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 › Credit Risk Measurement and Management
Free FRM Credit Risk Measurement and Management Practice Questions & Answers
400 exam-style Credit Risk Measurement and Management questions. Pick your answer, hit Check answer, and see the worked solution — free to start, no signup.
100% free · No login to startQuestion 121
A credit risk manager wants to test whether a PD model is well-calibrated across different risk bands. The most appropriate statistical test is:
Select an option first.
Correct answer: B — The Hosmer-Lemeshow test — it groups predicted probabilities into bands (deciles) and compares predicted vs. actual default rates in each band using a chi-squared test.
Explanation: Hosmer-Lemeshow: specifically designed to test calibration of probability models. Groups observations into deciles of predicted probability, then tests if predicted and actual default rates within each group are significantly different. A tests distributional fit; C tests normality; D tests stationarity of time series.
Question 122
A bank's credit risk team discovers that its PD model has a Gini coefficient of 0.72 on the development sample but 0.48 on a new out-of-sample validation. The most likely cause is:
Select an option first.
Correct answer: B — Model overfitting — the model was over-tuned to the development dataset's specific patterns, capturing noise rather than true credit risk signals, and fails to generalise to new data.
Explanation: A 24-point Gini degradation from development to validation is a strong signal of overfitting. Remedies: reduce model complexity, apply regularisation, re-specify variables using economic logic, validate on multiple time periods. A is possible but a 24-point drop is too large to attribute to sample size. C is possible but partial; D is wrong — logistic regression can generalise well when properly specified.
Question 123
The 'survival function' S(t) in credit risk modelling represents:
Select an option first.
Correct answer: B — The probability that a borrower has NOT defaulted by time t — it is the complement of the cumulative default probability.
Explanation: S(t) = P(Default time > t) = 1 − F(t), where F(t) is the cumulative distribution function of the default time. Survival functions are used in credit duration models, hazard rate models, and CDO pricing. A, C, D describe different concepts.
Question 124
In a 'hazard rate' (intensity) model for credit risk, the 'default intensity' λ(t) represents:
Select an option first.
Correct answer: B — The instantaneous conditional rate of default at time t, given survival to time t — the probability of defaulting in the next instant given no default so far.
Explanation: Hazard rate λ(t) = (−dS/dt) / S(t). In reduced-form credit models (Jarrow-Turnbull, Duffie-Singleton), λ(t) is modelled as a stochastic process, allowing the default process to be calibrated to CDS curves. A is a historical average; C is the market spread (related but not the same); D is a structural model metric.
Question 125
A reduced-form credit model (intensity model) differs from a structural model (Merton) in that:
Select an option first.
Correct answer: B — In reduced-form models, default is modelled as a surprise event driven by an exogenous intensity process, without explicitly modelling the firm's asset value — making them easier to calibrate to market CDS data.
Explanation: Structural models (Merton): default when assets hit a threshold — requires unobservable asset value. Reduced-form models: default arrives at a random time driven by a calibrated hazard rate process — directly calibrated to CDS market prices. A reverses the calibration advantage; C reverses the data source; D is wrong.
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 126
A credit risk model produces a ROC curve with an area under the curve (AUC) of 0.92. This indicates:
Select an option first.
Correct answer: B — The model has strong discriminatory power — it correctly ranks borrowers by default risk 92% of the time when comparing a random defaulter to a random non-defaulter.
Explanation: AUC (AUROC) = P(score of random defaulter < score of random non-defaulter). AUC = 0.92 means the model ranks the defaulter above the non-defaulter 92% of the time. AUC = 0.5 is random; AUC = 1.0 is perfect. A is accuracy (different metric); C is overfitting concern (requires comparison to OOS AUC); D is correlation (different).
Question 127
The 'credit loss distribution' of a large, well-diversified credit portfolio is typically:
Select an option first.
Correct answer: B — Positively skewed with a heavy right tail — many small losses (expected loss region) but rare catastrophic large losses (tail).
