Credit Risk Management

Credit Risk
 Basic risk in banking
• Banks are in the business of credit intermediation
• Credit risk is a result of being in the business
 “...the major cause of serious banking
problems continues to be directly related to lax
credit standards for borrowers and
counterparties...”
Basle Committee on Banking Supervision, 2000
2
 Credit Risk

► Sources of credit risk in different financial transactions

 Bank credits Non-payment by borrower

 Money market (bank deposits) Non-payment by borrower

Non-payment by issuer
 Securities transactions Failure to deliver / pay
Increase in probability of default

 Foreign exchange transactions Non-payment by seller

 Derivatives
►OTC transactions Failure to deliver / pay
►Exchange traded instruments No credit risk!

3
Credit Risk
•Definition
• Risk of loss due to
• default of the counterparty to fulfill its obligations in a financial
transaction (default risk)
• decrease in market value of an asset due to deterioration of the
borrower’s credit quality (increase in default probability/spread risk)

•Credit risk could be analyzed in two ways

• Direct vs. indirect credit risk
• Settlement vs. Pre-settlement credit risk

4
Credit Risk
Direct vs. Indirect Credit Risk
•Direct credit risk
• Credit risk arising from factors directly related to the
counterparty / borrower
• Risk of non-performance (default risk)
• Increase in the probability of default (downgrade risk)

•Indirect credit risk

• Probability of loss due to change in credit quality of a
third party
• e.g. increase in general credit spread in a particular rating class
without any change in a counterparty’s default probability

5
Credit Risk
Settlement vs. Pre-Settlement Risk

►All financial transactions involve a settlement process

 Counterparties fulfill their obligations in financial contracts, e.g.
make payment, deliver assets etc.

►There is a possibility that the settlement in a financial

transaction may not go through as planned due to the failure of
the counterparty to fulfill its end of the bargain (default)

►Default can occur either before or at the settlement date

 Pre-settlement credit risk
 Settlement risk

6
Credit Risk
Settlement vs. Pre-Settlement Risk
►Pre-settlement credit risk
 Failure of a counterparty to perform on an obligation
before settlement, i.e. from the time it is agreed upon until
settlement (before the initiation of the settlement)
►Non-defaulting counterparty is informed about the failure of its
counterparty and does not have to go through the settlement
process
►There still may be a loss for non-defaulting counterparty as it may
have to replace the transaction with another counterparty at a
higher cost relative to the original transaction
 Replacement Cost Risk

7
Credit Risk
Settlement vs. Pre-Settlement Risk
►Replacement cost risk
If the counterparty defaults before the
settlement date, failed transaction needs to be
renewed with another counterparty
There is a risk that the transaction has to be
renewed at a worse price
Replacement cost risk is related to the market
risk of the underlying asset
►Higher price volatility means higher replacement
cost risk

8
Example: Interest Rate Swap
Example: Interest Rate Swap
Example: Interest Rate Swap

Credit Risk !
Credit Risk
Settlement vs. Pre-Settlement Risk
►Settlement risk
 Default occurs at the settlement date
 Amount of loss depends on the type of settlement
►No intermediary
 The risk of losing principal amount due to failure of the
counterparty (Principal Risk / Herstatt Risk)
►FX transactions
►Delivery-Versus-Payment (DvP Settlement)
 No principal risk since clearing/settlement agent makes sure
payment is received before delivering assets or vice versa
 But there is still a residual risk in DvP settlement
►Replacement Cost Risk as in the pre-settlement risk

12
Drivers of Credit Risk
►Quantification of credit risk requires
identification of key risk drivers
Default probability
Expected exposure at the time of default
►Exposure can vary during the life of a transaction
►Exposure at the time of default could be quite different than
current exposure
Loss given default
►Actual loss could be different than exposure at default due
to recoveries

13
Default Probability
•Default probability is an estimation for the
likelihood of default of a counterparty within a
certain time period
• Ratings given by external credit rating agencies
• Moody’s, Standard & Poors (S&P), Fitch Ratings

• Historical default statistics

• Internal ratings
• Transition matrices between ratings classes

14
Rating Scales

15
Rating Definitions
Moody’s Ratings for Long-term Obligations
►Investment grade ratings
 Aaa - Obligations rated Aaa are judged to be of the highest
quality, with minimal credit risk.
 Aa - Obligations rated Aa are judged to be of high quality
and are subject to very low credit risk.
 A - Obligations rated A are considered upper-medium grade
and are subject to low credit risk.
 Baa - Obligations rated Baa are subject to moderate credit
risk. They are considered medium-grade and as such may
possess certain speculative characteristics.

