, Malaysia,
Saudi Arabia, and Türkiye issued a significant amount of sukuk in
the domestic market, with their outstanding stock ranging between 10Measured on a rolling four quarter sum.
13 and 47 percent of their marketable domestic debt at the end of 11IMF staff regressed nonresident bond flows, as a percentage of
2024. Frontier markets like Pakistan also have sizable outstanding the previous month’s nonresident stock, against the change in the
sukuk (11 percent). Analysis in this chapter relating to domestic Federal Reserve’s advanced economy dollar index, the VIX index,
marketable bonds covers sukuk. and emerging market–US policy rate differentials, with controls on
9Unlike foreign concessional loans and international bonds, which commodity prices, emerging market and US inflation surprise, and
are typically longer term, ranging between 10 and 30 years but emerging market and US industrial production.
contingent upon access restrictions, domestic bond maturities in 12Cerutti and Claessens (2024) assert that only up to about

emerging markets could range between 1 and 30 years. Shifting from 25 percent of the variation in portfolio flows can be explained by the
external to domestic debt in the initial stages could therefore result global financial cycle.
in a reduction in the average maturity of the overall debt portfolio. 13This includes only funds reported by EPFR.

82 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Figure 3.4. Portfolio Flows, Nonresident Holdings, and Investor Returns for Selected Emerging Market and Developing
Economies
Portfolio flows to emerging markets have continued, albeit at a slower pace. Nonresident holdings of local currency debt are below their peaks but
remain significant in some cases.
1. Nonresident Flows to Local Currency Bond Markets, Excluding China 2. Nonresident Holdings of Local Currency Government Bonds,
(Percent of GDP on rolling 12-month sum) 2012–Present
(Percent of GDP)
3.5 Interquartile range Aggregate Median 20
Range Latest
3.0
2.5
2.0
1.5
10
1.0
0.5
0
−0.5
−1.0 0
2011 12 13 14 15 16 17 18 19 20 21 22 23 24 25

ZAF
MYS
EGY
HUN
PER
COL
BRA
MEX
THA
HUN
ROU
POL
IDN
IND
GHA
TUR
DOM
KEN
Total returns on emerging market local currency debt have been weak, Downside risks and risk-adjusted returns for emerging market local
undercut by currency performance. currency debt have lagged other asset classes.
3. Decomposition of Total Returns for Emerging Market Local Currency Debt 4. Risk-Adjusted Returns and Downside Risk
(Percent, two-year rolling returns, total return in US dollar terms) (Fifth percentile of monthly returns since 2015; Sharpe ratio since 2015)
EM LC FX return EM LC price return
40 EM LC coupon return EM LC total return (USD) 0
30 US HY return −0.01
US IG bonds

Fifth percentile of returns

20 −0.02
US HY bonds
10 Treasury −0.03
bonds EM hard currency bonds
0 −0.04
−10 EM LC total return (USD) −0.05
−20 US equities −0.06
World ex. US
−30 EM equities equities −0.07
−40 −0.08
2011 12 13 14 15 16 17 18 19 20 21 22 23 24 −0.2 0 0.2 0.4 0.6 0.8 1
Sharpe ratio since 2015

Sources: Bloomberg Finance L.P.; EPFR; J.P. Morgan; IMF World Economic Outlook database; IMF staff estimates; and national sources.
Note: Panels 1 and 2 include the same unbalanced panel of 17 countries, labeled in panel 2. Egypt includes only US Treasury bills; GDP is interpolated. Data labels in the
figure use International Organization for Standardization (ISO) country codes. Panel 3 displays returns from the J.P. Morgan EM Government Bond Index—Global
Diversified: a local currency government bond index with a maximum country weight of 10 percent. Panel 4 considers monthly returns since 2015; other asset class returns
are derived from benchmark indices. In panel 4, Treasury bills are used as the risk-free rate. EM = emerging market; ex. = excluding; FX = foreign exchange; HY = high
yield; IG = investment-grade; LC = local currency; USD = US dollar.

Among frontier markets, nonresident participation their presence in a number of markets (Figure 3.5,
in domestic local currency debt markets has been panel 3).
more varied and at times prone to large fluctuations,
although it can also be a significant part of some
markets. Domestic bank ownership has generally been Investor Base for Some Local Currency Bond
steady over time, indicating that bank absorption has Markets Has Shifted from Nonresident to
largely kept pace with increased issuance in recent Resident
years (see the “Sovereign-Bank Nexus Has Risen in The uncertainty and risks around nonresident
Recent Years” section), while NBFIs have increased inflows highlight the value of a strong domestic investor

International Monetary Fund | October 2025 83

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 3.5. Investor Base in Selected Emerging Market Local Currency Government Bond Markets
Nonresident share of LCBMs has declined and is near multiyear lows in … with nonresident holdings largely failing to keep pace with higher
many countries … domestic net issuance.
1. Range for Nonresident Share of LCBM since 2012 2. Decomposition of Change in Nonresident Share of LCBMs Since
(Percent share) December 2019
(Percentage points, contribution of nonresident stock and total debt
stock to change in nonresident share of debt)
Latest
70 Pre-COVID shock (December 2019) Nonresident contribution Total debt contribution Sum 40
60 Range since 2012 30

50 20
10
40
0
30
−10
20 −20
10 −30
0 −40
PER
ZAF
MYS
ROU
COL
HUN
IDN
MEX
POL
BRA

TUR
DOM
GHA
IND

ROU
IND
BRA
MYS

TUR
DOM
PER
COL
THA
POL
ZAF
MEX
GHA
IDN
THA

HUN
NBFIs have become a larger presence across most markets, while banks Investor composition varies considerably across countries, with larger
continue to be a sizable share of the investor base. NBFI ownership among major EMs, more banks among other EMs, and
more “other” ownership among frontier markets.
3. LCBM Investor Composition 4. Detailed Investor Composition, by LCBM and Country Type, 2024
(Percent share) (Percent share, interquartile range, median)
70 2012 Interquartile range Median 0.6
2018 0.5
60 2024
0.4
50
0.3
40 0.2
30 0.1

20 0.0
Major EMs

Other EMs
Frontier
markets
Major EMs

Other EMs
Frontier
markets
Major EMs

Other EMs
Frontier
markets
Major EMs

Other EMs
Frontier
markets
10
0
NBFIs Banks Nonresident Other NBFI Banks Nonresident Other

Sources: ASISA; central banks; EUROPACE AG/Haver Analytics; IMF Monetary and Financial Statistics; and ministries of finance.
Note: Panels 3 and 4 use a baseline sample of 29 countries, of which the sample in panels 1 and 2 are included. In panels 1 and 2, India includes only central government
securities. In panel 1, data on nonresident holdings for Hungary are not adjusted for repo transactions. In panel 3, × represents the average, and the horizontal line
represents the median. Data labels in the figure use International Organization for Standardization (ISO) country codes. EM = emerging market; LCBM = local currency
bond market; NBFI = nonbank financial institution.

base. However, investor composition varies considerably significant NBFI investor base,14 the sector is primarily
across EMDEs. Many emerging markets with more composed of long-term buyers such as pension funds
developed financial markets have been able to rely on a and insurance companies (see the “Vulnerabilities:
diverse set of resident NBFIs and banks. Emerging and Limited Absorption Capacity and the Sovereign-Bank
frontier markets with less developed local markets have Nexus” section). In a limited number of countries (Bra-
less consistent funding models. Debt absorption often zil, Mexico, and South Africa), mutual and investment
involves different types of investors, with central banks, funds hold more than 10 percent of government bonds.
public institutions, and other private buyers playing
more significant roles. Banks have a large presence in
most LCBMs, although less so in Latin America, with 14Availability and consistency of granular classification of investor
a median ownership share of close to 30 percent across categories, especially among nonbank investors, varies greatly and
countries (Figure 3.5, panel 4). Among countries with a presents analytical limitations.

84 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Figure 3.6. Local Stress Index and the Investor Base for Local Currency Bond Markets in Emerging Markets
Emerging market LCBMs experienced some stress in early 2022 as many ... with most emerging markets seeing an increase in domestic NBFIs’
emerging market central banks raised rates rapidly ... participation in the postpandemic era.
1. Bond Local Stress Index 2. Change in Domestic NBFI Participation, by Change in Local Stress Index
(Index) (Percent change, third quarter of 2022 to fourth quarter of 2024)
0.8 0.2
Median Interquartile range BRL CNY
ROU IDR
0
0.6 INR

Change in local stress index

HUF
PEN PLN
−0.2
ZAF
0.4 CLP COP
MYR −0.4
MXN
THB
0.2
−0.6

0 −0.8
Jan. 2018 Jul. 2019 Jan. 2021 Jul. 2022 Jan. 2024 −5 0 5 10
Change in domestic NBFI Holdings (percent of outstanding)

Sources: Bloomberg Finance L.P.; EUROPACE AG/Haver Analytics; and IMF staff calculations.
Note: Local stress index methodology is from the IMF’s October 2020 Global Financial Stability Report. Variables for emerging markets are bid-ask spreads, estimated
liquidation cost (Bloomberg’s liquidity assessment model), term premia (the ACM model), and three-month realized volatility and asset swap spread. Data labels in the
figure use International Organization for Standardization (ISO) country codes. LCBM = local currency bond market; NBFI = nonbank financial institution.

