RESEARCH REPORT

Picture credit to DALL-E 3, prompt based on report title and abstract

Climate scenarios in banks

a case study
Lena Sofie Buchwald and Isabel Tretow
October 2023
 Climate scenarios in banks – a case study

Lena Sofie Buchwald and Isabel Tretow

October 20, 2023

The report is based on a Master thesis, which was supervised and refined by Ulf
Moslener and Sebastian Rink. Part of SATISFY under the research project Sustainable
Finance and Climate Protection by the German Federal Ministry of Education and
Research (grant number: 01LA2210A)

We thank Alix Auzepy, Fernanda Ballesteros, and Alexandra Hüttel for their helpful
comments.

1
 Abstract

In this report, we develop and apply a typology that helps banks differenti-
ate climate scenarios based on their characteristics and discuss four use cases.
With the increasing importance of climate scenarios for assessing the risks and
opportunities of climate change in the financial system, banks face the question
of which climate scenarios to select for different use cases. We review academic
and grey literature to develop a typology that helps banks differentiate climate
scenarios using a heuristics approach. We differentiate commonly used scenarios
based on the typology. We then apply the typology through a case study in a
large German promotional bank. Being the first case study with a bank in this
field, the paper adds to the literature by providing a characterisation of climate
scenarios for banks and by supporting the selection of a climate scenario amongst
banks.

2
Contents

1 Introduction 7

2 Methodology 9

3 Theoretical Perspective on Climate Scenarios 11

3.1 Definition of Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3.2 The Role of Climate Scenarios . . . . . . . . . . . . . . . . . . . . . . . 15

3.3 Classification of Climate Scenarios . . . . . . . . . . . . . . . . . . . . . 18

4 Practical Perspective on Climate Scenarios 22

4.1 Scenario Use in the Banking Sector . . . . . . . . . . . . . . . . . . . . . 22

4.2 Climate Scenario Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4.3 Practical Usage of Climate Scenarios in the Banking Sector . . . . . . . 26

5 Model Development 36

5.1 Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

5.2 Model Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

6 Case Study 44

6.1 Refined Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

6.2 Challenges of Using Climate Scenarios . . . . . . . . . . . . . . . . . . . 53

7 Conclusions 56

3
References 58

4
List of Tables

1 Scenario characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2 Classification of the IEA, IPCC, NGFS, and Agora scenarios according

to van Notten et al. (2003) . . . . . . . . . . . . . . . . . . . . . . . . . 22

3 Types of climate scenarios identified among the IEA, IPCC, NGFS, and
Agora scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4 Model on the characteristics of climate scenarios . . . . . . . . . . . . . 40

5 Overview of experts interviewed . . . . . . . . . . . . . . . . . . . . . . . 46

6 Overview of relevant characteristics of climate scenarios for banks . . . . 47

5
List of Figures

1 Illustration of the applied methodology. . . . . . . . . . . . . . . . . . . 9

2 Overview of banks that have participated in the UNEP FI TCFD pro-

gramme. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3 Process for conducting a climate scenario analysis . . . . . . . . . . . . . 25

4 Use of different Scenario Providers among banks . . . . . . . . . . . . . 28

5 Use of different IEA Scenarios among banks . . . . . . . . . . . . . . . . 28

6 Use of different IPCC Scenarios among banks . . . . . . . . . . . . . . . 28

7 Use of different NGFS Scenarios among banks . . . . . . . . . . . . . . . 28

8 Split of adopted Perspective . . . . . . . . . . . . . . . . . . . . . . . . . 31

9 IEA split by Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

10 IPCC split by Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . 31

11 NGFS split by Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . 31

12 Risk Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

13 IEA split by Risk Type . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

14 IPCC split by Risk Type . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

15 NGFS split by Risk Type . . . . . . . . . . . . . . . . . . . . . . . . . . 33

6
1 Introduction

• The use of climate scenarios is recommended to better understand how the

risks and opportunities of climate change could affect a company’s business
model.

• According to the ECB (2021), especially banks experience difficulties in

using scenario analysis on climate- and environmental-related risks.

• To get a better understanding of what makes a climate scenario usable for

banks, we develop a model to identify characteristics of climate scenarios
which are relevant for different use cases.

About 190 countries signed the Paris Agreement in 2015 with the aim of lim-
iting global warming to well below 2°C, preferably 1.5°C (Nations, 2015). To reach
this goal, countries define nationally determined contributions (NDC), that is, how
they plan to contribute to achieving the temperature target. The European Union
(EU) aims to reduce emissions by 55% by 2030 compared to 1990 levels and reach
climate neutrality by 2050 (European Commission, n.d.-b, n.d.-c). The achievement
of these goals requires support from individuals and companies, which entails different
actors becoming Paris-aligned. The European Commission aims to foster this process
by increasing regulation to assess and report their exposure to climate-related risks
(European Commission, 2019).

The Task Force on Climate-Related Financial Disclosures (TCFD), consisting

of 31 members from across the G20 countries, released recommendations on climate-
related financial disclosure for companies in 2017 (on Climate Related Financial Dis-
closures (TCFD), 2022). These recommendations are a basis for developing mandatory
regulations on climate-related disclosures in various countries (Huiskamp, Brinke, &
Kramer, 2022). An essential part of the TCFD’s recommendations is using climate
scenarios to better understand how the risks and opportunities of climate change might
affect a company’s business (Huiskamp et al., 2022; on Climate Related Financial Dis-
closures (TCFD), 2017c). Progress in their implementation differs greatly depending
on, for example, the kind of company (Green and Sustainable Finance Cluster Ger-
many (GSFC Germany) (2019).

7
 Banks can mainly use climate scenarios in two ways: First, as part of risk man-
agement (i.e., how the climate affects the bank) and second, impact management, that
is, assess the impact of a bank on the climate. The appropriate implementation re-
quires the expansion of existing strategies and risk management processes. Especially
banks experience difficulties, with only 11% of the 113 banks surveyed already using
scenario analysis on climate- and environmental-related risks, according to a recent
study by the European Central Bank (ECB) (n.d.-a). For example, one reason is the
perception of climate scenario analysis as a “complex and daunting task” (Huiskamp
et al., 2022, p. 2).

Existing research has developed an implementation process for climate scenarios

(Huiskamp et al., 2022), providing an overview of environmental scenario studies (van
Vuuren, Kok, Girod, Lucas, & de Vries, 2012) or analysing climate scenarios from
the Intergovernmental Panel on Climate Change (IPCC) in detail (Bauer et al., 2017;
O’Neill et al., 2017; Riahi et al., 2017; Ritchie & Dowlatabadi, 2018). So far, no
study focusses on bridging the gap between the original scientific and policy-oriented
intent of climate scenarios and its use in the banking sector. Given the low adaptation
rates in this sector, as well as the fact that the Global Compact Network Germany
(GCNG) “identified basic challenges in the selection of suitable scenarios” (Global
Compact Network Germany (GCNG), 2019, p. 9), this report aims to provide a
better understanding of what makes a climate scenario usable from the perspective of
a bank. Thus, this report seeks to answer the following question:

Which characteristics of climate scenarios are relevant to the practical use of climate
scenarios in banks?

To answer the research question, a model is developed that combines characteris-

tics of scenarios outlined in the literature with the reasons given by banks for selecting
a particular climate scenario. The model is refined through a case study in a German
promotional bank. This report extends previous research in two ways. Firstly, we
provide a characterisation of climate scenarios in the context of the banking sector.
Secondly, the framework developed in the report supports banks in selecting appro-
priate climate scenarios.

The remainder of this report is structured as follows: Beginning with the general
methodology applied in this report in Section 2, Section 3 provides an overview of the
climate scenario literature and the characteristics of climate scenarios. This perspec-

8
tive is extended with the practical use of climate scenarios in Section 4. Subsequently,
a more detailed explanation of the definition of the underlying problem, the objectives
of the model, and the design of the model is given in Section 5. Section 6 discusses
the model in the context of a large German promotional bank. Section 7 concludes.

2 Methodology

The analysis of climate scenarios and the corresponding alignment of business

activities with climate goals have gained increasing relevance for banks over the last
years. This report aims to determine the relevant characteristics of climate scenarios
for their practical application in banks. Based on a review of existing literature in this
field and an extensive analysis of TCFD reports of 52 banks in the following sections,
we develop a model which combines scenario characteristics from both theoretical and
practical perspectives. Afterward, this model is refined through a case study in a
German bank to identify which characteristics are essential in practise. A graphical
illustration of the methodology is provided in Figure 1.

Figure 1: Illustration of the applied methodology.

9
 The subsequent sections, Sections 3 and 4, create a basis for developing the model
by reviewing the existing literature (point A in Figure 1). Precisely, Section 3 focusses
on climate scenarios from a theoretical perspective. Initially, the literature on scenar-
ios and their properties is reviewed, in general. Then, a scenario typology developed
by van Notten et al. (2003) is applied to climate scenarios from four different insti-
tutions to gain insights into their characteristics from a theoretical point of view. As
previous research recognises scenario typologies as a valuable tool to guide the selec-
tion of an appropriate climate scenario (Huiskamp et al., 2022), it is expected that
the application of the models will reveal relevant information on climate scenarios for
the subsequent model development. While Section 3 aims to create a common under-
standing of climate scenarios, Section 4 provides information on their practical use.
We analyse the use of climate scenarios in the banking sector (Section 4.3) by exam-
ining the TCFD reports of 52 banks participating in the UNEP FI TCFD Banking
Pilots (called “pilot banks” in the following). Therefore, each bank’s TCFD report,
or, if no TCFD report is available, the bank’s annual report is examined by keyword
search. With the information extracted, banks’ use of climate scenarios along the three
items perspective covered, risk type addressed, and sectors is analysed. These insights
support the development of the model and ensure its practical relevance.

To answer the research question, a model is developed on the characteristics of

climate scenarios relevant to banks, bringing together theoretical and practical perspec-
tives. This generates more insightful results than building a model from a theoretical
or practical perspective (Ven & Johnson, 2006).

To build the model, we apply an heuristics approach, a method that helps to

quickly find a solution to an existing problem (Hertwig, 2006). Precisely, this means
that we start from an existing set of criteria to derive more detailed characteristics
of climate scenarios. Cognitive heuristics “offer possibilities to inspire collective intel-
ligence, creativity, and the immersion into diverse futures” (Schirrmeister, Göhring,
& Warnke, 2020, p.15). Combining them with scenarios fits well as they “seek to go
beyond trends and extrapolations”. Thus, for example, Schoemaker (1991) already
analyses a heuristic approach to scenario planning. In this report, heuristics support
the identification of characteristics of climate scenarios. Therefore, a common heuristic
from climate science is used, consisting of the criteria ’legitimacy’, ’credibility’, and
’salience’ (see point B in Figure 1) (Cradock-Henry & Frame, 2021).

After the model has been developed, we refine it and derive characteristics par-

10
ticularly relevant for banks through a case study in a German bank (see point C in
Figure 1). This approach is suitable since case studies allow in-depth investigation of
a real-life setting (Ridder, 2017). Specifically, it provides a better understanding of
which and why specific characteristics are essential to make climate scenarios useable
for banks. The case study helps to differentiate four use cases of climate scenarios for
banks. Each use case represents a combination of ’risk type - risk perspective’ and
implies a different set of relevant characteristics. As part of the case study, we conduct
interviews with six employees from different departments in one German promotional
bank to gather various perspectives on the use of climate scenarios.

3 Theoretical Perspective on Climate Scenarios

• The literature review shows that climate scenarios can be applied in many
different contexts. A characterisation could help to differentiate and apply
them appropriately.

• Climate scenarios have become increasingly relevant for conducting climate

scenario analysis in companies.

• e show how climate scenarios are structured by categorising 17 climate sce-

narios from the IEA, IPCC, NGFS and Agora according to a model devel-
oped by van Notten et al. (2003).

• Although there are several similarities among the climate scenarios (e.g., as
they are using qualitative and quantitative data), differences occur between
and within climate scenario providers (e.g., whether they are outlining pe-
ripheral or trend dynamics).

The following Section 3.1 explains the broad applicability of scenarios and a pos-
sible way of differentiating them. Next, we describe the role of scenarios in assessing
climate-related risks and opportunities (Section 3.2). Lastly, 17 climate scenarios from
four different institutions are examined in more detail (Section 3.3).

11
3.1 Definition of Scenarios

Scenarios play an important role in the research field of future studies (Bishop,
Hines, & Collins, 2007): They outline possible future states and potential development
paths leading to a future situation. Thus, scenarios are a valuable tool to understand
what the future might look like based on a particular set of factors (Kosow & Gaßner,
2008). These factors are usually the result of making assumptions about the underlying
drivers. Therefore, scenarios should not be considered as exact predictions or forecasts,
but instead aim to illustrate hypothetical future pathways (Kosow & Gaßner, 2008).
Evaluating “alternative futures” (Amer, Daim, & Jetter, 2013, p. 23) allows organi-
sations to be prepared even for uncertain times, enhancing their ability to cope with
difficult circumstances. Kosow and Gaßner (2008) stress the usefulness of scenarios by
deriving four functions:

• the explorative and/or knowledge function,

• the communication function,

• the goal-setting function, and

• the decision-making and strategy formation function.

The exploratory function refers to the fact that scenarios can extend the scope of
thinking beyond conventional alternatives and challenge existing views by considering
a wide range of options (Greeuw et al., 2000; Kosow & Gaßner, 2008). Therefore, they
support the identification of opportunities and the generation of new insights (Amer
et al., 2013).

Scenarios can be used either as part of the communication or to foster commu-

nication. In the former, scenarios promote a commonly understood problem and the
exchange of ideas to direct the focus on communication processes and potential im-
provements. Regarding the latter aspect, scenarios help to convey information on a
topic, and thus increase understanding and stimulate discussions on this topic.

When setting goals, scenarios are a valuable method for developing an idea of
what the future should look like. Additionally, decision-making processes and strategic
planning are facilitated based on scenarios as various options can be tested (Kosow &
Gaßner, 2008).

12
 Given the broad applicability of scenarios, it is not surprising that they have
already been used “in many industries such as energy and sustainability” (Tiberius,
Siglow, & Sendra-Garcı́a, 2020, p.240) and across functions such as strategy, oper-
ations, and finance. Especially in times of increased uncertainty, as occurred after
the oil price shocks in 1973 and 1979, scenario planning received increasing atten-
tion from companies (Schoemaker, 1993). This phenomenon also holds more recently,
with scenario research increasing considerably since 2009, partially explained by the
financial crisis in 2007/2008, which affected people’s attitude towards risk and un-
certainty. Therefore, Tiberius et al. (2020) expect an increasing use of scenarios in
various industries.

Research on scenarios can be differentiated between applied and methodologi-

cal research (Tiberius et al., 2020). As an example of the former, Liu et al. (2015)
examined the withdrawal and consumption of water for electricity generation in the
United States by applying seven scenarios to explore the impact of growing electric-
ity demands and socioeconomic developments. Unlike this applied scenario research,
Ramirez et al. (2015) focused on the methodological aspect of scenario research. The
scenario methodology allows producing ’interesting research’, that is, ’research that
develops theory, is innovative and less formulaic’ (Ramirez et al., 2015, p.70). Specif-
ically, using scenarios as a scholarly method facilitates the identification of new areas
of research and the understanding of complex situations.