16
Rating Definitions
Moody’s Ratings for Long-term Obligations
• Speculative grade ratings
• Ba - Obligations rated Ba are judged to have speculative elements and are
subject to substantial credit risk.
• B - Obligations rated B are considered speculative and are subject to high credit
risk.
• Caa - Obligations rated Caa are judged to be of poor standing and are subject to
very high credit risk.
• Ca - Obligations rated Ca are highly speculative and are likely in, or very near,
default, with some prospect of recovery of principal and interest.
• C - Obligations rated C are the lowest rated class of bonds and are typically in
default, with little prospect for recovery of principal or interest.
• S&P and Fitch uses an additional rating grade called “D” implying a default
has already occurred through either
• Non-payment of interest or principal when these payments are due
• Upon filing for bankruptcy or similar legal action
• S&P also assigns “SD” or selective default
17
Rating Definitions
•Moody’s assigns further numerical classifiers 1, 2 or 3 to
distinguish relative credit quality within a certain rating grade
except for “AAA”
• e.g. Aa1, Aa2, Baa3, Ba1 etc.
•S&P and Fitch use “+” or “-” signs for the same purpose
• e.g. AA+, AA, BBB-, BB+ etc.
•All rating agencies also assign an opinion, which is called
“Outlook”, regarding the likely direction of any medium-term
(around 1 year) rating actions
• “Positive” indicating the next move is likely to be an upgrade
• “Stable” indicating a change is not expected in the near to medium-
term
• “Negative” indicating the next move is likely to be a downgrade

18
Rating Definitions
•Credit-watch or Watch-list
• Rating agencies place borrowers under review by giving an indication
about their next move, which is likely to come in less than three
months
• S&P places borrowers under review with “positive”, “negative” or
“developing” implications
• Moody’s places borrowers under review (on Watch-list) for possible
upgrade, downgrade or direction uncertain
•Different rating classes for different sectors
• Sovereigns, financials, corporates, structured finance, insurance,
counterparty ratings
• They usually have similar rating scales
• But, does that mean similar risk levels for different instruments with
same credit ratings?
• AAA government vs. AAA CDO

19
Criticism Against Rating Agencies
►Rating agencies are usually late to change their
ratings, especially on downgrades
►There is an inherent conflict of interest in their
business model
 They are paid by the institutions that they rate
►They have played a role in the current crisis by
getting involved in the design of the structured
products, which are the main cause of the current
credit crunch
 They are accused of advising borrowers on how to
structure borrowing instruments to achieve a certain
credit rating

20
Criticism Against Rating Agencies

04/04/2012 Copyright (C) 2012 - Buket İmir / BA 4514 - Risk Management 21

Transition Matrix

Sovereign Foreign Currency Average One-Year Transition Rates

1975-2007*
Rating one year later (%)
Rating as of
January 1 # of counts AAA AA A BBB BB B CCC/CC SD
AAA 394 97.97 2.03 0 0 0 0 0 0
AA 228 3.51 94.3 1.32 0 0.44 0.44 0 0
A 213 0 4.23 93.43 2.35 0 0 0 0
BBB 187 0 0 8.56 87.7 2.67 1.07 0 0
BB 224 0 0 0 6.25 87.05 4.46 1.34 0.89
B 179 0 0 0 0 10.06 84.92 3.35 1.68
CCC/CC 18 0 0 0 0 0 22.22 38.89 38.89

*Source: Standard & Poor's Sovereign Ratings

22
Transition Matrix
Corporate Transition Matrix – Moody’s

23
Expected Exposure
•Exposure is amount at risk
• Refers to the maximum amount that can be lost if the risk
materializes
• Different than “Value-at-Risk” which is a risk measure for market
risk
• It changes depending on the type of transaction
•Expected exposure is the estimated exposure at the time
of default
• Expected exposure is usually different than current exposure
• For calculating expected loss, expected exposures are more
relevant than current exposures
• But they are not easy to estimate in some cases

24
Recovery Rate
•In some cases, credit losses are less than the
exposure at the time of default
• Seniority of loans – some creditors get paid
before others
•Historical recovery rates
• Risk mitigating techniques
•Netting agreements
•Receiving collateral and other guarantees
•Other risks need to be considered
• Legal risk, market and liquidity risk of collateral, etc.

