MBFI

Credit Risk Management

Credit Rating Based Models

Ashok Thampy
Professor, Finance and Accounting Area, IIMB
 Credit Rating
• Credit rating is a process by which securities of
differing levels of credit risk are put in different
rating buckets which indicate different levels of
default risk
• Credit rating done by – banks (internal) and rating
agencies (external)
• Rating agencies specialize in evaluating the credit
worthiness of corporate, municipal, and sovereign
issuers of debt securities.
 Credit Rating
• Rating agencies – very important in markets with
well developed debt markets (US) where capital
markets are the primary source of debt capital.
• Fixed income mutual funds – provide quality of
portfolio in terms of credit ratings
• Corporations – depend on credit ratings to
establish their credit worthiness which affects their
cost of borrowing.
• Regulators use credit rating by specifying
minimum rating for specific kinds of investments.
Example: Pension funds are generally required to
invest only in BBB rating and above instruments.
 Rating Agencies
• International:
– Moody’s Investors Service
– Standard and Poor’s (S&P)
– Fitch IBCA
• Local:
– Australian Rating (S&P)
– Canadian Bond Rating Service (CBRS)
– Agence d’Evaluation Financiere (S&P)
– Credit Rating Services of India Ltd (CRISIL)
– Information and Credit Research Agency (ICRA)
– Japan Bond Research Institute
– Korean Inverstors Services
 Long Term Senior Debt Rating Symbols
S&P Moody's Interpretation
AAA Aaa Highest quality; extremely strong
INVESTMENT GRADE

AA+ Aa1
AA Aa2
AA- Aa3 High quality
A+ A1
A A2
A- A3 Strong payment capacity
BBB+ Baa1
BBB+ Baa2
BBB- Baa3 Adequate payment capacity
BB+ Ba1
BB Ba2
SPECULATIVE GRADE

BB- Ba3 Likely to fulfill obligations; ongoing uncertainty

B+ B1
B B2
B- B3 High risk obligations
CCC+ Caa1
CCC Caa2
CCC- Caa3 Current vulnerability to default
C Ca In bankruptcy or default, or other marked shortcomings
D
Source: Caouette, J.B., Altman, E.I,. Altman, and P. Narayanan: Managing
Credit Risk, John Wiley & Sons. Pg69.
 S&P ratings definition

https://www.standardandpoors.com/en_US/web/guest/article/-
/view/sourceId/504352
 Average Cumulative15-Year Default Rates (%)
40