EMDE Bond Market Sensitivity to Global More Resident Investors Is Associated with
Shocks Smaller Effect of Global Shocks on Local Bond
Markets
Local Market Stress Has Receded in Recent Years
Regressions confirm that increased resident bank
LCBMs in emerging markets experienced periods
holdings are associated with a decline in the transmis-
of heightened stress during the COVID-19 pandemic
sion of global shocks to LCBMs.16 Global shocks are
in 2020 and later in 2022 when many central banks
generally accompanied by an increase in local market
rapidly hiked interest rates, as measured by the IMF
strains, as measured by rising bond yields or widening
Local Stress Index (Figure 3.6, panel 1). Greater par-
bid-ask spreads (see Figure 3.7, panels 1 and 2, yellow
ticipation by domestic NBFIs in LCBMs (as detailed
bars). However, the presence of nonresident investors
in the previous section) appears to have coincided
is associated with an amplification of such pressure
with the normalization of market functioning during
(blue bars), and increased resident bank participation is
the 2022 episode (Figure 3.6, panel 2). In contrast, in
associated with a dampening of these pressures (green
2020, domestic banks absorbed the bulk of issuance,
bars), particularly when investor participation is above
while central bank purchases were also associated with
the sample average. Moreover, the attenuation effects
a reduction in market stress (see the October 2020
persist and are larger in some instances during peri-
Global Financial Stability Report).15 The empirical
ods of financial market stress; the latter suggests that
models in the next section provide more granular
analysis of the stabilizing role of domestic banks and
NBFIs during global shocks. 16Cross-country panel regressions are used to quantify the effects

of investor participation on LCBMs. The dependent variables are

15Eckhold and others (2024) note that the increased presence of changes in five-year yield spreads and bid-ask spreads for 14 emerg-
central banks in LCBMs after the COVID-19 shock was in line with ing markets, while the key independent variable is the change in the
their financial stability mandates, effectively addressing issues related Chicago Board Options Exchange Volatility (VIX) Index, a proxy
to market dysfunction. However, in some EMDEs, the size of the for global shocks, and its interaction with shares held by nonresident
interventions may have significantly increased risks to central banks’ and resident investors (further segmented into banks and NBFIs),
balance sheets, raising issues of policy solvency, operational indepen- controlling for economy level macroeconomic fundamentals (see
dence, fiscal dominance, and moral hazard. Online Annex 3.1 for data and model discussions).

International Monetary Fund | October 2025 85

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 3.7. Effect of Global Risk Factors on Local Currency Bond Markets in Emerging Markets and the Role of Investor
Composition
Above-average nonresident (domestic bank) shares increase (reduce) the Above-average nonresident (domestic) shares increase (reduce) the impact
impacts of VIX on yield spreads. of VIX on market liquidity.
1. Effects of 10-Percentage-Point VIX Increase on Five-Year Yield Spreads 2. Effects of 10-Percentage-Point VIX Increase on Five-Year Bid-Ask Spreads
(Basis points) (Basis points)
25 1.5

20 1.2

15 0.9

10 0.6

5 0.3

0 0
All Stress All Stress All Stress All Stress All Stress All Stress All Stress All Stress
VIX +10 ppt Nonresident Resident Banks Resident NBFIs VIX +10 ppt Nonresident Resident banks Resident NBFIs
average impacts (mean and +1 SD) (mean and +1 SD) (mean and +1 SD) average impacts (mean and +1 SD) (mean and +1 SD) (mean and +1 SD)
Yield spreads (basis points) Bid ask spread (basis points)

Source: IMF staff calculations.

Note: Bars indicate the estimated impact of a 10-percentage-point increase in the VIX, along with the effects of a one-standard-deviation increase in investor participation
for nonresidents, resident banks, and resident NBFIs. Solid bars signal an amplification effect; hollow bars indicate attenuation. Shaded bars indicate statistical
insignificance. See Online Annex 3.1 for more information. “Stress” refers to a subsample in which the VIX is above its 75th historical percentile. The sample is Brazil,
China, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Peru, Poland, Romania, South Africa, Thailand, and Türkiye. 1 SD = one standard deviation;
NBFIs = nonbank financial institutions; ppt = percentage point; VIX = Chicago Board Options Exchange Volatility Index.

LCBMs’ response to global shocks could be dispropor- under “All”). Online Annex 3.1 contains more details
tionately higher in stress scenarios. on the model, results, and robustness checks.
More specifically, the model results show that By contrast, increased resident bank bond hold-
increased nonresident ownership is accompanied by ings are associated with a mitigation of the impacts
an amplification of the effects of a VIX shock on local of a VIX shock. A one-standard-deviation increase
currency bond yield spreads to the same-maturity in ownership by resident banks from the average of
US Treasury yield and bond bid-ask spreads.17 A 29 percent to 44 percent is associated with a damp-
10-percentage-point increase in the VIX is associated ening of the sensitivity of yield and bid-ask spreads,
with a 19 basis point increase of the five-year local respectively, from 19 to 11 basis points and from 0.8
currency yield spread and a 0.7 basis point increase to 0.7 basis points (Figure 3.7, panels 1 and 2, left
in the bid-ask spread, when nonresident ownership of green bars under “All”).
local currency bonds is at the cross-country average For sovereign bond investors, such effects are
level of 22 percent (Figure 3.7, panels 1 and 2, light meaningful. For example, given the average monthly
blue bars under “All”). Should this ownership increase change in emerging market yield spreads of approxi-
by one standard deviation (to 34 percent), the sensi- mately 1.4 basis points, a 4 basis point increase in yield
tivity of yield and bid-ask spreads to the increase in spreads—based on a 34 percent nonresident ownership
VIX rises, respectively, to 23 basis points and 0.9 basis and a 10-percentage-point increase in VIX—represents
point (Figure 3.7, panels 1 and 2, dark blue bars nearly three times the typical monthly movement. On
market liquidity, given the average monthly change in
17These results are qualitatively consistent with the literature (for
bid-ask spreads of about 0.02 basis points, an impact
example, Ebeke and Lu 2015; Ho 2022; BIS 2024; October 2024
Fiscal Monitor, Chapter 1), which documents the procyclical nature of 0.1 to 0.2 basis points is about 5 to 10 times the
of nonresident flows and the stabilizing role of banks. average movement. These results are qualitatively

86 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

similar when other proxies of global shocks, such as This nonlinearity is also seen with regard to the size
the Merrill Lynch Option Volatility Estimate (MOVE) of government debt, whereby higher volatility would be
Index, are used.18 expected in more highly indebted countries. For those
The role of resident NBFIs is more nuanced. high government debt economies (that is, with debt-to-
Increased NBFI bond holdings do not statistically alter GDP ratios above the sample median of 47 percent),
the impact of a VIX shock on yield spreads in this domestic bank holdings are associated with smaller
sample (Figure 3.7, panel 1, left red bars), but they are pass-throughs of global shocks, while the presence of
associated with an attenuation of the shock’s impact on nonresidents is accompanied by larger impacts. This
bid-ask spreads (Figure 3.7, panel 2, left red bars). This highlights the importance of domestic investors in
lack of statistical significance in the yield spread regres- LCBMs to help weather periods of stress. That said,
sion may be driven by the heterogeneity of NBFIs greater nonresident participation is also associated with
across countries, given the diversity of their investment narrower average yield spreads in these high-debt econ-
mandates, investment horizons, and funding stability. omies, suggesting a supportive role in reducing financ-
To disaggregate this effect, impacts were examined by ing costs (see Online Annex 3.1 for more details).
region. In some economies in emerging Asia, where Notably, the effects of global shocks and investor
pension funds and insurers account for a dominant participation on domestic bond markets persist. Effects
share of NBFI local currency bond holdings, increased typically peak within one quarter before gradually
NBFI participation is accompanied by an attenuation receding. A larger nonresident investor share is asso-
of the impacts of a VIX shock (Online Annex Table ciated with an amplification in both the magnitude
3.1.4).19 This is consistent with pension funds and and the duration of the spread response (Figure 3.8,
insurers having more stable funding and typically being panels 1 and 4), likely underscoring the procyclical
regarded as “safe hands” with long-term investment nature of nonresident flows during risk-off episodes.
decisions. By contrast, in Latin America (notably Brazil By contrast, greater participation from domestic banks
and Mexico), where mutual funds play a larger role, a is accompanied by a dampening in the initial impact
larger presence of NBFIs does not appear to have the and an acceleration in the normalization of spreads,
same effect. That said, greater market participation suggesting domestic banks’ role in stabilizing and their
of NBFIs does appear to help deepen market liquid- market‑making function (Figure 3.8, panels 2 and 5).
ity more broadly, as seen in a narrowing of bid-ask For domestic resident NBFIs, although increased
spreads. participation does not statistically alter the response of
The effect of global shocks on LCBMs appears yield spreads to the VIX shock (Figure 3.8, panel 3),
nonlinear and tends to be larger in volatile times and it does support market liquidity by dampening the
for more indebted economies. Focusing on periods response of bid-ask spreads in significant and durable
when the VIX is above its 75th historical percentile ways, underscoring the sector’s role in deepening mar-
(Figure 3.7, panels 1 and 2, bars labeled “Stress”), a ket liquidity (Figure 3.8, panel 6).
10-percentage-point increase in the VIX raises local
currency yield and bid-ask spreads much more (right
yellow bars). The amplification and attenuation effects Vulnerabilities: Limited Absorption
of higher nonresident and resident holdings, respec- Capacity and the Sovereign-Bank Nexus
tively, are also larger, particularly for market liquidity. Resident Investors’ Absorption Capacity May
Become More Challenged
18It could be that pure time series variables (for example, monthly

changes in VIX or MOVE) are picking up variations in other global

Although the regressions show that high resident
conditions or shocks not considered in the panel regressions. Online participation is associated with smaller impacts of
Annex 3.1 shows the results of a specification whereby yearly fixed global shocks, the capacity of resident investors to
effects are added to the regression in both level and interaction
continue absorbing supply might be waning. Net
terms, in addition to monthly VIX changes. The coefficients on the
VIX regressors decline in magnitude in this specification, suggest- local currency government bond issuance continues to
ing that there may indeed be other global forces that need to be grow at a pace faster than prepandemic rates in several
considered. emerging markets (Figure 3.9, panel 1). Concerns that
19The regional approach was taken due to the lack of more gran-

ular data on the types of NBFIs that hold local currency bonds at an resident investors may not fully absorb rapid issuance
economy level. could be a reason for the widening of local currency

International Monetary Fund | October 2025 87

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 3.8. Investor Composition and the Long-Term Effect of a VIX Increase on Local Currency Bond Markets in
Emerging Markets
Greater nonresident participation significantly Greater domestic bank participation significantly Greater domestic NBFI participation significantly
amplifies the effects of global shocks on local mitigates the effects of global shocks on local mitigates the effects of global shocks on local
currency bond markets. currency bond markets. currency bond market liquidity.
1. Impacts of VIX and Nonresident Participation 2. Impacts of VIX and Resident Banks’ 3. Impacts of VIX and Resident NBFIs’
on Five-Year Yield Spreads Participation on Five-Year Yield Spreads Participation on Five-Year Yield Spreads
(Basis points) (Basis points) (Basis points)
Average impacts of 10 ppt VIX spike Average impacts of 10 ppt VIX spike Average impacts of 10 ppt VIX spike
40 Nonresident share +1 SD to Resident bank share +1 SD to 40 Resident NBFI share +1 SD to 40
34 percent 44 percent 43 percent
30 30 30

20 20 20

10 10 10

0 0 0

−10 −10 −10

0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
Months Months Months
4. Impacts of VIX and Nonresident 5. Impacts of VIX and Resident Banks’ 6. Impacts of VIX and Resident NBFIs’
Participation on Five-Year Bid-Ask Spreads Participation on Five-Year Bid-Ask Spreads Participation on Five-Year Bid-Ask Spreads
(Basis points) (Basis points) (Basis points)
Average impacts of 10 ppt Average impacts of 10 ppt VIX spike Average impacts of 10 ppt VIX spike
1.5 VIX spike Resident bank share +1 SD to 1.5 Resident NBFI share +1 SD to 1.5
Nonresident share +1 SD to 44 percent 43 percent
34 percent
1 1 1

0.5 0.5 0.5

0 0 0

−0.5 −0.5 −0.5

0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12
Months Months Months

Sources: IMF staff calculations.

