Ensae 3A/MS/Master SFA 2023-2024 (Semester 1).
Exam “Foundations of Risk Management”
(session principale)
1.5 hour. No document. A calculator is allowed.
QCM. Answer the following questions. A correct answer yields 3 points ; no ans-
wer provides 0 point ; an uncorrect answer yields a negative -1 point. For each question,
onlyone answer is correct. Maximum = 60 points. Final note = score / 3.
Q1. Which type of risks is not intended to be covered by regulatory capital ?
1. inversions of some not-risky yield curves
2. credit spread widenings
3. loss of part of the customer base
4. defaults of some counterparties
Q2. Which type of risk is not part of the Pillar 1 solvency ratio ?
1. Credit Risk
2. Market Risk
3. Operational Risk
4. Interest Rate Risk
Q3. Which financial agregate is not part of the Net Banking Income ?
1. Net Interest Margin
2. Net commission
3. Net trading activities (gains / losses)
4. Salaries & Operating Expenses
Q4. The numerator and denominator of the NSFR are respectively :
1. Available Stable Funding (ASF) and Required Stable Funding (RSF)
2. Required Stable Funding (RSF) and Available Stable Funding (ASF)
3. HQLA and Net Outflows
4. Net Outflows and HQLA
1
Q5. When a bank grants a loan and if C denotes the interest rate paid by the client,
C is the sum of which of the following components :
1. Risk Free rate + Credit Risk margin + Commercial Margin
2. Risk Free rate + Credit Risk margin
3. Risk Free rate + Commercial Margin
4. Credit Risk Margin + Commercial Margin
Q6. Which type of loss should be covered by regulatory capital ?
1. Unexpected Loss occuring in 0.01% (99.9% percentile) over 1 year
2. Unexpected Loss occuring in 0.1% (99 percentile) over 1 year
3. Expected Loss over 1 year
4. Liquidity Risk
Q7. The Basic Approach, TSA or AMA approaches are used to measure :
1. Credit Risk
2. Market Risk
3. Operational Risk
4. Counterparty Risk
Q8. Concerning the main Value-at-risk methods, which assertion is wrong ?
1. Monte-Carlo VaR methods suffer from some "Ghost effects".
2. Parametric VaR calculations generally require the availability of some Greeks
associated to every portfolio position.
3. Historical VaR calculations are relatively easy and quick.
4. Historical VaR requires full repricing of all portfolio positions.
Q9. Concerning Portfolio Credit Risk models, which assertion is wrong ?
1. CreditMetrics is not able to manage time-varying exposures.
2. CreditMetrics manages credit spread as dynamic risk factors.
3. Moody’s-KMV requires that all firms in the portfolio are listed.
4. The correlation models of CreditMetrics and Moody’s-KMV share the same
logic and structure.
2
Q10. Concerning the regulatory model of Portfolio credit risk (the "Basel 2 mo-
del"), which assertion is wrong ?
1. every individual exposure in the portfolio has to be negligible compared to the
total portfolio notional.
2. asset correlations are fixed and the same for every couple of counterparties.
3. default probabilities given the systemic factor are calculated in closed-form for-
mulas.
4. the law of the portfolio loss is approximately the law of the loss given the sys-
temic factor.
NB : please enter your answers in the table at the end of this document.
Exercise (questions may be answered in English or French).
An asset manager has got a simple portfolio (called "Portfolio 1") whose compo-
nents are
(i) NA stocks of firm A. The stock price and the stock return at time t are denoted
St and rA,t respectively ;
(ii) NB zero-coupon bonds that have been issued by another firm B. They have the
same maturity T and their price at time t is Pt = 1/(1 + Ht )T −t . The rate Ht
is called the Yield-to-maturity of these bonds.
The two underlying risk factors will be given by the random processes (rA,t ) and (Ht ).
This asset manager would like to evaluate the risk of the latter portfolio.
E1. What is the portfolio value at time t ? What is the relationship between Ht and
the credit spread curve of B ? Deduce the (risk-neutral) default probability of firm B
between t and T from its yield-to-maturity and some discount factors.
E2. Write the Profit and Loss (called P &L(t, dt)) of Portfolio 1 between the two
dates t and t+dt,
for some
a small dt.
Assume of (drA,t, dHt ) is centered
the joint law
Gaussian, E (drA,t )2 = σ12 dt, E (dHt )2 = σ22 dt, E drA,t .dHt = ρσ1 σ2 dt. By
neglecting its deterministic part, what is the law of P &L(t, dt) ? Calculate the Value-
at-Risk at level α ∈ (0, 1) associated to Portfolio 1, between t and t + dt, dt << 1.
E3. Consider a random loss L whose law is centered and Gaussian N (0, σ 2 ). Prove
that its expected shortfall at level α ∈ (0, 1) is ESα (L) = cα σ, for some constant cα
that does not depend on σ. Calculate cα . Deduce the expected shortfall associated to
Portfolio 1 (between t and t + dt, at level α).
E4. Calculate the marginal Value-at-risk and the marginal expected shortfall at level
α, for firm A exposure.
E5. The asset manager considers that its exposure with respect to firm B is too
large. Propose three ways of reducing the risk associated to the bonds in Portfolio 1.
E6. At t = 0, bond B was rated "AA" by Standard and Poor’s. The latter agency
suddenly decides to change its rating to "A". What is the impact of this rating change
3
on Portfolio 1’s value ? What do you expect concerning the credit spread of B and its
Yield-to-maturity ?
E7. To reduce the credit risk exposure related to firm B, the portfolio manager
decides to enter into a Credit Default Swap with a counterparty C. The underlyings of
the CDS will be a certain notional amount of bonds issued by B. Which side the CDS
deal will be chosen by the portfolio manager : protection buyer or protection seller ?
Why choosing firm A as the counterparty C is not a good idea ? Is the portfolio exposed
to new sources of risk ?
E8. Portfolio 1 is enriched with several new financial instruments : other stocks,
vanilla stock options, FX forward contracts, barrier options, interest rate swaps with
triggers. We get a significantly larger portfolio, called Portfolio 2. Our manager would
like to calculate a risk measure for Portfolio 2 that would integrate all sources of mar-
ket risk and credit risk (an “integrated” risk measure). Which methodology can we
recommend ?
4
Name :
First name :
Locker number :
Results
Answer 1 Answer 2 Answer 3 Answer 4 Score
Q1
Q2
Q3
Q4
Q5
Q6
Q7
Q8
Q9
Q10