35 Investment Grade

30 Specualtive Grade

25

20

15

10

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Source: Standard & Poor

 Standard and Poor's Financial Ratios Across Ratings
Leverage (%) Cash flow coverage
Rating Total debt/Capital LT Debt/Capital EBITDA/Interest EBIT/Interest
AAA 23 13 26.5 21.4
AA 38 28 12.9 10.1
A 43 34 9.1 6.1
BBB 48 43 5.8 3.7
BB 63 57 3.4 2.1
B 75 70 1.8 0.8
CCC 88 69 1.3 0.1
Source: S&P's Corporate Ratings Criteria (2002), based on median financial ratios
over 1998 to 2000 for industrial corporations.
 Moody's Cumulative Default Rates (Percent), 1920-2002
Rating Year
1 2 3 4 5 6 7 8 9 10
Aaa 0.00 0.00 0.02 0.09 0.19 0.29 0.41 0.59 0.78 1.02
Aa 0.07 0.22 0.36 0.54 0.85 1.21 1.60 2.01 2.37 2.78
A 0.08 0.27 0.57 0.92 1.28 1.67 2.09 2.48 2.93 3.42
Baa 0.34 0.99 1.79 2.69 3.59 4.51 5.39 6.25 7.16 7.99
Ba 1.42 3.43 5.60 7.89 10.16 12.28 14.14 15.99 17.63 19.42
B 4.79 10.31 15.59 20.14 23.99 27.12 30.33 32.36 34.37 36.10
Caa-C 14.74 23.95 30.57 35.32 38.83 41.94 44.23 46.44 48.42 50.19
Investment 0.17 0.50 0.93 1.41 1.93 2.48 3.03 3.57 4.14 4.71
Speculative 3.83 7.75 11.41 14.69 17.58 20.09 22.28 24.30 26.05 27.80
All 1.50 3.09 4.62 6.02 7.28 8.41 9.43 10.38 11.27 12.14
Source: Financial Risk Manager Handbook, Wiley
 S & P Cumulative Default Rates (Percent), 1981-2002
Rating Year
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
AAA 0.00 0.00 0.03 0.07 0.11 0.20 0.30 0.47 0.54 0.61 0.61 0.61 0.61 0.75 0.92
AA 0.01 0.03 0.08 0.17 0.28 0.42 0.61 0.77 0.90 1.06 1.20 1.37 1.51 1.63 1.77
A 0.05 0.15 0.30 0.48 0.71 0.94 1.19 1.46 1.78 2.10 2.37 2.60 2.84 3.08 3.46
BBB 0.36 0.96 1.61 2.58 3.53 4.49 5.33 6.10 6.77 7.60 8.48 9.34 10.22 11.28 12.44
BB 1.47 4.49 8.18 11.69 14.77 17.99 20.43 22.63 24.85 26.61 28.47 29.76 30.98 31.70 32.56
B 6.72 14.99 22.19 27.83 31.99 35.37 38.56 41.25 42.90 44.59 45.84 46.92 47.71 48.68 49.57
C 30.95 40.35 46.43 51.25 56.77 58.74 59.46 59.85 61.57 62.92 63.41 63.41 63.41 64.25 64.25
 Problems with ratings data
• Small sample
• Difference between AAA and AA may not be very
significant.
• Default rate over 1 year period for AA=0.01%
• From the binomial process,
Variance = p(1-p)/N
• For N=8000, SD = 0.0001 = 0.01% = average
• Cannot really distinguish AA credit from AAA
credit
• If probability of default = 5% and N=100
– SD = 2.18% (very large)
Estimation of default rates for low probability events with
few observations can be very imprecise.
Probability of default = 0.01% Binomially distributed

Probability of no default = 99.99%

Variance = p(1-p)/N

where p is the probability of default

and N is the number of trials

N SD

1 1.000%

10 0.316%

100 0.100%

1000 0.032%

8000 0.011% ~ average

10000 0.010%
 Rating agencies and the Crisis
• “…..financial crisis have cost investors dear. As
always on such occasions, the search is on for
somebody to blame. Among the leading suspects
are the firms that rate the credit worthiness of
bond issuers. The raters, firms such as Moody’s
Investor’s Service, Standard and Poor’s, Duffs and
Phelps and IBCA, are supposed to be the financial
market’s early warning system. Instead, the
agencies have spent the past few months belatedly
reacting to events.”
Rating agencies and the Asian Crisis

• “Asia’s financial crisis have cost investors dear.

As always on such occasions, the search is on for
somebody to blame. Among the leading suspects
are the firm that rate the credit worthiness of bond
issuers. The raters, firms such as Moody’s
Investor’s Service, Standard and Poor’s, Duffs and
Phelps and IBCA, are supposed to be the financial
market’s early warning system. Instead, the
agencies have spent the past few months belatedly
reacting to events.”
» Economist , 1997
 Example: Based on the historical data from S&P, what is the
Approximate historical 1 year probability of default for a B rated
Obligor?
S & P Cumulative Default Rates (Percent), 1981-2002
Rating Year
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
AAA 0.00 0.00 0.03 0.07 0.11 0.20 0.30 0.47 0.54 0.61 0.61 0.61 0.61 0.75 0.92
AA 0.01 0.03 0.08 0.17 0.28 0.42 0.61 0.77 0.90 1.06 1.20 1.37 1.51 1.63 1.77
A 0.05 0.15 0.30 0.48 0.71 0.94 1.19 1.46 1.78 2.10 2.37 2.60 2.84 3.08 3.46
BBB 0.36 0.96 1.61 2.58 3.53 4.49 5.33 6.10 6.77 7.60 8.48 9.34 10.22 11.28 12.44
BB 1.47 4.49 8.18 11.69 14.77 17.99 20.43 22.63 24.85 26.61 28.47 29.76 30.98 31.70 32.56
B 6.72 14.99 22.19 27.83 31.99 35.37 38.56 41.25 42.90 44.59 45.84 46.92 47.71 48.68 49.57
C 30.95 40.35 46.43 51.25 56.77 58.74 59.46 59.85 61.57 62.92 63.41 63.41 63.41 64.25 64.25
Based on historical evidence, a B rated counterparty is approximately
18 times more likely to default over a 1 year time period than a
BBB rated counterparty. Over a 10 year time period, a B rated
Counterparty is how many times more likely to default than a
BBB rated counterparty?