The scenarios and their use have shortcomings. For example, making appropriate
scenario assumptions is crucial. Even if the outcome of the scenario seems likely or
desirable, this might not be confused with forecasts or predictions, as the scenarios only
aim at showing a range of future developments (Kosow & Gaßner, 2008). Furthermore,
presenting them as possibilities makes them “psychologically less threatening to those
holding different worldviews” (Schoemaker, 1993, p.209).

Previous research outlines various characteristics and types of scenarios (Börjeson,

Höjer, Dreborg, Ekvall, & Finnveden, 2006; Girod & Mieg, 2008; Greeuw et al., 2000;
Kishita, Hara, Uwasu, & Umeda, 2016; Kosow & Gaßner, 2008; van Notten, Rotmans,
van Asselt, & Rothman, 2003). Most models show some unique aspects. However,
there are also several overlaps between the scenarios. Although, so far, no consensus
has been reached on a core set of characteristics or scenario typologies (Little, Hester,
& Carey, 2016), van Notten et al. (2003) aim to provide a ’broadly shared scenario
typology’ (p. 423).

13
 While other researchers already use scenario typologies to guide the selection of an
appropriate climate scenario (Huiskamp et al., 2022), they did not use the model from
van Notten et al. (2003). Instead, Huiskamp et al. (2022) develop an implementation
process called the climate resilience cycle to simplify and standardise the application
of climate scenario analysis for companies. As part of the first step of the cycle, the
selection of suitable reference scenarios, they use a scenario typology to facilitate the
decision-making process. The typology is based on the user’s knowledge needs and
clusters scenarios in predictive, explorative, and normative scenarios (Huiskamp et al.,
2022).

The comprehensibility of the van Notten et al. (2003) model stands out when
comparing it to other models (Börjeson et al., 2006; Huiskamp et al., 2022). Addition-
ally, their work is appreciated by other researchers (Amer et al., 2013; Bishop et al.,
2007; Börjeson et al., 2006). Thus, their typology is applied in this report to analyse
the characteristics of climate scenarios in more detail (see Section 3.3).

To build their model, van Notten et al. (2003) reviewed and analysed about
70 scenario studies and, afterwards, derive a typology consisting of three overarching
themes, the “project goal”, the “process design”, and the “scenario content”. First, the
“project goal” might vary from exploration, which is supposed to raise awareness and
foster creative thinking, to decision support, which means the possibility of analysing
future pathways regarding their desirability. Second, the “process design” focusses
on differentiating between an intuitive or a formal approach. The former refers to
qualitative knowledge used to develop scenario storylines; the latter describes scenario
development as a “rational and analytical exercise” (van Notten et al., 2003, p. 427).
The last theme, the “scenario content”, takes a closer look at the composition of a
scenario, for example, the variable types included, to distinguish between complex and
simple scenarios. These three themes are broken down into 14 scenario characteristics.
These characteristics, including more detailed explanations, can be found in Table 1.

14
Table 1: Scenario characteristics

Overarching theme Scenario characteristics Explanation

Project goal: exploration Inclusion of norms: descriptive vs nor- Influence of norms and values in
vs decision support mative scenario development
Vantage point: forecasting vs back- Starting point which is used to
casting develop a scenario
Subject: issue-based, area-based, insti- Focus of the scenario
tution-based
Time scale: long-term vs short-term Time horizon of the scenario
Spatial scale: global/supranational vs Geographical areas covered by a
national/local scenario

Process design: intuitive Data: qualitative vs quantitative Data conveyed in the scenario
vs formal Method of data collection: participa- Process for deriving information
tory vs desk research when developing the scenario
Resources: extensive vs limited Financial/research resources, in-
vested time in scenario project
Institutional conditions: open vs con- Leeway available during scenario
strained development

Scenario content: complex Temporal nature: claim vs snapshot Data availability during covered
vs simple time horizon
Variables: heterogeneous vs homo- Types and numbers of variables
geneous in a scenario
Dynamics: peripheral vs trend Type of future pathways de-
scribed by scenario
Level of deviation: alternative vs Variety of possible futures con-
conventional sidered in scenario
Level of integration: high vs low Degree to which study compo-
nents are united
Note. Adapted from “An updated scenario typology”, by van Notten et al., 2003

3.2 The Role of Climate Scenarios

Currently, the world moves at a critical junction from which various climate de-
velopments are possible. In light of these uncertainties associated with climate change
and the responses of societies to it, scenarios are a valuable tool for assessing different
future outcomes. Especially when considering the multiple challenges and possibilities
of climate change, scenarios help to examine them thoroughly. When applying scenar-
ios, organisations need information to comprehensively evaluate “climate-related risks
and opportunities on strategy, business model and financial planning” (Global Com-
pact Network Germany (GCNG), 2019, p. 2). To help overcome this challenge, the
TCFD, initiated by the Financial Stability Board (FSB) in 2015, provides guidelines

15
on the use of climate scenarios (on Climate Related Financial Disclosures (TCFD),
2017b).

Unlike traditional risk management, where the underlying models are based on
historical data, climate scenarios must deal with uncertainty, as they cannot rely on
past sources, as climate change-related risks have rarely been observed (Allen et al.,
2020; Chenet, Ryan-Collins, & van Lerven, 2021). Each ’what if’ scenario represents
a possible future state of the world, but does not provide accurate forecasts (Eis &
Schafer, 2019; Institutional Investors Group on Climate Change (IIGCC), 2019). To
be precise, climate scenarios describe various pathways of CO2e emissions leading to
different temperature increases based on assumed probabilities of occurrence. To de-
termine how a CO2e emission path develops, factors such as economic and population
growth, cost of technology, and sectoral-/country-specific CO2e emissions, are relevant
(GSFC Germany, 2019). TCFD defines an appropriate climate scenario as plausible,
distinctive, consistent, relevant, and challenging (TCFD, n.d.). Climate scenarios refer
to the two interdependent categories of climate-related risks, physical and transition
risks (TCFD, 2017b). As both risk categories are likely to impact organisations in the
future, we provide a short definition of both (GCNG, 2019).

Physical risks arise from climate variability, individual severe weather events, and
long-term climate changes and can be defined as their hazardous effect on human and
natural systems (Batten, Sowerbutts, & Tanaka, 2017; Eis & Schafer, 2019). They
can be divided into categories of acute and chronic physical risks. Acute physical risks
result from increased exposure to extreme weather events that cause immediate and
localised effects such as floods or hurricanes. Chronic physical risks are related to
long-term climate changes and changes in environmental conditions that might lead
to chronic heat waves or rising sea levels (Bundesanstalt für Finanzdienstleistungsauf-
sicht (BaFin), 2020; Eis & Schafer, 2019; Global Compact Network Germany (GCNG),
2019; on Climate Related Financial Disclosures (TCFD), 2017b). In addition to direct
effects, supply chains, operations and employees of organisations can be impacted by
indirect consequences of physical risks such as extreme changes in temperature and
water availability (BaFin, 2020; GCNG, 2019; TCFD, 2017b). These impacts might
affect the financial performance of organisations (TCFD, 2017b). The exposure of a
company to physical risks depends on various factors, including the geographic loca-
tion of the value chains and sites, internationalisation, and reliance on infrastructure
(GCNG, 2019). Although the insurance industry has already shifted the emphasis

16
to possible implications of physical risks, the banking sector has not yet extensively
analysed its consequences on its credit and lending portfolios (Eis & Schafer, 2019).

Currently, the world moves at a critical junction from which various climate de-
velopments are possible. In light of these uncertainties associated with climate change
and the responses of societies to it, scenarios are a valuable tool for assessing different
future outcomes. Especially when considering the multiple challenges and possibilities
of climate change, scenarios help to examine them thoroughly. When applying scenar-
ios, organisations need the information to comprehensively evaluate “climate-related
risks and opportunities on strategy, business model and financial planning” (GCNG,
2019, p. 2). To help overcome this challenge, TCFD, initiated by the Financial Sta-
bility Board (FSB) in 2015, provides guidelines on using climate scenarios (TCFD,
2017b).

Besides the differentiation in physical and transition risks, the risk perspective
needs to be considered, describing the risk’s direction of impact. 2 types of impact are
differentiated in the following: First, the outside-in perspective considers risks arising
from climate change that may impact a company. Second, the inside-out perspective
covers a company’s (negative) impacts on the climate like the use of climate scenar-
ios for aligning a bank’s portfolio (Green and Sustainable Finance Cluster Germany
(GSFC Germany), 2019; Hahnkamper-Vandenbulcke, 2021). Differentiating between
both, the risk type and the risk perspective, is crucial as different combinations of
both lead to other relevant characteristics of climate scenarios.

Primarily, climate scenarios have been designed to provide policymakers and sci-
entists with information on potential climate-related impacts from macroeconomic
analyses. Now that TCFD proposed the high relevance of climate scenarios for cli-
mate scenario analysis in businesses, the application of this tool has changed, and the
usability of climate scenarios needs to be assessed form this angle.

Transition risks are associated with the change transition to a lower-carbon econ-
omy and can materialise in financial losses or economic dislocation (Batten et al.,
2017; John, Khaykin, Pyanet, & Westheim, 2018). The financial and reputational
risks that organisations face depend on the velocity and focus of the transformation
process (TCFD, 2017b). Transition risks are related to changes in policy, legal, technol-
ogy, market (price) and reputation to meet climate-related adaptation and mitigation
needs (GCNG, 2019; TCFD, 2017b). Risks related to policy actions that promote

17
the transition to a low-carbon society are referred to as political risks. In contrast,
legal risks include the risks of litigation related to climate change. Technology risks
arise from the development of low-emission technologies with the potential to disrupt
existing systems and products. During the transition to a lower carbon economy,
changing markets and social expectations can materialise in market and reputation
risks, respectively (GCNG, 2019; TCFD, 2017b).

3.3 Classification of Climate Scenarios

As recommended by TCFD, most organisations use existing climate scenarios pub-

lished by various (inter)national institutions. These are often referred to as reference
scenarios since they “describe plausible future states of physical climate conditions
and transition pathways” (Huiskamp et al., 2022, p. 4).

To better understand the characteristics of climate scenarios, we apply the van

Notten et al. (2003) typology to climate scenarios which are relevant for the Ger-
man banking sector. This includes international climate scenarios provided by the
International Energy Agency (IEA), IPCC, and NGFS, as well as German climate
scenarios from Agora Energiewende / Verkehrswende, together with the Stiftung Kli-
maneutralität (Agora).

To carry out the classification, we analyse the most recently published climate
scenarios for each institution. This means the IEA scenarios used for the World En-
ergy Outlook 2021, the IPCC scenarios from the Sixth Assessment Report, the NGFS
scenarios published in June 2021, and the Agora scenarios from 2021. Each climate
scenario is examined separately to illustrate individual differences among them. We re-
view the literature on publicly available primary and secondary data from the agencies’
websites and previous researchers. In Table 2, the detailed results of the classification
are shown. Appendix A provides a more detailed introduction to each scenario and
explains why each scenario is categorised in this way.

In general, the typology of van Notten et al. (2003) reveals several similarities
among the climate scenarios. For example, all climate scenarios can be classified as
issue-based as they deal with climate change, a phenomenon that affects society. Be-
sides using a variety of resources, including qualitative and quantitative data, it is also
striking that all climate scenarios show the development of different variables like tem-

18
perature, GDP etc. over time. The largest differences become visible when comparing
the different climate scenarios of each institution. While some scenarios rather re-
flect existing trends and illustrate how the future might look like when following these
trends, other climate scenarios rather take a specific future outcome as a starting point
and outline a pathway on how to achieve this state.

19
20
21
4 Practical Perspective on Climate Scenarios

• Although scenarios have already played an essential role for banks, e.g., as
part of stress tests, they are now becoming increasingly relevant to assessing
climate-related impacts.

• Despite existing guidelines on using climate scenarios, the practical applica-

tion differs among banks: Given the variety of climate scenarios available,
banks not only choose different climate scenario providers, but may also se-
lect various climate scenarios from each provider, depending on the analysis.

The following section aims to understand how climate scenarios are used in prac-
tise by banks. Section 4.1 sets the stage for the general use of scenarios by banks.
Section 4.2 provides information on the regulatory context for using climate scenarios.
Afterwards, we examine the current status of their implementation in banks in Section
4.3.

4.1 Scenario Use in the Banking Sector

Although climate scenarios are a tool relevant for different industries, a cross-
sector analysis shows that the banking sector is behind other sectors in its imple-
mentation (TCFD, 2021b). As opposed to companies operating in specific industries,
banks can be exposed to many different industries and numerous customers through
their portfolios.

Although traditional risk analyses in banks, such as decision trees and Bayesian
models, are valuable tools for assessing risks in stable environments, they are rather
inappropriate for fundamental risks (e.g., political risks or uncertainties regarding the
economy/industry structure). Under these circumstances, scenario planning is more
suitable for “overcoming corporate blindspots and myopic thinking frames” (Schoe-
maker, 1993, p. 209). In the banking industry, scenarios are widely used as part
of stress tests (Cortés, Demyanyk, Li, Loutskina, & Strahan, 2020; Fernandes, Igan,
& Pinheiro, 2020; Sahin, de Haan, & Neretina, 2020). They illustrate what a bank
“might lose during a forward-looking, hypothetical severe economic downturn which

22
then gets translated into a forecast of regulatory capital ratios conditional on various
stress scenarios” (Cortés et al., 2020, p. 261). Generally, an adverse range of scenar-
ios is included to ensure consideration of tail risks. Additionally, the application of
common scenarios across banks leads to consistency in the supervisory standards used
across banks. Finally, publishing the results of the bank stress tests restores trust and
decreases market uncertainty (Sahin et al., 2020). Therefore, it seems reasonable that
the disclosure of the results of stress tests moves the stock and credit markets and
reduces the systematic risks of banks in the years following publication (Sahin et al.,
2020).

These findings are also supported by Fernandes et al. (2020), who find evidence
that stress tests contain relevant new information, especially during crises. A recent
example of such a crisis is the COVID-19 pandemic. In particular at the beginning
of the pandemic, scenarios were used to determine, for example, the impact of the
virus on the liquidity insurance function of banks (Acharya & Steffen, 2020) or to
estimate the size of the capital shortfall of European banks and to derive appropriate
countermeasures (Schularick, Steffen, & Tröger, 2020).

After the 2008 financial crisis, the banking sector faced increased regulation
(Cortés et al., 2020). As part of this, stress testing has also gained importance
in detecting potential risks to the banking sector and defining preventive measures
(Hernández, Garcı́a, Suárez, & Tarancón, 2022). Thus, in the EU, the ECB conducts
stress tests with its supervised banks at least once a year, which is a requirement
by EU law (European Central Bank (ECB), 2022c). In 2022, the ECB conducted a
climate risk stress test for the first time to analyse how well banks manage climate-
related risks and to derive best practises on handling climate-related risks. Although,
for now, this exercise does not affect banks’ capital levels, banks and supervisors are
expected to learn from it and become aware of climate-related vulnerabilities. The
climate stress test has shown that the use of a variety of climate scenarios reveals that
a bank’s exposure to physical risks could vary depending on the respective portfolio of
the bank. For example, banks with a large real estate portfolio could be more severely
hit by a flooding scenario than banks with a smaller real estate portfolio (European
Central Bank (ECB), 2022b).

