Riskmetrics model:
• FIXED INCOME SECURITIES :
Daily price volatility = (Price sensitivity to a small
change in yield) * (Adverse daily yield move)
OR
Daily price volatility = Duration/(1+R) * (Adverse
daily yield move)
Knowing initial market value, Daily earnings at risk
(DEAR):
DEAR = Daily price volatility * Dollar market value
position = Duration/(1+R)* P *dR
Riskmetrics model:
• FIXED INCOME SECURITIES (EXAMPLE):
– From statistics, we know that probability that deviation in
interest rates will be more than 1.65
is 10%. Thus, for adverse deviation probability is 5%
– Trading portfolio of $1mln
– Standard deviation ( ) is 10 basis points
– Duration of portfolio is 7 years
– Current total yield on this portfolio = 7.243%
– What are potential loss in earnings on this portfolio if the
1 bad day in 20 occurs tomorrow?
PLEASE SEE “STATISTIC RECAP” SLIDES AT THE END OF PPT
Riskmetrics model:
• FIXED INCOME SECURITIES (EXAMPLE cont-d):
– Interest rate change = 1.65*Standard deviation =
1.65*10bp = 16.5 bp or 0.165%
– Daily price volatility= -Duration/(1+R)*(Adverse daily
move) = - 7/(1+0.07243)*(0.00165)=1.077%
– DEAR = Dollar market value of position*Price
volatility = 1,000,000*1.077% = 10,770$
Riskmetrics model:
• FIXED INCOME SECURITIES :
We can measure 1 day VAR, which is DEAR
calculations shown on previous slides
What if we need to measure N day VAR?
VAR = DEAR * N
Riskmetrics model:
• FOREIGN EXCHANGE
VAR = (Dollar value of position)*(Forex volatility)
where,
Dollar value of position = FX position*$/1 foreign currency
Forex volatility = expected adverse change in US dollar value
Riskmetrics model:
• FOREIGN EXCHANGE (EXAMPLE):
– Trading position in Euro (e.g. bond denominated
in Euro) 1.6 million Euro
– Current $/Euro FX rate = 0.625
– Standard deviation of $/Euro FX rate is 56.5bp
– What is the potential daily earnings exposure to
adverse euro to dollar exchange rate changes for
FI from the 1.6 million euro trading portfolio?
*Adverse euro to dollar exchange rate of 1 bad day
in 20 occurs
Riskmetrics model:
• FOREIGN EXCHANGE (EXAMPLE cont-d):
– Current dollar value position = Euro position*FX
rate = 1.6million Euro*0.625 = $1 million
– FX rate change = 1.65*Standard deviation =
1.65*56.5 bp = 93.2 basis points or 0.932%
– DEAR = Dollar value of position*FX rate change =
$1million*0.00932 = $9,320
Riskmetrics model:
• EQUITIES
DEAR = Dollar value of position * Stock market
return volatility
Individual stock market return volatility =
Systematic risk + Unsystematic risk
OR
Individual stock market return volatility:
β 2 σ 2market σ firm
2
Riskmetrics model:
• EQUITIES (EXAMPLE):
– FI holds portfolio of stocks which replicates the
stock market index S&P100 in amount $1million
– FI portfolio is very well diversified
– Standard deviation of market return of S&P100 is
200bp
– What loss will FI incur if adverse stock market
returns materialize tomorrow?
*Adverse stock market returns of 1 bad day in 20
occurs
Riskmetrics model:
• EQUITIES (EXAMPLE cont-d):
– Very well diversified portfolio => Unsystematic
risk ( firm ) =O
– Portfolio replicates S&P100 => portfolio returns
replicate Market index, thus β 1 and
σ portfolio σ S&P100
– Change in market return of portfolio = Standard
deviation *1.65 = 1.65*200bp = 330bp or 3.3%
– DEAR = Dollar amount of portfolio*Change in
market return = $1mln*0.033 = $33,000
Riskmetrics model:
• PORTFOLIO AGGREGATION:
We cannot simply sum up all DEARs, because that ignores
any degree of offsetting or correlation amount fixed
income (FI), FX and Equity (EQ) trading positions.
DEAR Portfolio formula:
1/ 2
DEAR DEAR DEAR
2
fi
2
fx
2
eq
2 * Corrfi_fx * DEAR fi * DEAR fx
DEAR_Portfolio
2 * Corrfi_eq * DEAR fi * DEAR
eq
2 * Corrfx_eq * DEAR fx * DEAR eq
where
DEAR = Daily earnings at risk from FI, EQ or FX
Corr = correlation between FI, EQ or FX
Riskmetrics model:
• PORTFOLIO AGGREGATION (EXAMPLE)
• DEAR (fixed income securities) = $10,770
• DEAR (Forex) = 9,320$
• DEAR (Equities) = 33,000$
• Correlation between returns on FIXED INCOME
SECURITIES and EQUITIES = 0.4
• Correlation between returns on FIXED INCOME
SECURITIES and FOREX = -2
• Correlation between returns on FOREX and
EQUITIES = 0.1
• WHAT is DEAR of Portfolio?
