Professional Documents
Culture Documents
✗ C) A higher concentration ratio and higher correlation coefficient both increase the joint probability of
default.
✓ D) A higher concentration ratio and lower correlation coefficient both reduce the joint probability of
default.
Explanation
A lower concentration ratio and lower correlation coefficient both reduce the joint probability of default. A decrease in the
concentration ratio results in a decrease in the default correlation.
Which type of distribution has the best fit for bond correlations?
✗ A) Lognormal distribution.
✗ B) Johnson SB distribution.
✗ C) Beta distribution.
✓ D) Generalized extreme value (GEB) distribution.
Explanation
The Generalized extreme value (GEB) distribution is the best fit for bond correlation distributions.
Which of the following most accurately describes the parameters of a generalized Pareto distribution (GPD)?
✗ A) β The scale parameter: 0 > β. The shape (tail) index: ξ, can be any real number.
✗ B) The scale parameter: β, which can be any real number. The shape (tail) index: ξ > 0.
✓ C) The scale parameter: 0 < β. The shape (tail) index: ξ, can be any real number.
✗ D) The scale parameter: β, can be any real number. The shape (tail) index: ξ, can be any real number.
Explanation
The two parameters are a scale parameter, which must be positive, and a shape parameter that can take on any value.
1 of 24
Question #4 of 69 Question ID: 439851
The International Bank has backtested its VAR models and has found four exceptions. Under the Basel Committee Penalty Zone
rules, how would this be classified and what would be the associated VAR multiplier?
Explanation
The Green Zone is for up to 4 exceptions, Yellow being 5 - 9, and Red 10 or more. Green has a multiplier of 3, Yellow ranges
from 3.4 to 3.85, and Red is 4.
Remember the multiplier multiplies the VAR estimate to provide an increased degree of caution in the estimate. The more
exceptions the greater the degree of caution needed with the VAR estimate because the model is less accurate.
A risk manager that wants to incorporate the effect of outliers into her statistical correlation measures would most likely use:
✓ A) Kendall's τ correlation.
✗ B) Pearson correlation.
✗ C) Ordinal correlation.
Explanation
The Pearson correlation coefficient is preferred to ordinal measures when outliers are a concern. Spearman rank and Kendall τ
are ordinal correlation coefficients that should not be used with cardinal financial variables because they underestimate risk by
ignoring the impact of outliers.
A value at risk (VaR) model is most likely to be limited in usefulness most likely because of assumptions concerning the:
Explanation
Financial models require assumptions regarding the underlying distribution of the asset returns. Value at risk (VaR) models often
2 of 24
assume that asset returns follow a normal distribution. However, empirical studies actually find higher kurtosis in return
distributions, which suggest a distribution with fatter tails than the normal distribution.
Model risks that lead to incorrect pricing of securities are most likely caused by:
✓ A) mathematical inconsistencies.
✗ B) stress testing.
✗ C) return distribution inconsistencies.
✗ D) volatility smiles.
Explanation
Financial models at times may fail to accurately measure risk due to mathematical inconsistencies. Risk managers and traders
need to be aware of the possibility of mathematical inconsistencies causing model risk that leads to incorrect pricing and the
inability to properly hedge risk.
Suppose mean reversion exists for a variable with a value of 20 at time period t - 1. Assume that the long-run mean value for this
variable is 40 and ignore the stochastic term included in most regressions of financial data. What is the expected change in
value of the variable for the next period if the mean reversion rate is 0.4?
✗ A) -4.
✓ B) 8.
✗ C) 10.
✗ D) -10.
Explanation
The mean reversion rate, a, indicates the speed of the change or reversion back to the mean. If the mean reversion rate is 0.4
and the difference between the last variable and long-run mean is 20 (= 40 - 20), then the expected change for the next period is
8 (i.e., 0.4 × 20 = 8).
Suppose that 25 days ago the observed market variable percentage change was 2.3% with a daily volatility estimate of 2%.
What is the sample percentage change using the Hull and White (HW) approach if the current daily volatility is estimated at
2.8%?
✗ A) 0.8%.
