You’re on a roll
Create a free account to save your FRM progress, track your accuracy by topic, and pick up right where you left off.
Free Questions › FRM › Valuation and Risk Models
Free FRM Valuation and Risk Models Practice Questions & Answers
400 exam-style Valuation and Risk Models questions. Pick your answer, hit Check answer, and see the worked solution — free to start, no signup.
100% free · No login to startQuestion 31
Variance in a return distribution measures:
Select an option first.
Correct answer: B — The average squared deviation of returns from their mean.
Explanation: Variance = E[(R − μ)²], the average squared deviation from the mean. It is the square of standard deviation. A is the mode; C is the worst-case loss; D is a separate statistical concept. Variance is the core risk measure in the mean-variance framework.
Question 32
A portfolio of two assets with a correlation of −1 between their returns has a portfolio standard deviation of:
Select an option first.
Correct answer: B — Zero (if weights are chosen correctly).
Explanation: With correlation = −1, assets move perfectly in opposite directions. By choosing weights such that the weighted standard deviations exactly offset each other [w₁σ₁ = w₂σ₂], the portfolio variance = 0. This represents perfect diversification — a risk-free portfolio from two risky assets.
Question 33
The covariance matrix of asset returns is used in parametric portfolio VaR to:
Select an option first.
Correct answer: B — Capture the variance of each asset and the pairwise correlations between all assets simultaneously.
Explanation: The covariance matrix Σ contains: diagonal elements = individual asset variances, off-diagonal elements = pairwise covariances. Portfolio variance = w'Σw. This captures both individual risks and diversification effects in one compact matrix calculation. A is historical sim; C is optimisation; D is credit risk.
Question 34
Which measure of central tendency is most relevant for characterising the 'expected' portfolio return in mean-variance analysis?
Select an option first.
Correct answer: C — Arithmetic mean
Explanation: In the mean-variance framework, 'mean' refers to the arithmetic expected return. Geometric mean is relevant for compounding over time but is not the basis of standard mean-variance optimisation. Median and mode are used for non-normal distribution analysis.
Question 35
Which of the following correctly describes a fat-tailed (leptokurtic) distribution relative to the normal distribution?
Select an option first.
Correct answer: A — More probability mass near the mean and in the tails, less in the shoulders.
Explanation: Leptokurtosis (excess kurtosis > 0): more peaked around the mean and heavier tails than the normal — the 'shoulders' of the distribution have less mass to compensate. This means extreme events (tail losses) occur more frequently than normal VaR models predict. B, C, D are incorrect.
Get the full FRM question bank — free
Drop your email and we'll send you fresh FRM practice questions, fully worked solutions and exam-deadline reminders. No spam — unsubscribe in one click.
Want to save your score and take a full mock exam? Create a free account →
Question 36
A GARCH(1,1) model has α = 0.08 and β = 0.90. The long-run average daily variance is 0.000144 (daily vol ≈ 1.2%). What is the ω parameter?
Select an option first.
Correct answer: B — 0.00000288
Explanation: ω = Long-run variance × (1 − α − β) = 0.000144 × (1 − 0.08 − 0.90) = 0.000144 × 0.02 = 0.00000288. A uses (1−0.90)=0.10 then divides by 10; C is an arithmetic error; D doubles the correct answer.
Question 37
Under the EWMA model with λ = 0.94, if yesterday's estimated variance was 0.0001 and yesterday's return was 2%, today's estimated variance is approximately:
Select an option first.
Correct answer: B — 0.000118
Explanation: EWMA: σ²_t = λ × σ²_(t-1) + (1−λ) × r²_(t-1) = 0.94 × 0.0001 + 0.06 × (0.02)² = 0.000094 + 0.06 × 0.0004 = 0.000094 + 0.000024 = 0.000118. A is only the persistence term; C adds a rounding error; D is unchanged (ignores today's return).
Question 38
Which of the following volatility models explicitly incorporates a mean-reversion component toward a long-run variance level?
Select an option first.
Correct answer: C — GARCH(1,1) model.
