4.5 Market Risk Measurement and Management

4.5 Market Risk Measurement and Management

Question 1

PE2018Q84 / PE2019Q84/ PE2020Q84 / PE2021Q84 / PE2022Q84
Assume that portfolio daily returns are independently and identically normally distributed. A new quantitative analyst has been asked by the portfolio manager to calculate portfolio $\text{VaRs}$ for 10-, 15-, 20-, and 25-day periods. The portfolio manager notices something amiss with the analyst’s calculations displayed below. Which one of following Vas on this portfolio is inconsistent with the others?

A. ${\text{VaR}}_{10-day}=\text{USD}316\text{million}$
B. ${\text{VaR}}_{15-day}=\text{USD}465\text{million}$
C. ${\text{VaR}}_{20-day}=\text{USD}537\text{million}$
D. ${\text{VaR}}_{25-day}=\text{USD}600\text{million}$

Answer: A
Learning Objective: Explain and apply approaches to estimate long horizon volatility/VaR and describe the
process of mean reversion according to a $\text{GARCH}(1,1)$ model.

Calculate VaR(1-day) from each choice:
${\text{VaR}}_{10-day}=316\to {\text{VaR}}_{1-day}=316/\sqrt{10}\approx 100$
${\text{VaR}}_{15-day}=465\to {\text{VaR}}_{1-day}=465/\sqrt{15}\approx 120$
${\text{VaR}}_{20-day}=537\to {\text{VaR}}_{1-day}=537/\sqrt{20}\approx 120$
${\text{VaR}}_{25-day}=600\to {\text{VaR}}_{1-day}=600/\sqrt{25}\approx 120$

${\text{VaR}}_{1-day}$ from A is different from those from other answers

---

Question 2

What is a limitation of the mean-variance framework for measuring financial risk?

A. The mean-variance approach ignores the first two moments of the underlying distribution.
B. The mean-variance approach ignores the skewness and kurtosis of the underlying distribution.
C.The mean-variance approach restricts the underlying distribution to a non-negative Fishburn measure.
D. The mean-variance approach requires that the underlying distribution have an entropy measure between $0$ and $1$.

---

Question 3

A risk manager is measuring the $\text{VaR}$ of a portfolio of investment securities for a regional bank. The portfolio has a current market value of USD $10$ million with a mean daily return of $0\%$ and variance of daily returns of $0.0005$. Assuming there are $250$ trading days in a year and that the portfolio returns a normal distribution; the estimated annual $90\%$ $\text{VaR}$ is closest to:

A. USD $101,000$
B. USD $287,000$
C. USD $4,531,000$
D. USD $5,815,000$

Answer: C

$\text{VaR}(X\%{)}_{\text{dollar}}=|E(R)-Z_{X\%}\times \sigma |\times \text{Asset Value}$

$\text{Annual VaR}=|0-1.28\times \sqrt{0.0005}|\times \sqrt{250}\times 10,000,000\approx 4,525,483\text{million}$

---

Question 4

There exist two portfolios A and B. Each has their individual VaR. When putting them together in a new portfolio C, which of the following will be always true?

A. ${\text{VaR}}_{\text{C}}<{\text{VaR}}_{\text{A}}+{\text{VaR}}_{\text{B}}$
B. ${\text{VaR}}_{\text{C}}>{\text{VaR}}_{\text{A}}+{\text{VaR}}_{\text{B}}$
C. ${\text{VaR}}_{\text{C}}={\text{VaR}}_{\text{A}}+{\text{VaR}}_{\text{B}}$
D. None of the above

Answer: D
One important drawback of VaR is that it is not sub-additive.

---

Question 5

A risk manager oversees the risk measurement of two portfolios. Portfolio A has a $\text{VaR}$ of USD $5$ million and an $ES$ of USD $10$ million. Portfolio B has a $\text{VaR}$ of USD $7$ million and an $ES$ of USD $15 $million. When combining portfolios A and B, the risk manager observes that the $\text{VaR}$ of the aggregate portfolio is USD $15$ million and the $ES$ is USD $20$ million. This is because:

A. $ES$ is subadditive, while $\text{VaR}$ is not subadditive.
B. $\text{VaR}$ is subadditive, while $ES$ is not subadditive.
C. $\text{VaR}$ satisfies positive homogeneity, while $ES$ does not satisfy positive homogeneity.
D. $ES$ satisfies positive homogeneity, while $\text{VaR}$ does not satisfy positive homogeneity.