✗ B) 2.2%.
✓ C) 3.2%.
✗ D) 0.3%.
Explanation
3 of 24
The historical percentage change is adjusted based on the ratio of current daily volatility to historically observed daily volatility 25
days ago. The sample percentage change is 3.2% and is calculated as follows:
✗ A) t-distribution.
✓ B) standard normal distribution.
✗ C) binomial distribution.
✗ D) uniform distribution.
Explanation
A Gaussian copula maps the marginal distribution of each variable to the standard normal distribution.
Within the Basel penalty zones, which of the following multipliers would most likely apply to a yellow zone with five to nine
exceptions?
✗ A) 3.00.
✗ B) 4.00.
✓ C) 3.65.
✗ D) 3.35.
Explanation
The yellow zone applies for five to nine exceptions with multiplier ranges from 3.40 to 3.85. The green zone applies for zero to
four exceptions with a multiplier of 3.00. The red zone applies for ten or more exceptions with a multiplier of 4.00.
The process of comparing losses predicted by a VAR model to those actually experienced over the test period is called:
✗ A) verification.
✓ B) backtesting.
✗ C) authentication.
✗ D) validation.
Explanation
Backtesting is the process of comparing losses predicted by a VAR model to those actually experienced over the testing period.
4 of 24
Question #13 of 69 Question ID: 439850
Explanation
The Kupiec log-likelihood ratio is used to backtest VAR models based on the probability of exceptions, the number of exceptions,
and the number of observations. The model is rejected if the ratio is greater than 3.84.
In the process of integrating various risks into value at risk (VaR) models, what are endogenous liquidity risks?
Explanation
Endogenous liquidity is an adjustment for the price effect of liquidating positions. It depends on trade sizes and is applicable when
market orders are large enough to move prices. Endogenous liquidity is the elasticity of prices to trading volumes and is more
easily observed in instances of high liquidity risk.
The generalized extreme value (GEV) distribution is useful for: I. estimating VAR. II. stress testing. III. estimating correlation. IV.
backtesting.
Explanation
The GEV distribution describes the distribution of the maximums from a large sample of identically distributed observations. It's
not particularly useful for VAR estimation since VAR does not consider the distribution of the maximum, but it is useful for stress
testing. GEV also has nothing to do with correlations and would not be used for backtesting to see if a VAR model was effective.
Which of the following statements regarding disadvantages of non-parametric methods is least accurate?
5 of 24
✓ A) Volatile data periods lead to VAR and ES estimates that are too low.
✗ B) Analysis depends critically on historical data.
✗ C) Difficult to detect structural shifts/regime changes in the data.
✗ D) Cannot accommodate plausible large impact events outside of the sample period.
Explanation
One of the disadvantages of non-parametric methods is that volatile data periods lead to VAR and ES estimates that are too
high.
The risk of a stock or bond which is NOT correlated with the market (and thus can be diversified) is known as:
Explanation
Specific risk is uncorrelated with market movements (e.g., the risk that ABC will enter into unprofitable business and eventually
default). This type of risk can be minimized through diversification of the portfolio.
Which of the following causes for exceptions established by the Basel Committee would be considered more serious so that a
penalty should apply?
✗ A) II and III.
✗ B) I only.
✓ C) I and III.
✗ D) II only.
Explanation
Where the basic integrity of the model is lacking or model accuracy needs improvement, the penalty should apply.
Suppose an individual buys a correlation swap with a fixed correlation rate of 0.4 and a notional value of $1 million for one year.
6 of 24
The realized pairwise correlations of the daily log returns at maturity for three assets are ρ2,1 = 0.8, ρ3,1 = 0.4, and ρ3,2 = 0.3.
What is the correlation swap buyer's payoff at maturity?
✓ A) $100,000.
✗ B) $200,000.
✗ C) $300,000.
✗ D) $400,000.
Explanation
Which of the following statements regarding generalized extreme value (GEV) and peaks-over-threshold (POT) is CORRECT?
✗ A) Both POT and GEV focus on the distribution of extreme values above a specified threshold.