Explanation: GARCH(1,1): σ²_t = ω + α·r²_(t-1) + β·σ²_(t-1). The ω term combined with α + β < 1 pulls variance toward the long-run mean ω/(1−α−β). EWMA has no mean-reversion (equivalent to IGARCH with α + β = 1). Simple rolling and implied vol do not explicitly model mean reversion.
Question 39
Which of the following is the primary reason regulatory frameworks (Basel) moved toward Expected Shortfall as the standard risk measure over VaR?
Select an option first.
Correct answer: C — ES captures the severity of tail losses beyond the VaR threshold, is sub-additive, and provides a more complete picture of extreme risk.
Explanation: The FRTB (Fundamental Review of the Trading Book) replaced VaR with ES at 97.5% as the primary market risk measure. Key reasons: ES captures tail severity (VaR ignores it), ES is coherent (sub-additive), and it reduces incentives to accumulate tail risk just outside the VaR threshold. A, B, D are wrong.
Question 40
In the context of measuring volatility, the term 'volatility of volatility' (vol of vol) refers to:
Select an option first.
Correct answer: B — How much volatility itself changes or fluctuates over time.
Explanation: Vol of vol is the variability of volatility itself — how much σ changes from day to day. High vol of vol indicates unstable, unpredictable volatility regimes. This concept is related to the Vanna and Volga sensitivities in options and matters for the accuracy of GARCH-type models.
Question 41
The term 'correlation breakdown' in risk management refers to:
Select an option first.
Correct answer: B — The empirical observation that correlations between assets tend to spike toward +1 during market crises, eliminating the diversification benefits observed in normal times.
Explanation: Correlation breakdown: in normal markets, assets may have low correlations. During crises (2008, 2020 COVID), most asset classes move together as risk-off sentiment dominates — correlations jump toward +1. This means diversification, which was modelled using calm-period correlations, fails exactly when needed most.
Question 42
Which of the following describes 'worst case analysis' in risk measurement?
Select an option first.
Correct answer: B — Identifying the single scenario that produces the largest possible portfolio loss.
Explanation: Worst case analysis seeks the maximum possible loss across a defined set of scenarios. It does not assign probabilities — it identifies the extreme boundary of risk. This differs from VaR (probabilistic threshold) and stress testing (specific plausible scenarios). A is expected shortfall-like; C is high-confidence VaR; D is a specific simulation approach.
Question 43
A portfolio manager runs a GARCH(1,1) model and finds α = 0.08 and β = 0.90. The long-run average daily volatility is 1.2%. What is ω?
Select an option first.
Correct answer: B — 0.000024
Explanation: ω = Long-run variance × (1 − α − β) = (0.012)² × (1 − 0.08 − 0.90) = 0.000144 × 0.02 = 0.00000288. Wait: (0.012)² = 0.000144. (1−0.08−0.90)=0.02. ω = 0.000144 × 0.02 = 0.00000288. Actually answer B is 0.000024 — let me recheck. Long-run var = ω/(1-α-β). If LR vol=1.2%=0.012, LR var=(0.012)²=1.44e-4. ω = 1.44e-4 × 0.02 = 2.88e-6 ≈ 0.00000288. Correct answer is C. Marking C as correct.
Question 44
A portfolio has 60% in Asset A (σ=10%) and 40% in Asset B (σ=20%), with correlation = 0.5. The portfolio standard deviation is approximately:
Select an option first.
Correct answer: B — 12.2%
Explanation: Portfolio variance = (0.6)²(0.10)² + (0.4)²(0.20)² + 2(0.6)(0.4)(0.5)(0.10)(0.20) = 0.0036 + 0.0064 + 0.0048 = 0.0148. σ_p = √0.0148 ≈ 12.2%. A uses simple weighted average without correlation; C and D have computational errors.
Question 45
A risk manager uses a 250-day rolling window to compute historical volatility. A market crisis occurred 300 days ago with a very high daily return. How does this affect the current volatility estimate?
Select an option first.
Correct answer: B — It has no effect because it occurred outside the 250-day window.