Note. Adapted from “UNEP FI TCFD Banking Pilot Projects”, by UNEP FI, n.d.

23
Figure 2: Overview of banks that have participated in the UNEP FI TCFD
programme.

4.2 Climate Scenario Analysis

Understanding the effects of climate change on the business of a bank is key

to enhancing its resilience (TCFD, 2017c). Appropriate disclosure of the impact for
climate change gains importance to “enable stakeholders to better understand the con-
centrations of carbon-related assets in the financial sector and the financial system’s
exposures to climate-related risks” (Financial Stability Board (FSB), 2015, November
9, p. 4). Therefore, the TCFD provides guidelines on reporting climate-related is-
sues. Standardised procedures for disclosing climate-related information are crucial to
enhance transparency and comparability among organisations. In particular, climate
scenario analysis is a useful “tool for forward-looking assessment of financial climate-
related risks and opportunities” (Institutional Investors Group on Climate Change
(IIGCC), 2019, p. 9) in organisations. Its adoption helps to understand the dynamics
of climate change and allows companies to analyse future potential impacts to support
better decision making (Eis & Schafer, 2019; Huiskamp et al., 2022; John et al., 2018).

Scenario analysis is often mistakenly used in connection with projections and

sensitivity analysis. Technically, these two are not scenario analyses because they
update an existing parameter and not an entire set of variables (GSFC Germany,
2019). To support organisations in applying climate-related scenario analysis, the
TCFD (2017) provides an implementation framework with six steps (Figure 2):

24
 1. Governance structure: assess which stakeholders should be part of the approach
to identify the organisation’s climate-related opportunities and risks. In addition,
evaluate the participation of an organisation’s management. The TCFD recommends
direct and active management participation in the implementation process of climate
scenario analysis (Global Compact Network Germany (GCNG), 2019; on Climate Re-
lated Financial Disclosures (TCFD), 2017c).

2. Analysis of climate-related opportunities and risks: determine significant

climate-related opportunities and risks. Both physical and transition risks might af-
fect organisations in the future. Thus, it is recommended to analyse the company’s
exposure to physical and transition risks.

Figure 3: Process for conducting a climate scenario analysis

Note. Adapted from “A Process for Applying Scenario Analysis to Climate-Related

Risks and Opportunities”, by TCFD, 2017c

3. Selection of scenario and supporting tools: select climate scenarios and sup-
porting tools (GCNG, 2019). The TCFD recommends the use of different climate
scenarios since “the selection of a set of scenarios (not just one) that covers a reason-
able variety of future outcomes” (TCFD, 2017b, p. 27) is a crucial aspect of climate
scenario analysis. Two or three scenarios related to physical climate-related scenarios
or NDCs can be added to account for the organisation’s current circumstances (TCFD,
2017b). It is highlighted that the chosen scenarios should cover both favourable and
unfavourable future outcomes (TCFD, 2017b). According to the TCFD banks should
describe the resilience of their strategy in different climate scenarios, including a tran-
sition to a “2°C or lower scenario” (on Climate Related Financial Disclosures (TCFD),
2017a, p. 25). Scenarios must be tailored to the organisation’s business environment
(Huiskamp et al., 2022; TCFD, 2017b). Tools can be used to examine physical and
transition risks and thus facilitate applying climate scenario analysis (GCNG, 2019).

25
 4. Assessment of business impact: identify the main climate-related risks and
opportunities of the organisation and their financial impacts (GCNG, 2019; TCFD,
2017c). Implications on “revenue, expenditures, assets and liabilities or capital and
financing” (GCNG, 2019, p. 3) can be drawn from the results of the climate scenario
analysis.

5. Identification of countermeasures: consider possible actions to be taken in

response to the results (GCNG, 2019; TCFD, 2017b). Here, the double materiality
perspective, outlined by the European Commission in its Non-Financial Reporting
Directive, has to be taken into account, referring to two different perspectives for
reporting of climate-related information (2019), the outside-in and inside-out perspec-
tive. Institutions should incorporate the respective findings in two ways: First, in risk
management and second, in strategic planning. The former aims at better understand-
ing to what extent climate-related changes impact, for example, market, credit, and
operational risks (outside-in). The latter refers to identifying the changes required in a
company’s business model to transition to a low carbon economy (inside out) (Koberle,
Ganguly, & Ostrovnaya, 2021). Changes in business strategy can be implemented to
reduce identified risks and exploit detected opportunities (GCNG, 2019).

6. Disclosure: disclose and document critical information about the climate sce-
nario analysis. As the choice of climate scenarios is crucial to scenario analysis, the
selection process is particularly relevant for external stakeholders and other financial
institutions. Comprehensive external reporting with a high level of transparency is rec-
ommended (Global Compact Network Germany (GCNG), 2019; on Climate Related
Financial Disclosures (TCFD), 2017b).

4.3 Practical Usage of Climate Scenarios in the Banking Sector

Based on the reports of banks, we assess the current use of climate scenarios in the
banking sector. In particular, this includes examining the TCFD reports of 52 pilot
banks that have participated in Phase I, II or III of the UNEP FI TCFD programme
(Figure 3). We will use this analysis to build our model.

Since 2018, the TCFD has published annual status reports describing the “progress
on climate-related disclosure and TCFD implementation efforts, insights, and chal-
lenges” (on Climate Related Financial Disclosures (TCFD), 2021a). The TCFD pro-

26
vides information on the use of climate scenarios in “more than 1,600 companies” (on
Climate Related Financial Disclosures (TCFD), 2021b, p. 1), of which more than 280
can be allocated to the banking industry as of 2021 (on Climate Related Financial
Disclosures (TCFD), 2021b).

The pilot banks’ most recent TCFD reports available in English or German as of
31 May 2022 are downloaded and examined to obtain sufficient information on these
aspects. If no TCFD report was published, we derived the corresponding information
from the bank’s annual report. As only specific sections of the reports are important
for the analysis of practical usage of climate scenarios, the reports are investigated by
keyword search (“climate scenario”, “scenario”, “scenario analysis”, “portfolio align-
ment”, “physical”, “transition”, “risk”, “TCFD”, “sector”) to collect relevant infor-
mation. The analysis focusses on the use of the chosen scenarios with respect to the
perspective of the three items covered, the type of risk addressed, and the sectors anal-
ysed. Initially, we inspect the distribution of the scenarios and the scenario providers
among the 52 pilot banks. Afterwards, we analyse the scenarios regarding the three
items to visualise the different usages of climate scenarios in banks. In both steps,
the quantitative data are complemented by qualitative reasonings extracted from the
reports. Some pilot banks use scenarios by more than one provider and cover more
than one perspective, risk type, or sector. Banks that do not provide information on
any of the items are marked as “NA”.

First, we analyse which climate scenario providers the pilot banks select (Figure
4). For a better overview, the providers selected by less than five banks are labelled
’Other’. In general, international climate scenarios, such as the IEA, IPCC and NGFS
scenarios, are used most frequently (by more than 20 pilot banks each). Local climate
scenarios such as the Agora scenario (n.d.) are rarely used. A breakdown of the
respective scenarios of the three most commonly used scenario providers is shown in
Figure 5, Figure 6, and Figure 7. For analysing the use of the NGFS scenarios, a
different level of granularity is applied as many banks only provide information on the
scenario cluster (orderly, disorderly, hot house world) and not on the specific scenario
used.

27
Figure 4: Use of different Scenario Figure 5: Use of different IEA Scenar-
Providers among banks ios among banks

Figure 6: Use of different IPCC Sce- Figure 7: Use of different NGFS Sce-
narios among banks narios among banks

The frequent use of climate scenarios by the IEA, IPCC and NGFS is explained
by the perception of the banks of them as “reliable and established climate scenario
provider[s]” (K. Group, 2022, p. 88). As the reasons why banks select a specific cli-
mate scenario provider vary, we assume that the choice of a climate scenario provider
depends on the purpose pursued. Although the scenarios provided by the IEA are
chosen due to their ”sectoral and regional coverage, as well as recognition amongst
policy makers and financial institutions” (Commerce International Merchant Bankers
Group Holdings Berhad (CIMB Group), 2022, p. 86), IPCC scenarios are selected as
they “provided best scientific consensus at the time of the report” (Bank of America,
2019, p. 6). Sovcombank selected the NGFS scenarios because they “provide a com-
mon starting point for analysing climate risks to the economy and financial system”
(Sovcombank, 2022, p. 8). Another advantage of them is that they are “widely used

28
by industry regulators” (ING Groep (ING), 2021, p. 42) and “allow stakeholders to
explore transition risks, as well as physical risks” (Santander, 2021, p. 14). Apart
from these favourable characteristics, Citi criticises that “NGFS scenario[s] do [. . . ]
not have a sector-specific CO2 emissions pathway for oil and gas” (Citi, 2021, p. 57)
which would have been necessary since Citi’s energy portfolio includes oil and gas
companies.

In addition to these three widely used climate scenario providers, mainly British
banks (Barclays PLC, 2022; N. Group, 2022) and, due to their involvement in the UK,
Swiss banks (Credit Suisse, 2021; UBS, 2022) used the Climate Biennial Exploratory
Scenarios (CBES) developed by the Bank of England. Therefore, these scenarios are
applied by five of the banks included in the analysis (Figure 4). Since these scenarios
are “built upon a subset of the Network for Greening the Financial System (NGFS)
climate scenarios and expanded on the NGFS scenarios in certain aspects” (Barclays
PLC, 2022, p. 23), they are consistent with NGFS (UBS, 2022). Furthermore, it can be
derived that some banks based their climate scenario analysis on scenarios developed
by consulting firms for bank purposes (Canadian Imperial Bank of Commerce (CIBC),
2019; L. B. Group, 2022). Although they are specially tailored for the bank, they are
based on existing climate scenarios. In the case of Lloyds Banking Group, new climate
scenarios were designed “to reflect that the Group is a UK focused institution that
serves global clients” (2022, p. 66).

As climate scenarios from the IEA, IPCC, and NGFS are the most frequently used
among banks, we examine their application in more detail. Precisely, the scenario by
different providers seem to be deployed for different purposes. To understand the
underlying reasons for this and to derive relevant characteristics for the selection of
specific climate scenarios, the analysis focusses on perspective, risk type, and sector.

Perspective: The perspective can be either inside-out or outside-in for a given

scenario (reflecting double materiality). In the data from this study, 47 banks include
the outside-in perspective, while only 27 include the inside-out perspective (Figure 8).
Thus, pilot banks focus more on assessing how climate change affects the bank than
on how the bank affects the climate. When comparing the three scenario providers
regarding the perspective, both IPCC and NGFS cover outside-in risks predominantly,
while more than half of the analysed banks using IEA scenarios adopt an inside-out
perspective (Figure 9, Figure 10, and Figure 11). This is not surprising as “IEA sce-
narios describ[e] the efforts needed to reduce carbon dioxide emissions” (BNP Paribas,

29
2020, p. 43). Therefore, these were selected by the Paris Agreement Capital Transition
Assessment (PACTA) as applicable scenarios for the alignment of financial portfolios,
which is part of the inside-out perspective ((2DII), 2021). PACTA, developed by 2º
Investing Initiative together with several partners, including Frankfurt School, is a
methodology that helps banks align their portfolios with the goals of the Paris Agree-
ment.

Banks applying scenarios with a focus on the inside-out perspective usually aim
to fulfil similar purposes. Wells Fargo states that the use of climate scenarios “helps
[. . . ] [to] determine how best to optimize [their] financing activities to meet partic-
ular targets, which can be business- or science-based” (2021, p. 28). Especially the
alignment of a bank’s portfolio with the goals of the Paris Agreement or the impact
assessment of climate change in specific sectors is crucial (ING, 2021; Itaú Unibanco
Holding S.A., 2021). Therefore, climate scenarios that have an inside-out perspec-
tive help to “translate the Paris Agreement into carbon budgets and sector-specific
transition pathways, or ‘technology roadmaps’“ (Generale, 2022, p. 35).

30
Figure 8: Split of adopted Perspective Figure 9: IEA split by Perspective

Figure 10: IPCC split by Perspective Figure 11: NGFS split by Perspective

For the outside-in perspective, climate scenarios help to “gain a better overview
of the climate risk associated with the portfolio” (Den norske Bank (DNB), 2022, p.
80) and to identify the segments most affected by climate-related risks (CaixaBank,
2021). Additionally, these scenarios are a “credible basis to explore possible impacts
on the economy and financial systems” (FirstRand, 2021, p. 38) and to “develop a set
of overarching climate principles, to guide climate-related decisions and support the
group strategy” (S. B. Group, 2022, p. 12). TSKB highlights that from the outside-
in perspective, “climate scenario analysis aim[s] to develop and expand the level of
awareness as well as resilience, foresight and financial planning on how climate-related
physical and transition risks and opportunities could affect institutions over time”
(2021, p. 21). Additionally, DNB underlines that this “method enables projections of
probability of default adjusted for climate risk” (2022, p. 79).

Risk type: The impacts of climate change on banks can be distinguished into

31
transition and physical risks (outside-in perspective). However, this differentiation is
not applicable to the impact of banks on the climate (inside-out perspective) (Gourdel,
Monasterolo, Dunz, Mazzocchetti, & Parisi, 2022). Thus, the risk type can only be
analysed from an outside-in perspective. In general, the analysis of the data shows
that transition risks are addressed slightly more frequently by pilot banks using climate
scenarios than physical risks (Figure 12). With 21 banks using them, physical risks
constitute the majority of risk types addressed by IPCC scenarios (Figure 14). On the
contrary, IEA and NGFS, with 15 and 24 pilot banks, respectively, cover transition
risks in practise (Figure 13, Figure 15).

Physical risks are mainly analysed for “estimating the potential financial impact
of extreme weather events in the future” (Australia and New Zealand Banking Group
(ANZ), 2021, p. 7). Additionally, scenarios that include physical risks are selected to
“evaluate the resilience of sectors which have weight in the Bank’s loan portfolio being
vulnerable to climate change” ((TSKB), 2021, p. 21). Transition risks are essential
to assess how the counterparty’s financials could be affected by transitioning to a low
carbon economy (Credit Suisse, 2021).

32
 Figure 12: Risk Types Figure 13: IEA split by Risk Type

Figure 14: IPCC split by Risk Type Figure 15: NGFS split by Risk Type

Sectoral coverage: even though banks can decide whether they prefer using a
climate scenario analysis on, for example, a portfolio, sector, or asset level, analysing
a bank’s portfolio along its sectoral split is recognised as a best practise in the market.
This is because banks’ portfolios usually cover a variety of sectors whose exposure
to climate risks differs (European Central Bank (ECB), 2022a; Green and Sustainable
Finance Cluster Germany (GSFC Germany), 2019). Based on this, so-called heatmaps
can be derived, illustrating the vulnerability of the bank in various sectors. Therefore,
it is worthwhile to examine whether specific scenarios are especially useful for assessing
one sector.