✗ B) POT requires the estimation of one more parameter than GEV.
✗ C) Only one of the approaches has a tail parameter denoted ξ.
✓ D) POT approach may introduce additional uncertainty.
Explanation
The POT approach requires a choice of a threshold, which may introduce additional uncertainty.
GEV theory focuses on the distributions of extremes, whereas POT focuses on the distribution of values that exceed a certain
threshold.
✗ A) equities, only
✗ B) bond, equities and default correlations.
✓ C) The Johnson SB distribution is the best fit for:
✗ D) bond and default correlations, only.
Explanation
The Johnson SB distribution is the best fit for equities and default correlations.
7 of 24
Question #22 of 69 Question ID: 439823
Which of the following statements accurately describe filtered historical simulation? Filtered historical simulation:
Explanation
The filtered historical simulation is the most comprehensive, and hence most complicated, of the non-parametric estimators. The
process combines the historical simulation model with conditional volatility models (like GARCH or asymmetric GARCH). Thus,
the method contains both the attractions of the traditional historical simulation approach with the sophistication of models that
incorporate changing volatility. In simplified terms, the model is flexible enough to capture conditional volatility and volatility
clustering as well as a surprise factor that could have an asymmetric effect on volatility. From a computational standpoint, this
method is very reasonable even for large portfolios, and empirical evidence supports its predictive ability.
Which statement about Pearson correlation coefficient financial models is least accurate?
✗ A) The Pearson correlation coefficient requires that the variance calculations of the variables X
and Y are finite.
✗ B) The joint distribution between variables in Pearson correlation coefficient model must be elliptical in
order to have a meaningful interpretation.
✗ C) The Pearson correlation coefficient measures the linear relationships between two variables.
✓ D) A Pearson correlation of zero implies independence between two variables.
Explanation
A Pearson correlation of zero does not imply independence between the two variables. It does mean that there is not a linear
relationship between the variables.
Explanation
The Kendall τ correlation and Spearman rank correlation are ordinal correlation measures.
8 of 24
Question #25 of 69 Question ID: 496367
Which of the following statements is least likely accurate about copula functions?
Explanation
Copulas are created from distributions that have unique shapes and are then mapped to a known distribution with well-defined
properties, such as the normal distribution.
The risk measure that extracts the dependence structure from the joint distribution function created from several continuous
marginal distribution functions is known by what name?
✗ A) Comonotonic.
✗ B) Multivariate correlation.
✗ C) Correlation.
✓ D) Copula.
Explanation
A copula is a risk measure that extracts the dependence structure from the joint distribution function created from several
continuous marginal distribution functions.
When ξ = 0, the generalized extreme value distribution (GEV) becomes which of the following distributions?
✗ A) Weibull distribution.
✗ B) Gaussian distribution.
✗ C) Frechet distribution.
✓ D) Gumbel distribution.
Explanation
When ξ = 0, the GEV becomes a Gumbel distribution and the tails are "light" as is the case for the normal and log-normal
distributions.
When ξ > 0, the GEV becomes a Frechet distribution and the tails are "heavy" as is the case for the t-distribution and the Pareto
distribution.
When ξ < 0, the GEV becomes a Weibull distribution and the tails are "lighter" than a normal distribution.
9 of 24
Question #28 of 69 Question ID: 439847
Explanation
If a risk model is rejected when it is accurate, a Type I error has occurred. A Type II error occurs when a model is accepted when
it is inaccurate.
A risk manager uses the past 400 months of correlation data from the Wilshire 5000 to estimate the long-run mean correlation of
common stocks and the mean reversion rate. Based on this historical data the long-run mean correlation of the Wilshire 5000
was 29% and the regression output estimates the following regression relationship: Y = 0.247 - 0.77X. Suppose that in June
2014, the average monthly correlation for all Wilshire 5000 stocks was 42%. What is the estimated one-period autocorrelation
for this time period based on the mean reversion rate estimated in the regression analysis?
✗ A) 23%.
✓ B) 32%.
✗ C) 26%.
✗ D) 30%.