---

Question 6

A risk manager is reviewing the benefits of diversification with her firm’s board of directors. She explains that if two securities have a 1-day $95\%$ $\text{VaR}$ of $X$ and $Y$. then the 1-day $95\%$ $\text{VaR}$ of the combined portfolio will always be less than or equal to $X+Y$. Which of the following statements is consistent with the risk managers explanation?

A. The risk manager assumed that the values of the two securities are jointly normally distributed random variables.
B. The risk manager assumed that the values of the two securities are uncorrelated.
C The risk manager assumed that the values of the two securities are independent and identically distributed random variables.
D. The risk manager made no assumptions because $\text{VaR}$ is a coherent risk measure. which supports her statement

---

Question 7

PE2018Q36 / PE2019Q36 / PE2020Q36 / PE2021Q36 / PE2022Q36
Bank A and Bank B are two competing investment banks that are calculating the 1-day $99\%$ $\text{VaR}$ for an at-the-money call on a non-dividend-paying stock with the following information:

• Current stock price: USD $120$
• Estimated annual stock return volatility: $18\%$
• Current Black-Scholes-Merton option value: USD $5.20$
• Option delta: $0.6$

To compute $\text{VaR}$, Bank A uses the linear approximation method, while Bank B uses a Monte Carlo simulation method for full revaluation. Which bank will estimate a higher value for the 1-day $99\%$ $\text{VaR}$?

A. Bank A.
B. Bank B.
C. Both will have the same $\text{VaR}$ estimate.
D. Insufficient information to determine.

Answer: A
Learning Objective: Compare delta-normal and full revaluation approaches for computing $\text{VaR}$.

The option’s return function is convex with respect to the value of the underlying. Therefore, the linear approximation method will always underestimate the true value of the option for any potential change in price. Therefore the $\text{VaR}$ will always be higher under the linear approximation method than a full revaluation conducted by Monte Carlo simulation analysis. The difference is the bias resulting from the linear approximation, and this bias increases in size with the change in the option price and with the holding period.

---

Question 8

The hybrid method is a combination of historical simulation and

A. Historical Standard Deviation
B. MDE
C. EWMA
D. GARCH

Answer: C
The hybrid approach combines the two simplest approaches, historical simulation and risk metrics, by estimating the percentiles of the return directly (similar to historical simulation), and using exponentially declining weights on past data (similar to risk metrics).

---

Question 9

The historical simulation approach is more likely to provide an accurate estimate of the $\text{VaR}$ than the Risk Metrics approach for a portfolio that consists of:

A. A small number of emerging market securities.
B. A small number of broad market indexes.
C. A large number of emerging market securities.
D. A large number of board market indexes.

Answer: A
The Risk Metrics approach is a delta-normal model that requires the returns to be approximately normally distributed, while the historical simulation model requires much less stringent assumptions. The returns on a portfolio with small number of securities is less likely to be normally distributed than a larger portfolio and an emerging markets index is less likely to be normally distributed than a broad market index. Therefore the historical simulation approach will most likely provide a better $\text{VaR}$ estimate than Risk Metrics for a portfolio with a small number of emerging market securities.

---

Question 10

Which of the following is a disadvantage of the historical simulation method over the Risk Metrics model? The historical method requires:

I. A worst-case scenario as an input.
I. The future is determined by the past.
Ill. Standard deviations and correlations.
IV. The assumption of normal distributions for asset returns.

A. l and Ill only
B. ll only
C. Il and IV only
D. Ill only

---

Question 11

Which of the following is not a disadvantage of nonparametric methods compared to parametric methods?

A. Data is used more efficiently with parametric methods than nonparametric methods.
B. Identifying market regimes and conditional volatility increases the amount of usable data as well as the estimation error for historical simulations.
C. MDE may lead to data snooping or over-fitting in identifying required assumptions regarding an appropriate kernel function.
D. MDE requires a large amount of data that is directly related to the number of conditioning variables used in the model.

Answer: B
The use of market regimes and identifying conditional means and volatility actually reduces not increases the amount of data from the full sample. The full sample of data is split into subgroups used to estimate conditional volatility. Therefore, the amount of data available for estimating future volatility is significantly reduced.