25
Recovery Rate
•Recovery rates published by Moody’s

Source: European Corporate Default and Recovery Rates, 1985-2006 by Moody’s

26
Loss Given Default (LGD)
•LGD refers to the expected amount of loss given
a counterparty’s default
• LGD = Exp. Exposure * (1 – Recovery Rate)

•LGD and default probability are the inputs for

calculating “Expected Loss”
• Expected Loss = LGD * Default Probability

Risk Severity of loss Probability of occurrence

27
Measuring Credit Risk
► A simple measure of credit risk
Expected loss (EL)
►EL = D * X * (1 - R)
where
D: Probability of default
X: Expected exposure
R: Recovery rate
►The formula is quite simple, but getting the
parameters is not !

28
Expected Loss
•How good as a measure for credit risk ?
• Expected loss is not really a risk measure since it is the
amount of loss you would expect on average
• The argument is that if you expect it, you should account for it
• Loan loss reserves
• Pricing credit products

• Once expected losses are reserved for, the capital is required

only for losses beyond expected loss
•So we need a better measure that can help us
estimate the “Unexpected Loss ” rather than the
expected loss

29
 Unexpected Loss
Credit Loss Distribution Probability

EL = E * D * (1 - R)

A simple approximation for Unexp. Loss:

WC(EL) = WC(E) * WC(D) * (1 – WC(R))

Confidence Level depends on the

targeted credit rating

Credit Losses
Unexpected Loss Expected Loss

Economic Capital

30
Unexpected Loss
►Using worst case scenarios is quite conservative
and unrealistic since it assumes
Correlations are 1
►Meaning that all borrowers default at the same time
 When exposures are the highest
 Recovery rate is the lowest
►A better risk measure for unexpected loss would be
related to return volatility similar to market risk
measure of Value-at-Risk

31
Loss Volatility
Portfolio of 3 Loans
Loss P robability P rob. Wtd. Loss P rob. Wtd. Sqrd Dev.
No Default o f A 0 95% 0 1.48
Default of A 25 5% 1.25 28.20
To tal 100% 1.25 29.69
Vo latility of A 5.45

No Default o f B 0 90% 0 8.10

Default of B 30 10% 3 72.90
To tal 100% 3 81.00

Vo latility of B 9.00

No Default o f C 0 80% 0 64.80

Default of C 45 20% 9 259.20
To tal 100% 9 324.00

Vo latility of C 18.00
 Loss Volatility
Portfolio Loss Distribution
(Assuming defaults are independent)

Lo ss Pr o b ab i li t y C um. Pr o b . Pr o b . W t d . Lo ss Pr o b . W t d . Sq r d D ev.

No Default 0 68.4% 68.4% 0.00 120.08

Default of A o nly 25 3.6% 72.0% 0.90 4.97
Default of B o nly 30 7.6% 79.6% 2.28 21.32
Default of C o nly 45 17.1% 96.7% 7.70 172.38
Default of A & B only 55 0.4% 97.1% 0.22 6.97
Default of A & C only 70 0.9% 98.0% 0.63 28.99
Default of B & C only 75 1.9% 99.9% 1.43 72.45
Default of A, B & C together 100 0.1% 100.0% 0.10 7.53
T otal 100.0% 13.25 434.69

P ortfolio Loss Vo latility 20.85

Loss Volatility
Credit Loss Distribution
Expected loss: 13.25 mill $
80%
70%
60%
50%
95 percentile of cumulative loss distribution
40% Unexpected loss: 45 mill $

30%
20%
10%
0%
0 -25 -30 -45 -55 -70 -75 -100