Credit Risk: Estimating

Default and Migration

Probabilities
CHAPTER 6
Introduction
Credit risk is the risk of loss that may arise due to either the default or due to
decline in the credit standing of a borrower, bond issuer, or counterparty in a
derivative transaction
Default can be defined in a number of ways, viz.,
◦ Missing of a contractual payment
◦ Bankruptcy
◦ Restructuring
◦ Fall in the value of the assets to less than the value of debt
◦ Breach of a covenant of securitization structure
Example
ABC Ltd. had issued bonds a few years back.
The bonds are due to mature in the near future.
Meanwhile, the company has become financially distressed.
It has made an offer to its bondholders to repurchase the bond at 80% of the
face value.
The company mentions in the offer letter that it will have to file for bankruptcy if
the bondholders insist on redemption of bonds at par.
The action of the company amounts to default.
Credit Risk VaR
Credit risk VaR is central to determination of regulatory capital for credit risk and
for credit risk management.
It is the level of loss from defaults and migrations that will not be exceeded X%
of the time over T years.
Some credit risk VaR models consider only losses from defaults.
Other models also consider losses from downgrades or credit spread changes.
Credit Risk VaR
The key inputs for the calculation of credit risk VaR are:
[1] Probability of default (PD) or migration in rating category
[2] Exposure at default (EAD)
[3] Loss given default (LGD) (based on recovery data)
[4] Credit risk correlations
Measuring Probability of Default
Measurement of default and migration probabilities is essential for measuring
credit risk
The techniques include:
[1] Credit scoring
[2] External ratings, internal ratings
[3] Use of financial statements of obligors
[4] Use of historical data on defaults and migrations
[5] Use of market data on bond prices, bond spreads and equity prices
Credit Scoring
Mainly used to measure the credit risk of retail banking exposures [Details]
Features of retail credit risk:
[1] Loss due to default by a single customer is not very high
[2] Default by retail customers are independent of one another
[3] Retail loan portfolios are well-diversified. Hence, easier to estimate expected
loss, and price loans accordingly.
[4] Banks can get warning signals in advance of likely default by customers based
on their loan servicing behavior. Banks can take preventive actions: reduce
exposure to such customers/change the terms of the loan
Credit Scoring for Retail Exposures
Credit scoring models estimate the probability of default based on
characteristics of the obligor
Banks have large amount of historical data on the default behavior of obligors
and the characteristics of defaulting obligors
Credit scoring models convert this information into numbers
Higher score ⟹ lower credit risk
Models can be used to decide whether to accept or reject a credit application/to
estimate probability of default
Credit Scoring for Retail Exposures
Credit scoring models: Benefits
They make the credit assessment process less time consuming and more
efficient
Banks can identify risky customers and businesses they would like to
avoid/curtail
Credit scoring can be used to automate the process of credit decision making for
small credits and credit cards, to a large extent
The credit decision becomes more objective and consistent
Credit Risk Assessment for Non-Retail
Exposures
Non-retail exposures (Commercial credit risk) is related to:
◦ Corporate exposures
◦ Bank exposures
◦ Sovereign exposures
Assessment is relatively difficult as the likelihood of default and the loss given
default are affected by a multitude of factors
Assessment of creditworthiness of non-retail customers requires analysis of
both financial (profitability, liquidity, solvency) and non-financial factors (quality
of management, industry analysis, macroeconomic factors)
Credit rating systems consider all these factors in arriving at a credit rating for
the potential borrower
The External Rating Agency
A credit rating given by an external agency is an opinion on
◦ either the general creditworthiness of an obligor
◦ or the creditworthiness of an obligor in relation to a specific debt security
It is an opinion on the obligor’s overall capacity to meet its financial
obligations/to make timely payments on principal and interest on that security
In theory, a credit rating is an attribute of a bond issue, not a company. However,
in many cases all bonds issued by a company have the same rating. A rating is
therefore often referred to as an attribute of a company.
The External Rating Agency
The rating of a specific issue is based on:
◦ Assessment of the issuer
◦ Terms of the issue
◦ Collateral
◦ Guarantors
Assessment is done from quantitative, qualitative and legal perspectives
Credit Rating Agencies: S&P, Moody’s
The External Rating Agency
The ratings are reviewed every year
Ratings change relatively infrequently: One of the objectives of rating agencies
when they assign ratings is stability. They want to avoid ratings reversals, where
a company’s bonds are downgraded and then upgraded a few weeks later.
Ratings therefore change only when there is reason to believe that a long-term
change in the company’s creditworthiness has taken place.
The External Rating Agency
The reason for this is that bond traders are major users of ratings.
◦ Often they are subject to rules governing what the credit ratings of the bonds they
hold must be.
◦ If these ratings changed frequently, they might have to do a large amount of trading
(and incur high transaction costs) just to satisfy the rules.
Further, rating agencies try to “rate through the cycle.”
◦ Suppose that the economy exhibits a downturn and this has the effect of increasing
the probability of a company defaulting in the next six months, but makes very little
difference to the company’s probability of defaulting over the next three to five
years.
◦ A rating agency would not usually change the company’s credit rating in these
circumstances.
Internal Risk Rating Systems
Most banks have procedures for rating the creditworthiness of their corporate
and retail clients.
This is a necessity.
The ratings published by rating agencies are usually available only for companies
that have issued publicly traded debt.
As a result, many small and medium-sized companies do not have credit ratings
from rating agencies.
Further, the internal-ratings-based (IRB) approach allows banks to use their
internal ratings in determining the probability of default, PD.
Prediction Approach
Historical data from financial statements is analyzed statistically to predict
default behavior
A well-known model is Altman’s Z model
Edward Altman has pioneered the use of accounting ratios to predict default.
In 1968, he developed what has become known as the Altman’s Z-score.
Altman’s Z-Score
Using statistical technique, he attempted to predict defaults from five
accounting ratios:

For publicly traded manufacturing companies, the original Z-score was:

Altman’s Z-Score
If the Z-score was greater than 3.0, the company was considered unlikely to
default. If it was between 2.7 and 3.0, there was reason to be “on alert.” If it was
between 1.8 and 2.7, there was a good chance of default. If it was less than 1.8,
the probability of a financial embarrassment was considered to be very high.
Variations on the model have been developed for manufacturing companies that
are not publicly traded and for non-manufacturing companies.
The Z-score methodology has been revised and extended since Altman’s original
research and can now be used to produce probabilities of default.
Example
Consider a company for which working capital is 170,000, total assets are
670,000, earnings before interest and taxes is 60,000, sales are 2,200,000, the
market value of equity is 380,000, total liabilities is 240,000, and retained
earnings is 300,000.
In this case, X1 = 0.254, X2 = 0.448, X3 = 0.0896, X4 = 1.583, and X5 = 3.284.
The Z-score is
1.2 × 0.254 + 1.4 × 0.448 + 3.3 × 0.0896 + 0.6 × 1.583 + 0.999 × 3.284 = 5.46
The Z-score indicates that the company is not in danger of defaulting in the near
future.
Calculating Default Probability using
Historical Data
Data on defaults, migrations, and recoveries are available from historical records
or from rating agencies
 Historical Default Probabilities

Source: https://www.spglobal.com/ratings/en/research/articles/240328-default-transition-and-recovery-2023-annual-global-corporate-default-and-rating-
transition-study-13047827#:~:text=These%20conditions%20heightened%20the%20credit,chart%201%20and%20table%201).
Historical Default Probabilities
The table is typical of the data that are produced by rating agencies.
It shows the default experience through time of companies that started with a
certain credit rating.
Example: A bond with an initial credit rating of BBB has a 0.14% chance of
defaulting by the end of the first year, a 0.39% chance of defaulting by the end
of the second year, and so on.
The probability of a bond defaulting during a particular year can be calculated
from the table.
Example: the probability that a bond initially rated BBB will default during the
second year of its life is 0.39 − 0.14 = 0.25%.
Historical Default Probabilities
For investment-grade bonds, the probability of default in a year tends to be an
increasing function of time.
◦ Example: AA (0.02%, 0.03%, 0.06%, 0.08%, 0.09% for years 1/2/3/4/5)
◦ Reason: The bond issuer is initially considered to be creditworthy and the more time
that elapses, the greater the possibility that its financial health will decline.
For bonds with a poor credit rating, the probability of default is often a
decreasing function of time.
◦ Example: CCC/C (25.98%, 9.97%, 5.47%, 3.07%, 2.16% for years 1/2/3/4/5)
◦ Reason: For a bond with a poor credit rating, the next year or two may be critical. If
the issuer survives this period, its financial health is likely to have improved.
Historical Default Probabilities: Hazard
Rate
From the given table, the probability of a CCC/C bond defaulting during the third
year is 41.42 − 35.95 = 5.47%.
This is referred to as the unconditional default probability.
It is the probability of default during the third year as seen at time zero.
The probability that the CCC/C-rated bond will survive until the end of year two
is 100 − 35.95 = 64.05%.
The probability that it will default during the third year conditional on no earlier
default is therefore 0.0547∕0.6405 or 8.54%.
This is a conditional default probability for a one-year time period.
Unconditional and Conditional Default
Probabilities
The unconditional default probability:
◦ The probability of default for a certain time period measured at the time of
origination of a loan
The conditional default probability:
◦ Risk of default after origination of the loan
◦ The probability of default in a certain time period given that a default has not
occurred so far
Historical Default Probabilities: Hazard
Rate
When we consider a conditional default probability for a short time period of
length Δt, we get a measure known as the hazard rate or default intensity.
The hazard rate, λ(t), at time t is defined so that λ(t)Δt is the probability of
default between time t and t +Δt conditional on no default between time zero
and time t.
If V(t) is the cumulative probability of the company surviving to time t (i.e., no
default by time t), and Q(t) is the probability of default by time t:
𝑄 𝑡 = 1 − 𝑉(𝑡)
ഥ 𝑡 𝑡
𝑄 𝑡 = 1 − 𝑒 −𝜆
where λത 𝑡 is the average hazard rate between time zero and time t.
Example
Suppose that the hazard rate is a constant 1.5% per year.
The probability of a default by the end of the first year is:

The probability of a default by the end of the second year is:

The probabilities of a default by the end of the third, fourth, and fifth years are
similarly 0.0440, 0.0582, and 0.0723.
The unconditional probability of a default during the fourth year is 0.0582 −
0.0440 = 0.0142. The probability of default in the fourth year, conditional on no
earlier default, is 0.0142∕(1 − 0.0440) = 0.0149.
Cumulative Default Probability
The probability that a borrower will default over a multiyear period
𝐶𝑝 = 1 − (𝑝1 ∗ 𝑝2 ∗ ⋯ ∗ 𝑝𝑛 )
𝑝𝑛 is the marginal probability of no default for year n
In the first year, the cumulative and marginal default probabilities are equal
Cumulative Default Probability
If the marginal probabilities of default for years 1 and 2 are 0.06 and 0.08
respectively, the cumulative probability that the default will occur over a 2-year
period is:
1-(0.94*0.92) = 13.52%

OR
P(default in year 1) = 0.06
P(default in year 2) = 0.94*0.08 = 0.0752
P(default) = 0.06+0.0752 = 0.1352
Cumulative Default Probability
From the following transition matrix, calculate the probability of default of a
“B”-rated entity over a 2-year period:
Cumulative Default Probability
For a B-rated entity, the year 2 default probability is given by:
(0.04*0.00)+(0.89*0.01)+(0.06*0.12) = 0.0161

Two period cumulative probability of default is:

0.01+0.0161 = 0.0261 = 2.61%
Calculation of Default Probability Using
Market Data
Corporate bonds offer an extra yield (risk premium) to compensate for the
probability of default
Similarly, loans to risky borrowers carry a higher interest rate to cover the
default risk
Example: The risk-free return for 1-year is 10%
The expected return from a corporate bond of Grade B is 15.8%
The probability of repayment perceived by the market is 1.10/1.158 = 0.95
The probability of default is 0.05
The risk premium for a default probability of 5% is 5.8%
Calculation of Default Probability Using
Market Data: Example
The risk-free return for 1-year is 10%
The expected return from a corporate bond of Grade B is 15.8%
Assume that the PD = x
The following relationship should hold: Return from risk-fee investment should
be equal to the return from risky investment taking risk into account
1.10 = (x*0 assuming no recovery)+((1-x)*1.158)
1-x = 1.10/1.158 = 0.95
The probability of default is 0.05
The risk premium for a default probability of 5% is 5.8%
Estimating Default Probabilities from
Bond Spreads: Example
The yield on a 1-year Treasury bill is 2.5%
The yield on a 1-year Corporate bond is 3.5%
The recovery in the case of default is zero
The probability of default (PD)?
Answer: 1-(1.025/1.035) = 0.00966 = 0.97%
If the Recovery Rate (RR) is 50% :
1.025 = 1.035(1-PD) + 1.035(PD)(RR)
0.5175 PD = 0.01
PD = 0.0193 = 1.93%
Estimating Default Probabilities from
Credit Spreads: Approximate Calculation
Bond spread is a specific example of credit spread (Details)
Average probability of default per year (over the life of an instrument) is given
by:

where s(T) is the credit spread for a maturity of T,

R is the recovery rate, and
λ is the average hazard rate between time zero and time T.
Estimating Default Probabilities from
Credit Spreads: Example
Suppose that a five-year credit spread for a company is 240 basis points and that
the expected recovery rate in the event of a default is 40%.
The average probability of a default per year over the five-year period,
conditional on no earlier default, is:
0.024∕(1 − 0.4) or 4%.
Estimating Default Probabilities from
Credit Spreads: Example
Suppose that the credit spreads for 3-,5-,and 10-year instruments are 50, 60,and
100 basis points, respectively, and the expected recovery rate is 60%.
The average hazard rate over three years is approximately 0.005∕(1 − 0.6) =
0.0125.
The average hazard rate over five years is approximately 0.006∕(1 − 0.6) = 0.015.
The average hazard rate over 10 years is approximately 0.01∕(1 − 0.6) = 0.025.
From this we can estimate that the average hazard rate between year 3 and year
5 is (5 × 0.015 − 3 × 0.0125)∕2 = 0.01875.
The average hazard rate between year 5 and year 10 is (10 × 0.025 − 5 × 0.015)∕5
= 0.035.
Example
The 3-year credit spread is 60 basis points
The 5-year credit spread is 70 basis points
The expected recovery rate is 70%
Average hazard rate over 3 years = 0.006/(1-0.7)=0.02
Average hazard rate over 5 years = 0.007/(1-0.7)=0.0233
Average hazard rate between years 3 and 5 is:
[(0.0233*5)-(0.02*3)]/2 = 0.0283
Recovery Rates
When a company goes bankrupt, those who are owed money by the company
file claims against the company.
Sometimes there is a reorganization in which these creditors agree to a partial
payment of their claims.
In other cases, the assets are sold by the liquidator and the proceeds are used to
meet the claims as far as possible.
Some claims typically have priorities over other claims and are met more fully.
The recovery rate for a bond is normally defined as the price at which it trades
about one month after default as a percent of its face value.
The percentage recovery rate is 100 minus the percentage loss given default
(LGD).
Recovery Rates
Exposure at Default (EAD) is the amount that is owed by the counterparty at the
time of default.
Loss Given Default (LGD) is the proportion of EAD that is expected to be lost in
the event of default.
◦ It is calculated as 1 minus the recovery rate.
◦ For example, if a bank expects to recover 30% of the amount owed in the event of
default, LGD = 0.7
Probability of Default (PD) is the probability of the counterparty defaulting
within a given time period.
Recovery Rates: The Dependence of
Recovery Rates on Default Rates
The average recovery rate on corporate bonds exhibits a negative dependence
on default rates.
In a year when the number of bonds defaulting is low, economic conditions are
usually good and the average recovery rate on those bonds that do default
might be as high as 60%.
In a year when the default rate on corporate bonds is high, economic conditions
are usually poor and the average recovery rate on the defaulting bonds might be
as low as 30%.
The result of the negative dependence is that a bad year for defaults is doubly
bad for a lender because it is usually accompanied by a low recovery rate.
Recovery Rates: The Dependence of
Recovery Rates on Default Rates
Why?
One of the lessons from the credit crisis of 2007–2008 is that the average
recovery rate on mortgages is negatively related to the mortgage default rate.
As the mortgage default rate increases, foreclosures lead to more houses being
offered for sale and a decline in house prices.
This in turn results in a decline in recovery rates.
Thus,
More defaults lead to more assets for liquidation.
Higher supply leads to fall in prices.
Migration Probabilities
Credit ratings assigned by rating agencies do not remain constant
The ratings are revised from time to time
The evolution of ratings over a period of time is captured by migration matrices
or transition matrices prepared by rating agencies
The migration matrices depict the probability of migration of an entity from one
credit-risk category to other credit-risk categories over a period of time
The migration probabilities are based on historical data
Over a period, the risk may improve or deteriorate
Further Reading
Retail banking exposure
Retail banking: Mainly concerned with acceptance of deposits from customers
and granting credit to them
Home mortgage loans
◦ Secured by mortgage of the house property that is financed
◦ Credit risk of the bank depends on the ratio of the loan to the value of the underlying
property (loan to value ratio)
Home equity loans
◦ Loan against the equity of the borrower in his residential property
◦ Equity is the difference between the current value of the residential property and the
balance of the mortgage loan due on that property
Retail banking exposure
Instalment loans
◦ Include educational loans, vehicle loans, and other similar loans
◦ Generally secured
Credit card loans
◦ Revolving in nature
◦ Unsecured
Loans to small business enterprises
◦ Business loans of small amount: considered as retail loans
◦ Secured loans (secured by assets owned by the business)
[Back]
Credit Scoring for Retail Exposures
Approaches for credit scoring in retail portfolios:
Application scoring model: Assign scores based on the credit history of
applicants gathered from the loan application and credit bureaus.
Example: CIBIL, CRIF High Mark
Behavioral Scoring: Based on history of credit behavior of the loan applicant
Dynamic Scoring: Makes use of statistical techniques and makes predictions for
different time horizons
Application Scoring Models
Assign scores based on the credit history of applicants gathered from the loan
application and credit bureaus
Application scoring done at the time of the loan application to take acceptance
or rejection decision
Step 1: Set the definition of default
Step 2: Identify information relevant to prediction of default
Step 3: Collect relevant information (age, gender, income, history of
delinquencies, outstanding amounts, etc.)
Step4: Assign scores to different attributes and reach to a final score
Application Scoring Models
A number of decisions are involved in building the scorecard, viz.,
◦ What characteristics (variables) to include
◦ Number of categories to use for a characteristic
◦ Number of points to assign to an attribute