Note: The figure shows the long-term impact of a 10-percentage-point increase in VIX on five-year yield spreads to US Treasuries and five-year bid-ask spreads. The black
lines show the effects when nonresidents, resident banks, and resident NBFIs holdings are at their averages of 22, 29, and 31 percent, respectively. The blue, green, and
red, dashed lines show the effects when holding share increases by 1 standard deviation to 34 percent for foreign institutions, 44 percent for domestic banks, and 43
percent for domestic NBFIs. The shaded areas represent 90 percent confidence intervals. The emerging markets covered in the sample are Brazil, China, Colombia,
Hungary, India, Indonesia, Malaysia, Mexico, Peru, Poland, Romania, South Africa, Thailand, and Türkiye. 1 SD = 1 standard deviation; NBFIs = nonbank financial
institutions; ppt = percentage point; VIX = Chicago Board Options Exchange Volatility Index.

bond yields relative to interest rate swap rates in recent continue to rely on banks to be the main buyers of
months for some emerging markets—large issuance sovereign debt. On average, NBFIs in emerging and
tends to be a significant driver of wider swap spreads frontier markets hold roughly 22 and 40 percent of
(Figure 3.9, panel 2). their assets in sovereign debt, respectively.20 While
Financial assets of NBFIs in EMDEs increased there is substantial heterogeneity across jurisdictions,
by about 11 percent of GDP since 2013, as mea- emerging and frontier market pension funds allocate
sured by an equal-weighted average across a selection
of 20 economies (Figure 3.9, panel 3). However, 20Average NBFI sovereign debt holdings cover only jurisdictions
NBFI presence in the frontier markets in this sam- with data available in the IMF Monetary and Financial Statistics
ple remains low. This suggests frontier markets will data set.

88 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Figure 3.9. Financial Sector Assets and Absorption Challenges in Selected Emerging Markets
Local currency issuance has expanded faster than prepandemic ... ... contributing to wider spreads against swap rates.
1. Local Currency Debt Issuance 2. Model Decomposition of Five-Year Swap Spreads
(Year-over-year percent change) (Percent)
Short-term liquidity premium Du and Schreger credit risk premium
30 Interquartile range Median Net marketable issuances Slope of swap curve 1.6
Actual Fitted (model)

20 0.8

10 0

0 −0.8
Jan. 2018 Jan. 2020 Jan. 2022 Jan. 2024 Jan. 2018 Jan. 2020 Jan. 2022 Jan. 2024

Nonbank financial institution assets are limited in most frontier markets, Emerging and frontier market pension fixed-income allocations are higher
although they have generally expanded in most countries over the past than those of advanced economies.
decade.
3. Emerging and Frontier Market Financial Sector Asset Growth 4. Pension Fund Asset Allocation across Countries, 2024 or Latest Available
(Percent of GDP) (Percent of pension fund assets)
Bank NBFI Growth indicator Bills and bonds Equity Cash
CIS (when look-through unavailable) Other
2013
CHN 2023
MYS 2013 AE average
2023
2013
USA
BRA 2023 DOM
ZMB 2013
2023 IND
HNU 2013 ROU
2023
2013 IDN
THA
MEX
Emerging markets

2023
Emerging markets

2013
CHL 2023 HUN
PHL 2013
2023
THA
IND 2013 PHL
2023
2013
TUR
POL 2023 MYS
MEX 2013
2023 COL
SAU 2013 PER
2023
2013 ZAF
TUR 2023 POL
2013
COL 2023 KAZ
ROU 2013
2023
NGA
JAM 2013 CRI
Frontier markets

2023
2013
EGY
ARM
Frontier markets

2023 KEN
KEN 2013
2023 PAK
KAZ 2013 JAM
2023
2013 ARM
ZAF 2023
2013
ZMB
GHA 2023 BWA
0 100 200 300 400 0 20 40 60 80 100

Sources: Bloomberg Finance L.P.; Fitch Ratings; OECD 2024; IMF Monetary and Financial Statistics and World Economic Outlook database; and IMF staff calculations.
Note: Panel 1 shows growth in marketable local currency debt. In panel 2, estimates are derived from ordinary least squares regression on individual economies, with
attribution to only relevant variables for individual countries. The slope of the swap curve captures issuers’ incentives to adjust maturities; when the curve is steep, issuers
may swap longer-term obligations for shorter-term payments by receiving fixed longer-term rates, and paying short or floating rates, thus widening the bond-swap spread.
Sample emerging market economies for panels 1 and 2 are confined to Brazil, Chile, Colombia, Hungary, Indonesia, Malaysia, Mexico, Poland, Romania, and South Africa
because of data availability. Panels 3 and 4 focus on the same [29]-country sample as Figure 3.5, subject to data availability. In panel 3, countries without 2013 data are
represented using 2014–15 figures instead. In panel 3, × indicates that NBFI assets grew at a faster pace than that of banks between 2013 and 2023. In panel 4, “AE
average” is the average pension fund asset allocation across all advanced economies. Asset allocation includes direct and indirect holdings via CIS. Where look-through is
available, CIS are decomposed into underlying asset classes; otherwise, the data fall within “CIS (when look-through unavailable).” “Other” refers to real estate, loans,
derivatives, and other alternative investments. Data come from both defined-benefit and defined-contribution pension plans. The figure uses OECD data for members,
while data for nonmember countries are compiled from national authorities or the largest pension funds directly. Data labels in the figure use International Organization
for Standardization (ISO) country codes. AE = advanced economy; CIS = collective investment schemes; NBFI = nonbank financial institution; OECD = Organisation for
Economic Co-operation and Development.

International Monetary Fund | October 2025 89

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

roughly half their assets to fixed-income securities21 significantly increase their holdings of sovereign debt
(Figure 3.9, panel 4), materially higher than the during periods of fiscal stress or sovereign distress.
average advanced economy fixed-income allocation of Furthermore, among state-owned banks, those with
30 percent. In many cases, high fixed-income alloca- weaker capitalization levels typically increase their
tions reflect that investable alternatives are limited and sovereign exposures the most, which could erode
that allocations face regulatory constraints. For frontier their vulnerable capital base and lead to the mispric-
markets in particular, small NBFI sectors and high ing of sovereign debt and crowding out of private
shares of assets held in sovereign debt indicate that sector credit (see the “Both Strengths and Weaknesses
some countries could already be in the state of high of LCBMs Are Relevant in Emerging and Frontier
domestic debt and low absorption capacity. Markets” section).
Since 2014, the rapid growth in local currency debt
issuance has coincided with a growing sovereign-bank
Sovereign-Bank Nexus Has Risen in Recent Years nexus, as reflected in the increase in banks’ govern-
Although large resident banks’ presence in LCBMs ment debt holdings as a share of their total assets
helps mitigate the impact of global shocks, excessive (Figure 3.10, panel 1). This may have been driven by a
government bond holdings by banks can exacerbate the combination of liquidity management needs, attractive
sovereign-bank nexus. The nexus involves three chan- yields—especially in high interest rate environments—
nels through which stress in one sector can propagate and, in some cases, moral suasion from authorities.
to the others (Chapter 2, April 2022 Global Financial Economies with higher debt burdens tend to have a
Stability Report). The first channel is through direct greater concentration of government bonds on their
exposure, specifically the impact of banks’ realized banks’ balance sheets (Figure 3.10, panel 2). The nexus
losses on large government debt holdings during a fiscal is particularly pronounced in emerging and frontier
crisis. The second relates to the safety net channel, markets with smaller and less developed capital mar-
whereby contingent liabilities from implicit government kets, where domestic banks often serve as the primary
guarantees of the banking system occur. The third vehicle for absorbing sovereign debt (Chapter 2, April
involves the macroeconomic channel, whereby weaken- 2022 Global Financial Stability Report).23
ing economic fundamentals simultaneously undermines As a result of the sovereign-bank nexus, default risks
sovereign creditworthiness and erode banks’ asset qual- of sovereigns and banks tend to move closely together,
ity through rising defaults and slower credit growth. and there is potential for a two-way causality. From
Banking and sovereign debt crises have frequently the perspective of international credit rating agencies,
occurred at the same time or in quick succession banks’ credit ratings are generally constrained by the
(Chapter 2, April 2022 Global Financial Stabil- sovereign’s “country ceiling,”24 with exceptions granted
ity Report). Sovereign-bank linkages can trigger only in rare cases. The cap reflects rating agencies’
self-fulfilling crises: As fears of a sovereign default rise, transfer and convertibility criteria, which assess the risk
banks with significant exposure to the sovereign are that a government might impose capital or exchange
seen as riskier. Furthermore, the failure of a domestic controls that restrict payments to nonresident creditors
bank heavily invested in domestic sovereign debt may for debt service. Consequently, a sovereign downgrade
result in wider spillovers to corporate lending and often triggers ratings downgrades for these “bound
other sectors of the economy. firms” that are subject to the cap (Chapter 2, April
In the context of financial repression, moral 2022 Global Financial Stability Report).
suasion has been recognized as a key reason for The close interconnectedness between sovereign
domestic banks to hold government securities (Deghi risk and banking sector risk is evident from the
and others 2022).22 These pressures are particularly co-movements of implied default risk. Observations
pronounced for state-owned banks, which generally from monthly data since 2010 indicate that during

21“Fixed-income securities” include both local and foreign cur- 23However, there is significant heterogeneity across economies.

rency instruments, encompassing domestic and foreign government Rising holdings of local currency debt by domestic banks has
and corporate issuers. occurred in jurisdictions with both deteriorating and improving capi-
22Financial repression may manifest through various channels, tal adequacy ratios, suggesting that government debt accumulation
including the directed placement of government securities with by banks could be driven by considerations other than capital.
state-owned banks, public enterprises, or government-controlled 24Country ceilings are not credit ratings but serve as a reference

institutional investors or the administrative setting of government that can limit the foreign currency ratings assigned to entities within
security yields at below-market levels (IMF 2021). a sovereign’s jurisdiction.

