Chapter 63: Estimating Market Risk Measures: An Introduction

and Overview

1. Q.1475: If Profit/Losses (P/L) are distributed normally with a

standard deviation of 18 and a mean of 12, then what is the value of
the corresponding VaR using a 95% confidence interval?

 A. 9.87

 B. 17.61

 C. 13.956

 D. -13.956

2. Q.1479: A generally coherent risk measure tends to involve

increasingly sophisticated weighting functions. Which of the
following is a suitable replacement for the equal weights in the
'average VaR' to estimate any risk measure?

 A. 9.87

 B. 17.61

 C. 13.956

 D. -13.956

3. Q.1480: Coherent risk measures are fundamental in the field of risk

management, providing a consistent framework for assessing the
risk of financial portfolios. The key to estimating coherent risk
measures lies in the:

 A. Ability to assign weights to assets in a portfolio accurately.

 B. Ability to calculate exponential value accurately.

 C. Ability to estimate quantiles.

 D. Ability to approximate risk exposure.

4. Q.1481: The precision of a risk measure estimate is evaluated using

the corresponding standard error(s). On which of the following does
the quantile (VaR) standard error depend?

 A. f(q),f(q), Sample size n,n, and pp

 B. p,p, Standard error s,s, and variance of qq

 C. Sample size n,n, p,p, and the square root of the error.

 D. Variance of q,q, sample size n,n, and f(q)f(q)

5. Q.1482: A portfolio has a beginning period value of $200. The
arithmetic returns follow a normal distribution with a mean of 5%
and a standard deviation of 10%. Calculate VaR at both the 95% and
99% confidence levels, respectively:

 A. $23, $36.6

 B. $43, $56.6

 C. $1.65, $2.33

 D. $23, $43

6. Q.2628: You are assigned to calculate the monthly VaR for the stock
of Apex Inc. You are provided with the following data for the ten
worst returns of the stock during the last 100 months: -12%, -7%, -
32%, -26%, -24%, -20%, -19%, -17%, -15%, -14%. Which of the
following is closest to the monthly VaR for Apex, using a confidence
level of 95%?

 A. -32%

 B. -17%

 C. -12%

 D. -14.5%

7. Q.2632: An analyst has gathered the following information about a

portfolio that has normally distributed geometric returns: Mean 10%,
Standard Deviation 40%, Portfolio 100 million. What is the 95%
lognormal VAR for this portfolio?

 A. $74.7 million

 B. $35.3 million

 C. $42.8 million

 D. $113.4 million

8. Q.2636: Jacob Watson is a risk manager for a large bank. Presently,

he is estimating the VaR for the equities portfolio of the bank. He is
considering estimating the VaR using normal and lognormal
distribution assumptions. He has gathered the following information
about the portfolio: Value of the portfolio USD 1 million, Mean 15%,
Volatility 25%. What would be the 1-year 99% VaR for the portfolio
under the two assumptions?

 A. Normal distribution: $495,000; Lognormal distribution:

$460,000
  B. Normal distribution: $460,000; Lognormal distribution:
$495,000

 C. Normal distribution: $432,500; Lognormal distribution:

$351,000

 D. Normal distribution: $499,000; Lognormal distribution:

$432,500

9. Q.2817: Assume that the P/L over a specified period is normally

distributed and has a mean of 14.1 and a standard deviation of
28.2. What is the 95% VaR and the corresponding 99% VaR?

 A. The 95% VaR is 32.289 and the 99% VaR is 51.4932

 B. The 95% VaR is 36.495 and the 99% VaR is 51.556

 C. The 95% VaR is 55.236 and the 99% VaR is 36.49551

 D. The 95% VaR is 36.225 and the 99% VaR is 41.586

10. Q.2818: Over time, the arithmetic returns rtrt are normally
distributed with a mean of 1.55 and a standard deviation of 1.07.
The portfolio is currently worth 1 unit. Calculate the 95% VaR and
the 99% VaR.

 A. The 95% VaR is 2.3658 and the 99% VaR is 3.6588

 B. The 95% VaR is 1.4542 and the 99% VaR is 0.0652

 C. The 95% VaR is 0.6742 and the 99% VaR is 3.00896

 D. The 95% VaR is 0.21015 and the 99% VaR is 0.93882

11. Q.2819: Let’s assume that the geometric returns RtRt are
normally distributed with a 0.079 mean and 0.312 standard
deviations. Further assumption is that the portfolio is currently worth
1 unit. Calculate the 95% and 99% lognormal VaR.