Explanation
There is a -13% difference from the long-run mean correlation and June 2014 correlation (29% - 42% = -13%). The inverse of
the X coefficient in the regression relationship implies a mean reversion rate of 77%. Thus, the expected correlation for July
2014 is 32.0%:
Looking at the Basel regulatory framework, how would the Basel approach be best described?
✓ A) Non-integrated.
✗ B) Integrated.
✗ C) Non-compartmentalized.
✗ D) Idiosyncratic.
Explanation
The Basel approach is a "building block" approach, which is a non-integrated approach to risk measurement.
10 of 24
Question #31 of 69 Question ID: 439854
The Basel market risk charges require VAR to be computed over a horizon of:
Explanation
The Basel Accord requires VAR to be calculated over a 2-week period or 10 trading days.
The specific risk of an equities position can be defined as the risk that:
Explanation
The risk that disappears in the portfolio construction process is called the asset's unsystematic risk (also called unique,
diversifiable, or firm-specific risk). In a well-diversified portfolio, the risk that is left cannot be diversified away. The risk that
remains is called the systematic risk (also called beta, nondiversifiable risk, or market risk).
The process of mapping a fixed-income portfolio to a set of risk factors is primarily associated with all of the following EXCEPT:
✗ A) duration mapping.
✗ B) cash flow mapping.
Explanation
Fixed-income portfolio risk factors are usually mapped to principal, duration, and cash-flow risk factors.
Which of the following is NOT a required step in using the delta-normal method to determine VAR for a fixed-income portfolio?
11 of 24
✗ C) Decompose and map the portfolio.
✗ D) Determine the changes in the values of the market factors.
Explanation
Convexity is not a required input in applying the delta-normal method to fixed-income portfolios, although it might prove useful if
interest rates, rather than bond prices were used as market factors.
Which of the following methods of mapping portfolios of fixed income securities is the most precise?
Explanation
Cash flow mapping is the most precise method of the three because we map the present value of the cash flows (face amount
discounted at the sport rate for that maturity) onto the risk factors for zero coupon bonds of the same maturities and include the
intermaturity correlations.
Principal mapping includes only the risk of repayment of the principal amounts and considers the average maturity of the portfolio.
Duration mapping involves the risk of the bond being mapped to a zero-coupon bond of the same duration. Duration mapping uses
the duration of a portfolio (the duration measure is only an approximation) to calculate the VAR.
If a portfolio of assets are all perfectly positively correlated, the portfolio VAR will equal which of the following?
✗ A) Marginal VAR.
✓ B) Undiversified VAR.
✗ C) Diversified VAR.
✗ D) Component VAR.
Explanation
If all the assets are perfectly positively correlated there will be no diversification in that all assets will move together in the same
direction to the same magnitude. This, therefore, reflects the undiversified VAR (i.e., no diversification)—the worst possible
scenario of all assets defaulting together.
12 of 24
✓ A) A full covariance matrix.
✗ B) The beta model.
✗ C) The component VAR.
✗ D) A diagonal matrix.
Explanation
Using a full covariance matrix will result in more precise estimates of VAR.
✓ A) default risk.
✗ B) mark-to-market risk.
✗ C) liquidity risk.
✗ D) market risk.
Explanation
Which of following statements is(are) correct regarding the cyclical feedback loop from value at risk (VaR) constraints on
leveraged investors?
I. When net worth rises, leverage decreases, and when net worth declines, leverage increases.
II. Asset purchases increase when asset prices are rising, and assets are sold when asset prices are declining.
✗ D) II only.
Explanation
Leverage (measured as total assets to equity) is inversely related to the market value of total assets. When net worth rises,
leverage decreases, and when net worth declines, leverage increases. This results in a cyclical feedback loop: asset purchases
increase when asset prices are rising, and assets are sold when asset prices are declining.
Suppose a creditor makes a $1,000,000 loan to company X and a $1,000,000 loan to company Y. Based on historical
information of companies in this industry, companies X and Y each have a 9% default probability and a default correlation
coefficient of 0.5. What is the expected loss for this creditor under the worst case scenario?