To obtain a consistent breakdown, the sectors are extracted from the TCFD re-
ports in the first step and manually categorised according to the Global Industry

33
Classification Standard (GICS) in the second step. GICS is a renowned standard for
classifying industries and, due to the wide variety of sectoral breakdowns, helps to
aggregate the sectors appropriately to enhance the comparability of the sectors cov-
ered by the scenarios (Phillips & Ormsby, 2016). Some pilot banks mention within
their reports that they cover several sectors with climate scenarios, but do not provide
detailed information on which sectors. In this case, the covered sectors are classified
as ’Other’.

The analysis of the complete data set shows that energy (34 banks), real estate
(25 banks), and utilities (24 banks) are the three sectors most often covered by the
pi-lot banks. Compared to the general distribution, banks using the IEA scenarios
include consumer discretionary (10 banks), energy (17 banks), and utilities (15 banks)
disproportionately often, while the sectors real estate (7 banks) and financials (2 banks)
are covered relatively less frequently. In contrast, banks using the IPCC scenarios
mainly focus on the two sectors real estate (9 banks) and financials (5 banks), while
the sectors energy (6 banks) and utilities (1 bank) are less covered compared to the
overall sector breakdown. Banks using the NGFS scenario have a sectoral coverage
very similar to the overall sector distribution, but with relatively higher coverage of
the sector energy (15 banks) and slightly less containment of real estate (7 banks) and
utilities (6 banks).

Two relevant factors for choosing a specific sector can be identified. First, several
banks state that the vulnerability of a sector to climate-related risks was critical in the
selection process (outside-in) (Absa Group Limited, 2022; Carbon Disclosure Project
(CDP), 2022; Canadian Imperial Bank of Commerce (CIBC), 2019; Danske Bank,
2021). Second, the pilot banks (inside out) chose the sectors with the highest im-
pact on climate change (inside-out) (Bank of Montreal (BMO), 2021; Sanpaolo, 2021;
K. F. Group, 2021; K. Group, 2022). Shinhan Bank even goes one step further, ex-
plaining that “among high-emitting industries [. . . ] industries with the highest priority
– power generation and Oil and Gas – [were chosen] for analysis“ (S. F. Group, 2021,
p. 68). KBC Group also highlights the need to choose sectors “wherein KBC’s finan-
cial leverage is largest to support the transition to a low-carbon economy” (K. Group,
2022, p. 96). This is similar to the Standard Bank Group report stating that the
“client sectors in [. . . ] [the] portfolio which present the greatest potential for climate-
related opportunities” were chosen (S. B. Group, 2022, p. 12). Santander outlines that
besides “climate relevant sectors [. . . ] [also] individuals, SCF [Santander Consumer

34
Finance] and corporates customers” (2021, p. 20) were focused on the climate scenario
analysis.

The analysis of the practical application of climate scenarios in the banking in-
dustry shows that the three most frequently used climate scenario providers, IEA,
IPCC and NGFS, are all relevant in practise (Figure 4). Many banks even use cli-
mate scenarios by several institutions. However, national climate scenarios like the
on from Agora seem rarely important in practise. When evaluating the results of the
analyses focussing on the covered perspective, risk types, and sectors in Section 4.3,
different reasons for using climate scenarios of one of the three leading providers are
identified. IEA scenarios are used primarily to align the bank portfolio to a defined
temperature goal; therefore, they are applied mainly to adopt an inside-out perspec-
tive (Figure 9). Consistent with the perspective taken, carbon-intensive sectors are
selected to bring bank activities with the highest carbon-saving potential in line with
the chosen emission pathway (ABN AMRO, 2022; Danske Bank, 2021; ING Groep
(ING), 2020). Unlike IEA, IPCC and NGFS focus on an outside-in perspective (Fig-
ure 10 and Figure 11). IPCC scenarios focus on examining physical risks (Figure
14) and include sectors such as real estate, which are especially vulnerable to these
risks (Chatain, Ghosh, Preudhomme, & Mazzacurati, 2021). NGFS scenarios, on the
contrary, primarily cover transition risks and sectors most affected by this risk type,
such as energy (NGFS, 2021). Concludingly, while analysing the practical use of cli-
mate scenarios shows their broad applicability, certain tendencies are discernible as to
which climate scenario appears to be best suited in practise for which perspective or
risk type.

35
5 Model Development

• We extend the model of van Notten et al. (2003) to integrate characteris-

tics relevant for practical applications of climate scenarios in the banking
context.

• is should help banks in the climate scenario selection process by highlighting

particularly relevant characteristics.

• The model on the characteristics of climate scenarios relevant for banks is

based on the common climate science heuristic of the three criteria: ’legiti-
macy’, ’credibility’ and’salience’.

In this chapter, we build on the literature review and practical perspective and
develop a model of characteristics for climate scenarios that combines theory and
practise. Section 5.1 compares the results of both analyses. Section 5.2 presents the
model.

5.1 Problem Definition

Section 3.3 shows that the model of van Notten et al. (2003) can be applied
to analyse the climate scenarios by IEA, IPCC, NGFS, and Agora. However, the
typology allows only for differentiating three types of climate scenarios. These scenario
types are primarily derived from four characteristics that vary across climate scenarios:
normative / descriptive, forecasting / forecasting, trend / peripheral and conventional
/ alternative scenario. Looking at the three types of climate scenarios identified (Table
3), it is striking that the classification of a scenario as normative or descriptive seems
to impact the latter characteristics, as normative climate scenarios are also peripheral
and alternative scenarios, whereas descriptive scenarios are trend and conventional
scenarios. Table 2 shows further differences for the Agora scenarios, which can be
explained by the fact that it is a national climate scenario.

In general, most of the scenario characteristics are the same among the scenarios
analysed. Therefore, although the typology helps to better understand climate sce-
narios and how they are constructed, the concept of the model by van Notten et al.

36
Table 3: Types of climate scenarios identified among the IEA, IPCC, NGFS, and
Agora scenarios
Scenario characteristics Scenarios
Nature of
Inclusion Vantage Level of
Type the dynam- NGFS IEA IPCC AGORA
of norms point deviation
ics
1 normative backcast- peripheral alternative NZ; B2; DNZ; NZE; SSP1-2.6 KN2045;
ing DT SDS KN2050
(Orderly & Disorderly)
2 descriptive forecasting trend conventional NDCs; CP APS; SSP2-4.5 -
(Hot House World) STEPS
3 normative forecasting peripheral alternative - - SSP1-1.9; -
SSP3-7.0;
SSP5-8.5
Note. The scenario characteristics are adopted from “An updated scenario typology” by van Notten et al., 2003

(2003) does not allow us to draw any conclusions about unique characteristics since
several climate scenarios (also between different providers), leading to challenges when
differentiated on the model as seen in the practical analysis where the model is of
limited use. The model does not seem to sufficiently differentiate the characteristics
of climate scenarios and can be extended to fully capture their variety.

5.2 Model Design

Legitimacy, Credibility, and Salience. To develop a new model on the charac-

teristics of climate scenarios relevant for banks, a common heuristic of climate science
is applied. This refers to the three criteria ’legitimacy’, ’credibility’ and ’salience’.
Other researchers have already used these attributes as, for example, a framework
for “co-producing knowledge for action [. . . ] for many different sustainability issues”
(Cash & Belloy, 2020, p. 9) or to examine the evolution of IPCC scenarios against
these criteria (Girod, Wiek, Mieg, & Hulme, 2009). Specifically, ’legitimacy’ refers to
the transparency and fairness involved in the scenario construction process (Girod et
al., 2009), which is required to ensure trust in the knowledge provided by the scenario
(Cash & Belloy, 2020). Furthermore, ’credibility’ focusses on whether the informa-
tion inherent in a climate scenario is “meeting standards of scientific plausibility and
technical adequacy” (Cash et al., 2002, p. 4). The last criterion, ’salience’, some-
times called relevance, refers to the relevance of information for decision making. High
salience is given when information addresses, for example, problems relevant to the
decision-maker (Cash et al., 2002).

37
 Using the three criteria from climate science as a starting point allows us to
identify various characteristics of climate scenarios. “Their cohering function for as-
sessments of climate change impacts and adaptation options” (Hulme & Dessai, 2008,
p. 66) ensures the completeness of the model. Starting from these criteria, character-
istics specific to climate scenarios are identified. In the first step, the 14 characteristics
of van Notten et al. (2003) are assigned to a respective criterion, since the validity of
the van Notten et al. (2003) model has been recognised by various researchers (Amer
et al., 2013; Bishop et al., 2007; Börjeson et al., 2006). Afterwards, 12 criteria derived
from the practical analysis are added based on the insights gained from analysing the
TCFD reports (see Section 4.3). This ensures that the reasons considered by banks
during the selection of a climate scenario are included in the model. The complete
model can be found in Table 4.

Characteristics derived from theory. Van Notten et al. (2003)’s “project

goal”-theme aims at answering the question “why” a scenario is developed. Having
this question answered provides the fundamental basis for constructing a scenario and
ensures transparency on the objectives pursued with the scenario. Thus, the five
characteristics outlined under this overarching theme by van Notten et al. (2003) are
included under ’legitimacy’ (inclusion of norms (L1), vantage point (L2), subject (L3),
time scale (L4), spatial scale (L5)).

The theme ’process design’ contains two characteristics, the nature of the data and
the data collection method, which also belong to ’legitimacy’ (L6 and L7, respectively).
The last two, nature of the resources and institutional conditions, are assigned to
the second criterion, ’credibility’ (C1 and C2, respectively). This is because open
institutional conditions mean that more freedom is given to the scenario project, i.e.,
developing the scenario is not impacted by external interventions such as political or
personal relations. Being independent is a “key ingredient [. . . ] of scientific credibility”
(Ruggiero, 2007, p. 1) and, thus, the institutional conditions impact whether a scenario
is perceived as credible. In addition, the development of a scenario can be considered
more reliable through the extensive use of resources, since investing a great deal of
time or resources in scenario development allows to include various aspects relevant to
the scenario (Cradock-Henry & Frame, 2021). Hence, scenario users are expected to
perceive it as more credible.

All characteristics of the theme ’scenario content’ by van Notten et al. (2003)
are assigned to ’legitimacy’ (temporal nature (L8), nature of variables (L9), nature

38
of dynamics (L10), level of deviation (L11), and level of integration (L12)) as they
provide information on the general construction of the scenario and thereby increase
transparency.

39
Table 4: Model on the characteristics of climate scenarios

Criterion No. Scenario characteristics Form of characteristic Source

Legitimacy L1 inclusion of norms descriptive/normative van Notten et al. (2003), Greeuw et al. (2000)

L2 vantage point forecasting/backcasting van Notten et al. (2003), Greeuw et al. (2000)

L3 subject issue-/area-/institution- van Notten et al. (2003)

based

L4 time scale short-/long-term van Notten et al. (2003), Greeuw et al. (2000)

L5 spacial scale global/international/ van Notten et al. (2003), Greeuw et al. (2000)
national/regional /local

L6 nature of the data qualitative/quantitative van Notten et al. (2003), Greeuw et al. (2000),
Kishita et al. (2015)

L7 method of data collection participatory approach / van Notten et al. (2003), Greeuw et al. (2000)
desk research

L8 temporal nature snapshop/chain van Notten et al. (2003)

L9 nature of the variables homogeneous/heterogeneous van Notten et al. (2003)

L10 nature of the dynamics peripheral/trend van Notten et al. (2003), Greeuw et al. (2000)

L11 level of deviation alternative/conventional van Notten et al. (2003)

L12 level of integration low/high van Notten et al. (2003), Greeuw et al. (2000)

Credibility C1 inclusion of norms descriptive/normative van Notten et al. (2003), Greeuw et al. (2000)

C2 vantage point forecasting/bastcasting BNP (2022), CIMB Group (2022)

C3 subject issue-/area-/institution- Bank of America (2019), BMO (2021), TSKB (2021)

based

C4 time scale short-/long-term Banorte (2022), Citi (2021), ING (2021)

C5 spacial scale global/international/ ANZ (2021)

national/regional /local

C6 nature of the data qualitative/quantitative Deutsche Bank (2022), ING (2021)

Salience S1 inclusion of norms descriptive/normative BNP (2022), Credit Suisse (2021), Sovcombank
(2022)

S2 vantage point forecasting/bastcasting BMO (2021), Citi (2021), DanskeBank (2021), ING
(2021)

S3 subject issue-/area-/institution- Banorte (2022), Credit Suisse (2021)

based

S4 time scale short-/long-term Nomura (2021)

S5 spacial scale global/international/ BBVA (2020), ING (202)

national/regional /local

S6 nature of the data qualitative/quantitative Bank of Ireland (2022), ING (2020)

S7 method of data collection participatory approach / BMO (2021), Citi (2021), Credit Suisse (2021)
desk research

S8 temporal nature snapshop/chain AIB (2022)

Note. Criteria are adopted from “Salience, Credibility, Legitimacy and Boundaries: Linking Research, Assessment and Decision
Making” from Cash et al., 2002

40
 According to the assignment of the scenario characteristics identified by van Not-
ten et al. (2003), 12 characteristics are found under ’legitimacy’, two under ’credi-
bility’, but none under ’salience’. This observation can be explained by the fact that
“salience” refers to the transmission of information relevant to the users of the sce-
nario. Although the van Notten et al. (2003) model was built to identify scenario
types from various industries, it cannot include detailed criteria applicable to climate
scenarios as this was not part of the researchers’ objectives.

Characteristics derived from practise. We extend the model by van Notten

et al. (2003) with additional characteristics relevant to climate scenarios by including
findings from the practical perspective. Based on the analysis of the TCFD reports
(as outlined in Section 4.3), the reasons for selecting a particular climate scenario
are assessed to derive the related scenario characteristics. The practical examination
also supports many of the characteristics under ’legitimacy’, for example, the point of
view (Citi, 2021), the time scale (Chartered, 2022) and the spatial scale (KBC Group,
2022; Lloyds Banking Group, 2022) but no other characteristics are added here; while
’credibility’ and ’salience’ are extended by four and eight characteristics, respectively.

Additions to “credibility”: the first identified criterion is the use of a science-based

approach (C3) (Bank of America, 2019; Bank of Montreal (BMO), 2021; (TSKB),
2021) that can be applied to a limited or extensive degree. This criterion aims at
evaluating whether banks value the fact that a climate scenario is built based upon
scientific standards to a large extent, and thus ensures credibility in the selected sce-
nario.

The second criterion, the usage of a climate scenario by other banks (C4), ex-
amines whether banks consider the approach taken by other banks regarding climate
scenario analysis. If many other banks use a particular climate scenario, this implies
frequent use, while the usage of a few banks implies infrequent use. This aspect gains
importance as NGFS aims to provide “a common reference framework for financial
institutions” (Banorte, 2022, p. 17). Using the same climate scenario in an industry
brings consistency and comparability (ING Groep (ING), 2021), which also contributes
to credibility (Abernathy, Stefaniak, Wilkins, & Olson, 2017).