---

Question 12

The hybrid approach for estimating $\text{VaR}$ is the combination of a parametric and a nonparametric approach. It specifically combines the historical simulation approach with:

A. The delta normal approach.
B. The exponentially weighted moving average approach.
C. The multivariate density estimation approach.
D. The generalized autoregressive conditional heteroskedasticity approach.

Answer: B
Learning Objective: Compare and contrast different parametric and non-parametric approaches for estimating conditional volatility.

The hybrid approach combines two approaches to estimating $\text{VaR}$, the historical simulation and the exponential smoothing approach (i.e. an EWMA approach). Similar to a historical simulation approach, the hybrid approach estimates the percentiles of the return directly, but it also uses exponentially declining weights on past data similar to the exponentially weighted moving average approach.

---

Question 13

A risk manager is considering switching from using historical volatility to using implied volatility in $\text{VaR}$ calculations. Which of the following statements about implied volatility is correct?

A. Implied volatility estimates are model dependent and a misspecified model can result in erroneous forecasts.
B. Implied volatility estimates require that historical returns are indicative of future returns.
C. Implied volatility estimates tend to underestimate future volatility as a result of mean reversion
D. Implied volatility estimates are generally accurate even if there is only one trade in the option used to calculate an estimate.

---

Question 14

PE2018Q15 / PE2010Q15 / PE2020Q15 / PE2021Q15/ PE2022Q15
A portfolio manager bought $600$ call options on a non-dividend-paying stock, with a strike price of USD $60$, for USD $3$ each. The current stock price is USD $62$ with a daily stock return volatility of $1.82\%$, and the delta of the option is $0.5$. Using the delta-normal approach to calculate $\text{VaR}$, what is an approximation of the 1-day $95\%$ $\text{VaR}$ of this position?

A. USD $54$
B. USD $557$
C. USD $787$
D. USD $1,114$

Answer: B
Learning Objective: Describe the delta-normal approach for calculating $\text{VaR}$ for non-linear derivatives.

The delta of the option is $0.5$.

The 1-day $95\%$ $\text{VaR}$ of 1 share of the underlying equals to $1.82\%\times 1.645\times 62=1.8562$

Therefore, the $\text{VaR}$ of one option equals to $0.5\times 1.8562=0.9281$, and multiplying by $600$ units provides the 1-day $95\%$ $\text{VaR}$ of the entire position $556.86$.

A is incorrect. USD $53.8902$ is the result obtained by ignoring delta and using the call option price, not stock price, to determine $\text{VaR}$ of position: $\text{VaR}=0.0182\times 1.645\times 600\times 3=53.8902$.

C is incorrect. USD $787.40$ is the result obtained when the $\text{VaR}$ of the position is incorrectly calculated at the $99\%$ confidence level ($\text{VaR}=0.0182\times 2.326\times 62\times 0.5\times 600=787.3975$).

D is incorrect. USD $1,113.72$ is the result obtained when delta is not applied to the formula ($\text{VaR}=1.8562\times 600=1,113.72$).

---

Question 15

PE2018Q52
An at-the-money European call option on the DJ EURO STOXX 50 Index with a strike of $2,800$ and maturing in 1 year is trading at EUR $350$, where contract value is determined by EUR $10$ per index point. The risk-free rate is $3\%$ per year, and the daily volatility of the index is $2.05\%$. If we assume that the expected return on the DJ EURO STOXX 50 Index is $0\%$, the $99\%$ 1-day $\text{VaR}$ of a short position on a single call option calculated using the delta-normal approach is closest to:

A. EUR $8$
B. EUR $53$
C. EUR $84$
D. EUR $669$

Answer: D
Learning Objective: Compare delta-normal and full revaluation approaches for computing $\text{VaR}$,

The $99\%$ VaR of the index is ${\text{VaR}}_{\text{index}}=|0-2.33\times 0.0205|\times 2800\times 10\approx 1,337.42$

Since the option is at-the-money, the delta is close to $0.5$.