13 of 24
✗ A) $28,150.
✗ B) $53,650.
✗ C) $76,430.
✓ D) $49,060.
Explanation
The worst case scenario is the joint probability that both loans default at the same time. The joint probability of default is
computed as:
Thus, the expected loss for the worst case scenario for the creditor is:
The Peaks Over Threshold (POT) approach serves as a basis for an expanded model of risk estimation. Which of the following
statements are false regarding POT?
I. Under the POT method, in the case of "fat" tails, not all moments are defined.
II. POT is often estimated with a Generalized Pareto Distribution.
✗ A) II only.
✗ B) Both I and II.
✗ C) I only.
Explanation
Explanation
Correlation risk applies to market risk, credit risk, systemic risk and concentration risk.
14 of 24
You are mapping risk factors in a large portfolio of 270 assets. How many covariance terms would be needed if the risk factors
were not reduced?
✗ A) 72,900.
✗ B) 36,450.
✗ C) 72,630.
✓ D) 36,315.
Explanation
Covariance
= n × (n − 1) / 2
terms
= 270 × 269 / 2
= 36,315
Extreme value theory can assist with VAR calculations by providing better probability estimates of
extreme losses than those indicated by a standard normal distribution. Using the generalized Pareto
distribution (GPD), the parameter that indicates the fatness of tails is the:
✗ A) scaling parameter, b.
✗ B) slope coefficient, b.
✓ C) shape parameter, ξ.
✗ D) threshold level, µ.
Explanation
✓ A) None of these.
✗ B) S&P Futures Contract.
✗ C) German Government Bond.
✗ D) Japanese Government Bond.
Explanation
Specific risk exposure refers to securities that undergo adverse price movements as a result of idiosyncratic factors related to
individual issuers. Thus, government debt issues and futures contracts based on diversified indexes are not exposed to specific
risk.
15 of 24
Relative to return-based risk measures, position-based risk measures are:
✓ A) more difficult to implement but can detect style drift more quickly.
✗ B) easier to implement and can detect style drift more quickly.
✗ C) easier to implement but can detect style drift more slowly.
✗ D) more difficult to implement and can detect style drift more slowly.
Explanation
Position-based risk measures evaluate the manager's holdings on a current basis and can thus detect style drift more quickly. It
is more difficult to implement though. Return-based risk measures evaluate risk using historical returns whereas position-based
risk measures require an examination of each position a manager holds.
Under the Extreme Value Theorem (EVT), which of the following is (are) TRUE regarding the modeling of market risk?
I. The three key resulting distributions are: Gumbel, Weibull, and Frechet.
II. EVT permits the analysis of maxima and minima distributions.
III. EVT is does not account for "heavy" tails observed in the market place.
IV. EVT is dependent upon the normal distribution.
✗ A) None of these.
✗ B) I only.
✗ C) I and III only.
✓ D) I and II only.
Explanation
Statement III is incorrect because EVT allows for "heavy" tails as we see in the market place. Statement IV is incorrect because
EVT is not dependent upon the normal distribution.
You are backtesting a VAR model using analysing exceptions using failure rates. Which of the following statements is (are)
CORRECT?
I. The probability of rejecting an accurate VAR model is a Type II error.
II. The probability of accepting an inaccurate model is a Type I error.
✗ A) I only.
✗ B) II only.
Explanation
The probability of rejecting an accurate model is a Type I error. The probability of accepting an inaccurate model is a Type II
error.
16 of 24
Question #49 of 69 Question ID: 439844
Extreme value theory (EVT) can assist with value at risk (VAR) calculations by providing better probability estimates of observing
extreme losses than that indicated by a standard normal distribution because empirical distributions exhibit fat tails. If one uses
the generalized Pareto distribution (GPD) method to generate parameter estimates for the shape parameter, fat tails will indicate
a:
✗ A) positive parameter estimate and VAR calculations that are too large.
✓ B) positive parameter estimate and VAR calculations that are too small.
✗ C) negative parameter estimate and VAR calculations that are too small.