The following characteristic deals with the degree to which a climate scenario is
based on recent data and is expected to be updated regularly (C5). Depending on
the frequency of updates, this characteristic might be classified as one-time, meaning

41
that a climate scenario is not updated after being published, or as continuous, refer-
ring to regular updates. Considering that climate scenarios illustrate possible future
developments, they must use recent data so that the model has a solid foundation. If
out-dated data were used, this would reduce the validity of the analysis.

When analysing the impact of climate change from a risk management perspective
(outside-in), compliance with regulatory guidelines (C6) gains importance. Although
the use of a particular climate scenario is not currently prescribed (TCFD, 2017a),
adhering to the regulator’s guidelines is essential to avoid adverse consequences. Here,
climate scenarios can be distinguished based on whether they are recognised by the
regulator (TCFD, 2017c) or published and used by industry regulators themselves,
like the NGFS scenarios (Network for Greening the Financial System (NGFS), 2021).
Given that fulfilling regulatory requirements is of highest importance for banks, it
is expected that using a climate scenario that is accepted or even applied by the
regulatory authorities is credible.

Additions to “salience”: eight characteristics are derived that compare the content
of the scenario in detail. The first scenario characteristic refers to the temperature
tar-get outlined by a scenario (S1). Since the various climate scenarios show future
pathways leading to different temperature outcomes in the final year, this may be a
relevant decision criterion, especially to ensure Paris alignment (BNP Paribas, 2022;
Sovcombank, 2022). Therefore, this could require a Paris-aligned temperature target
(that is, below 2 ° C). For other purposes, for example, when assessing the impact
of climate change on a bank, it might be worthwhile to include a non-Paris-aligned
temperature target to assess the impact of current policies.

Secondly, the practical analysis shows that climate scenarios apply to various
sectors. They may differ on the extent to which a sector is covered. For example,
Danske Bank (2021) criticises NGFS scenarios for insufficient sector-specific assump-
tions. Thus, another characteristic of the scenario is the granularity of the sectors
covered (S2), which can be low, medium or high, where low refers to providing little
information on a specific sector, and high implies the availability of precise sectoral
information. In addition to the varying level of detail on a sectoral level, climate
scenarios can also differ regarding the degree of detail to which various regions are
covered.

To precisely determine the impact of climate change on a particular country or

42
region, banks may need data on a regionally disaggregated level (Banorte, 2022).
Thus, the third characteristic focusses on the granularity of the regions covered by
a climate scenario (S3). Depending on the degree of disaggregation required, this
characteristic might take the form of low, medium or high, where low implies a strong
aggregation level (e.g., on a global scale) and high refers to a large disaggregation (e.g.,
on the postcode or even the address level) (Network for Greening the Financial System
(NGFS), 2020).

The next criterion (S4) focusses on the type of risk considered in a climate scenario,
that is, whether physical, transition or both risks are included in the climate scenario.
For example, Nomura (2021) chose the NGFS scenarios as these include both types
of risk within one scenario. However, other banks use different climate scenarios to
examine the impact of transition and physical risks in the same sector (Citi, 2020;
Sanpaolo, 2021; K. Group, 2022). Focussing on transition risks, technological change
is required for decarbonisation (Banco Bilbao Vizcaya Argentaria (BBVA), 2022; ING
Research, 2020).

Scenarios illustrating the required technological changes to achieve, for example,

net zero in 2050, help banks to adjust their portfolios accordingly (inside-out). There-
fore, the provision of a technology guideline is another relevant characteristic of the
scenario for bank decision making (S5) and can be distinguished depending on whether
the scenario conveys this information in detail or at a high level. The assignment of
this characteristic to ’salience’ is supported by Cash et al. (2002), who recognise
the availability of appropriate technology for the environmental context as a relevant
aspect of ’salience’.

In addition to technological development, policy changes can be a major driver of

scenario development. For example, “policies that restrict flying to reduce emissions
from aviation” (ING Research, 2020, p. 9) would significantly contribute to achieving
net zero. Depending on whether these policy reactions occur immediately or delayed,
the impact on banks can vary (S6). Within this report, delayed means any policy
reaction that occurs after 2030 while immediately covers the period until 2030. This
is because the EU developed a separate Climate Target Plan for 2030, which serves as
a milestone for achieving climate neutrality by 2050 (European Commission, n.d.-a).

To achieve emission reductions, climate scenarios also differ in their assumptions

for carbon dioxide removal (Citi, 2021). This means that some scenarios allow offset-

43
ting, for example, by planting forests to achieve emission targets. Although net zero
emissions can be achieved in this way (CO2 removals offset anthropogenic CO2 emis-
sions), this should not be confused with absolute zero emissions, which means that no
CO2 emissions are produced at all (Citi, 2021; Intergovernmental Panel on Climate
Change (IPCC), 2018). Consequently, banks could include whether carbon dioxide re-
moval is prohibited or permitted in a certain scenario in their decision-making process
for a climate scenario (S7).

The last characteristic of the scenario refers to the specific variables used in a
scenario (S8). For example, AIB (2022) considers precipitation, heavy rainfalls, and
rise in sea level as key elements in the analysis of the physical environment. Hence,
whether a climate scenario contains weather-related and/or economic variables is also
relevant for deciding upon a scenario.

6 Case Study

• Based on a case study in a German promotional bank, we identify 14 char-

acteristics of climate scenarios that are most relevant to banks.

• We identified four use cases based on different combinations of perspective

(inside-out / outside-in) and risk type (physical and transition).

• Different climate scenario providers have strengths and weaknesses along

those dimensions.

• Despite the general usability of climate scenarios for banks, there are still
some challenges in using climate scenarios, related to insufficient data gran-
ularity, a difficult translation of a standard climate scenario to bank require-
ments and the variety of disclosure formats

We refine the model derived in the previous section by conducting a case study.
We begin by explaining the context of the case study. Subsequently, we explain the
14 characteristics of the model that are applicable to four use cases and derive the

44
corresponding implications for the use of climate scenarios in banks (Section 6.1). The
second part differentiates challenges related to the data provided by climate scenarios,
their respective implementation, and the disclosure of the results of the climate scenario
application (Section 6.2).

Case study setting. To refine the model and identify which characteristics are
particularly relevant to banks when selecting a climate scenario and why, we carry
out a case study in a German promotional bank. For confidentiality, the name of the
Bank is disguised and hereafter referred to as the “Bank”. Since the adoption of the
TCFD guidelines, climate scenarios from IEA, IPCC, and NGFS have been applied
in the Bank. These are also the most frequently used climate scenarios among the
pilot banks (Section 4.3). Thus, the Bank represents a suitable case for analysing the
reasons for choosing a specific climate scenario.

We choose semi-structured interviews to reveal the relevant characteristics of cli-

mate scenarios as they give the opportunity to slightly adapt the questions asked
according to the specific knowledge and organisational context of the interviewee. At
the same time, they leave enough space to explain the derived characteristics in more
detail (Saunders, Lewis, & Thornhill, 2016). Having flexibility for follow-up questions
allows diving into the details of relevant characteristics, which is an essential base
for understanding why specific scenario characteristics are suitable for banks (Adams,
2018).

Since the selection of appropriate interviewees constitutes an ’iterative process’

(Bogner & Menz, 2009, p. 55), we conducted four preliminary talks to gain an under-
standing of where the knowledge on climate scenarios is distributed among the Bank.
Subsequently, we selected six interviewees, as they have worked with climate scenarios
for at least one year. For case studies, the applicable experts must be identified ”in
relation to the concrete field of operation in which the expert acts” (Bogner & Menz,
2009, p. 54). An overview of the interviewees can be seen in Table 5, where the last
two experts have only been interviewed as part of preliminary talks (due to a lack of
technical expertise on the matter).

In the interview, we discuss all the characteristics of the scenarios as presented

in Table 4. Each interviewee discusses the relevant characteristics during the climate
scenario selection process. Criteria are randomly selected, that is, the order of char-
acteristics is varied across all interviews, to avoid bias (Baehring, Thommes, Hauff, &

45
Sossdorf, 2008).

Table 5: Overview of experts interviewed

Name Division Responsibility Persepective Risk Type

Expert 1 Risk controlling Climate-related stress testing Outside-in Transition
Expert 2 Risk controlling Climate-related stress testing Outside-in Transition
Expert 3 Risk manage- Climate impact on sovereign ratings Outside-in Transition/Physical
ment
Expert 4 Risk manage- ESG-risks (TCFD-reporting, cli- Outside-in Transition/Physical
ment mate scenario analysis)
Expert 5 Strategy Paris-compatible alignment of port- Inside-out NA
folio
Expert 6 Strategy Country-specific Climate-Factsheets Outside-in Physical
Expert 7 Risk controlling Climate-related stress testing Outside-in Transition
Expert 8 Bank’s sub- Impact management Inside-out NA
sidiary

6.1 Refined Model

Relevant characteristics. To derive a list of relevant characteristics for banks,

we decided to differentiate them according to four combinations of the ’perspective risk
type’ (see Table 6 for more details). Section 4.3 already shows that not only the outside-
in and inside-out perspective may lead to choosing different climate scenario providers
but also the risk type examined under the outside-in perspective. This observation is
confirmed by the case study in which the interviewees perceived different characteristics
as relevant. Table 6 shows the characteristics of the relevant climate scenario and their
respective criteria (“legitimacy”, “credibility”, and “salience”), as well as a split by four
combinations of ’perspective risk type’. The former differentiates between the outside-
in and inside-out perspectives. The latter applies only to the outside-in perspective
and distinguishes physical, transition, or both risks. The expression ’both risks’ is
included since banks may want to assess physical and transition risks simultaneously
and do not want to examine them independently.

46
Table 6: Overview of relevant characteristics of climate scenarios for banks

Use Cases
Criteria Characteristic
Outside-in – Outside-in – Outside-in –
Inside out
Physical risk Phys. & trans. risks Transition risk
Spatial scale Global
Temporal nature Chain
Legiti-
Time scale Long-term Short-/Long-term
macy
Nature of the data Quantitative Quantitative & Qualitative
Level of integration High NA
Science-based approach Extensive
Institutional conditions Open
Credi-
Usage of scenario by banks Frequent
bility
Up-to-dateness NA Continuous
Acknowledged by Acknowledged by
Regulatory compliance Published by the regulator
the regulator the regulator
Temperature target Paris-aligned/non-Paris-aligned Paris-aligned
Sali- Granularity of re-
High Medium NA
ence gions covered
Granularity of sec-
NA high
tors covered
Weather-related Technology
Required data Weather-related Economic
and economic guidelines
IPCC
Climate scenarios NGFS NGFS IEA
(RCP 2.6, RCP
used in the Bank (NZ, B2, CP) (NZ, DT) (NZE, SDS)
4.5, RCP 8.5)

Some characteristics share the same expression across all combinations of ’per-
spective risk type’ for the Bank, while others show different expressions. The broadly
shared characteristics occur mainly among the first two criteria, ’legitimacy’ and ’cred-
ibility’. For ’salience’, the expressions vary. Based on this, we conclude that charac-
teristics under ’legitimacy’ and ’credibility’ are required to ensure general acceptance
of a climate scenario. Thus, scenario providers should comply with these characteris-
tics when building climate scenarios. On the contrary, the characteristics listed under
’salience’ illustrate the different aspects depending on the combination of ’perspective
- risk type’. These are particularly relevant for banks when selecting a climate scenario
and help to distinguish various climate scenario providers. Subsequently, we discuss
the individual characteristics.

47
 Legitimacy

The first criterion, spatial scale, takes the same form in all combinations of ’per-
spective risk type’: a climate scenario must have a global scale when bank operations
are spread throughout the world. This ensures that one climate scenario covers large
parts of the portfolio, which is crucial to receiving meaningful results in stress tests.
Combining various local climate scenarios, such as the Agora scenarios, to increase
the geographical coverage might not be possible due to systematic differences among
the climate scenarios. However, local climate scenarios can complement global climate
scenarios by providing more in-depth information for specific countries/regions.

Secondly, only having chain data (i.e., time series data) makes climate scenarios
usable for all perspectives, for example, to model risk-bearing capacity under the
outside-in perspective. For the inside-out perspective, aligning a portfolio with the
goals of the Paris agreement is a continuous process where the changing financing
restrictions must be outlined over several years.

In contrast to the previous criteria, the relevant time scale of a climate scenario
varies across the use cases. When analysing the impact of physical risks, long-term
horizons usually yield stronger effects, given that physical risks are not expected to
hit the Bank soon (and thus are less relevant for stress testing as of now). Although
this holds for the Bank, the exposure of other banks to physical impact may differ.
Therefore, they must determine for themselves whether a short-term analysis of phys-
ical risks is required. For transition risks, the short- and long-term views are essential.
Supervisory authorities demand a risk assessment for the next 12 months (short-term).
However, more significant impacts can only be seen under a long-term risk analysis.
Concerning the inside-out perspective, an appropriate portfolio steering is only pos-
sible when having a short-term view. However, as Paris compatibility becomes only
apparent in the long term, the climate scenario must continue for a longer period.

Analysing the impact of physical risks requires quantitative data on the occurrence
and intensity of extreme weather events. For transition risks, further explanations by
qualitative statements on the underlying assumptions are necessary to understand the
quantitative data. The same holds for the inside-out perspective, where quantitative
data is used for calculating quotas on, for example, investments in gas-fired power
plants. Still, qualitative data conveys information on technologies that could become
relevant in the future (e.g., carbon capture and storage technologies).

48
 The last characteristic under “legitimacy” shows that a high level of integration
within a scenario is required for the outside-in perspective. A coherent climate scenario
facilitates understanding the applied scenario and communicating its results to the
management. Thus, such a scenario is more likely to be accepted. However, the level
of integration has not been mentioned as relevant from the inside-out perspective.

Credibility

The first three criteria are all shared among the various combinations of ’perspec-
tive risk type’: When the climate scenario uses a science-based approach extensively,
this contributes to credibility, as it conveys durability (application of a consistent ap-
proach over the years) and is a quality feature given that scientific work has been
peer-reviewed. Secondly, open institutional conditions ensure the independence of the
scenario publishing institution and provide the basis for meeting the standards of sci-
entific work as outlined before. Thirdly, looking at how other banks deal with the
topic before or during the implementation of climate scenarios allows to derive best
practises and fosters acceptance of the scenario. Therefore, the frequent use of the
climate scenario by other banks provides a solid guideline on whether a bank is on the
right track. Still, this does not imply that banks should solely follow the approach
other banks take, but could also be a pioneer.

Regarding the up-to-dateness characteristics, climate scenarios should contain the

most current information to obtain the correct results of their implementation. When
climate scenarios are used over a longer time horizon (e.g., for portfolio alignment), con-
tinuous updates are required. Similarly to software updates, scenario updates ensure
keeping up with recent market trends, such as technological developments. Choosing
a climate scenario that is expected to be updated regularly also avoids additional costs
from switching to another scenario after a few years as the selected scenario is out-
dated. However, updates mainly focus on the underlying economic model and not the
climate pathways as these are usually not exposed to significant changes over time.
Consequently, continuous updates are less relevant for analysing physical risks, which
mainly depend on a specific climate pathway.