The $99\%$ VaR of the call option is ${\text{VaR}}_{\text{option}}=0.5\times 1,337.42=668.71$

---

Question 16

PE2018Q91 / PE2019Q91 / PE2020Q91/ PE2021Q91 / PE2022PSQ23 / PE2022Q91
A portfolio of investment securities for a regional bank has a current market value equal to USD $7,444,000$ with a daily variance of $0.0002$. Assuming there are $250$ trading days in a year and that the portfolio returns follow a normal distribution, the estimate of the annual $\text{VaR}$ at the $95\%$ confidence level is closest to which of the following?

A. USD $32,595$
B. USD $145,770$
C. USD $2,297,507$
D. USD $2,737,737$

Answer: D
Learning Objective: Explain and apply approaches to estimate long horizon volatility/$\text{VaR}$, and describe the process of mean reversion according to a $\text{GARCH}(1,1$) model.

Daily standard deviation: $\sigma =\sqrt{0.0002}=0.01414$

$\text{VaR}(X\%{)}_{\text{dollar}}=|E(R)-Z_{X\%}\times \sigma |\times \text{Asset Value}$

$\text{Annual VaR}=|0-1.645\times 0.01414|\times \sqrt{250}\times 7,444,000=2,737,737$

---

Question 17

PE2018Q83 / PE2019Q83 / PE2020Q83 / PE2021Q83 / PE2022Q83
An analyst have been asked to estimate the $\text{VaR}$ of an investment in Big Pharma Inc. The company’s stock is trading at USD $26$ and the stock has a daily volatility of $1.5\%$. Using the delta-normal method, the $\text{VaR}$ at the $95\%$ confidence level of a long position in an at-the-money put on this stock with a delta of $-0.5$ over a 1-day holding period is closest to which of the following choices?

A. USD $0.28$
B. USD $0.32$
C. USD $0.57$
D. USD $2.84$

Answer: B
Learning Objective: Describe the delta-normal approach to calculating VaR for non-linear derivatives.

${\text{VaR}}_{\text{stock}}=|0-1.645\times 0.015|\times 26\approx 0.642$

${\text{VaR}}_{\text{option}}=|\Delta |\times {\text{VaR}}_{\text{stock}}=0.5\times 0.642\approx 0.32$

---

Question 18

PE2020Q41 / PE2021Q41 / PE2022Q41
A junior risk analyst is modeling the volatility of a certain market variable and is trying to decide between $\text{EWMA}$ and $\text{GARCH}(1,1)$ models. Which of the following statements about the two models is correct?

A. The $\text{EWMA}$ model is a special case of the $\text{GARCH}(1,1)$ model with the additional assumption that the long-run volatility is zero.
B. A variance estimated from the $\text{GARCH}(1,1)$ model is a weighted average of the prior day’s estimated variance and the prior day’s squared return.
C.The $\text{GARCH}(1,1)$ model assigns a higher weight to the prior day’s estimated variance than the $\text{EWMA}$ model.
D. A variance estimated from the $\text{EWMA}$ model is a weighted average of the prior day’s estimated variance and the prior day’s squared return.

Answer: D
Learning Objective: Apply the exponentially weighted moving average ($\text{EWMA}$) approach and the $\text{GARCH}(1,1)$ model to estimate volatility.

The $\text{EWMA}$ estimate of variance is a weighted average of the prior day’s variance and prior day’s squared return.

A is incorrect. $\text{EWMA}$ is a particular case of $\text{GARCH}(1,1)$ with the weight assigned to the long-run average variance rate as zero and the sum of the weights of the other two parameters equal to 1.

B is incorrect because there is also weight assigned to the long-run average variance rate.

C is incorrect because such a comparison can only be done under specific parameter configurations.

---

Question 19

PE2020Q50 / PE2021Q50 / PE2022Q50
A risk analyst is estimating the variance of stock returns on day $n$, given by ${\sigma }_n^2$, using the equation:

${\sigma }_n^2=\gamma V_L+\alpha {\mu }_{n-1}^2+\beta {\sigma }_{n-1}^2$,

where ${\mu }_{n-1}$ and ${\sigma }_{n-1}$ represent the return and volatility on day $n-1$, respectively.

If the values of $\alpha$ and $\beta$ are as indicated below, which combination of values indicates that the variance follows a stable $\text{GARCH}(1,1)$ process?

A. $\alpha =0.073637$ and $\beta =0.927363$
B. $\alpha =0.075637$ and $\beta =0.923363$
C. $\alpha =0.084637$ and $\beta =0.916363$
D. $\alpha =0.086637$ and $\beta =0.914363$

Answer：B
Learning Objective: Apply the exponentially weighted moving average ($\text{EWMA}$) approach and the $\text{GARCH}(1,1)$ model to estimate volatility.