✗ D) negative parameter estimate and VAR calculations that are too large.
Explanation
Fat tails will generate a positive shape parameter, which indicates that VAR estimates are probably too small.
Which of the following non-parametric estimators combines the historical simulation model with conditional volatility models?
Explanation
The filtered historical simulation is the most comprehensive and most complicated of the non-parametric estimators. It contains
both the attractions of the traditional historical simulation approach with the sophistication of models that incorporate changing
volatility.
For which of the following options can the delta-normal VAR method be expected to provide an accurate estimate of true VAR?
I. Deep-out-of-the-money options.
II. Deep-in-the-money options.
III. At-the-money options.
✗ A) I only.
✗ B) III only.
✓ C) I and II.
✗ D) II only.
Explanation
The delta-normal VAR method cannot be expected to provide an accurate estimate of true VAR over ranges where deltas are
17 of 24
unstable. That would occur when options are at-the-money. Deep-out-of-the-money and deep-in-the-money options have
relatively stable deltas. Over those ranges, the relationship between the value of the underlying instrument and the value of the
option is very much like a forward currency contract. The delta-normal VAR can be calculated by assessing the volatility of the
underlying spot prices and the correlation between the price of the option and the spot price.
Which of the following is TRUE comparing VAR and extreme value theory (EVT)?
Explanation
The principal shortcoming of VAR is that it does not consider losses beyond a specified threshold.
Which of the following statements about extreme value theory (EVT) is FALSE?
Explanation
EVT models are appropriate for low probability, high impact events; not everyday occurrences.
There is a short position in 1 year bonds with a $150m face value and a 6% annual interest rate with interest paid semi-annually.
The annualized interest rate on zero coupon bonds is 3% for a 6 month maturity and 4.1% for a 12 month maturity. Decompose
the bond into the cash flows of the two standard instruments and then determine the total present value of all the cash flows of
the standard instruments?
✗ A) −$153,848,482.
✗ B) −$155,848,482.
✗ C) −$154,848,482.
✓ D) −$152,848,483
Explanation
We need to separate the individual cash flows of the bond are as follows:
18 of 24
2nd Coupon = −$4.5m after 12 months
Principal = −$150m = −$150m after 12 months
Next, we find the present value of these cash flows using the spot rates given in the question. Be careful to adjust the annual
rate/2 to find the 6 month rate.
Which statement most accurately describes equity correlations and correlation volatilities throughout various economic states?
Historical correlation levels for stocks are:
Explanation
Historical correlation levels for common stocks in the Dow are highest during recessions. Correlation volatility for Dow stocks is
high during recessions but highest during normal economic periods.
Extreme value theory (EVT) helps quantify two key measures of risk. The magnitude of:
Explanation
Extreme value theory (EVT) looks at the value of losses beyond an identified cutoff point.
With of the following items is not one of the advantages of non-parametric simulation methods?
19 of 24
✗ A) Not hindered by parametric violations of skewness.
✓ B) Data that requires adjustments is often readily available.
✗ C) Intuitive and often computationally simple.
✗ D) Can accommodate more complex analysis.
Explanation
An advantage of non-parametric methods is that data is often readily available and does not require adjustments (e.g., financial
statements adjustments).
Which of the following is NOT a required step in determining VAR for a fixed-income portfolio?
Explanation
Regression analysis against a hypothetical but identical portfolio is not used in the process of choosing appropriate market
factors.
In a Gaussian copula, the marginal distribution of each variable is mapped to the new distribution on:
✗ A) a marginal basis.
✓ B) a percentile basis.
✗ C) a random basis.
✗ D) a distributed basis.
Explanation
A Gaussian copula maps the marginal distribution of each variable to the standard normal distribution on a percentile basis.
The process of identifying variables that influence the value of an asset is called:
✗ B) value quantification.
✗ C) influence loading.
✗ D) variable decomposition.
20 of 24
Explanation
The process of identifying variables, or risk factors, that influence the value of an asset is known as risk factor mapping.