Prob. Of default within 1 year for BBB = 0.36%

Prob. Of default within 1 year for B = 6.72%

Prob. Of default in 10 year horizon for BBB = 7.60%

Prob. Of default in 10 year horizon for B = 44.59%
No. of times = approx 6 times

Cumulative Default Frequency (CDF) for B rises at a lower rate

compared to BBB.
Cumulative and Marginal Default
Rates
• Cumulative default rates = total frequency
of default at any time between starting date
and year T
• Marginal default rate is the frequency of
default during year T.
Cumulative and Marginal Default Rates
Default

d1 di=marginal default rate

Default

d2 Default

1-d1
d3
No default
1-d2
No default
1-d3
No default

Year 0 1 2 3
Cumulative and marginal default
rates
Cumulative default rate is Ci.

C1 = d1
C2 is the cumulative default rate of defaulting within 2 years
C2 = d1 + (1-d1).d2

C3 is the cumulative default rate of defaulting within 3 years:

C3 = d1 + (1-d1).d2 + (1-d1).(1-d2).d3
Cumulative and marginal default
rates
Cumulative default rate is Ci.

C1 = d1
C2 is the cumulative default rate of defaulting within 2 years
C2 = d1 + (1-d1).d2
1- C2 = 1- [d1 + (1-d1).d2]
= 1 – d1 - (1-d1).d2 = (1-d1).(1-d2)
C2 = 1 – (1-d1).(1-d2)
Survival Rate: This is the proportion of issuers initially rated R
That will not have defaulted by time T:

SN(R ) = П(1-di(R ))

Average Default Rate: The cumulative default rate can be expressed

In the form an average per period default rate, d, as follows:
Cumulative Default Rate within N periods =CN , which is given by:

CN = 1 - П(1-di(R )) = 1- (1-d)N
Example: Consider a “B” rated firm that has default rates of
d1 = 5% and d2 = 7%.
In the first year, c1 = d1 = 5%

The probability of defaulting in year 2, d2 = (1-5%).7% = 6.65%

2 years survival rate is = (1-5%).(1-7%) = 88.35%

The cumulative probability of defaulting in year 1 and 2 =

5% + 6.65% = 11.65%

Alternatively, we can get the cumulative probability of default

In year 2 = 1- (1-d1)(1-d2) = 1-88.35% = 11.65%
Example: The probability of an AA rated counterparty defaulting
Over the next year is 0.06%. What is the probability of it
defaulting over the next 3 months ?

Let dq be the prob. of defaulting in any quarter.

(1-dq)4 = (1-0.0006) = 0.9994

(1-dq) = (0.9994)1/4 = 0.99985

dq = 0.00015 = 0.015%
 Transition Probabilities
• Migration is a discrete process that consists
of credit ratings changing from one period
to the next.
• Transition matrix gives the probability of
moving to one rating conditional on the
rating at the beginning of the period. The
common assumption is that these moves
follow a Markov process, i.e., that
migrations across states are independent
from one period to the next.
 Transition Probabilities
Starting Ending State Total
State Probability
A B C D
A 0.97 0.03 0.00 0.00 1.00
B 0.02 0.93 0.02 0.03 1.00
C 0.01 0.12 0.64 0.23 1.00
D 0 0 0 1.00 1.00

Consider a company in year 0 in the B category.

In year 1, with probability D(t1)| B(t0) = P(D1|B0) = 3%
In year 2, the probability of default is =
P(D2|A1)P(A1)+P(D2|B1).P(B1)+P(D2|C1)P(C1)
= 0 x 0.02 + 0.03 x 0.93 + 0.23 x 0.02 = 3.25%

Cum. Prob. Of default over the 2 years = 3% + 3.25% = 6.25%

 Transition Probabilities
Consider a company in year 0 in the B rating category.
Paths to default
Time 1: B → D
Time 3: B → A → A →D
Time 2: B → A → D B → A → B →D
B→B→D B → A → C →D
B→C→D B → B → A →D
B → B → B →D
B → B → C →D
B → C → A →D
B → C → B →D
B → C → C →D
 Transition Probabilities
Consider a company in year 0 in the B rating category.