◦ Setting the cut-off
A score measures relative risk and is useful for the purpose of ranking customers
on the basis of risk
For regulatory requirements, the scores, however, is not sufficient.
What is required is the probability of default.
Behavioral Scoring Model
Is based on the history of credit behavior of the loan applicant
Used to assess the credit risk of an already existing obligor based on his behavior
◦ Average of account balance
◦ Delinquency in payment
◦ Change in job status
A 24-month time frame is used: the past 12-months information is used to
predict the default status over the next 12 months
Is dynamic as it is based on dynamic variables
◦ Changing balances in the customer account
Behavioral Scoring Model
It uses a much larger number of variables
◦ Maximum/average/minimum balance in account
◦ History of missed payments
The different roles that a customer can take should be considered
◦ Borrower/guarantor
◦ Savings account/term deposit
Variables are measured at different points in time
◦ Average measures/ratios are used
Dynamic Scoring Model
Makes use of statistical techniques
Makes predictions for different time horizons
Measures risk at different points of time in the future, for example,
3/4/5/6/12/18 months and so on, unlike the other two scoring approaches
Difficult to implement as special statistical techniques are required (example,
survival analysis)
In addition to providing information on whether a borrower will default, also
tells when a borrower is likely to default
Credit Scoring for Retail Exposures
Treatment of Reject and Withdrawal Cases:
Withdrawal cases are those which do not accept the bank’s offer because they
find a better offer elsewhere
Application scorecards are built on the basis of historical accept data and not
reject or withdrawal data
This creates a bias, that can be addressed through different approaches, viz.,
◦ Obtaining their information from credit bureaus
◦ Grant credit to some rejects also and observe their behavior
◦ Nearest neighbour method
Credit Risk Assessment for Non-Retail
Exposures: Expert-based approach
A subjective approach when data is not available
Company characteristics such as industry position, market share, management
quality, etc. are assessed and score is assigned based on experience/intuition
Credit Risk Assessment for Non-Retail
Exposures: Shadow Rating Approach
It starts with a data set with ratings for a particular set of obligors.
In addition, information is collected for each obligor such as accounting ratios,
firm characteristics, stock price behavior, etc. that may have an influence on the
rating.
This information is combined in one data set and an analytical model is built to
predict the rating.
This model provides information about how different firm characteristics
contribute to the rating and provides advice to corporates on how to improve
their rating.
Credit Default Swaps
A credit default swap (CDS) is a derivative that allows market participants to
trade credit risks.
The simplest type of CDS is an instrument that provides insurance against the
risk of a default by a particular company.
The company is known as the reference entity and a default by the company is
known as a credit event.
The buyer of the insurance obtains the right to sell bonds issued by the company
for their face value when a credit event occurs and the seller of the insurance
agrees to buy the bonds for their face value when a credit event occurs.
The total face value of the bonds that can be sold is known as the credit default
swap’s notional principal.
Credit Default Swaps
The buyer of a CDS makes periodic payments to the seller until the end of the
life of the CDS or until a credit event occurs.
These payments are usually made in arrears every quarter.
The regular payments from the buyer of protection to the seller of protection
cease when there is a credit event.
However, because these payments are made in arrears, a final accrual payment
by the buyer is usually required.
The total amount paid per year, as a percent of the notional principal, to buy
protection is known as the CDS spread.
Credit Default Swaps: Example
Suppose that two parties enter into a five-year credit default swap on December
20, 2018.
Assume that the notional principal is $100 million and the buyer agrees to pay
90 basis points per year (quarterly in arrears) for protection against default by
the reference entity.
If the reference entity does not default (that is, there is no credit event), the
buyer receives no payoff and pays approximately $225,000 (= 0.25 × 0.0090 ×
100,000,000) on March 20, June 20, September 20, and December 20 of each of
the years 2019, 2020, 2021, 2022, and 2023.
If there is a credit event, a substantial payoff is likely.
Credit Default Swaps: Example
Suppose that the buyer notifies the seller of a credit event on May 20, 2021 (five
months into the third year).
If the contract specifies physical settlement, the buyer of protection has the
right to sell to the seller of protection bonds issued by the reference entity with
a face value of $100 million for $100 million.
If, as is now usual, there is a cash settlement, a two-stage auction process is
used to determine the mid-market value of the cheapest deliverable bond
several days after the credit event.
Suppose the auction indicates that the cheapest deliverable bond is worth $35
per $100 of face value. The cash payoff would be $65 million.
Credit Default Swaps: Example
The regular payments from the buyer of protection to the seller of protection
cease when there is a credit event.
However, because these payments are made in arrears, a final accrual payment
by the buyer is usually required.
In our example, where there is a default on May 20, 2021, the buyer would be
required to pay to the seller the amount of the annual payment accrued
between March 20, 2021, and May 20, 2021 (approximately $150,000), but no
further payments would be required.
The total amount paid per year, as a percent of the notional principal, to buy
protection is known as the CDS spread.
In our example, the CDS spread is 90 basis points.
Credit Default Swaps
A key aspect of a CDS contract is the definition of a credit event (i.e., a default).
Usually, a credit event is defined as a failure to make a payment as it becomes
due, a restructuring of debt, or a bankruptcy.
CDS in India
SBI: Annual Report 2024