90 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Figure 3.10. Recent Trends in Sovereign-Bank Nexus Risks in Selected Emerging Market and Developing Economies
The strength of the nexus has increased since the onset of the COVID-19 Economies with a larger debt burden tend to see their banking sectors
pandemic. hold more sovereign debt.
1. Evolution of Banks’ Holdings of Domestic Government Debt 2. Banks’ Sovereign Debt Holdings, by Sovereign Debt to GDP for 2023
(Percent of total assets) (Percent)
25 Interquartile range Emerging market average Emerging markets 60
Frontier market average Frontier markets

Banks’ holdings (percent of assets)

20
40

15

20
10

5 0
Jan. Jan. Jan. Jan. 0 40 80 120 160
2014 2017 2020 2023 General government debt to GDP (percent)

A shock to sovereign implied default risk could affect banks’ expected Banks in emerging market and developing economies could face capital
default frequency, especially during stressed periods. shortfalls in the event of a hypothetical domestic debt-restructuring
scenario.
3. Sensitivity of Emerging Market Banks’ Expected Default Frequency to 4. Reported and Simulated Capital Ratio in Emerging Market and
Sovereign Implied Risk Developing Economies
(Beta, January 2010 to June 2025) (Percent of risk-weighted assets)
2.5 Current (interquartile range) Simulated (interquartile range) 25
Current (median) Simulated (median)
2.0 20
1.5
15
1.0
10
0.5

0 5

−0.5 0
Average Beta stress Beta stress 2014 15 16 17 18 19 20 21 22 23
(2 standard deviations) (maximum)

Sources: Bloomberg Finance L.P.; Moody’s; IMF Financial Soundness Indicators and Public Sector–Bank Nexus databases; and IMF staff calculations.
Note: For panels 1, 2, and 4, the emerging markets are Argentina, Brazil, Chile, Colombia, Dominican Republic, Hungary, Indonesia, Malaysia, Mexico, the Philippines,
Peru, Poland, South Africa, Thailand, and Türkiye, while frontier markets are Algeria, Botswana, Costa Rica, Ghana, Jordan, Kazakhstan, Kenya, Namibia, Nigeria, Pakistan,
Sri Lanka, Uganda, and Zambia. In panel 2, data are from the end of 2023 or latest available. In panel 3, sensitivity is on two-year rolling monthly observation for emerging
markets only, because of data availability constraints. Stressed data points represent observations that exceed 2 standard deviations, while the maximum bars indicate the
highest beta values recorded during the period. Sovereign implied default risk is from major emerging markets’ five-year credit default swaps. In panel 4, the scenario
analysis assumes a 40 percent haircut on government bond holdings.

stress periods, a one-standard-deviation increase debt restructuring event25 that haircuts local currency
in emerging market sovereigns’ implied default bond prices by 40 percent results in more than half
probability rate is associated with around half of a of banking sectors seeing regulatory capital ratios fall
standard deviation rise in banks’ expected default
frequency. This effect also appears to intensify during
periods of extreme stress, reaching near a one-for-
one relationship, on average, for emerging markets 25The data sets used for the hypothetical effect of domestic debt

(Figure 3.10, panel 3). restructuring consist of the ratio of the domestic banking sector’s
holdings to domestic government securities, risk-weighted assets, and
Using aggregated bank data from 15 emerging mar- regulatory capital from the IMF’s Financial Soundness Indicators and
kets and 13 frontier markets, a hypothetical domestic Public Sector–Bank Nexus databases.

International Monetary Fund | October 2025 91

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

below the critical 10 percent threshold26 (Figure 3.10, emerging markets,” “other emerging markets,” and
panel 4). A reverse simulation shows that most bank- “frontier markets” on the basis of the relative size of
ing systems with more than 20 percent of assets in their LCBM and availability of benchmark bonds.29
domestic government bonds are unlikely to withstand While findings provide actionable insight, data gaps
haircuts of 30 percent or more without breaching the limit comparability across dimensions, and results may
10 percent regulatory threshold.27 While the analysis not generalize to EMDEs where macrofinancial or
likely underestimates the extent of a sovereign distress market structures differ materially.
event (credit risk) by not considering other amplifica-
tion channels, the accounting effect alone highlights
the vulnerability of the banking systems in this sample Both Strengths and Weaknesses of LCBMs Are
of economies.28 Such fragilities could also be exposed Relevant in Emerging and Frontier Markets
during a noncredit risk event, such as an upward Macroeconomic and institutional conditions vary
shift in local yield curves (market risk) or forced sales widely across EMDEs, shaping the depth and resilience
during dash-for-cash episodes (liquidity risk). of LCBMs. Major emerging markets with more devel-
oped LCBMs tend to exhibit stronger economic funda-
mentals, including deeper domestic institutional investor
Deepening Local Currency Bond Markets bases and lower financial dollarization. These features
to Enhance Financial Stability help anchor investor confidence, lower sovereign risk pre-
Developing LCBMs requires not only sound mia, and support the formation of a yield curve (Figure
macroeconomic fundamentals and adequate domestic 3.11, panel 1). In contrast, frontier markets often display
financial savings but also a strong policy framework weaker and more volatile macroeconomic conditions,
and robust financial market structure to channel these limiting their capacity to price risk and sustain demand
financial savings into a well-functioning local market. for long-term local currency bonds. Financial systems
The IMF–World Bank Local Currency Bond in these markets remain bank dominated, with con-
Market Framework (2021) provides a structured, centrated investor holdings and limited intermediation.
data-driven approach to identify market development These structural weaknesses, when combined with heavy
gaps, assess absorption capacity constraints, and guide reliance on local currency debt issuance, can heighten
sequencing of reforms to develop LCBMs. This frame- the risk of financial repression, lead to crowding out of
work evaluates the stage of development of four core private sector credit, and increase financial stability risks
building blocks—money markets, primary market issu- (Chapter 2, April 2022 Global Financial Stability Report).
ance, secondary markets, and investor base—alongside Flexible exchange rate regimes and inflation-­targeting
two supporting blocks related to financial market frameworks have supported bond market development
infrastructure (FMI) and legal-regulatory systems, in many major emerging markets by anchoring expecta-
against the backdrop of macroeconomic and institu- tions and reducing volatility (Figure 3.11, panel 2).30 In
tional enabling conditions. contrast, frontier markets often face higher inflation vol-
This section applies the LCBM framework to assess atility, reflecting weaker policy anchors and the absence
market structure in 37 EMDEs with sizable LCBMs, of credible inflation targeting frameworks alongside
drawing on available data and recent technical assis- greater exchange rate pass-through. The exchange rate
tance experience. Economies are grouped as “major can affect bond yields in two ways. In some econ-
omies, central banks adjust policy rates to stabilize
26Minimum capital ratios vary between countries, and the
the exchange rate, and this directly moves short-term
10 percent of risk-weighted assets threshold assumed may exceed the
minimum ratio required in some jurisdictions. However, a decline yields. In others, expectations of depreciation affect
below this threshold is likely to trigger corrective supervisory action yields indirectly by lifting short-term rates through
(Barrail, Dehmej, and Wezel, forthcoming). the inflation channel and pushing up long-term yields
27Unlike the fixed losses-given-default assumption of 40 percent

used in the simulation presented in Figure 3.10, panel 4, a reverse

simulation calculates the maximum losses that banks can sustain on 29The list of economy groupings is indicated in Online

their local currency government bond holdings while still maintain- Annex 3.2.
ing regulatory ratios above the 10 percent threshold, based on their 30As seen in many EMDEs, inflation-targeting frameworks do

initial capital ratio. not automatically guarantee a reliable yield curve. The operating
28Countries that undertook domestic debt restructuring in recent framework should include well-defined goals, robust decision
years were also characterized by a limited capacity of their domestic making, a coherent strategy, operational procedures, and effective
banking systems to transmit shocks to the wider economy (IMF 2021). communication.