Lastly, choosing a climate scenario accepted or recognised by the regulator is es-

sential across all perspectives. Although the use of a specific climate scenario provider
is not yet prescribed, the interviewees mention the importance of choosing a climate
scenario that is acknowledged by the regulator to avoid additional effort in switching

49
to a new climate scenario provider. However, for the outside-in perspective combined
with transition and with both risks, several interviewees also state that they selected
a climate scenario as it is published by the regulator (that is, referring to the NGFS
scenarios). Intuitively, using this scenario would be preferable for every perspective,
as it leaves very limited to no room for being criticised by the regulator for using a
specific climate scenario. Still, Section 4.3 and the case study show that also other
climate scenarios gained recognition. Thus, we conclude that other climate scenarios
contain additional information, which keeps banks from using the NGFS scenarios for
all application cases.

Salience

The inside-out perspective requires Paris-aligned temperature targets. Steering a

portfolio to Paris-compatibility is only possible when the underlying climate scenario
outlines a corresponding path. On the contrary, for the outside-in perspective, the
consideration of various climate scenarios, i.e., Paris-aligned and non-Paris-aligned
pathways, is relevant. Despite the TCFD recommendation to assess the impacts of
a 2 ° C climate scenario (on Climate Related Financial Disclosures (TCFD), 2017b),
TCFD also suggests using “favorable and unfavorable [scenarios]” (TCFD, 2017b, p.
27) to cover various future outcomes.

Assessing physical risks requires a high regional granularity since the impact of
extreme weather events might already differ at low distances. Therefore, even the
address level might sometimes be needed to comprehensively examine the physical
risks. On the contrary, when analysing the impact on the Bank for the remaining
two outside-in perspectives, medium granularity of the regions is sufficient, that is,
country-level data. Here, less granularity is needed as physical risks are not analysed
as precisely as when examined together with transition risks. However, the level of
granularity required in a bank depends on the bank’s risk assessment approach and,
therefore, could differ for other banks.

For the inside-out perspective, a high sectoral granularity is essential for the Bank
as it decided to steer its portfolio along technology-based pathways. This requires a
sectoral breakdown in the first step. Afterwards, the Bank can identify sector-specific
technologies that are aligned with the chosen climate scenario. As the Bank’s risk
assessment is structured along various sectors, a high sectoral granularity is relevant
for the transition risk assessment under the outside-in perspective. This should also

50
hold for other banks according to our TCFD analysis.

A sectoral disaggregation was not required for the physical assessment nor for the
combined (physical and transition) risk assessment. Still, this view largely depends on
the fact that the Bank aims to determine the impact of climate change on sovereign
ratings under the combined perspective and to analyse the exposure of countries to
physical risks. If other banks apply climate scenarios differently, this classification may
change.

The criterion required data captures the data requirements for each combination
of “perspective risk type”. After the case study, we created this criterion combining
the characteristics “technology guidelines” and “variables used” shown in Table 4.
The interviews show that depending on the combinations of ’perspective risk type’,
specific data are required to model the impacts of climate change. Technological guide-
lines can be seen as one form of data needed. Thus, both characteristics are merged.
Assessment of physical risks requires various weather-related data, for example, the
magnitude of change in weather events such as floods and storms. On the contrary,
when examining transition risks, economic and social data on gross domestic product
(GDP), CO2 price, or population are needed. Weather-related and economic data are
essential for the combined perspective. However, as explained above, less granular-
ity is sufficient, especially regarding weather-related data. Concerning the inside-out
perspective, technological guidelines are relevant to appropriately steer the portfolio.
Precisely, the scenario should show which technologies may still be financed, which
may only be financed to a limited extent, and which must not be financed anymore.

Excluded characteristics. Unlike the original model shown in Table 4, several

characteristics seem to be irrelevant to the Bank decision making or are (implicitly)
included in other characteristics. In the following, the reasons for excluding them
in the final model are briefly explained. As already outlined in the definition of the
problem (Section 5.1), the characteristics inclusion of norms (descriptive/normative),
dynamics (peripheral/trend), and level of deviation (alternative/conventional) seem to
be closely linked to each other. The first interviewees observed the same dependencies
among the criteria and considered none as relevant; therefore, all of them are excluded
as they do not provide additional information. Further, the characteristic vantage
point (forecasting/backcasting) is dropped in favour of the characteristic temperature
target, which may imply a backcasting approach if, for example, a 1.5 ° C pathway is
outlined. As this report focusses on the assessment of climate scenarios, the subject

51
of the scenarios is already issued-based. With the subject being specified in advance,
this characteristic involves little value-add, and hence can also be neglected.

Although the characteristic method of data collection also plays a minor role when
selecting a climate scenario, the characteristic is replaced by the use of a science-based
approach. A similar reasoning applies to the following characteristic, the variables
used : The differentiation between heterogeneous / homogeneous has received little
attention although the inclusion of specific variables in a climate scenario is relevant,
referred to as required data in the model. The number of resources is excluded since
the availability of sufficient resources is implicitly assumed when uses a science-based
approach. The last two excluded characteristics are carbon dioxide removal and policy
reaction. The interviews reveal a lack of importance of the former criterion. Instead
of using carbon dioxide removal as a decision criterion, the assumptions of the climate
scenario are simply used. Combined with the fact that the public and politicians in
Germany rarely accept carbon capture and storage technologies (Arning et al., 2019).
Hence, including the former characteristic would not be reasonable when examining a
German bank. For the latter, the experts perceived this characteristic to already be
included in transition risks or as irrelevant.

Despite excluding several characteristics from van Notten et al. (2003), the distri-
bution of characteristics in the final model supports the chosen approach to combine
a theoretical and practical perspective. Relevant characteristics are derived from both
perspectives in equal parts: The first seven characteristics in Table 6 (from spatial
scale to institutional conditions) originate from van Notten et al. (2003) and the lat-
ter seven characteristics come from the pilot banks’ reports (usage of scenario by banks
to required data).

Implications. While climate scenarios were not developed for banks, the derived
use cases bridge the gap by helping banks select a suitable climate scenario. Increasing
the adoption rates of climate scenarios is particularly relevant, as an early assessment
of climate-related risks and opportunities allows banks to improve their business re-
silience to climate change. The interviews reveal that banks should be aware of two
aspects when conducting scenario analysis. First, to maintain consistency in the cli-
mate scenario applied, and second, to allocate sufficient resources to the work with
climate scenarios. Banks should use consistent climate scenarios across their teams to
avoid a mismatch in the climate scenarios applied. Setting up a central unit that coor-
dinates the use of climate scenarios in a bank increases internal alignment. Many banks

52
hire external consultants to support them in working with climate scenarios (Barclays
PLC, 2022; Fargo, 2021). Although they are a valuable interim support, banks should
build up sufficient internal capacity and expert knowledge to be prepared to work with
climate scenarios in the future.

6.2 Challenges of Using Climate Scenarios

During the case study, we identified challenges related to using climate scenarios
within the three fields of data, implementation, and disclosure. These aspects directly
feed back into the use of our model.

Data. The availability of data is highly relevant for applying climate scenarios
in practise. Therefore, it is apparent that the challenge of missing data needs to
be overcome. The TCFD recommendations already outline that “the availability and
granularity of data can be a challenge for organizations” (TCFD, 2017b, p. 30), further
highlighting its importance. The case study reveals missing granularity of information
within three categories. First, the time intervals between the data points are too large.
Especially the stress testing team of the Bank needs a yearly provision of the data, as
shorter-term scenarios are essential for management decisions. Currently, only data
are available over five years. Thus, interpolation is used to provide a workaround for
this data lack. A yearly provision of data would make this additional task obsolete
and facilitate the work with climate scenarios. Second, the provision of regional data
by the IEA and IPCC scenarios is perceived as insufficient because it is covered within
the characteristic of granularity of the covered regions. Regarding the application
of climate scenarios for the outside-in perspective, several experts stress the need for
regional-level data. We identify different approaches to deal with this problem in
practise. To disaggregate data on regions, additional tools and datasets, for example,
from the Climate Service Centre and local climate models, are used. Another method
is to break down the data provided at the country level to a regional level by internal
experts. In addition to that, assumptions about missing data are made to cope with
the issue of data unavailability. Third, the granularity of the sectors covered provided
by the NGFS and IPCC scenarios is limited (Allen et al., 2020). As the effects of
climate change differ significantly across sectors, it is necessary to provide granular
sector information in climate scenarios. These are crucial to, for example, identify in
which sectors banks must reduce or omit their business activities to become net zero

53
and which sectors benefit from climate change (e.g., renewable energy technologies).
Thus, the supply of sector-based data by scenario providers could help banks generate
more insightful results from using climate scenarios.

In addition to a lack of data granularity, at least one scenario provider does not
provide three relevant variables. First, the results of the case study highlight that the
NGFS scenarios, so far, do not cover acute physical risks for all countries. As NGFS
scenarios do not allow banks to fully assess physical risks, additional sources of acute
physical risks must be applied (e.g., International Disaster Data Base). This is related
to a higher implementation effort, as they must be aligned and integrated with existing
systems and processes of organisations.

Second, IEA scenarios can be improved with regard to the granularity of sectors
covered: more information on the regionally disaggregated energy mix would be helpful
to facilitate aligning the portfolio. As the Paris Agreement differentiates between
developing, emerging, and industrialised countries, this should also be reflected in the
scenario data.

Third, the case study uncovers that the translation of the climate scenarios into
granular risk drivers such as oil prices (transition risk driver) or the strength of storms
(physical risk driver) is missing within the provided data. Thus, a provision of these
risk drivers and their intensity by the scenarios is desired for the application in the
Bank.

In addition, we find that a simplified overview of the data provided for each sce-
nario would facilitate the selection and comparison of climate scenarios. Furthermore,
finding the assumptions underlying the climate scenarios, precisely the NGFS scenar-
ios, is perceived as difficult, as they are provided in different documents. Thus, it
would unburden the application of climate scenarios if scenario providers conveyed all
assumptions connected to the scenarios in a bundled and aggregated form, for example,
on an overview page.

Implementation. Climate scenarios are usually provided in a standardised for-

mat to make them usable for all types of organisations. Therefore, when a bank
decides to apply a particular climate scenario, it must first adapt the scenario to
its requirements. The experts interviewed highlight that adjusting the scenarios to
company-specific systems and processes poses challenges. As little guidance is pro-

54
vided for translating the standardised climate scenarios into a bank-specific format,
this task is perceived as very difficult. For example, banks face the difficulty of trans-
lating the identified risks in companies into economic impacts. If a bank is insured
against the risk in question, its loss is reduced; however, taking the insurance into
account when assessing risk increases complexity. Therefore, extensive internal and
external resources are used to examine how the translation and adaptation process can
be performed.

Furthermore, if the Bank uses a specific sectoral split for risk assessment, this
sector mapping might be unsuitable for climate scenarios. For some industries like
energy, no further breakdown is provided, which prevents a distinction between, for
ex-ample, green- and non-green-energy sources. Consequently, existing industry splits
must be redefined to allow a practical application of climate scenarios.

Disclosure. The TCFD guidelines for climate-related financial disclosures also

include recommendations for climate scenarios (TCFD, 2017b). However, in practise,
the degree and format of information provided on climate scenarios within the anal-
ysed TCFD reports vary greatly between banks. Often, the specific scenario used,
or the underlying temperature goal, is not evident from the reports. Providing best
practises and more detailed regulations on the use of a particular climate scenario
would increase transparency and comparability between banks and thus support other
banks in the selection and implementation process (Green and Sustainable Finance
Cluster Germany (GSFC Germany), 2019; Jürgens et al., 2021). This also concerns
the handling of scenario updates: Although two banks use climate scenarios by the
same scenario provider, they might differ if one bank uses an older version of the cli-
mate scenario. Thus, we call for regulators to provide more guidance on the use of the
latest climate scenarios to increase comparability among banks. To guide banks when
climate scenarios are updated, scenario providers should inform about these updates
in advance.

Considering that TCFD recommends running climate scenarios that examine the
inside-out and outside-in perspective as well as physical and transition risks, the use of
climate scenarios from various scenario providers is necessary. Although IEA, IPCC,
and NGFS scenarios have several fields of application, neither of them is the most
appropriate provider for all use cases. It would help banks if scenario providers dis-
closed the purposes for which their climate scenarios are most suitable and published
concepts for integration with other scenario providers. The NGFS already started this

55
by mapping its scenarios against the IEA scenarios (NGFS, n.d.). However, this infor-
mation does not provide sufficient guidance for banks in choosing a climate scenario.
Alternatively, consolidating the three most frequently used scenario providers would
allow one to connect their areas of expertise and to ease the use of climate scenarios
for banks. This would also increase transparency and comparability between banks.

7 Conclusions

Climate scenarios were originally designed for policymakers and scientists (Bauer
et al., 2017; O’Neill et al., 2017; Riahi et al., 2017; Ritchie & Dowlatabadi, 2018).
Over the last years, the use of climate scenarios has gained increasing importance for
banks. However, little research has been done to analyse what makes climate scenarios
usable in practise. This report aims to bridge the gap between research and practise by
examining the practical use of climate scenarios and which characteristics of them are
particularly relevant for banks. We review existing literature from both the theoretical
and practical perspectives to develop a model using a heuristics approach. This model
is subsequently refined by conducting a case study in a German bank and, ultimately,
covers 14 characteristics of climate scenarios. The practical application of the model
shows its usability and allows us to derive challenges related to climate scenarios.

Some limitations inherent to the selected approach remain. First, the criteria
relevant from a practical point of view are derived based on analysing the TCFD
reports. Although the banks analysed can be assumed to be relatively advanced in
their use of climate scenarios due to their participation in the UNEP FI TCFD Banking
Pilots, other banks might also apply climate scenarios extensively. Therefore, future
researchers could examine whether other banks are also advanced in using climate
scenarios and refine the existing model based on their perspective on climate scenarios.

Second, the most recent data are used for both the classification of climate sce-
narios according to the van Notten et al. (2003) model and the analysis of the TCFD
reports. However, some climate scenarios classified in Section 3.3 were updated in 2021
and banks might not yet have implemented these updated versions. Although climate
scenarios are only updated and not completely changed, it may be worthwhile to ex-
amine the TCFD reports in the future to determine whether different characteristics
relevant to banks are mentioned.

56
 Third, this report focusses on the selection process of climate scenarios, as this is
expected to reduce barriers to entry into the work with climate scenarios. However,
the challenges outlined in Section 6.2 stress the need to also address problems during
the use of climate scenarios. Therefore, future researchers should aim to identify the
most significant difficulties in using climate scenarios in various banks, for example,
by conducting interviews with their employees.

The paper adds to the literature by providing a characterisation of climate sce-

narios for banks and by supporting the selection of a climate scenario amongst banks.
It is the first case study with a bank in this field. Our aim is to contribute to the con-
tinuous improvement of climate scenarios by outlining concrete actions for providers
and regulators. The case study shows that climate scenario providers should increase
regional and sectoral data granularity to become the “one-stop” solution for banks.
The analysis of TCFD reports highlights that the climate scenarios used and the level
of detail in each report vary widely.