 A. The 95% VaR is 0.8951 and the 99% VaR is 0.2351

 B. The 95% VaR is 0.88526 and the 99% VaR is 0.56898

 C. The 95% VaR is 0.3522 and the 99% VaR is 0.4762

 D. The 95% VaR is 0.8951 and the 99% VaR is 0.56898

12. Q.3011: Assume you are dealing with a stock “A” that displays
a highly negatively skewed distribution comprised of the past 260-
days returns. Suppose you have P1 = A and P2 = -A, meaning P1 is
long stock A and P2 is short stock A. Which statement is most likely
to be accurate about a 99% VaR?
  A. VaR(P1) > VaR(P2)

 B. VaR(P1) < VaR(P2)

 C. VaR(P1) = VaR(P2)

 D. Cannot be concluded from the given information.

13. Q.3036: What would be the 95% parametric VaR of a portfolio

made of two independently normally distributed stocks
– AA and BB,
with A∼N(0.5,1)A∼N(0.5,1) and B∼N(3,15)B∼N(3,15)? Assume
that P=A+BP=A+B.
 A. 56
 B. 4.87
 C. 3.08
 D. 1.54
14. Q.5289: An investment banker is evaluating the risks of a
portfolio of bonds. The portfolio is valued at CAD 150 million and
contains CAD 20 million in bond X. The annualized standard
deviations of returns of the overall portfolio and bond X are 12% and
9%, respectively. The correlation of returns between the portfolio
and bond X is 0.60. Assuming the investment banker uses a 1-year
99% VaR and the returns are normally distributed, what is the VaR
of bond X?
 A. CAD 1,453,879
 B. CAD 4,186,800
 C. CAD 5,813,777
 D. CAD 4,636,800
15. Q.5292: A data scientist is analyzing a dataset and wants to
determine the distribution of his data. The scientist decides to use a
QQ plot in his analysis. Which of the following statements about QQ
plots is correct?
 A. QQ plots are used to evaluate the precision of a statistical
estimator.
 B. QQ plots are useful in determining the statistical
significance of a hypothesis test.
 C. QQ plots are specifically used when the sample size is
greater than 100.
 D. QQ plots are useful in determining if a dataset follows a
normal distribution.
Chapter 64: Non-parametric Approaches

1. Q.1487: The non-parametric density estimation is based on the

assumption that a basic historical simulation does not get the best
out of the information at hand. Which of the following examples
demonstrates this drawback?

 A. If we have 100 P/L observations, the basic HS only permits

us to estimate VaR at discrete confidence levels, say, 95%.

 B. If we have 100 P/L observations, the VaR at the 95%

confidence level is given by the seventh-largest loss.

 C. If we have 100 P/L observations, the VaR at the 95%

confidence level is given by the fourth-largest loss.

 D. If we have 100 P/L observations, the VaR at the 95%

confidence level is given by the ninth-largest loss.

2. Q.1490: Estimating historical simulation ES or VaR does not have

any theoretical problems; however, it has a practical problem. Which
one is it?

 A. As the holding period decreases, the number of

observations decreases too.

 B. As the holding period increases, the number of observations

decreases.

 C. As the holding period decreases, the size of data decreases.

 D. As the holding period increases, the size of the data

decreases.

3. Q.1492: Bootstrapping is a common technique in financial modeling

and statistics, used to construct yield curves or to estimate the
distribution of a statistic. Understanding its correct application is
crucial for accurate analysis. Which one of the following statements
is most likely correct? A bootstrapping exercise:

 A. Resampling from our existing data set without replacement.

 B. Assumes that the distribution of returns will remain the

same in the past and in the future.
  C. Assumes that the distribution of returns in future will be
markedly different from past distributions.

 D. Results in a VaR estimate that is a sum of sample VaRs

after repeated sampling.

4. Q.1495: Even though bootstrapping has numerous advantages, the

bootstrap estimates are associated with a little bias or error. Which
of the following presents an error of bootstrapping?

 A. Un-sampling variability.

 B. Re-sampling variability.

 C. Dual-sampling variability.

 D. Bootstrapping variability.

5. Q.1496: One of the drawbacks of the historical simulation approach

is that the discreteness of the data rules out estimation of VaRs
between data points. For example, if there were 100 historical
observations, estimation of the VaR is a straightforward process at
the 95% or the 96% confidence levels, but it is impossible to
incorporate a confidence level of, say 95.5%. Which of the following
methods can solve this problem?

 A. Applying Brute Force

 B. Bootstrapping

 C. Non-parametric density estimation

 D. Use of a large number of re-samples

Chapter 65: Parametric Approaches (II): Extreme Value

6. Q.2175: Extreme events have a very low probability of occurrence

but are nonetheless taken very seriously in the financial world.
Which of the following best explains why?