Riskmetrics model:
• PORTFOLIO AGGREGATION (EXAMPLE cont-d)
1/ 2
DEAR DEAR DEAR
2
fi
2
fx
2
eq
2 * Corrfi_fx * DEAR fi * DEAR fx
DEAR_Portfolio
2 * Corrfi_eq * DEAR fi * DEAR
eq
2 * Corrfx_eq * DEAR fx * DEAR eq
DEAR of portfolio = (10,7702+93202+33,0002+
+2*(-2)*10,770*9,320+
+2*(0.4)*10,770*33,000+
+2*(0.1)*9,320*33,000)1/2= $39,969
Recap of statistics:
• 3 sigma Rule:
• Every normal curve (regardless of its mean or standard
deviation) conforms to the following "rule".
– About 68% of the area under the curve falls within 1 standard
deviation of the mean.
– About 95% of the area under the curve falls within 2 standard
deviations of the mean.
– About 99.7% of the area under the curve falls within 3 standard
deviations of the mean.
• To be used in this PPT:
– About 90% of the area under the curve falls within 1.65
standard deviations of the mean.
Effect of the credit risk on return of FI:
• The contractually promised return on a loan
depends on the following factors:
– Interest rate on the loan
– Any fees relating to the loan
– The credit risk premium on the loan
– The collateral backing of the loan
– Other non-price terms (e.g. compensating balance
and reserve requirement)
Effect of the credit risk on return of FI:
• The contractually promised return on a loan (1+k):
of (BR m)
1 k 1
1 [b(1 RR)]
k- gross return on the loan
of – direct fees
BR – base lending rate
m – credit risk premium
b – % of the loan held as non-interest bearing
compensating balance
RR – reserve requirement by Central Bank
Effect of the credit risk on return of FI:
• The contractually promised return on a loan
(1+k) EXAMPLE:
FI issued loan on the following terms:
- Base rate – 12%, credit risk premium – 2%
- Loan origination fee 0.125% of the loan amount
- Compensating balance of 10% of the loan amount
- Central bank reserve requirement 10%
What is the contractually promised return on a
loan?
Effect of the credit risk on return of FI:
• The contractually promised return on a loan
(1+k) EXAMPLE:
of (BR m)
1 k 1
1 [b(1 RR)]
1+k = [1+(0.00125+(O.12+0.02)]/[1-(0.1/(1-0.1)]
Contractually promised return on a loan =
15.52%
Effect of the credit risk on return of FI:
• The expected return on a Loan depends on
the following:
– Contractually promised return on a loan
– Probability of default
Expected return E(r) = p*(1+k) + (1-p)*Zero OR
Expected return E(r) = p*(1+k)
p – probability of the repayment of the loan
1+k – contractually promised return on a loan
Effect of the credit risk on return of FI:
• The expected return on a Loan EXAMPLE:
– FI contractually promised return on a loan is 12%
– Probability of default of the loan is 2.5%
Expected return?
• Expected return of the loan =
(1+0.12)*0.975+(1-0.975)*Zero = 1.092
i.e. Expected return is 9.2%
Measurement of Credit risk:
• Linear probability model calculates probability of
default using linear regression model
n
PD i β j *X ij Error
j1
PD – probability of default of loan i
Beta j – importance of factor j in past repayment
experience
X (ij) – observed factor j in loan i
Measurement of Credit risk:
• Linear probability model EXAMPLE:
– Based on historic data FI identified the following
PD pattern for the Borrowers:
PD =0.5(Leverage ratio)+0.1(Sales/Asset ratio)
– Perspective borrowers Debt/Equity ratio = 0.3 and
Sales/Assets = 2
– Assume Error = zero
– What is expected Probability of Default of the
Perspective borrower?
Measurement of Credit risk:
• Linear probability model EXAMPLE:
PD =0.5(Leverage ratio)+0.1(Sales/Asset ratio)
PD = 0.5*0.3+0.1*2 = 0.15+0.2 = 0.35 or 35%
Expected default of the perspective Borrower is 35%
Measurement of Credit risk:
• Linear discriminant model – divides borrowers to
high/low default classes
• Altman’s discriminant model:
Z 1.2 X 1 1.4 X 2 3.3 X 3 0.6 X 4 X 5
X1 – Working capital/Total assets
X2 – Retained earnings/Total assets
X3 – EBIT/Total assets
X4 – Market Value of equity/Book value of long-term debt
X5 – Sales/Total assets
Altman’s discriminant interpretation:
If Z<1.81 => high default risk
If Z>2.99 => low default risk
If 1.81>Z>2.99 =. Indeterminate default risk
Measurement of Credit risk:
• Term structure derivation of Credit risk
Probability of default on a one period debt instrument:
• Contractually promised return on corporate bond 1+k
• Contractually promised return on risk free bond 1+i
• FI is indifferent which one to give, when expected return
on corporate bond is equal to expected return on risk free
bond:
• Hence, (1+k)*p = 1+i, where p is probability of repayment
of the loan
Probability of default (1-p) = 1-(1+i)/(1+k)
Risk premium = k-I
As perceived by the market probability of default (1-p)
requires the FI to set risk premium of k-i
Measurement of Credit risk:
• Term structure derivation of Credit risk
Probability of default on a one period debt instrument
EXAMPLE:
• Interest rate on Treasury bills 10%
• Interest rate on Corporate bond grad A 15.8%
• Probability of default = 1-(1+0.1)/(1+0.158) = 5%
• Risk premium = 15.8%-10% = 5.8%
Market requires risk premium of 5.8% for default
probability of 5%
Measurement of Credit risk:
• Term structure derivation of Credit risk
Probability of default on multi-period debt instrument:
• Cumulative default probability = 1 –p1*p2*p3..