In conclusion, climate scenarios gained increasing relevance for the banking sector
over the last years, despite initially being developed for scientists and policy makers.
This report illustrates the versatile applicability of climate scenarios for banks and
helps to differentiate their use cases accordingly. Although the use of climate scenarios
involves various challenges, these can be overcome when climate scenario providers
and banks work together to benefit from each other’s knowledge. The successful use of
climate scenarios in the banking sector not only contributes to achieve the goals of the
Paris Agreement, but also to improve the resilience to climate change of the banking
sector.

57
References

(2DII), P. A. C. T. A. P. . I. I. (2021). Pacta for banks scenarios. Retrieved

from https://www.transitionmonitor.com/wp-content/uploads/2021/07/
20210706-PACTA-for-Banks Scenario-Supporting-document V11.pdf
Abernathy, J., Stefaniak, C., Wilkins, A., & Olson, J. (2017). Literature review and
research opportunities on credibility of corporate social responsibility reporting.
American Journal of Business, 32 (1), 22–41. Retrieved from https://doi.org/
10.1108/AJB-04-2016-0013 doi: 10.1108/AJB-04-2016-0013
ABN AMRO. (2022). Integrated annual report 2021. Online. Retrieved from https://
downloads.ctfassets.net/1u811bvgvthc/62067Em47HIFTg3hGxdQHY/
3fcd6a3ac35ba44be18fad01cc84b7c0/ABN AMRO Integrated Annual
Report 2021.pdf
Absa Group Limited. (2022). 2021 task force on climate-related financial disclosures
(tcfd) report. Online. Retrieved from https://www.absa.africa/content/dam/
africa/absaafrica/pdf/sens/2021/2021-Absa-Group-Limited-TCFD.pdf
Acharya, V. V., & Steffen, S. (2020). “stress test” for banks as liquidity insurers in
a time of covid. Online. Retrieved from https://voxeu.org/article/stress
-tests-banks-liquidity-insurers-time-covid
Adams, W. C. (2018). Conducting semi-structured interviews. In K. E. Newcomer,
H. P. Hatry, & J. S. Wholey (Eds.), Handbook of practical program evaluation
(4th ed., pp. 492–505). Jossey-Bass.
Agora Energiewende. (n.d.). Agora energiewende. Online. Retrieved from https://
www.agora-energiewende.de/ueber-uns/agora-energiewende/
Allen, T., Dees, S., Boissinot, J., Graciano, M. C., Chouard, V., Gaye, A. De, . . .
Vernet, L. (2020). Climate-related scenarios for financial stability assessment:
an application to france (Tech. Rep.). Banque de France. Working Paper Series
no. 774. Retrieved from https://publications.banque-france.fr/sites/
default/files/medias/documents/wp774.pdf
Allied Irish Banks (AIB). (2022). Sustainability report for the fi-
nancial year ended 31 december 2021. Online. Retrieved from
https://aib.ie/content/dam/frontdoor/personal/sustainability/
sustainability-full-report-2021.pdf
Amer, M., Daim, T. U., & Jetter, A. (2013). A review of scenario planning. Futures,
46 , 23–40. doi: 10.1016/j.futures.2012.10.003

58
Arning, K., Offermann-van Heek, J., Linzenich, A., Kaetelhoen, A., Sternberg, A.,
Bardow, A., & Ziefle, M. (2019). Same or different? insights on public perception
and acceptance of carbon capture and storage or utilization in germany. Energy
Policy, 125 , 235–249. doi: 10.1016/j.enpol.2018.10.039
Australia and New Zealand Banking Group (ANZ). (2021). 2021 climate-related finan-
cial disclosures. Online. Retrieved from https://www.anz.com/content/dam/
anzcom/shareholder/ANZ-2021-Climate-related-Financial-Disclosures
.pdf
Baehring, K., Thommes, K., Hauff, S., & Sossdorf, M. (2008). Methodologische
grundlagen und besonderheiten der qualitativen befragung von experten in un-
ternehmen: ein leitfaden. Die Unternehmung: Swiss Journal of Business Re-
search and Practice, 62 (1), 89–111.
Banco Bilbao Vizcaya Argentaria (BBVA). (2022). Report on tcfd 2021. Online.
Retrieved from https://shareholdersandinvestors.bbva.com/wp-content/
uploads/2022/03/BBVA TCFD Report December 2021 ENG.pdf
Bank of America. (2019). Responsible growth and a low-carbon economy. Online. Re-
trieved from https://about.bankofamerica.com/content/dam/about/pdfs/
task-force-climate-financial-disclosures-report.pdf
Bank of Montreal (BMO). (2021). 2021 climate report. Online. Retrieved
from https://our-impact.bmo.com/wp-content/uploads/2022/03/BMO 2021
Climate-Report EN FINAL.pdf
Banorte. (2022). Report on climate-related risks and opportunities 2021.
Online. Retrieved from
https://investors.banorte.com/~/media/
Files/B/Banorte-IR/financial-information/annual-reports/en/2021/
TCFDBANORTEeng.pdf
Barclays PLC. (2022). Barclays plc climate-related financial disclosures 2021. On-
line. Retrieved from https://home.barclays/content/dam/home-barclays/
documents/investor-relations/reports-and-events/annual-reports/
2021/Barclays-TCFD-Report-2021.pdf
Batten, S., Sowerbutts, R., & Tanaka, M. (2017). Let’s talk about the weather:
The impact of climate change on central banks (staff working paper no. 603)
(Tech. Rep.). Bank of England. Retrieved from https://doi.org/10.2139/
ssrn.2783753
Bauer, N., Calvin, K., Emmerling, J., Fricko, O., Fujimori, S., Hilaire, J., . . . van
Vuuren, D. P. (2017). Shared socio-economic pathways of the energy sector –
quantifying the narratives. Global Environmental Change, 42 , 316–330. doi:

59
 10.1016/j.gloenvcha.2016.07.006
Bishop, P., Hines, A., & Collins, T. (2007). The current state of scenario devel-
opment: An overview of techniques. Foresight, 9 (1), 5–25. doi: 10.1108/
14636680710727516
BNP Paribas. (2020). Tcfd report 2020. Online. Retrieved from https://group
.bnpparibas/uploads/file/tcfd 2020 bnpparibas en.pdf
BNP Paribas. (2022). In a changing world, let’s accelerate the energy transition and
the incorporation of climate related issues - 2021 tcfd report. Online. Retrieved
from https://cdn-group.bnpparibas.com/uploads/file/tcfd report 2021
eng.pdf
Bogner, A., & Menz, W. (2009). The theory-generating expert interview: Episte-
mological interest, forms of knowledge, interaction. In A. Bogner, B. Littig, &
W. Menz (Eds.), Interviewing experts (pp. 43–80). Palgrave Macmillan.
Bundesanstalt für Finanzdienstleistungsaufsicht (BaFin). (2020). Guid-
ance notice on dealing with sustainability risks. Online. Retrieved
from https://www.bafin.de/SharedDocs/Downloads/EN/Merkblatt/
dl mb Nachhaltigkeitsrisi%0Dken en.pdf? blob=publicationFile&v=5
Börjeson, L., Höjer, M., Dreborg, K. H., Ekvall, T., & Finnveden, G. (2006). Scenario
types and techniques: Towards a user’s guide. Futures, 38 (7), 723–739. doi:
10.1016/j.futures.2005.12.002
CaixaBank. (2021). Caixabank - climate change 2021. Online. Retrieved from
https://www.cdp.net/en/formatted responses/responses?campaign
id=74241094&discloser id=896524&locale=en&organization name=
Ansell&organization number=806&program=Investor&project year=
2021&redirect=https%3A%2F%2F2.zoppoz.workers.dev%3A443%2Fhttps%2Fcdp.credit360.com%2Fsurveys%2F2021%
2Fdbbr64m
Canadian Imperial Bank of Commerce (CIBC). (2019). Building a sustainable future.
Online. Retrieved from https://www.cibc.com/content/dam/about cibc/
inside cibc/environment/building-a-sustainable-future-report-en
.pdf
Carbon Disclosure Project (CDP). (2022). Caixabank - climate change 2021.
Online. Retrieved from https://www.caixabank.com/deployedfiles/
caixabank com/Estaticos/PDFs/Sostenibilidad/cuestionario CDP
cumplimentado final 2021.pdf
Cash, D., & Belloy, P. (2020). Salience, credibility and legitimacy in a rapidly shifting
world of knowledge and action. Sustainability, 12 (18), 7376. Retrieved from

60
 https://doi.org/10.3390/su12187376 doi: 10.3390/su12187376
Cash, D., Clark, W., Alcock, F., Dickson, N., Eckley, N., & Jäger, J. (2002). Salience,
credibility, legitimacy and boundaries: Linking research, assessment and de-
cision making (Tech. Rep.). Kennedy School of Government Working Paper
RWP02–046, Harvard University. Retrieved from https://dash.harvard.edu/
bitstream/handle/1/32067415/Salience credibility.pdf?sequence=4
Chartered, S. (2022). Climate-related financial disclosure report 2021.
Retrieved from https://av.sc.com/corp-en/content/docs/tcfd-climate
-change-disclosure.pdf
Chatain, L., Ghosh, R., Preudhomme, N. A., & Mazzacurati, E. (2021). Critical
industries have substantial exposure to physical climate risks. Online. Retrieved
from https://assets.website-files.com/5df9172583d7eec04960799a/
618872a58d35f2643cbcaef2 BX9770 ESG Critical%20industries%20have%
20substantial%20exposure 7Nov2021.pdf
Chenet, H., Ryan-Collins, J., & van Lerven, F. (2021). Finance, climate-change
and radical uncertainty: Towards a precautionary approach to financial policy.
Ecological Economics, 183 , 106957. Retrieved from https://doi.org/10.1016/
j.ecolecon.2021.106957 doi: 10.1016/j.ecolecon.2021.106957
Citi. (2020). Finance for a future climate-resilient future ii — citi’s 2020 tcfd report.
Online. Retrieved from https://www.citigroup.com/citi/sustainability/
data/finance-for-a-climate-resilient-future-2.pdf?ieNocache=111
Citi. (2021). Taskforce on climate-related financial disclosures report 2021 -
citi’s approach to climate change and net zero. Online. Retrieved from
https://www.citigroup.com/citi/sustainability/data/taskforce-on
-climate-related-financial-disclosures-report-2021.pdf
Commerce International Merchant Bankers Group Holdings Berhad (CIMB
Group). (2022). Sustainability report. Online. Retrieved from https://
www.cimb.com/content/dam/cimb/group/documents/investor-relations/
annual-general-meeting/2022/cimb-sr-2021.pdf
Cortés, K. R., Demyanyk, Y., Li, L., Loutskina, E., & Strahan, P. E. (2020). Stress
tests and small business lending. Journal of Financial Economics, 136 (1), 260–
279. Retrieved from https://doi.org/10.1016/j.jfineco.2019.08.008 doi:
10.1016/j.jfineco.2019.08.008
Cradock-Henry, N. A., & Frame, B. (2021). Advancing relevance, credibility, le-
gitimacy, and effectiveness as a heuristic for local-parallel scenarios. Frontiers
in Climate, 3 . Retrieved from https://doi.org/10.3389/fclim.2021.705229

61
 doi: 10.3389/fclim.2021.705229
Credit Suisse. (2021). Sustainablity report 2021. Online. Retrieved from https://
www.credit-suisse.com/media/assets/corporate/docs/about-us
Danske Bank. (2021). Climate and tcfd progress update. Online. Retrieved
from https://danskebank.com/-/media/danske-bank-com/file-cloud/
2021/6/climate-and-tcfd-progress-update---june-2021.pdf?rev=
920b3e9360e14a0fa2c2a81a9a7867dc
Den norske Bank (DNB). (2022). Annual report 2021. Online. Retrieved from
https://www.ir.dnb.no/sites/default/files/pr/202203097662-1.pdf
?ts=1648019291
Eis, J., & Schafer, J. (2019). Changing course: A comprehensive investor guide
to scenario-based methods for climate risk assessment, in response to the tcfd
(Tech. Rep.). UNEP FI. Retrieved from https://www.unepfi.org/wordpress/
wp-content/uploads/2019/05/TCFD-Changing-Course-Oct-19.pdf
European Central Bank (ECB). (2022a). 2022 climate risk stress test. Online.
Retrieved from https://www.bankingsupervision.europa.eu/ecb/pub/pdf/
ssm.climate stress test report.20220708~2e3cc0999f.en.pdf
European Central Bank (ECB). (2022b). Ecb banking supervision launches
2022 climate risk stress test [press release]. Online. Retrieved from
https://www.bankingsupervision.europa.eu/press/pr/date/2022/html/
ssm.pr220127~bd20df4d3a.en.html
European Central Bank (ECB). (2022c). Stress tests. Online. Re-
trieved from https://www.bankingsupervision.europa.eu/banking/tasks/
stresstests/html/index.en.html
European Commission. (2019). Guidelines on reporting climate-related in-
formation (Tech. Rep.). European Commission. Retrieved from
https://ec.europa.eu/finance/docs/policy/190618-climate-related
-information-reporting-guidelines en.pdf
European Commission. (n.d.-a). Klimazielplan für 2030. Online. Retrieved
from https://ec.europa.eu/clima/eu-action/european-green-deal/2030
-climate-target-plan de
Fargo, W. (2021). Task force for climate-related financial disclosures - tcfd re-
port 2020. Retrieved from https://www08.wellsfargomedia.com/assets/
pdf/about/corporate-responsibility/climate-disclosure.pdf
Fernandes, M., Igan, D., & Pinheiro, M. (2020). March madness in wall street: (what)
does the market learn from stress tests? Journal of Banking Finance, 112 ,

62
 105250. Retrieved from https://doi.org/10.1016/j.jbankfin.2017.11.005
doi: 10.1016/j.jbankfin.2017.11.005
Financial Stability Board (FSB). (2015, November 9). Proposal for a disclo-
sure task force on climate-related risks [press release]. Online. Retrieved
from http://www.fsb.org/2015/11/fsb-proposes-creation-of-disclosure
-task-force-on-climate-related-risks-2/
FirstRand. (2021). Climate-related financial disclosures 2021. Online. Retrieved
from https://www.firstrand.co.za/media/investors/annual-reporting/
FirstRand-TCFD-report-2021.pdf
Generale, S. (2022). Climate disclosure - societe generale’s climate re-
port aligned with the tcfd recommendations. Retrieved from https://
www.societegenerale.com/sites/default/files/documents/CSR/
Societe-Generale-Climate-Disclosure-Report-December-2021-EN.pdf
Girod, B., & Mieg, H. (2008). Wissenschaftliche und politische gründe für den wan-
del der ipcc-szenarien. GAIA - Ecological Perspectives for Science and Society,
17 (3), 302–311. doi: 10.14512/gaia.17.3.13
Girod, B., Wiek, A., Mieg, H., & Hulme, M. (2009). The evolution of the ipcc’s
emissions scenarios. Environmental Science Policy, 12 (2), 103–118. doi: 10
.1016/j.envsci.2008.12.006
Global Compact Network Germany (GCNG). (2019). Evaluating corporate cli-
mate risks. Online. Retrieved from https://www.globalcompact.de/
migrated files/wAssets/docs/Umweltschutz/Publikationen/DGCN
Discussion-paper TCFD screen 190528 k.pdf
Gourdel, R., Monasterolo, I., Dunz, N., Mazzocchetti, A., & Parisi, L. (2022). The
double materiality of climate physical and transition risks in the euro area (ecb
working paper series no 2665) (Tech. Rep.). European Central Bank. Retrieved
from https://doi.org/10.2139/ssrn.4115997 doi: 10.2139/ssrn.4115997
Green and Sustainable Finance Cluster Germany (GSFC Germany). (2019).
Anwender guidance zu den tcfd empfehlungen. Online. Retrieved from
https://gsfc-germany.com/wp-content/uploads/2019/08/190820 TCFD
Handreichungen DE-1.pdf
Greeuw, S. C. H., Asselt, M. B. A. V., Grosskurth, J., Storms, C. A. M. H., Rijkens-
Klomp, N., Rothman, D. S., . . . Ribeiro, T. (2000). Cloudy crystal balls: An as-
sessment of recent european and global scenario studies and models (Tech. Rep.).
European Environment Agency. Retrieved from https://www.eea.europa.eu/
publications/Environmental issues series 17/at download/file