 A. Extreme events tend to recur at rather regular time

intervals.

 B. Extreme events rarely have warning signs and thus markets

cannot prepare for them in any way.

 C. Extreme events are normally very costly and can create a

“ripple” effect on the global market.

 D. No models have been developed to accurately predict and

estimate the effects of extreme events in qualitative terms.
 7. Q.2176: Modeling extreme events in various fields, particularly in
finance and risk management, often presents a significant
challenge. This difficulty arises primarily due to certain factors that
hinder the effective modeling of these events. What is the primary
reason why modeling extreme events is typically difficult and
problematic?

 A. A lack of models that can estimate the effects of certain

extreme but possible events.

 B. A lack of qualified personnel to oversee the modeling

process.

 C. A lack of credible, reliable input data.

 D. The fact that extreme event modeling requires a

considerable investment of time and expertise.

8. Q.2177: In risk management and financial modeling, accurately

capturing the impact of extreme events is crucial for assessing
potential risks and safeguarding investments. Extreme events can
best be modeled through the application of:

 A. The central limit theorem

 B. The standard normal distribution

 C. Extreme-value theorems

 D. The exponential distribution

9. Q.2179: The following is the probability distribution function of the

generalized extreme value distribution:

\exp\left[-\left(1 + \frac{\xi(X - \mu)}{\sigma}\right)^{-\frac{1}{\xi}}\

right], & \xi \neq 0 \\ \exp\left[-\exp\left(-\frac{X - \mu}{\sigma}\right)\
right], & \xi = 0 \end{cases}$$ Where \(X\) satisfies the condition \(1 + \
frac{\xi(X - \mu)}{\sigma} > 0\). If \(\xi > 0\), the GEV becomes the: - A.
Frechet distribution - B. Pareto distribution - C. Gumbel distribution - D.
Weibull distribution

10. Q.2180: If ξ<0ξ<0, the GEV becomes the Weibull distribution,

but this distribution is rarely used to model financial returns mainly
because:

 A. Its cumulative distribution has heavier than normal tails and

very few empirical financial returns are heavy-tailed.

 B. Its cumulative distribution has lighter than normal tails and

very few empirical financial returns are light-tailed.
  C. It’s asymmetric.

 D. It’s symmetric.

11. Q.2182: For the standardized Frechet distribution

with ξ=0.3ξ=0.3, the 5% quantile is equal to:

 A. -0.9

 B. -0.834

 C. -0.4567

 D. -0.9349

12. Q.2957: Assuming that we are given the following parameters

based on the empirical values from contracts on futures clearing
companies; β=0.7β=0.7, ξ=0.12ξ=0.12, u=3u=3, Nu/n=5Nu/n=5.
Compute the VaR and the Expected Shortfall at 99.5%, respectively.

 A. VaR: 1.674; Expected Shortfall: 2.453

 B. VaR: 4.856; Expected Shortfall: 5.905

 C. VaR: 1.453; Expected Shortfall: 2.420

 D. VaR: 1.667; Expected Shortfall: 2.554

13. Q.3993: To retrieve the value at risk (VaR) for the U.S stock
market under the generalized extreme-value (GEV) distribution, a
risk analyst uses the following equation which characterizes a
heavy-tailed Fréchet distribution. The analyst uses the following
somewhat "realistic" parameters: Location, μ=3.0μ=3.0,
Scale, σ=0.80σ=0.80, Tail index, ξ=0.20ξ=0.20. If the sample
size, n=100n=100, then which is nearest to the implied 99.90%
VaR?

 A. 2.3651%

 B. 2.547%

 C. 3.521%

 D. 5.3389%

Chapter 66: Backtesting VaR

14. Q.1499: While conducting backtesting of VaR as an FRM

manager, an accurate model is one where:

 A. The number of exceptions should be greater than the VaR

significance level.
  B. The number of exceptions should be less than the VaR
significance level.

 C. The number of exceptions should be equal to the VaR

significance level.

 D. The number of exceptions should be zero.

15. Q.1502: Matthew Hopkins is invited to interview for a position

as a financial risk manager. After completing an initial set of
questions, the interviewer asks for the interpretation of the following
case: a $20 million 15-day VAR figure having a confidence level of
95%. Which of the following represents the CORRECT interpretation?

 A. There is a 5 percent chance that there will be a gain of

greater than $20 million in a time period of 15 days.

 B. The corresponding VAR of the following day is $20 million,

with a confidence interval of 95%.