92 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Figure 3.11. Building Blocks of Local Currency Bond Market Resilience in Emerging Market and Developing Economies
Major emerging markets typically display better macroeconomic Credible inflation targeting and FX regimes experience lower
fundamentals. macroeconomic volatility conducive for LCBMs.
1. Macroeconomic Preconditions, by Economy Group 2. Monetary Policy Regimes in EMDEs, End of 2024
(Normalized index: 0 = weakest; 1 = strongest) (Share of economies, left scale; median FX/inflation volatility (five-year rolling
window of standard deviation)
NBFI size AEs 1.0 6
1.0 Major EMs
Other EMs 0.8
0.8 FMs
0.6 4
0.6
Public debt 0.4 De-dollarization
0.2 0.4
2
0
0.2

0 0
Major EMs Other EMs FMs
FX Regime Monetary Volatility
Free floating Floating Policy Regime FX volatility
FX volatility Inflation volatility Soft peg Other managed IT Inflation
Non-IT volatility

Interest rate-based monetary policy framework anchors the short end of the Many EMDEs operate with wide interest rate corridors and weak repo
yield curve through better monetary policy transmission. contract enforceability, constraining money market depth.
3. Monetary Policy Frameworks in EMDEs 4. Policy Corridor Widths and Repo Contract Enforceability in EMDEs
(Share of economies, left scale; coefficient of determination, right scale) (Percent)
100 1.0 Percentage of economies with ICMA legal opinions
confirming enforceability (core coverage) (left scale)
80 0.8 Average width of monetary policy corridor (right scale)

60 0.6 100 2.5

80 2.3
40 0.4
2.0
20 0.2 60
1.8
0 0 40
Major EMs Other EMs FMs 1.5
Percentage of economies with O/N money market reference rate (left scale) 20 1.3
Percentage of economies with transaction-based O/N operational target for
monetary policy (left scale) 0 1.0
Consistency of monetary transmission to short end yields (R2) (right scale) Major EMs Other EMs FMs

(Figure continues next page)

through higher currency risk and term premia. In operate under interest-rate-based frameworks—often
practice, emerging markets are more often shaped by linked to formal inflation-targeting regimes in which
the direct policy rate channel, reflecting the stronger the policy rate and transaction-based overnight
credibility of inflation-targeting frameworks, while reference rate anchor the short end of the yield curve
frontier markets are more exposed to the risk premia (Figure 3.11, panel 3).31 Many frontier markets rely
channel. Both fixed and flexible exchange rate regimes on indicative or administratively set overnight rates,
can support bond market development if credible, but with weak links to underlying trades. Even where
flexible regimes reduce the possibility of abrupt cur-
rency depreciation. The latter also generate demand for 31Achieving
reliable reference rates for market participants would
hedging instruments, which can help deepen LCBMs. require the rates to be based on transparent computation method-
ology and provisions for periods when markets are volatile or under
Foundational money market features remain uneven stress (EBRD 2016). For detailed guidance on the transition to an
across EMDEs. Major emerging markets typically interest-based monetary policy framework, refer to IMF (2022).

International Monetary Fund | October 2025 93

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 3.11. Building Blocks of Local Currency Bond Market Resilience in Emerging Market and Developing Economies
(continued)
Real bond yields in some FMs, indicate a high-risk premium signaling Fiscal dominance is more pronounced in FMs, while sovereign-bank nexus
underlying stability risks in less-developed LCBMs. and nonresident exposures are higher in other emerging markets than in
peers.
5. Short-Term Real Yields on Government Bonds and Yield Curve Slope 6. Investor Base Composition of LCBMs in EMDEs, by Economy Group, 2024
in EMDEs (Percent)
(Percent)
Major EMs Other EMs FMs Major EMs Other EMs FMs
70
BWA CRI
3 60
UGA
Average 2s10s yield

50
curve spread (pp)

ZMB
2 HUN FM avg.
40
Major EM avg.
1 30
Other EM avg. 20
NGA
0 10
TUR
0
−2 0 2 4 6 8 Central Banks NBFI Nonbank Nonbank Nonresident
Average short-term real yield (percent) bank public others

Sources: Bloomberg Finance L.P.; Fitch Ratings; IMF Annual Report on Exchange Arrangements and Exchange Restrictions, International Financial Statistics, and Monetary
Operations and Instruments Database; ICMA, national sources; and IMF World Economic Outlook database.
Note: Panel 1 covers 46 countries: 13 AEs and 33 EMDEs (12 major EMs, 5 other EMs, and 16 FMs). De-dollarization is 100 minus the share of FX deposits in total deposits
(set at 100 percent for AEs); the prior year was used if the latest was unavailable; source: Fitch). The NBFI size is proxied by 2023 NBFI assets (percentage of GDP). FX and
inflation volatilities are five-year rolling standard deviations to the end of 2024. Public debt is general government gross debt (percentage of GDP to the end of 2024).
Indicators are normalized from 0 (weakest) to 1 (strongest), and group averages represent each classification. Panel 2 is based on 36 EMDEs (12 major EMs, 7 other EMs,
and 17 FMs). Data is based on IMF staff calculations using national sources for monetary policy regimes; IMF Annual Report on Exchange Arrangements and Exchange
Restrictions (AREAER); IMF World Economic Outlook database; Bloomberg; and Haver. FX and inflation volatility are five-year rolling averages of standard deviations,
calculated through end-2024. Panel 3 shows the share of jurisdictions with ICMA GMRA legal opinions confirming enforceability for core counterparties. Coverage may
exclude some entities. R2 values are from country-level regressions of annual changes in short-end government bond yields (one-year maturities, or the two-year where
one-year is unavailable) on annual changes in policy rates, 2016–25. Annual observations are constructed from monthly data averaged to yearly values. R2 indicates the
share of variation in yields explained by policy rate changes, that is, the consistency of monetary transmission. The sample covers 34 countries: 12 major EMs, 7 other EMs,
and 15 FMs. In panel 5, the dots show medians of annual averages over 2015–25. Real yields use the one-year maturity (or the two-year maturities if the one-year maturity
is unavailable). The y-axis shows the spread between 10-year and 2-year government bond yields. Outliers are defined as countries above the 95th percentile or below the
fifth percentile on either metric across the full sample (major EMs, other EMs, FMs). Group averages are shown as stars. Data labels use International Organization for
Standardization (ISO) country codes. Panel 6 covers 29 EMDEs (12 major EMs, 4 other EMs, 13 FMs). AEs = advanced economies; avg. = average; EMs = emerging
markets; EMDEs = emerging market and developing economy; FMs = frontier markets; FX = foreign exchange; GMRA = Global Master Repurchase Agreement; ICMA =
International Capital Markets Association; IT = inflation targeting; LCBM = local currency bond market; NBFI = nonbank financial institutional investors, excluding public
nonbank investors and other unclassified investors; non-IT = non–inflation targeting; pp = percentage point.

transaction-based rates exist, they are rarely used as haircuts, and operational limits on collateral circula-
operational targets, limiting price formation and weak- tion, reduce incentives to trade repos.33 While many
ening monetary policy transmission. emerging markets have adopted standardized legal
A deep and liquid repo market fosters interbank documentation supported by legal opinions, such as
and secondary bond market trading, anchors the the Global Master Repurchase Agreement, legal uncer-
short-term rate, and enhances financial stability by tainties around collateral enforcement and netting
reducing counterparty risk. Repo markets in EMDEs remain in several jurisdictions. Restrictions on short
lag advanced economies in scale and market depth.32 selling apply and fragmented settlement infrastructure
A key constraint for the interbank repo market is further limit the potential of interbank repo markets
collateral availability driven by a high degree of (Figure 3.11, panel 4). Gaps are larger in frontier
held-to-maturity portfolios. At the same time, high
33Another constraint is a divergence in collateral policy between
32Some major emerging markets have developed a deep repo central bank repos and interbank repos. The haircut determined
market. In Brazil, the central bank’s dominant role in liquidity should consider the maturity, quality, scarcity value, and price
management drives repo activity. Interbank repo activity in Mexico volatility of the underlying collateral; the term of the repo; and the
is a key driver of bond market liquidity. creditworthiness of the customer.

94 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

markets, where liquidity management is primarily a positive effect on the development of LCBM markets
quantity based, secured interbank markets are nascent, (Roldos 2004).36 Bond holdings by banks are shaped by
and repo transactions are mostly confined to central liquidity coverage requirements37 and preferential sover-
bank operations. eign risk weights, but in some economies also through
Most EMDEs are price takers in the primary mar- reserve or statutory liquidity requirements, which may
ket, paying positive real yields. Major emerging mar- result in financial repression. Central banks in frontier
kets generally sustain upward-sloping yield curves with markets hold a significant share of government secu-
moderate positive real yields, consistent with macro- rities, reflecting shallow investor bases and potentially
economic fundamentals, functioning price signals, and indicating elements of fiscal dominance, in addition
broader investor participation (Figure 3.11, panel 5). to high sovereign-bank nexus (Figure 3.11, panel 6).
In contrast, many frontier and smaller emerging Where prudential limits on foreign exchange (FX) posi-
markets record persistently high real yields and steep tions, caps on outward investment, and few alternative
yield curves, often reflecting elevated term premia from assets prevail, bank portfolios display strong “home bias”
inflation uncertainty, debt sustainability concerns, or and concentration in LCBMs.
liquidity and supply constraints. A few economies The availability of benchmark bonds is critical to
display flat or negative real yields, which may reflect the formation of deep and liquid markets. Annual bor-
financial repression, shallow investor participation, rowing plans anchored on credible medium-term fiscal
or credible disinflation episodes that compress term frameworks support predictable issuance and reduce
premia. The relationship between the slope of yield risks of fiscal dominance. Major emerging markets
curve and bank exposure to public sector debt is more maintain yield curve formation through predictable
negative in frontier markets.34 Similarly, the negative issuance, frequent reopening, and regular liability
real rate of returns during 2002–22 of pension funds, management operations, resulting in benchmark issues
which invest heavily in government bonds, have typically above $1 billion, that support market liquid-
been more pronounced in frontier markets relative to ity and index inclusion (Figure 3.12, panel 1). Many
emerging markets.35 Taken together, these persistent emerging markets have strengthened government cash
patterns over the past decade point to broader struc- flow forecasting and established cash buffers to support
tural differences across EMDEs, shaped by market buyback operations, while switch operations are usually
depth, the credibility of macroeconomic frameworks, cash neutral. Most other emerging markets issue in
and the scale of marketable debt in circulation. similar large sizes but with less consistency across ten-
Sovereign-bank links and exposure to nonresident ors. In contrast, many frontier markets issue smaller,
holders are more pronounced in other emerging markets irregular amounts, with mixed levels of transparency
than in major peers, highlighting relatively higher and weak auction discipline.38
vulnerability to funding shocks, rollover risks, and the Primary dealer (PD) frameworks in frontier markets
amplification of sovereign stress through the banking tend to prioritize auction participation over second-
system. In contrast, major emerging markets have larger ary market-making activities (Figure 3.12, panel 2).
NBFI participation, reflecting deeper financial systems High auction coverage, often above advanced econ-
and institutional investor bases. Pension reforms adopted omy norms, supports near-term funding but can
in some Latin American economies, which include vari- strengthen the sovereign-bank nexus, as banks end
ants of a funded, privately managed, and defined-con- up backstopping funding risks. In frontier markets,
tribution personal accounts retirement system, have had PD frameworks rarely include binding obligations for