where p1 is default probability in period 1
p2 is default probability in period 2 and so on
Measurement of Credit risk:
• Term structure derivation of Credit risk
Probability of default on multi-period debt instrument:
For 1st period: (1+k1)*p1 = 1+i
For 2nd period: (1+c1)*p2 = 1+forward rate
Assume no arbitrage profit, i.e. (1+i2)2=(1+f)*(1+i1)
1. Find return for the second period on Treasury bond:
(1+ 2-year interest rate)2=(1+forward rate)*(1+ 1-year interest rate)
OR (1+i2)2=(1+f)*(1+i1) => f = (1+i2)2/(1+i1) -1
[Link] return for the second period on Corporate bond:
(1+c1) = (1+return on 2-year bond)2/(1+return 1-year bond)
3. Find probability of default in the second period
1- p2= 1- (1+f)/(1+c1)
Measurement of Credit risk:
• Term structure derivation of Credit risk
Probability of default on multi-period debt instrument
EXAMPLE:
– Required yields for 1 year Treasury bond 10%
– Required yields for 2 year Treasury bond 11%
– Required yields for 1 year Corporate bond 15.8%
– Required yields for 2 year Corporate bond 18%
– What is Cumulative probability of default?
Measurement of Credit risk:
• Term structure derivation of Credit risk
Probability of default on multi-period debt instrument
EXAMPLE:
1. Find return for the second period on Treasury bond:
f = (1+i2)2/(1+i1) -1 => f=(1+11%) 2/(1+10%) -1 =12%
2. Find return for the second period on Corporate bond:
(1+c1) = (1+18%)2/(1+15.8%) =20%
3. Find probability of default in the second period 1- p2
1- p2= 1- (1+f)/(1+c1) = 1 – 93% = 7%
4. Cumulative default
Cumulative default = 1 –p1*p2 = 1-95%*93%=11.5%
Simple models of loan concentration risk
Concentration limits:
– On loans to individual borrower.
– Concentration limit = Maximum loss Loss rate.
• Maximum loss expressed as percent of capital.
– Some countries, such as Chile, specify limits by
sector or industry
Modern portfolio theory
• To apply Loan portfolio theory you need:
– (i) expected return on loan (measured by all-in-
spread);
– (ii) loan risk;
– (iii) correlation of loan default risks.
Modern portfolio theory
(formulas)
n
- Expected Return: R p X i Ri
i 1
Where:
- Ri – Mean return of i-th loan
- Xi – Proportion of i-th investment in total
portfolio
Modern portfolio theory
(formulas)
Variance: p2 X A2 A2 X B2 B2 2 X i X j i , j
X X 2 X i X j i , j A B
2
A
2
A
2
B
2
B
Where:
- Xi – Proportion of i-th investment in total portfolio
- Xj – Proportion of j-th investment in total portfolio
- δi – Standard deviation of i-th investment
- δj – Standard deviation of j-th investment
- δi,j – Covariance between the returns of i and j investments
- ρi,j – Correlation between the returns of i and j investments
Modern portfolio theory (Example)
• Expected return = 0.4*10%+0.6*12% = 11%
Modern portfolio theory (Example)
• Risk of portfolio:
Variance = 0.42*0.007344+0.62*0.009604+
+2*0.4*0.6*(-0.84)*0.0857*0.098 =
= 0.0012462
Risk (standard deviation) = 0.0353 = 3.53%
KMV portfolio manager model
Annual spread Annual fees
• Return on the loan = Annual all in spread (AIS)
– Expected loss (EL)
Expected Default frequency (PD) x Loss given default (LGD)
• Risk of the loan = PD(1 PD)
x LGD
• Correlation between systematic returns of
investment i and j
KMV portfolio manager model (Example)
• Loan A:
– Return = 5%+2% - 25%*3% = 6.25%
– Risk = [3%*(1-3%)]0.5x 25% = 4.265%
• Loan B:
– Return = 4.5%+1.5% - 20%*2% = 5.6%
– Risk = [2%*(1-2%)]0.5x 20% = 2.8%
KMV portfolio manager model (Example)
• Portfolio (A and B):
– Return = 60% x 6.25%+40% x 5.6% = 5.99%
– Risk:
Variance= 60%2 x(4.265%)2+ 40%2 x(2.8%)2 +
+2x40%x60%x(-0.25)x(4.265%)x(2.8%) = 0.06369%
Risk = (0.06369%)0.5 = 2.52%