 C. The amount of minimum loss spread over the next 15 days

is at least $20 million with a confidence of 95%.

 D. The amount of loss spread over the next 15 days is

expected to be less than $20 million in 95 percent of case
scenarios.

1. Q.2637: Jason Black, a risk analyst at a large multinational bank, is

backtesting the VaR model of the bank. The model being tested is a
daily, 98% VaR model. If the backtest is conducted for one year at a
two-tailed 95% test confidence level, and assuming that a year has
252 trading days, what is the least acceptable number of daily
losses that will lead Black to conclude that the model is calibrated
correctly?

 A. 12

 B. 9

 C. 10

 D. 11

Chapter 67: VaR Mapping

2. Q.3036: What would be the 95% parametric VaR of a portfolio made

of two independently normally distributed stocks – AA and BB,
with A∼N(0.5,1)A∼N(0.5,1) and B∼N(3,15)B∼N(3,15)? Assume
that P=A+BP=A+B.
  A. 56

 B. 4.87

 C. 3.08

 D. 1.54

3. Q.5289: An investment banker is evaluating the risks of a portfolio

of bonds. The portfolio is valued at CAD 150 million and contains
CAD 20 million in bond X. The annualized standard deviations of
returns of the overall portfolio and bond X are 12% and 9%,
respectively. The correlation of returns between the portfolio and
bond X is 0.60. Assuming the investment banker uses a 1-year 99%
VaR and the returns are normally distributed, what is the VaR of
bond X?

 A. CAD 1,453,879

 B. CAD 4,186,800

 C. CAD 5,813,777

 D. CAD 4,636,800

4. Q.5292: A data scientist is analyzing a dataset and wants to

determine the distribution of his data. The scientist decides to use a
QQ plot in his analysis. Which of the following statements about QQ
plots is correct?

 A. QQ plots are used to evaluate the precision of a statistical

estimator.

 B. QQ plots are useful in determining the statistical

significance of a hypothesis test.

 C. QQ plots are specifically used when the sample size is

greater than 100.

 D. QQ plots are useful in determining if a dataset follows a

normal distribution.

Chapter 68: Messages from the Academic Literature on Risk

Measurement for the Trading Book

5. Q.6431: Compared to basic (raw) historical simulation, which of the

following statements correctly characterizes what occurs during
bootstrapping?

 A. Bootstrapping compresses the original data, leading to

repeated observations.
  B. It replicates the most extreme historical returns in every
sample.

 C. It scales each historical return by a constant factor obtained

from the sample’s mean.

 D. It resamples (with replacement) many times, creating

multiple “alternative” data sets.

6. Q.6432: Which statement correctly captures the essence of a

correlation-weighted approach within the historical simulation
framework?

 A. It forces all pairwise asset correlations to zero by

partitioning returns into uncorrelated blocks while leaving
individual volatilities unchanged.

 B. It imposes a desired set of updated correlations among

assets by systematically adjusting historical returns, typically
without altering each asset’s overall volatility.

 C. It re-weights observed co-movements to ensure the sample

correlation matrix matches a diagonal matrix (i.e., only
variances remain, covariances are nullified).

 D. It replicates only those return paths whose correlations

exceed a specified threshold, discarding paths that deviate
from the target correlation profile.

7. Q.6433: Which of the following most closely characterizes the role

of filtered historical simulation in estimating risk?

 A. It fits a time-varying volatility model (e.g., GARCH) and then

adjusts historical returns before resampling, reflecting
evolving market conditions.

 B. It reorders past returns chronologically to match any short-

term volatility changes but leaves correlation unaltered.

 C. It discards all returns that fall outside a user-defined

confidence interval, ensuring that the final distribution is free
of extreme tail events.

 D. It transforms each asset’s average return into a

standardized zero mean but leaves the standard deviation
fixed at its historical level.

8. Q.6434: A small prop trading firm has only 80 days of historical

returns. They want to estimate VaR at 97.5% but find the traditional
historical simulation percentile approach too coarse. The head quant
 suggests a non-parametric density estimation method. What key
benefit does this non-parametric approach offer the firm?

 A. Increased sample size

 B. Flexible percentile selection

 C. Zero outliers

 D. Constant correlation

9. Q.6435: An analyst implements filtered historical simulation on

commodity returns. She first fits a GARCH model, generates
volatility forecasts, and standardizes each day’s return. Then she
resamples these standardized returns to estimate future risk. Which
aspect best defines “filtered” in this context?

 A. Zero-sum weighting

 B. Conditional volatility adjustment

 C. Ignoring non-positive returns

 D. Age-based discard of old data