34Based on a sample of 25 economies, with 14 emerging markets 36Many frontier markets and some emerging markets rely on a

and 11 frontier markets. Bank exposure to the public sector includes pay-as-you-go system. Therefore, pension assets remain shallow. Pen-
claims on the central government, local governments, and state- sion assets of 21 emerging markets averaged at 17 percent of GDP
owned enterprises, using data from Barrail, Dehmej, and Wezel in 2024, mainly contributed by Latin American emerging markets
(forthcoming). The correlation between bank exposure and the slope with an average of 27 percent, while for 11 frontier markets, pension
of the yield curve is –0.34 for frontier markets and –0.09 for emerg- assets stood at 14 percent (OECD 2024).
ing markets, indicating that frontier markets with larger bank hold- 37Given widespread bond illiquidity in many EMDEs, govern-

ings of public debt tend to exhibit flatter or inverted yield curves. ment bonds are treated as high-quality liquid assets because they are
35Based on a sample of 10 frontier markets and 16 emerging mar- eligible as collateral for central bank liquidity facilities.
kets, the average real rate of return during 2002–22 was found to be 38“Domestic Debt Securities Heat Map: 2023,” World

negative for 50 percent and 6 percent of these groups, respectively Bank, August 12, 2024, https://www.worldbank.org/en/data/
(OECD 2024). interactive/2024/08/12/domestic-debt-securities-heatmap.

International Monetary Fund | October 2025 95

 GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Figure 3.12. Local Currency Bond Market Development in Emerging Market and Developing Economies
Regular use of liability management operations supports building sizable PD systems support auction demand but often expose structural
benchmark bonds. weaknesses and deepen the sovereign-bank nexus.
1. Issuance of Benchmark Bonds and Regular Liability 2. PD System Support in Primary Markets
Management Operations (Percent, left scale; millions of dollars, right scale)
Percent of economies with a PD system
Regular LMOs BRA 100 Percent of each auction covered by PD obligations 2,000
40 Yes Average issuance size, millions of dollars
90
Average size of bonds outstanding

1,800
35 No 80 (right scale) 1,600
80 percent threshold
30 70 1,400
(US billion dollars)

MYS
CHL POL
25 60 1,200
THA
CHN 50 1,000
20 PHL ROU
TUR PER ZAF 40 800
15
SAU 30 600
10 DZA ARG IND
COL 20 400
KEN DOM TUN KAZ NGA
5 10 200
GHA EGY PAK HUN IDN
0 MEX 0
0
20 40 60 80 100 Major EMs Other EMs FMs
Percentage of bond universe above 1 billion US dollars

Quoting obligations for PDs and trading book activity remain limited, Bid-ask spreads and trade sizes reflect weaker liquidity in less developed
impeding bond market liquidity. LCBMs.
3. PD System Support in Secondary Markets 4. Secondary Market Liquidity and Average Trade Size
(Percent) (Basis points, left scale; billions of US dollars, right scale)
Percentage of countries where firm prices available Average bid-offer spread Average trade size (right scale)
100 60 10
Bank trading book as a percentage of total government
90 securities portfolio 9
80 50 8
70 40 7
60 6
50 30 5
40 4
30 20 3
20 10 2
10 1
0 0 0
Major EMs Other EMs FMs Major EMs Other EMs FMs

Sources: Bank of America; Bloomberg Finance L.P.; Deutsche Bank; JPMorgan; national sources; S&P Capital IQ; and IMF staff calculations.
Note: Panel 1 reports the average size of “benchmark bonds” based on the total outstanding bond universe and the number of outstanding bonds. The share of the bond
universe above US $1 billion equivalent is calculated from Bloomberg data, using exchange rates as of August 31, 2025. Excluded from the figure are countries with
0 percent of local currency government bonds outstanding above US $1 billion equivalent: Armenia, Botswana, Costa Rica, Jamaica, Jordan, Namibia, Sri Lanka, Uganda,
Vietnam, and Zambia. Panel 2 reflects staff assessments of PD frameworks in 37 EMDEs. Actual PD primary market coverage can be below 100 percent, but PDs are often
required to underwrite the full auction if demand is insufficient. The figure reports stated coverage obligations, not the 100 percent fallback underwriting. Panel 3 shows
the number of countries with firm secondary price obligations, based on 37 EMDEs, including those without PD systems. The analysis of trading books is based on banking
system averages for 2020–24 covering 24 EMDEs, measured as the share of total government securities portfolios, assuming most trading book assets are sovereign
securities. Panel 4 presents estimated bid-offer spreads and trade sizes for benchmark bonds in 26 EMDEs, based on typical market conditions. Data labels in the figure
use International Organization for Standardization (ISO) country codes. Data are as of August 31, 2025. EMDEs = emerging market and developing economies; LMO =
liability management operation; PD = primary dealer.

firm “two-way” quotes in the secondary market and books are significantly smaller than in many developed
often lack supporting infrastructure for PDs. This markets, impeding market liquidity. PDs in most
leaves few dealers with active trading books39 and cre- emerging markets include large global banks, which
ates shallow secondary market liquidity (Figure 3.12, help broaden market access, introduce high-frequency
panel 3). Even in major emerging markets, trading trading strategies, contribute to market liquidity, and
potentially narrow bid-ask spreads. However, these
39Shallow market depth, limited hedging instruments, and a
PDs can also act as conduits of increased sensitivity
predominance of buy-and-hold investors reduce trading incentives,
while regulatory costs and weak institutional support further con- during periods of market stress and sudden shifts in
strain dealer activity. global risk sentiment.
96 International Monetary Fund | October 2025
 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

High bid-offer spreads and weak pre- and post-trade access to domestic securities, reducing operational and
price transparency in frontier markets underscore per- legal barriers to entry.43
sistent secondary market inefficiencies. Wide spreads Efficient FMI44 enables safe, low-cost settlement
reflect small trade sizes, lack of benchmark securities, and supports investor confidence during stress. While
and concentrated buy-and-hold strategies, all of which most emerging markets have sound FMI in place,45
limit secondary market turnover (Figure 3.12, panel 4). FMI in frontier markets remains uneven, and oper-
Developing electronic interdealer platforms, ensuring ational gaps can amplify liquidity pressures and raise
pre- and post-trade transparency, and publishing a reli- risk premia by disrupting settlement and deterring
able yield curve have spurred trading activity in major investor participation, particularly during periods of
emerging markets.40 Some emerging markets (for market stress. Central clearing of repos, essential in
example, India, Malaysia, and South Africa) exhibit reducing counterparty risks, is present only in a few
bid-offer spreads comparable with advanced econo- major emerging markets, enhancing market liquidity
mies, but market liquidity is often concentrated in a by reducing risk-based trading costs and mitigating
few select benchmark bonds. investor uncertainty during market stress.46
Supporting market architecture (for example, hedg-
ing instruments, investor communication, and FMI)
shapes investor participation and market resilience. Policy Recommendations
Hedging markets supported by well-functioning Strengthening LCBMs is a crucial step toward
money markets enable both domestic and nonresi- reducing sovereign debt vulnerabilities and enhancing
dent investors to manage their exposure to interest financial resilience, alongside macroeconomic and fiscal
rate and exchange rate risk, lifting their participation stability. Where macrofinancial conditions are weak,
and liquidity. Emerging markets with deeper hedging rapid LCBM expansion that outpaces investor demand
markets have been able to weather shocks to liquid- can increase term premia, destabilize debt dynamics,
ity conditions better than others during observed and increase financial stability risks. Economies should
stress events (BIS 2024). Although hedging markets, therefore prioritize macroeconomic stability and strong
mostly FX derivatives, continued to grow in emerg- fiscal anchors that safeguard public debt sustainability.
ing markets, they have not always kept pace with Mobilizing adequate financial savings and channeling
issuance.41 Formal investor relations programs, timely them into LCBMs is key to strengthening absorption
transparency, and financial literacy programs have capacity and supporting LCBM development.
been effective in harnessing household and corporate Reform priorities should focus on strengthening
investments, including through mutual fund products market absorption capacity by developing the domestic
covering government bonds. institutional investor base. Deepening LCBMs must
While many EMDEs have relatively open capital be seen within broader financial sector development.
accounts, the absence of well-functioning LCBM and National pension system design, including shifts from
FX markets can deter direct nonresident investment.42
Establishing links with international central securities 43For example, the introduction of international central securities
depositories provides secure, standardized, cross-border depositories to Russia in 2013 allowed foreign investors to switch
from proxy instruments to direct holding of government bonds.
40Electronic interdealer platforms for government bonds were Issuance of credit-linked notes disappeared within two months, and
developed with public sector support in markets like Brazil, India, liquidity of government bonds surpassed that of cross-currency swaps
and South Africa. Malaysia, Mexico, and Thailand implemented (Lu and Yakovlev 2017).
dissemination portals to enhance pre- and post-trade transparency in 44FMIs provide services for LCBMs that facilitate the clearing,

over-the-counter markets to increase transparency. settlement, and recording of financial transactions, including the
41FX derivatives are generally more prevalent in emerging markets, transfer of securities and funds.
but Brazil, Chile, Mexico, and South Africa are among those with 45As of January 2025, large emerging markets have self-attested to

fairly balanced hedging markets. implementing the Committee on Payments and Market Infrastruc-
42Nonetheless, foreign investors may gain synthetic exposure tures and International Organization of Securities Commissions
via derivatives or proxies, for example, interest rate swaps (see also (CPMI-IOSCO) Principles for FMIs, although self-assessment does
Chapter 1), total return swaps, and credit-linked notes, which still not imply full compliance or guarantee sound operation in practice
influence yields and exchange rates. These instruments are attractive (Bank for International Settlements, Committee on Payments and
for nonresident investors who are either reluctant or not allowed by Market Infrastructures, and International Organization of Securities
their mandates to open an account in a local clearing and settlement Commissions, Update to the Level 1 online tracker, information as
infrastructure. Offshore over-the-counter cross-currency swaps or of January 2025).
nondeliverable forwards can also provide nonresident investors access 46Brazil, China, and India have implemented a central clearing

to local bond markets. counterparty for clearing of repos.