Time 3: B → A → A →D Prob = 0.02*0.97*0 = 0

B → A → B →D Prob = 0.02*0.03*0.03=0.00
B → A → C →D Prob = 0
B → B → A →D Prob = 0
B → B → B →D Prob = 0.0259
B → B → C →D Prob = 0.93*0.02*0.23=0.0043
B → C → A →D Prob = 0.000
B → C → B →D Prob = 0.0001
B → C → C →D Prob = 0.0029

Default probability in year 3 (marginal probability)= 0.0333

Cumulative default probability of default in 3 years
= 0.03 + 0.0325 + 0.0333 = 0.0958
 Time variation in defaults
• Defaults correlated with economic activity –
defaults more likely in an economic
slowdown than in a boom period
• Upgrades to downgrades ratio also changes
with the business cycle
• More upgrades to downgrades in a boom
period. In a slowdown, the more
downgrades than upgrades. So the rating
transition matrix will vary from year to
year. So it is the long term average rating
transition matrix that is used.
S&P Global Ratings Credit Research & Insights expects the
U.S. trailing-12-month speculative-grade corporate default rate
to reach 4.5% by June 2024, from 3.2% in June 2023.
In our base case, we expect defaults to continue rising as
corporates grapple with higher interest rates and slower growth
ahead.
17-Aug-2023
 Pecking Order of seniority

1. Secured creditors (upto the extent of collateral)

2. Priority creditors:
- Firms that lend money during bankruptcy period
- Providers of goods and services
(e.g., employees, lawyers, vendors
- Taxes (government)
3. General creditors:
- unsecured creditors before bankruptcy
4. Shareholders
Moody’s Recovery Rates for US
corporate debt
Seniority/Security Mean Std. Dev
Senior/Secured bank 69.91 23.47
loan
Equipment trust bonds 59.96 31.08
Senior/Secured bonds 52.31 25.15
Senior/Unsecured 48.84 25.01
bonds
Senior/Subordinated 39.46 24.59
bonds
Subordinated bonds 33.17 20.78
Junior/Subordinated 24.00 13.85
bonds
Preferred Stock 11.06 9.09
All 42.11 26.53
 Bank loans vs. bonds - seniority
• Bank loan obligations are generally rated
higher than a bond obligation and so
recovery rates for bank loans are higher.
• Recovery rate under IBC ~ 30%
 Bond prices, spreads and ratings
• Since bond prices incorporate information
very quickly, the credit spreads, the
premium paid by a risky bond over the risk
free rate will widen before the rating change
is made.
 CreditMetrics

• CreditMetrics asks, if next year is a bad

year, how much will I lose on my loan
portfolio?
 Credit VAR

• If you want that the loss in a year will

exceed a certain value only 5% of the times,
it means that such a loss should occur not
more than once in 20 years.
• If you would like to be more conservative
and want that the probability of loss
exceeding a certain value is only 1%, then
the loss should occur not more than once in
100 years.
 CreditMetrics methodology
• What are the “states” of an issuer’s credit
quality? What are the probabilities of
migrating between these states?
– credit rating
– transition matrics
• What is the value of a credit instrument in
all possible future states?
– mark to market
– revaluation of bonds, loans, swaps, etc.
• How do different issuers migrate together?
– correlations
Step1: Transition Matrix-how likely is an
issuer to change credit ratings or default
over a given time horizon?
Initial Rating at year end (%)
Rating AAA AA A BBB BB B CCC Default

AAA 90.81 8.33 0.68 0.06 0.12 0 0 0

AA 0.70 90.65 7.79 0.64 0.06 0.14 0.02 0

A 0.09 2.27 91.05 5.52 0.74 0.26 0.01 0.06

BBB 0.02 0.33 5.95 86.93 5.30 1.17 0.12 0.18

BB 0.03 0.14 0.67 7.73 80.53 8.84 1.0 1.06

B 0 0.11 0.24 0.43 6.48 83.46 4.07 5.20

CCC 0.22 0 0.22 1.30 2.38 11.24 64.86 19.79

 Step 2: Revaluation-in each
future credit rating, what is the
value of the instrument?
• Bonds: use average credit spreads to revalue in each non-default
state; use recovery rate estimate to revalue in default state
Recovery Rates Spread data by year