Explanation: Credit loss distributions: asymmetric (positively skewed) because: (1) most years have relatively small losses near EL; (2) rare recession years produce catastrophic tail losses. This skewness means simple normal distribution assumptions dramatically underestimate tail risk. A is wrong; C is simplistic; D has wrong skew direction.
Question 128
In credit portfolio modelling, 'asset correlation' captures:
Select an option first.
Correct answer: B — The degree to which the underlying business/economic conditions of two borrowers move together — high correlation means they tend to default simultaneously, increasing portfolio tail risk.
Explanation: Asset correlation ρ: how correlated are the systematic risk factors driving each borrower's default? High correlation → defaults cluster → fat portfolio loss tail → high credit VaR. Basel prescribes asset correlation in the IRB formula. A is rating correlation; C is PD-LGD correlation (also important but separate); D is spread-rate correlation.
Question 129
A portfolio manager notices that a bond's CDS-implied PD is significantly higher than its rating-implied PD. The most likely explanation is:
Select an option first.
Correct answer: B — The CDS market incorporates current market information and risk sentiment more rapidly than rating agencies, which tend to be slower to respond to changing conditions.
Explanation: CDS spreads are market-based — they reflect real-time investor sentiment, news, and market risk premium. Rating agencies update ratings infrequently and use TTC methodology. During stress periods, CDS-implied PDs often exceed rating-implied PDs (ratings lag). A is too absolute; C is possible but not the primary explanation; D is an empirical observation in some markets but not always.
Question 130
The 'conditional probability of default' in the Vasicek model, given a systematic factor Z, is:
Select an option first.
Correct answer: B — N((N⁻¹(PD) − √ρ × Z) / √(1−ρ)) — it increases when Z is negative (bad economic conditions).
Explanation: Vasicek conditional PD: PD(Z) = N[(N⁻¹(PD) − √ρ×Z) / √(1-ρ)]. When Z is very negative (severe recession), conditional PD approaches 1. When Z is very positive (strong boom), conditional PD approaches 0. This is the basis of the Basel IRB capital formula, which uses Z = N⁻¹(0.001) = -3.09 for the 99.9th percentile.
Question 131
A bank originating mortgages decides to retain the first-loss piece of a securitisation (5% of the pool) and sell the remaining tranches. The primary purpose from a regulatory capital perspective is:
Select an option first.
Correct answer: A — To comply with risk retention rules while maximising capital relief on the transferred portion.
Explanation: Risk retention (5% horizontal slice or first-loss): (1) satisfies regulatory requirements (EU Securitisation Regulation, Dodd-Frank); (2) aligns originator incentives with investors (skin in the game); (3) achieves significant risk transfer (SRT) capital relief on the remaining 95%. B is wrong (first-loss piece requires higher capital density); C is a secondary effect; D contradicts good practice.
Question 132
Which of the following best explains the 'pro-cyclicality' of IFRS 9 provisioning?
Select an option first.
Correct answer: B — In an expansion, PIT PDs and LGDs decline, reducing IFRS 9 provisions (building implicit reserves). In a recession, PIT PDs spike, causing a sudden large increase in provisions that hits earnings and capital simultaneously — amplifying the credit cycle.
Explanation: IFRS 9 pro-cyclicality: the forward-looking, PIT nature of IFRS 9 means provisions rise sharply in recessions when PIT PDs increase. While less procyclical than IAS 39 (incurred loss), IFRS 9 can still create cliff effects, particularly at Stage 1→2 migration. A is not always true; C is wrong; D is wrong (TTC would reduce, not eliminate, procyclicality).
Question 133
In credit risk modelling, the 'information coefficient' (IC) measures:
Select an option first.
Correct answer: B — The predictive correlation between the model's score and the binary default outcome, indicating how useful the score is for distinguishing future defaulters from non-defaulters.
Explanation: IC in credit scoring: correlation between predicted score and actual binary outcome (1 = default, 0 = non-default). Higher IC → more useful score. Related to AUROC: IC ≈ 2×(AUROC − 0.5) = Gini coefficient. A, C, D are different metrics.