International Monetary Fund | October 2025 97

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Table 3.1. Policies to Improve Resilience and Develop Deeper Local Currency Bond Markets
Country Type Primary Market Focus Investor Strategy Money and Secondary Market
Major emerging Consolidate benchmark issuance to reduce Deepen institutional investor base Expand use of interbank repos and encourage
markets fragmentation by promoting long-term savings larger trading books for primary dealers
institutions and banks
Establish central clearing counterparties where
appropriate
Other emerging Accelerate benchmark bond issuance Develop domestic institutional Improve the primary dealer framework
markets and initiate regular liability investor base to deepen local Encourage the use of repo contracts based
management operations currency issuance on internationally recognized master
agreements
Frontier markets Build benchmark bonds, establish issuance Reduce overreliance on banks and Implement interest rate–based monetary
rules and auction discipline, and facilitate nurture nascent institutional policy and encourage use of money market
greater coordination with monetary investors reference rates and repos
operation
Strengthen government cash management
Source: Authors.

pay-as-you-go system to funded systems, reflects social bond market reforms, integrating these priorities into
choices and takes time to implement. Complementary surveillance and IMF-supported lending programs to
funded arrangements, such as mandatory contributions ensure durable progress (Box 3.2). Table 3.1 sum-
to privately managed plans or provident funds can marizes the key priorities in LCBM development for
support LCBM development (BIS 2019). Additional different groupings of EMDEs.
instruments, including pension-like insurance prod- To improve LCBMs, appropriate steps should be
ucts, can help mobilize financial savings. Tax incentives taken across all the key LCBM building blocks:
can encourage participation in voluntary pension a. Sound monetary policy frameworks and deeper
systems, complementary programs, and life insurance money markets are vital (IMF 2015). FMs should
products. As the institutional investor base matures, adopt and operationalize interest rate–based
gradually relaxing mandatory investment requirements frameworks using credible policy instruments and
and adopting “prudent person rules” would reduce transaction-based reference rates. Developing repo
overexposure to government securities and broaden markets is critical for robust money markets (IMF
investment allocation choices. Finally, greater finan- and World Bank 2021). Major emerging markets
cialization of household savings through collective can further expand collateral reuse and term repos
investment schemes requires strong legal and regula- as well as facilitate access to NBFIs. Mitigating sys-
tory frameworks to protect investors and a neutral tax temic risks from repo markets during market stress
regime that avoids double taxation (IMF and World episodes will be critical, supported by consistent
Bank 2021). application of haircuts and margin requirements,
Shallow liquidity and concentrated investor bases enhanced transparency, and other risk management
can amplify spillovers from shocks, underscoring controls.
the need to deepen market liquidity. Strengthening b. Issuance strategies should emphasize predictabil-
LCBMs reinforces the Integrated Policy Framework47 ity and transparency to sustain demand and build
by improving monetary transmission, reducing cur- benchmark bonds to enhance market liquidity.
rency mismatches, and mitigating capital flow volatil- Aligning issuance with monetary operations can
ity. To this end, the IMF and World Bank are scaling stabilize systemic liquidity and reduce issuance vola-
up capacity development on money markets, FMI, and tility. To build a robust yield curve, frontier markets
should focus on a limited set of standardized bench-
47The
marks, while emerging markets can consolidate
Integrated Policy Framework has been developed by the
IMF to guide the joint use of monetary, exchange rate, macropru- liquidity through greater reopenings and regular use
dential, and capital flow management policies by considering policy of liability management operations.
trade-offs to manage external shocks, along with economy-specific c. Effective primary dealer frameworks and trading
frictions such as shallow markets, currency mismatches, foreign
investors’ limited appetite for emerging markets’ local currency debt, infrastructure remain essential for market liquid-
and poorly anchored inflation expectations. ity (Adrian, Fleming, and Nikolaou 2025). PD

98 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

obligations should be balanced and tailored to the insurance sector would be critical to expanding the
stage of market development, from indicative quotes institutional investor base in many EMDEs. Over
in frontier markets to firm quoting obligations in the medium term, institutional investor demand
major emerging markets. PD frameworks in frontier could be enhanced in frontier markets by align-
markets must balance privileges with obligations, ing investment mandates, solvency rules, and tax
while in major emerging markets the emphasis treatment. Clear and regular issuance communica-
should shift to enforcing quoting obligations and tion can help anchor investor expectations. Large
participation thresholds for PDs. emerging markets can build on pension reforms and
d. Policymakers should strengthen market micro- promote pooled investment vehicles like mutual
structure and systemic safeguards to enhance bond funds and voluntary pensions.
market resilience. Trading activity can be improved
in less liquid markets through implementation Nonresident participation in LCBMs should be
of electronic interdealer trading platforms and carefully considered, particularly in nascent frontier
enhanced market transparency through publication markets.
of reliable yield curves and better dissemination of a. The appropriate degree of nonresident participation
pre- and post-trade information. In major emerging in a domestic bond market is difficult to establish,
markets where repo activity has grown significantly, so both benefits and risks must be considered.
more robust clearing arrangements may be required. Where adequate levels of financial development
Establishment of central clearing counterparty could have not been attained and where financial market
reduce counterparty risks and dealer balance sheet structure—particularly FX and money markets—are
strain (Adrian, Nikolaou, and Wu 2025). shallow and macroeconomic stability is weak, a
e. Prudential treatment of sovereign bonds should avoid gradual and phased approach may be useful to open
reinforcing the sovereign-bank nexus. This can be participation of foreign investment in the LCBM.
done by gradually reducing incentives for held-to-­ Reliance on short-term debt instruments should be
maturity holdings and aligning liquidity coverage phased out. These can increase rollover risks and
requirements with global standards. Policymakers amplify volatility during stress, particularly when
should also remove legal and structural impediments they offer high real yields. Improving FX hedging
to secondary market trading of government securities tools in emerging markets can attract longer-term,
and support the development of hedging instruments, noncarry-trade flows, thereby mitigating capital
thereby enabling banks to hold more securities in outflows.
trading books and improving market liquidity. b. Managing high nonresident participation requires
f. Domestic sovereign issuance should incorporate strong institutions, especially in the context of an
sound contractual provisions for debt restructurings. integrated global financial environment. Appropriate
For economies, particularly at a nascent stage of FMI systems should support systematic monitoring
LCBM development, it can be useful for domestic of nonresident holdings and flows. Periodic assess-
bonds to include provisions relating to the negotia- ments of risks associated with nonresident holdings
tion process and restructuring mechanics to facilitate are important for formulating appropriate policy
orderly and predictable resolution if restructuring responses and building buffers. As an exception,
becomes necessary. where macroeconomic and prudential tools are
g. Investor base diversification to strengthen market insufficient, temporary and narrowly targeted capital
resilience should be a long-term priority and con- flow management measures may need to be con-
tingent upon broader financial sector development sidered in line with the IMF’s Institutional View
supported by coherent financial sector policies. (IMF 2012) to reduce excessive vulnerabilities with
Pension reforms and greater penetration of the life nonresident flows.

International Monetary Fund | October 2025 99

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 3.1. Local Currency Debt and Domestic Investors in the Corporate Sector in
Emerging Market and Developing Economies
Emerging market corporate debt has not risen (excluding China) corporate bonds and loans shows
strongly in recent years, in contrast with emerging that issuance has declined significantly from 2022
market and developing economy (EMDE) sover- through the third quarter of 2024, as the postpan-
eign debt. Aggregation of deal-level data for EMDE demic surge in issuance in 2021 waned (Figure 3.1.1,
panel 1). Bonds comprise around 80 percent of
This box was prepared by Jason Wu based on the work of emerging market corporate debt issuance, having
Jiayi Li. outgrown loans since the global financial crisis.

Figure 3.1.1. Corporate Debt in Emerging Markets

Issuance has declined, but local currency debt has gained ... especially local currency bonds ...
ground ...
1. Emerging Market Corporate Debt Issuance 2. Share of Local Currency Bonds in Total Emerging Market
(Billions of dollars, left scale, percent share, right scale) Corporate Bonds
(Percent)
Local currency Euros Before the global financial crisis (2003–06)
US dollars Local currency share After the global financial crisis (2013–16)
(right scale) Recent (2022–24) 100
90
600 60 80
500 50 70
60
400 40 50
40
300 30 30
200 20 20
10
100 10 0
Thailand
Malaysia
China
India
Philippines
Brazil
Indonesia
South Africa
Mexico
Colombia
Saudi Arabia
Chile
United Arab
Emirates
Argentina
0 0
2010
11
12
13
14
15
16
17
18
19
20
21
22
23
24

... and, to a lesser extent, corporate loans. Domestic investors entered emerging market corporate debt
markets during past crises.
3. Share of Local Currency Loans in Total Emerging Market 4. Cumulative Emerging Market Corporate Loan Issuance
Corporate Loans during the Global Financial Crisis and COVID-19
(Percent) (Billions of dollars)
Before the global financial crisis (2003–06)
100 GFC foreign investor GFC domestic investor
After the global financial crisis (2013–16)
90 COVID foreign COVID domestic
Recent (2022–24) 300
Cumulative change in debt volume

80 investor investor
70 200
60 100
50 0
40
30 −100
20 −200
10 −300
0 −400
Russia
China
South Africa
Saudi Arabia
India
Thailand
Malaysia
Brazil
Indonesia
United Arab
Emirates
Egypt
Mexico
Colombia
Philippines
Argentina
Chile

−500
1 2 3 4 5 6 7 8 9
Quarter since the onset of crisis

Sources: Dealogic; and IMF staff calculations.

Note: Data are aggregations from almost 90,000 bond and loan issuances for about 14,000 nonfinancial corporations in emerging market and
developing economies from 2000 to the third quarter of 2004. All shares are computed on a volume basis rather than on a deal count basis.
GFC = global financial crisis.