63
Group, K. (2022). Sustainability report 2021. Retrieved from https://www.kbc.com/
content/dam/kbccom/doc/sustainability-responsibility/PerfRep/2021/
csr-sr-2021.pdf
Group, K. F. (2021). Financing for net zero futures - 2021 kb financial group tcfd re-
port. Retrieved from https://www.kbfg.com/common/jsp/fileDownUtil.jsp
?filepath=/Kor/esg/attachFile/2021%20KBFG%20TCFD%20Report.pdf
Group, L. B. (2022). Financing a green future - climate report 2021. Re-
trieved from https://www.lloydsbankinggroup.com/assets/pdfs/
investors/financial-performance/lloyds-banking-group-plc/2021/
q4/2021-lbg-climate-report.pdf
Group, N. (2022). Climate-related disclosures report. Retrieved from https://
investors.natwestgroup.com/~/media/Files/R/RBS-IR-V2/results
-center/18022022/2021-climate-related-disclosure-report.pdf
Group, S. B. (2022). Climate-related financial disclosures (tcfd aligned reporting) 2021.
Retrieved from https://thevault.exchange/?get group doc=18/1648721862
-SBGClimateRelatedFinancialDisclosures2021.pdf
Group, S. F. (2021). Tcfd report. Retrieved from https://
carbonaccountingfinancials.com/files/institutions downloads/
shinhan-financial-group-tcfd-report-2021.pdf?c64aec34dc
Hahnkamper-Vandenbulcke, N. (2021). Non-financial reporting directive - brief-
ing (Tech. Rep.). European Parliament Research Service. Retrieved from
https://www.europarl.europa.eu/RegData/etudes/BRIE/2021/654213/
EPRS BRI(2021)654213 EN.pdf
Hernández, J., Garcı́a, F. J. P., Suárez, N., & Tarancón, J. (2022). A study on
the eba stress test results: influence of bank, portfolio and country-level char-
acteristics (ecb working paper series no 2648) (Tech. Rep.). European Cen-
tral Bank. Retrieved from https://www.ecb.europa.eu/pub/pdf/scpwps/
ecb.wp2648~e5d429eea8.en.pdf
Hertwig, R. (2006). Strategien und heuristiken (J. Funke & P. A. Frensch, Eds.).
Huiskamp, U., Brinke, B., & Kramer, G. J. (2022). The climate resilience cycle:
Using scenario analysis to inform climate-resilient business strategies. Business
Strategy and the Environment, 31 (4), 1763–1775. doi: 10.1002/bse.2982
Hulme, M., & Dessai, S. (2008). Negotiating future climates for public policy: a critical
assessment of the development of climate scenarios for the uk. Environmental
Science Policy, 11 (1), 54–70. doi: 10.1016/j.envsci.2007.09.003
ING Groep (ING). (2020). Climate risk report 2020: Our approach to climate risk

64
 management. Online. Retrieved from https://www.ing.com/MediaEditPage/
ING-Climate-Risk-report-2020.htm
ING Groep (ING). (2021). Climate report 2021. Online. Retrieved from https://
www.ing.com/2021-Climate-Report.htm
ING Research. (2020). Break it or make it - long term impact of tech-
nology and policy developments on the global energy transition. On-
line. Retrieved from https://think.ing.com/uploads/reports/ING Energy
Scenarios Report - FINAL3.pdf
Institutional Investors Group on Climate Change (IIGCC). (2019). Navigating
climate scenario analysis. a guide for institutional investors (Tech. Rep.).
IIGCC. Retrieved from https://www.iigcc.org/download/navigating
-climate-scenario-analysis-a-guide-for-institutional-investors/
?wpdmdl=1837&refresh=6304f3ac78fa91661268908
Intergovernmental Panel on Climate Change (IPCC). (2018). Summary for policymak-
ers. In V. Masson-Delmotte et al. (Eds.), Global warming of 1.5°c. an ipcc special
report on the impacts of global warming of 1.5°c above pre-industrial levels and
related global greenhouse gas emission pathways, in the context of strengthening
the global response to the threat of climate change (p. 3–24). Cambridge Uni-
versity Press. Retrieved from https://doi.org/10.1017/9781009157940.001
doi: 10.1017/9781009157940.001
John, C., Khaykin, I., Pyanet, A., & Westheim, J. (2018). Extending our horizons,-
assessing credit risk and opportunity in a changing climate: Outputs of a working
group of 16 banks piloting the tcfd recommendations (Tech. Rep.). Oliver Wyman
UNEP FI. Retrieved from https://doi.org/10.1007/978-3-319-07656-0 4
doi: 10.1007/978-3-319-07656-0 4
Jürgens, I., Kempa, K., Klein, C., Marchewitz, C., Moslener, U., Neuhoff, K., . . .
Schütze, F. (2021). The eu sustainable finance strategy – implications for the
future german federal government [policy brief ] (Tech. Rep.). Research Platform
Sustainable Finance. Retrieved from https://wpsf.de/wp-content/uploads/
2021/11/WPSF PB 9-2021 EU Strategy-1.pdf
Kishita, Y., Hara, K., Uwasu, M., & Umeda, Y. (2016). Research needs and challenges
faced in supporting scenario design in sustainability science: a literature review.
Sustainability Science, 11 (2), 331–347. doi: 10.1007/s11625-015-0340-6
Koberle, A., Ganguly, G., & Ostrovnaya, A. (2021). A guide to building climate-
financial scenarios for financial institutions (grantham institute briefing paper
no. 35) (Tech. Rep.). Imperial College London. Retrieved from https://doi

65
 .org/10.25561/85497 doi: 10.25561/85497
Kosow, H., & Gaßner, R. (2008). Methods of future and scenario analysis: Overview,
assessment, and selection criteria. Retrieved from https://nbn-resolving
.org/urn:nbn:de:0168-ssoar-193660
Little, J. C., Hester, E. T., & Carey, C. C. (2016). Assessing and enhancing en-
vironmental sustainability: A conceptual review. Environmental Science and
Technology, 50 (13), 6830–6845. doi: 10.1021/acs.est.6b00298
Liu, L., Hejazi, M., Patel, P., Kyle, P., Davies, E., Zhou, Y., . . . Edmonds, J. (2015).
Water demands for electricity generation in the u.s.: Modeling different scenarios
for the water – energy nexus. Technological Forecasting Social Change, 94 ,
318–334. doi: 10.1016/j.techfore.2014.11.004
Nations, U. (2015). Paris agreement. Retrieved from https://unfccc.int/sites/
default/files/english paris agreement.pdf
Network for Greening the Financial System (NGFS). (2020). Guide to
climate scenario analysis for central banks and supervisors. Retrieved
from https://www.ngfs.net/sites/default/files/medias/documents/ngfs
guide scenario analysis final.pdf
Network for Greening the Financial System (NGFS). (2021). Ngfs cli-
mate scenarios for central banks and supervisors. Retrieved from
https://www.ngfs.net/sites/default/files/media/2021/08/27/
ngfs climate scenarios phase2 june2021.pdf
Nomura. (2021). Nomura tcfd report 2021 - implementation of tcfd recommendations.
Retrieved from https://www.nomuraholdings.com/investor/library/tcfd/
2021/pdf/all.pdf
on Climate Related Financial Disclosures (TCFD), T. F. (2017a). Implement-
ing the recommendations of the task force on climate-related financial disclo-
sures. Retrieved from https://assets.bbhub.io/company/sites/60/2020/
10/FINAL-TCFD-Annex-Amended-121517.pdf
on Climate Related Financial Disclosures (TCFD), T. F. (2017b). Recom-
mendations of the task force on climate-related financial disclosures. Re-
trieved from https://assets.bbhub.io/company/sites/60/2021/10/FINAL
-2017-TCFD-Report.pdf
on Climate Related Financial Disclosures (TCFD), T. F. (2017c). Technical supplement
- the use of scenario analysis in disclosure of climate-related risks and opportuni-
ties. Retrieved from https://www.fsb-tcfd.org/wp-content/uploads/2017/
06/FINAL-TCFD-Technical-Supplement-062917.pdf

66
on Climate Related Financial Disclosures (TCFD), T. F. (2021a). Publications. Re-
trieved from https://www.fsb-tcfd.org/publications/
on Climate Related Financial Disclosures (TCFD), T. F. (2021b). Task force
on climate-related financial disclosures - 2021 status report. Retrieved from
https://www.fsb.org/wp-content/uploads/P141021-1.pdf
on Climate Related Financial Disclosures (TCFD), T. F. (2022). About - our goal.
Retrieved from https://www.fsb-tcfd.org/about/#our-goal
O’Neill, B. C., Kriegler, E., Ebi, K. L., Kemp-Benedict, E., Riahi, K., Rothman, D. S.,
. . . Solecki, W. (2017). The roads ahead: Narratives for shared socioeconomic
pathways describing world futures in the 21st century. Global Environmental
Change, 42 , 169–180. doi: 10.1016/j.gloenvcha.2015.01.004
Phillips, R. L., & Ormsby, R. (2016). Industry classification schemes: An analysis
and review. Journal of Business and Finance Librarianship, 21 (1), 1–25. doi:
10.1080/08963568.2015.1110229
Ramirez, R., Mukherjee, M., Vezzoli, S., & Kramer, A. M. (2015). Scenarios as a
scholarly methodology to produce “interesting research”. Futures, 71 , 70–87.
doi: 10.1016/j.futures.2015.06.006
Riahi, K., van Vuuren, D. P., Kriegler, E., Edmonds, J., O’Neill, B. C., Fujimori,
S., . . . Tavoni, M. (2017). The shared socioeconomic pathways and their en-
ergy, land use, and greenhouse gas emissions implications: An overview. Global
Environmental Change, 42 , 153–168. doi: 10.1016/j.gloenvcha.2016.05.009
Ridder, H.-G. (2017). The theory contribution of case study research designs. Business
Research, 10 (2), 281–305. Retrieved from https://doi.org/10.1007/s40685
-017-0045-z doi: 10.1007/s40685-017-0045-z
Ritchie, J., & Dowlatabadi, H. (2018). Defining climate change scenario characteristics
with a phase space of cumulative primary energy and carbon intensity. Envi-
ronmental Research Letters, 13 (2). Retrieved from https://doi.org/10.1088/
1748-9326/aaa494 doi: 10.1088/1748-9326/aaa494
Ruggiero, L. F. (2007). Scientific independence: A key to credibility. Retrieved from
https://www.fs.fed.us/rm/pubs other/rmrs 2007 ruggiero l001.pdf
Sahin, C., de Haan, J., & Neretina, E. (2020). Banking stress test effects on returns
and risks. Journal of Banking and Finance, 117 , 105843. doi: 10.1016/j.jbankfin
.2020.105843
Sanpaolo, I. (2021). 2020/2021 tcfd report. Retrieved from https://group
.intesasanpaolo.com/content/dam/portalgroup/repository-documenti/
sostenibilt/inglese/2021/TCFD%20Report 2021.11.pdf

67
Santander. (2021). Climate finance report. Retrieved from https://www.santander
.com/content/dam/santander-com/en/documentos/informe-anual-de
-sostenibilidad/2020/ias-2020-climate-finance-2020-21-en.pdf
Saunders, M., Lewis, P., & Thornhill, A. (2016). Research methods for business
students (7th ed.). Pearson Education Limited.
Schirrmeister, E., Göhring, A., & Warnke, P. (2020). Psychological biases and heuris-
tics in the context of foresight and scenario processes. Futures Foresight Science,
2 (3), 1–18. doi: 10.1002/ffo2.31
Schoemaker, P. J. H. (1991). When and how to use scenario planning: A heuristic
approach with illustration. Journal of Forecasting, 10 (6), 549–564. doi: 10.1002/
for.3980100602
Schoemaker, P. J. H. (1993). Multiple scenario development: Its conceptual and
behavioral foundation. Strategic Management Journal , 14 (3), 193–213. doi:
10.1002/smj.4250140304
Schularick, M., Steffen, S., & Tröger, T. (2020). Bank capital and
the european recovery from the covid-19 crisis (safe white paper no. 69)
(Tech. Rep.). Leibniz Institute for Financial Research SAFE. Retrieved
from https://safe-frankfurt.de/fileadmin/user upload/editor common/
Policy Center/SAFE White Paper 69.pdf
Sovcombank. (2022). Climate risk management report - in accordance with tcfd.
Retrieved from https://sovcombank.ru/document?id=10125
Tiberius, V., Siglow, C., & Sendra-Garcı́a, J. (2020). Scenarios in business and
management: The current stock and research opportunities (Vol. 121). doi: 10
.1016/j.jbusres.2020.08.037
(TSKB), T. S. K. B. (2021). Climate risk report on task force on climate-related finan-
cial disclosures (tcfd). Retrieved from https://www.tskb.com.tr/i/assets/
document/pdf/TCFD-eng-2021-05-24.pdf
UBS. (2022). Climate report 2021. Retrieved from https://www.ubs.com/
global/de/sustainability-impact/net-zero/ jcr content/mainpar/
toplevelgrid 2069146059/col3/tile copy copy.1278436393.file/
bGluay9wYXRoPS9jb250ZW50L2RhbS9hc3NldHMvY2Mvc3VzdGFpbmFiaWxpdHktYW5kLWltcGFjdC9kb2MvMjAy
ubs-climate-report-2021-en.pdf
van Notten, P. W. F., Rotmans, J., van Asselt, M. B. A., & Rothman, D. S. (2003).
An updated scenario typology. Futures, 35 (5), 423–443. doi: 10.1016/S0016
-3287(02)00090-3
van Vuuren, D. P., Kok, M. T. J., Girod, B., Lucas, P. L., & de Vries, B. (2012).

68
 Scenarios in global environmental assessments: Key characteristics and lessons
for future use. Global Environmental Change, 22 (4), 884–895. doi: 10.1016/
j.gloenvcha.2012.06.001
Ven, A. H. V. D., & Johnson, P. E. (2006). Knowledge for theory and prac-
tice. Academy of Management Review , 31 (4), 802–821. doi: 10.5465/amr.2006
.22527385

69