For a $\text{GARCH}(1,1)$ process to be stable, the sum of parameters $\alpha$ and $\beta$ need to be below $1$.

---

Question 20

PE2023Q2
A risk analyst at a hedge fund is conducting a historical simulation to estimate the $\text{ES}$ of a portfolio. The value of the portfolio at market close of any given day depends on the price of a stock and the level of an interest rate at the close of that day. The analyst uses closing values of these variables on the most recent $501$ trading days as the historical dataset for the simulation and collects the following data, with Day $0$ representing the first data point and Day $500$ representing the last data point of the historical period:

Day	Stock price (HKD)	Interest rate (%)
$0$	$76.00$	$2.50\%$
$1$	$72.00$	$2.60\%$
$\dots$	$\dots$	$\dots$
$500$	$94.00$	$3.80\%$

What stock price and interest rate would be most appropriate for the analyst to use in the scenario of the historical simulation for Day $501$?

A. The stock price would be HKD $89.05$, and the interest rate would be $3.90\%$.
B. The stock price would be HKD $89.05$, and the interest rate would be $3.95\%$
C. The stock price would be HKD $92.00$, and the interest rate would be $3.90\%$
D. The stock price would be HKD $92.00$, and the interest rate would be $3.95\%$

Answer: A
Learning Objective: Describe and explain the historical simulation approach for computing $\text{VaR}$ and $\text{ES}$.

A is correct. In a historical simulation with a $500$-day historical reference period, the $500$ historical daily changes (from Day $0$ through Day $500$) are used to create $500$ scenarios for what might happen between today and tomorrow (on Day $501$).

In practice, the risk factors that may be used in a historical simulation are divided into two categories: those where the percentage change in the past is used to define a percentage change in the future, and those where the actual change in the past is used to define an actual change in the future. Stock prices are usually considered to be in the first category, while interest rates are usually considered to be in the second category.

The historical change in the stock price from Day $0$ to Day $1$ should therefore be measured as a $\frac{72}{76}-1=-5.263\%$ change, while the change in the interest rate should be measured as a $2.60\%-2.50\%=0.10\%$ change.

Applying these changes to the current stock price and interest rate of HKD $94$ and $3.8\%$, respectively,produces a scenario for the historical simulation with a stock price of $94\times (1-0.05263)=89.05263$, and an interest rate of $3.80\%+0.10\%=3.90\%$.

---

Question 21

PE2022Q88
The CRO of a major bank is reviewing a new risk measure, $W$, with the risk team. The CRO runs a test on the new risk measure to determine if the measure is coherent and satisfies the property of translation invariance. Which of the following tests would correctly determine that the risk measure $W$ exhibits translation invariance?

A. When cash is added to a portfolio, the value of $W$ for that portfolio should decrease by the amount of cash that is added.
B. When $W$ is used to measure the risk of two portfolios A and B, then $W(A)+W(B)$ should be less than or equal to $W(A+B)$.
C. When $W$ is used to measure the risk of two portfolios A and B, and if portfolio A always produces a worse outcome than portfolio B, then $W(A)$ should always be higher than $W(B)$.
D. When $W$ is used to measure the risk of portfolio A, and if all exposures in portfolio A are increased by a constant factor, then $W(A)$ should increase proportionally by that factor.

Answer: A
Learning Objective: Define the properties of a coherent risk measure and explain the meaning of each property.

A is correct. According to the property of translation invariance, adding an amount of cash, $K$, into a portfolio will decrease the risk measure by $K$. Therefore, choice A correctly describes translation invariance.

B is incorrect. This is a test for subadditivity. According to the property of subadditivity, given two portfolios A and B, the risk measure for the portfolio formed by merging A and B will be less than or equal to the sum of the risk measures for A and B.

C is incorrect. This is a test for monotonicity. According to the property of monotonicity, a portfolio that produces consistently worse results in comparison to another portfolio will have a higher risk measure.

D is incorrect. This is a test of homogeneity. According to the property of homogeneity,
changing the size of a portfolio by multiplying the amount of all components by $\lambda$ results in the risk measure being multiplied by $\lambda$.

---