100 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

Box 3.1 (continued)

However, mirroring the sovereign bond market, by EMDE corporations having bolstered borrow-
the local currency share of emerging market ing relationships with banks headquartered in
corporate debt (by volume) has increased in foreign jurisdictions that prefer foreign currency
recent years, having jumped from 34 percent in loans.
2021 to almost 45 percent in the third quarter Domestic investors in corporate debt may have
of 2024 (Figure 3.1.1, panel 1, blue line). Across played a stabilizing role when markets were under
jurisdictions, corporations from emerging Asia strain. Following the onsets of the global financial
were especially active in issuing local currency crisis and the pandemic, domestic investors have
bonds. For example, on a volume basis, Malaysian increased their holdings of EMDEs’ corporate loans,
and Thai corporations have almost exclusively boosting debt volumes, while nonresident investors
issued bonds in local currency in the 2022 to have retraced their holdings notably (Figure 3.2.1,
2024 period (Figure 3.1.1, panel 2). The same panel 4). This finding resonates with the empirical
trend is observed qualitatively for corporate loans findings of this chapter: Higher domestic investor
(Figure 3.1.1, panel 3), although the growth of shares in EMDE sovereign bonds attenuate the
local currency loans may be somewhat attenuated adverse impacts of global shocks.

International Monetary Fund | October 2025 101

GLOBAL FINANCIAL STABILITY REPORT: Shifting Ground Beneath the Calm

Box 3.2. Case Studies of Local Currency Bond Market Reforms in Emerging Market and
Developing Economies Supported by IMF Technical Assistance
Deepening Local Currency Bond Markets and Laying the Foundations for a Robust Local
Mitigating Sovereign Debt Portfolio Risks in Currency Bond Market in Bangladesh
Georgia Confronted with higher financing needs and
Supported by a joint programmatic technical assis- falling concessional flows, Bangladesh identified
tance on debt management (2018–22), Georgia—a local currency bond market development as a policy
highly dollarized economy—made significant progress priority. A joint IMF–World Bank local currency
in deepening its domestic bond market and reduc- bond market diagnostic mission in 2023 identified
ing foreign exchange risk. While government debt major distortions—including interest rate caps, central
averaged around 40 percent of GDP between 2018 bank participation in auctions, and reliance on costly
and 2024, the share of domestic marketable debt nonmarketable domestic debt in the form of national
increased with tenors extending up to 11 years, savings certificates; all of which hampered price dis-
lowering the foreign exchange debt share from 81 to covery and market development.
70 percent. Foundational reforms followed, supported by condi-
Benchmark issuance underpinned market growth, tionality in the context of the IMF program. Con-
while liability management operations, including a ditions included transition to an interest rate–based
2024 switch operation, raised the average time to monetary policy framework, removal of the lending
maturity for domestic securities from 2.6 years to rate cap, elimination of central bank government bond
3.5 years (2018–24). A 2021 Eurobond ensured purchases, quarterly issuance calendars, publication of
refinancing and preserved international market a daily secondary market yield curve, and expanded
access. access through over-the-counter and stock exchange
The Ministry of Finance, with the support of the trading. National savings certificate rates were linked
National Bank of Georgia, launched the Market Mak- to market yields from 2025 to reduce market fragmen-
ers Pilot Program in 2020. This program improved tation. Follow-up technical assistance guided reforms
price discovery on benchmark bonds (approximately on primary dealer framework guidelines in June 2025,
$1.2 billion), although banks remain dominant inves- removing underwriting obligations and emphasizing
tors. Transparency enhancements aim to attract more market making activities by primary dealers.
nonbank and foreign investors, while diversification These efforts doubled the nominal stock of market-
remains a priority. able bonds between 2019 and 2024, with benchmark
bonds exceeding $500 million, securing FTSE Frontier
Emerging Market Bond Index inclusion. While
this may attract foreign investment, reducing the
This box was prepared by Arindam Roy and Bryan Gurhy. sovereign-bank nexus remains a key challenge

102 International Monetary Fund | October 2025

 CHAPTER 3 Global Shocks, Local Markets: The Changing Landscape of Emerging Market Sovereign Debt

References and Pitfalls to Market Stability.” Journal of International

Adrian, Tobias, Michael J. Fleming, and Kleopatra Nikolaou. Money and Finance 128 (July). https://doi.org/10.1016/
2025. “U.S. Treasury Market Functioning from the GFC to j.jimonfin.2022.102699
the Pandemic.” Staff Report No. 1146, Federal Reserve Bank International Finance Corporation (IFC). 2025. Financing
of New York. https://doi.org/10.59576/sr.1146 Firm Growth: The Role of Capital Markets in Low- and
Adrian, Tobias, Kleopatra Nikolaou, and Jason Wu. 2025. “Fos- Middle-Income Countries. Washington, DC: World Bank
tering Core Government Bond Market Resilience.” IMF Blog, Group. https://www.ifc.org/en/insights-reports/2025/
May 21. https://www.imf.org/en/Blogs/Articles/2025/05/21/ financing-firm-growth
fostering-core-government-bond-market-resilience International Monetary Fund (IMF). 2012. “The Liberalization
Arslanalp, Serkan, and Takahiro Tsuda. 2014. “Tracking Global and Management of Capital Flows: An Institutional View.”
Demand for Emerging Market Sovereign Debt.” IMF Work- IMF Policy Paper, Washington, DC. https://www.IMF.org/
ing Paper 14/39, International Monetary Fund, Washington, en/Publications/Policy-Papers/Issues/2016/12/31/
DC. https://doi.org/10.5089/9781484326541.001 The-Liberalization-and-Management-of-Capital-Flows
Bank for International Settlements (BIS). 2019. “Establishing -An-Institutional-View-PP4720
Viable Capital Markets.” Committee on the Global Financial International Monetary Fund (IMF). 2015. “Evolving Monetary
System Paper No. 62, BIS, Basel, Switzerland. Policy Frameworks in Low Income and Other Developing
Bank for International Settlements (BIS). 2024. Towards Liquid Countries.” IMF Policy Paper, Washington, DC.
and Resilient Government Debt Markets in EMEs. BIS https://www.imf.org/en/Publications/Policy-Papers/
Quarterly Review, March. Issues/2016/12/31/Evolving-Monetary-Policy-Frameworks-in
Barrail, Zulma, Salim Dehmej, and Torsten Wezel. Forthcoming. -Low-Income-and-Other-Developing-Countries-PP5000
“The Sovereign-Bank Nexus in Emerging Markets and Devel- International Monetary Fund (IMF). 2021. “Issues in
oping Economies: Trends, Determinants, and Macrofinancial Restructuring of Sovereign Domestic Debt.” IMF
Implications.” IMF Working Paper. International Monetary Policy Paper 2021/071, Washington, DC.
Fund, Washington, DC. https://doi.org/10.5089/9781616358112.007
Cerutti, Eugenio M., and Stijn Claessens. 2024. “The Global International Monetary Fund (IMF). 2022. “Transitioning
Financial Cycle: Quantities versus Prices.” IMF Working Operational Targets—From Reserve Money to Interest Rates.”
Paper 2024/158, International Monetary Fund, Washington, In Monetary and Capital Markets Department: Technical
DC. https://doi.org/10.5089/9798400282867.001 Assistance Handbook. Washington, DC: International
Deghi, Andrea, Salih Fendoglu, Tara Iyer, Hamid Reza Monetary Fund. https://www.IMF.org/en/Publications/
Tabarraei, Yizhi Xu, and Mustafa Y. Yenice. 2022. “The miscellaneous-publications/Issues/2022/01/18/monetary
Sovereign-Bank Nexus in Emerging Markets in the Wake of -and-capital-markets-department-technical-assistance
the COVID-19 Pandemic.” IMF Working Paper 22/223, -handbook
International Monetary Fund, Washington, DC. https://doi. International Monetary Fund (IMF) and World Bank. 2021.
org/10.5089/9798400222382.001 “Guidance Note for Developing Government Local
Ebeke, Christian, and Yinqiu Lu. 2015. “Emerging Market Currency Bond Markets.” Analytical Note 2021/001,
Local Currency Bond Yields and Foreign Holdings: A International Monetary Fund, Washington, DC.
Fortune or Misfortune?” Journal of International Money and https://doi.org/10.5089/9781513573922.064
Finance 59 (December): 203–19. https://doi.org/10.1016/ Lu, Yinqiu, and Dmitry Yakovlev. 2017. “Exploring
j.jimonfin.2015.07.006 the Role of Foreign Investors in Russia’s Local
Eckhold, Kelly, Julia Faltermeier, Darryl King, Istvan Currency Government Bond (OFZ) Market.”
Mak, and Dmitri Petrov. 2024. “EMDE Central Bank IMF Working Paper 17/28, Washington, DC.
Interventions during COVID-19 to Support Market https://doi.org/10.5089/9781475577587.001
Functioning.” IMF Working Paper 24/101, International Organisation for Economic Co-operation and
Monetary Fund, Washington, DC. https://doi.org/ Development (OECD). 2024. “Statistical Annex.”
10.5089/9798400275845.001 In Pension Markets in Focus 2023. OECD.
European Bank for Reconstruction and Development (EBRD). https://doi.org/10.1787/28970baf-en
2016. “A Framework for Developing Money Markets in Roldos, Jorge E. 2004. “Pension Reform, Investment
Frontier and Emerging-Market Economies.” EBRD, Paris. Restrictions, and Capital Markets.” IMF Policy Discussion
Ho, Edmond. 2022. “Foreign Participation in Local Currency Paper 04/4, International Monetary Fund, Washington, DC.
Government Bond Markets in Emerging Asia: Benefits https://doi.org/10.5089/9781451973730.003

International Monetary Fund | October 2025 103

Timely.
Topical.
Free.
Global economics at your fingertips
IMF.org/pubs | bookstore.IMF.org | eLibrary.IMF.org
IN THIS ISSUE:
CHAPTER 1
Shifting Ground beneath the Calm:
Stability Challenges amid Changes
in Financial Markets
CHAPTER 2
Risk and Resilience in the Global
Foreign Exchange Market
CHAPTER 3
Global Shocks, Local Markets:
The Changing Landscape of
Emerging Market Sovereign Debt

GLOBAL FINANCIAL STABILITY REPORT OCTOBER 2025