Explanation: With a fixed 250-day window, observations older than 250 days are simply excluded. This creates a 'ghost effect' — when the crisis day eventually falls out of the window, volatility estimates suddenly drop even though market conditions haven't changed. This is a key limitation of the rolling window approach.
Question 46
The 'volatility term structure' refers to:
Select an option first.
Correct answer: B — The pattern of implied volatilities across different option expiry dates for the same underlying asset.
Explanation: Volatility term structure: implied vol plotted against option maturity (all at the same strike). Typically upward-sloping in calm markets and inverted (short-dated vol > long-dated) during crises. This reflects the market's expectation of how volatility will evolve over time.
Question 47
Portfolio VaR is generally less than the sum of individual position VaRs because of:
Select an option first.
Correct answer: B — Diversification — correlations between positions below +1 reduce portfolio variance below the sum of individual variances.
Explanation: Portfolio variance = w'Σw. When correlations < 1, the portfolio variance (and VaR) is less than the sum of individual variances — the diversification benefit. At correlation = +1, no diversification; portfolio VaR = weighted sum. A, C, D are not the mathematical reason for VaR diversification benefit.
Question 48
Which of the following is the correct interpretation of a VaR that uses a 10-day holding period?
Select an option first.
Correct answer: B — The VaR estimates the loss threshold not expected to be exceeded over any 10-day period with the stated confidence.
Explanation: A 10-day VaR reflects the loss distribution over a 10-day horizon — the loss level that should not be exceeded over any 10-day period with the specified confidence. Basel III's market risk framework uses 10-day VaR (or scaled from 1-day). A, C, D mischaracterise what the holding period means.
Question 49
The minimum variance portfolio on the efficient frontier is characterised by:
Select an option first.
Correct answer: B — The single portfolio that has the lowest possible variance across all combinations of the risky assets.
Explanation: The global minimum variance portfolio (GMVP) is the leftmost point on the efficient frontier — it has the lowest possible variance achievable by any combination of the risky assets. A is the upper frontier; C is naive diversification; D is the tangency portfolio (on the CML).
Question 50
The efficient frontier shifts inward (represents lower return for a given risk) when:
Select an option first.
Correct answer: B — Correlations between assets increase toward +1, reducing diversification benefits.
Explanation: As correlations increase toward +1, the diversification benefit shrinks. The efficient frontier moves inward (toward the single-asset frontier), offering less return for each level of risk. Decreasing correlations (A) expand the frontier outward, improving the risk-return tradeoff.
Question 51
An analyst uses a 95% confidence interval for VaR. Approximately how many exceptions would be expected per year (252 trading days)?
Select an option first.
Correct answer: B — About 12-13
Explanation: At 95% VaR, exceptions occur 5% of the time. 5% × 252 ≈ 12.6 trading days per year. At 99%, expected exceptions ≈ 2.5/year. At 90%, ≈ 25/year. This is a standard back-testing calculation that risk managers must know.
Question 52
Which of the following most accurately describes 'incremental VaR'?
Select an option first.
Correct answer: B — The change in portfolio VaR resulting from adding a small additional position to the existing portfolio.
Explanation: Incremental VaR = ∂VaR_portfolio / ∂position_i × position_i. It measures the marginal contribution of each position to total portfolio VaR — essential for risk attribution, identifying the largest risk contributors, and making hedging decisions. A is standalone VaR; C is average VaR per position; D is a confidence level difference.
Question 53
Component VaR is useful for risk managers because:
Select an option first.
Correct answer: A — It sums to total portfolio VaR and shows each position's diversified contribution to overall risk.
Explanation: Component VaR = ρ_i,p × Standalone_VaR_i (correlation-adjusted). Crucially, the sum of all component VaRs equals total portfolio VaR. This decomposition enables risk attribution — identifying which positions drive total risk — while accounting for diversification. B is max loss; C and D are incorrect.
Question 54
The term 'phantom P&L' or 'ghost effect' in historical simulation refers to:
Select an option first.
Correct answer: B — The sudden drop in estimated volatility when a high-return period (like a crisis) rolls out of the fixed historical window.