HDFC: Annual Report 2024

ICICI: Annual Report 2024

Credit Spreads
The credit spread is the extra rate of interest per annum required by investors
for bearing a particular credit risk.
CDS spreads provide one measure of the credit spread.
Another is the bond yield spread. This is the amount by which the yield on a
corporate bond exceeds the yield on a similar risk-free bond.
The two should be approximately equal.

(Back)
Credit Spreads: CDS Spreads and Bond
Yields
Suppose that an investor buys a five-year corporate bond yielding 7% per year for its
face value and at the same time enters into a five-year CDS to buy protection against
the issuer of the bond defaulting.
Suppose that the CDS spread is 200 basis points or 2% per annum.
The effect of the CDS is to convert the corporate bond to a risk-free bond (at least
approximately).
If the bond issuer does not default, the investor earns 5% per year (when the CDS
spread is netted against the corporate bond yield).
If the bond issuer does default, the investor earns 5% up to the time of the default.
Under the terms of the CDS, the investor is then able to exchange the bond for its face
value.
This face value can be invested at the risk-free rate for the remainder of the five years.
Credit Spreads: The Risk-Free Rate
CDSs provide a direct estimate of the credit spread.
To calculate a credit spread from a bond yield, it is necessary to make an
assumption about the risk-free rate.
A number of researchers have compared bond yields to CDS spreads to imply a
risk-free rate.
This involves matching the maturities of CDSs and bonds and implying a risk-free
rate.
For example, if the five-year bond yield is 4.7% and the five-year CDS spread is
80 basis points, the implied five-year risk-free rate is 3.9%.
Credit Spreads: CDS–Bond Basis
The CDS–bond basis is the excess of the CDS spread over the bond yield spread
for a company.
CDS–Bond Basis = CDS Spread − Bond Yield Spread
CDS–Bond Basis = CDS Spread − (Bond Yield – Risk-Free Rate)
Ideally, the CDS–bond basis should be close to zero.
However, in practice it deviates from zero (can be positive or negative).
Reasons: Example:
◦ There is counterparty default risk in a CDS. (This pushes the basis in a negative
direction.)
◦ The restructuring clause in a CDS contract may lead to a payoff when there is no
default. (This pushes the basis in a positive direction.)
Option Theoretic Approach
This approach estimates the probability of default by comparing the value of
assets of a company with the amount of its outstanding debt
The company is expected to default when the value of its assets falls below the
value of its debt
The default probability or the “expected default frequency (EDF)” has an inverse
relationship with the gap between the expected value of assets and a threshold
value of debt that triggers default
This gap is called “distance to default”