Senior Secured 54% 1 2 3 4 5

Senior Unsecured 51% AAA 11.67 15.0 18.33 21.67 25.00

Senior Subordinated 39% AA 16.67 20.00 23.33 26.67 30.00

Subordinated 33% A 25.00 30.00 35.00 40.00 45.00

Junior Subordinated 17% BBB 57.67 62.00 66.33 70.67 75.00

BB 183.33 200.00 216.67 233.33 250.00
B 275.00 300.00 325.00 350.00 375.00
CCC 1208.33 1100.00 991.67 883.33 775.00
 Step 3: Probability and value
under different states
Assuming a flat government T-bill yield curve at 6%.
Original bond: A 5 year BBB bond, senior secured, paying a
coupon of 6.75%
BBB
Face value = $100m.

Defa-
Rating AAA AA A BBB BB B CCC
ult
Probability 0.02% 0.33% 5.95% 86.93% 5.3% 1.17% 0.12% 0.18%
Value 108.61 108.43 107.98 106.92 101.63 97.39 83.12 51
 Step 4: Expected value and
Ye ar
e nd
rating
Standard Deviation
Probability
of s tate %
Ne w loan
value +
coupon, $
Prob.
we ighte d
value , $

AAA 0.02% 108.61 0.022

AA 0.33% 108.43 0.358

A 5.95% 107.98 6.425

BBB 86.93% 106.92 92.946

BB 5.30% 101.63 5.386

B 1.17% 97.39 1.139

CCC 0.12% 83.12 0.10

De fault 0.18% 51 0.092

Mean 106.467

Standard Deviation 2.954

 Value at Risk and Capital

• Mean = $106.467m
• SD = $2.95m
• Assuming normal distribution:
– 5% VAR = 1.65 x SD = 1.65 x 2.95 = $4.868m
– 1% VAR = 2.33 x SD = 2.33 x 2.95 = $6.874m
A 1% VAR of $6.874m means that there is only a
1% chance that the loss will exceed $6.874m, in a
year.
 Value at Risk and Capital

• Assuming actual distribution with

approximations:
– 5% VAR (approximated by 6.77%)
• = 106.467 - 101.63 = $4.837m
– 1% VAR (approximated by 1.47%)
• = 106.467 - 97.39 = $9.077m
 Loan concentration risk

• Concentration limit =
• Maximum loss as a percent capital x 1/[Loss rate]
• If management is not willing to permit losses
exceeding 10% of an FI’s capital to a particular
borrower, and loss rate on a $1 of loan is 50cents,
then:
– Concentration limit = 10% x 1/[0.5] = 20%
– This is the maximum loan to a single borrower as a
percent of lender’s capital.
 Altman’s Discriminant Function
Z = 1.2X1 + 1.4X 2 + 3.3X 3 + 0.6X 4 + 1.0X5
Critical value of Z = 1.81
X1 = Working capital/to tal assets ratio
X2 = Retained earnings/t otal assets ratio
X3 = EBIT/total assets ratio
X 4 = Market value equity/ book value of total liabilitie s
X5 = Sales/tota l assets ratio

Firms with Z score below 1.81 are considered high default risk,
And above 2.99 are considered low default risk.
Altman’s Z score is a default indicator and not a probability of
Default.
Amtex Auto 2015 Jan
Market value of Amtek Auto
(Rs. Crores) 3,833
Book value (Rs. Crores) 78,131
Total assets (Rs. Crores) 2,70,444
Sales (Rs. Crores) 1,54,546
MV/BV 0.05
Working capital (Rs. Crores) 63,069
X1 (Working capital /TA) 0.23
X2 (Retained earnings/TA) 0.29
X3 (EBIT/TA) 0.04

X4 (MV/BV of long term debt) 0.02

X5 (Sales/TA) 0.57
Z score 1.41