Extreme value theory (EVT) can assist with value-at-risk (VAR) calculations by providing better probability estimates of observing
extreme losses than that indicated by a standard normal distribution because:
✓ A) extreme losses appear to occur more frequently than indicated by a normal distribution.
✗ B) EVT is the most efficient method for estimating extreme losses.
✗ C) extreme losses appear to occur less frequently than indicated by a normal distribution.
✗ D) the observed empirical distribution of most asset returns tends to be platykurtic.
Explanation
Extreme losses appear to occur with a higher frequency than indicated by a normal distribution. EVT has been shown to generate
more realistic probability estimates for extreme losses than a normal distribution.
Which of the following statements is incorrect regarding bootstrap historical simulation? The bootstrapping technique:
Explanation
Empirical analysis demonstrates that the bootstrapping technique consistently provides more precise estimates of coherent risk
measures than historical simulation on raw data alone.
Which of the following is least likely to be identified as one of the primary types of stress testing exercises?
Explanation
Scenarios do not typically incorporate the possibility that traders will re-hedge positions during times of market shocks.
21 of 24
Question #64 of 69 Question ID: 496368
How many unknown distributions can be mapped to a known distribution in a correlation copula?
✗ A) One, only.
✗ B) Two, only.
✓ C) Two or more.
✗ D) Three, only.
Explanation
A correlation copula is created by converting two or more unknown distributions that may have unique shapes and mapping them
to a known distribution with well-defined properties, such as the normal distribution.
In using a log-likelihood ratio to backtest a VAR model, the reason to measure the conditional rather than the unconditional
coverage of the model is to consider the:
✓ A) timing of exceptions.
✗ B) the influence of the positions of individual traders.
✗ C) size of the portfolio.
✗ D) number of assets in the portfolio tested.
Explanation
The unconditional coverage test statistic allows for potential time variation of the data. A bunching of exceptions may indicate
that market correlations or trading positions have changed.
✓ A) Difficult to estimate losses significantly larger than the maximum loss within the data set
✗ B) Analysis depends critically on forecasted data.
✗ C) Volatile data periods lead to VAR and ES estimates that are too low.
✗ D) Quiet data periods lead to VAR and ES estimates that are too high.
22 of 24
Explanation
• Volatile data periods lead to VAR and ES estimates that are too high.
• Quiet data periods lead to VAR and ES estimates that are too low.
• Cannot accommodate plausible large impact events if they did not occur within the sample period.
• Difficult to estimate losses significantly larger than the maximum loss within the data set (historical simulation cannot; volatility-
weighting can, to some degree).
• Need sufficient data, which may not be possible for new instruments or markets
Which of the following statements is least accurate regarding non-parametric density estimation?
✗ A) existing data points can be used to "smooth" the data points to allow for VAR calculation at
all confidence levels.
✗ B) One of the advantages of non-parametric density estimation is that the underlying distribution is free
from restrictive assumptions.
✓ C) The major downfall of the non-parametric approach compared to the traditional historical simulation
approach is that VAR can only be calculated for a continuum of points in the data set.
✗ D) Makes an adjustment that connects the midpoints between successive histogram bars in the original
data set's distribution.
Explanation
The major improvement of the non-parametric approach over the traditional historical simulation approach is that VAR can be
calculated for a continuum of points in the data set.
The relationship of correlation risk to credit risk is an important area of concern for risk managers. Which of the following
statements regarding default probabilities and default correlations is correct?
✓ A) The probability of default is higher in the short-term time horizon for non-investment grade
bonds.
✗ B) Creditors benefit by diversifying exposure across industries to increase the default correlations of
debtors.
✗ C) Changes in the concentration ratio are not directly related to changes in default correlations.
✗ D) The default term structure decreases with time to maturity for most investment grade bonds.
Explanation
The probability of default is higher in the immediate time horizon for non-investment grade bonds. The probability of default
23 of 24
decreases over time if the company survives the near-term distressed situation.
In large, diversified equity portfolios, it is often reasonable to ignore what type of risks in determining VAR?
✗ A) Beta.
✓ B) Specific.
✗ C) Market.
✗ D) Correlation.
Explanation
24 of 24