Explanation: Ghost/phantom effect: when a crisis day exits the rolling window, measured volatility drops abruptly even though current conditions haven't improved. This creates artificially low risk estimates. The opposite happens when the crisis day enters the window. This is a key limitation of fixed-window historical simulation.
Question 55
Under Basel's Internal Models Approach for market risk, which of the following multipliers is applied to the calculated VaR to determine the capital charge?
Select an option first.
Correct answer: B — A minimum multiplier of 3 (plus a qualitative add-on of up to 1 for back-testing exceptions).
Explanation: Basel IMA: Capital = max(VaR_t-1, multiplier × VaR_avg_60days) × (3 + add-on). The minimum multiplier is 3 (to provide a safety margin over model VaR). The add-on (0-1) is for back-testing exceptions. This was subsequently revised under the FRTB framework.
Question 56
Which of the following is a key advantage of GARCH volatility models over simple rolling-window models for VaR estimation?
Select an option first.
Correct answer: B — GARCH explicitly models volatility clustering and mean reversion, producing more responsive and theoretically grounded volatility forecasts.
Explanation: GARCH advantages: (1) explicitly models volatility clustering (α captures reaction to shocks), (2) incorporates mean reversion toward long-run variance (β determines persistence), (3) statistically estimated parameters. A is wrong — GARCH is complex; C is incorrect; D is wrong — all window-based models have some form of this issue.
Question 57
What happens to the estimated correlation between two assets under the EWMA approach when both assets experience a large return shock simultaneously?
Select an option first.
Correct answer: B — The estimated correlation increases, reflecting the co-movement observed in the recent data.
Explanation: EWMA applies to the covariance matrix: Cov_t = λ·Cov_(t-1) + (1−λ)·r_(i,t-1)·r_(j,t-1). A simultaneous large positive return in both assets increases the recent covariance term, raising the estimated correlation. This captures correlation breakdown (correlations rising in crises) dynamically.
Question 58
Which of the following statements about historical simulation is correct?
Select an option first.
Correct answer: C — It implicitly captures fat tails and skewness through the actual historical data.
Explanation: Historical simulation's core advantage: it uses actual historical returns, which naturally contain fat tails, skewness, and other non-normalities without any parametric assumption. A is wrong — it makes no distributional assumption. B and D are both incorrect — it can over- or under-estimate depending on the window.
Question 59
Two assets have standard deviations of 10% and 15% and a correlation of 0.4. The covariance is:
Select an option first.
Correct answer: B — 0.006
Explanation: Cov(A,B) = ρ × σ_A × σ_B = 0.4 × 0.10 × 0.15 = 0.4 × 0.015 = 0.006. A uses σ² rather than σ; C multiplies 0.25×0.01; D is just σ_B. The covariance is 0.006.
Question 60
A portfolio has 60% in Asset A (σ=10%) and 40% in Asset B (σ=20%), with correlation = 0.5. The portfolio standard deviation is approximately:
Select an option first.
Correct answer: C — 13.6%
Explanation: Portfolio variance = w_A²σ_A² + w_B²σ_B² + 2w_Aw_Bρσ_Aσ_B = 0.36×0.01 + 0.16×0.04 + 2×0.6×0.4×0.5×0.10×0.20 = 0.0036 + 0.0064 + 0.0048 = 0.0148. σ_p = √0.0148 ≈ 12.2%. Closest to C at 13.6% — let me recalculate: 2×0.6×0.4×0.5×0.10×0.20 = 2×0.024×0.5×0.02 = actually: 2×0.6×0.4×0.5×0.02 = 2×0.0024 = 0.0048. Total = 0.0036+0.0064+0.0048=0.0148. σ=√0.0148=12.2%. Correct answer is B (12.6% is closest after rounding). Marking B.
More free FRM topics
Ten questions in
- The ones you miss are saved as a drill you can repeat
- Your place is kept, on this device and any other
- A streak, if that is the thing that gets you back tomorrow
Every question on this page stays free and open either way.