Question 134
A forward-looking credit stress test incorporating a 'severe adverse' macroeconomic scenario is most useful for:
Select an option first.
Correct answer: B — Assessing whether the bank's capital is sufficient to absorb credit losses under a plausible but severe economic downturn — identifying vulnerabilities and informing capital planning.
Explanation: Credit stress testing: applies a severe scenario (e.g., GDP −5%, unemployment +5pp) to the portfolio PD, LGD, and EAD, estimating cumulative credit losses and checking if capital remains above regulatory minima. Identifies sector, name, and geographic concentrations that are most vulnerable. A is market risk; C is pricing (different purpose); D is incorrect — stress testing is an input to ICAAP.
Question 135
The 'expected shortfall' approach to credit risk is preferred over the 'VaR' approach because:
Select an option first.
Correct answer: B — ES captures the average severity of losses beyond the VaR threshold, providing a more complete picture of tail credit risk rather than just the loss level at a given confidence threshold.
Explanation: Credit ES = average loss given loss > credit VaR threshold. VaR only tells you the loss level at a given percentile; ES tells you how bad losses are in the tail beyond that point. ES is coherent (sub-additive) while VaR may not be. A is wrong; C is wrong (more complex); D is wrong (IFRS 9 uses ECL, not ES).
Question 136
Under Basel III, the 'minimum total capital ratio' is:
Select an option first.
Correct answer: B — 8.0% of risk-weighted assets (4.5% CET1 + up to 1.5% AT1 + up to 2.0% Tier 2).
Explanation: Basel III minimum: 8% total capital / RWA = 4.5% CET1 + 1.5% AT1 (max) + 2.0% Tier 2 (max). Plus Conservation Buffer (2.5%) = 10.5% with buffer. Plus CCyB and G-SIB surcharge add further. A is CET1 minimum only; C includes conservation buffer; D is a higher threshold.
Question 137
A bank calculates its 'risk-weighted assets' (RWA) by multiplying each credit exposure by its regulatory risk weight. For a loan to an unrated corporate borrower under the Basel SA, the risk weight is typically:
Select an option first.
Correct answer: C — 100%
Explanation: Basel SA corporate: unrated corporates receive 100% risk weight (one unit of capital per unit of exposure × 8%). Rated corporates may receive lower weights depending on external rating. Sovereign debt from OECD countries = 0-20%; covered bonds = 10-50%; retail = 75%. A is sovereign (OECD); B is bank/public sector; D is high-risk (below B-/defaulted).
Question 138
The 'Supervisory Review and Evaluation Process' (SREP) under Basel Pillar 2 enables supervisors to:
Select an option first.
Correct answer: B — Review each bank's ICAAP, impose institution-specific capital add-ons for risks not captured by Pillar 1, and require remediation of identified risk management weaknesses.
Explanation: SREP: regulators evaluate: (1) ICAAP quality; (2) adequacy of capital vs all risks (Pillar 2 add-on if needed); (3) risk management and governance quality; (4) stress test results. Bank-specific Pillar 2 requirements reflect each institution's unique risk profile. A, C, D are incorrect.
Question 139
The 'credit valuation adjustment' (CVA) charge in Basel III was introduced because:
Select an option first.
Correct answer: B — Banks suffered large mark-to-market CVA losses during the 2008 crisis without holding capital against CVA volatility — Basel III introduced a specific capital charge to cover the risk of CVA losses.
Explanation: During 2008, ~2/3 of credit losses from CCR were from MTM CVA losses (spread widening of counterparties like Lehman, Bear Stearns), not actual defaults. Basel III introduced CVA capital (SA-CVA or BA-CVA) to cover this previously uncapitalised risk. A is wrong; C and D are incorrect.
Question 140
The 'leverage ratio buffer' for G-SIBs under Basel requires:
Select an option first.
Correct answer: B — G-SIBs to maintain an additional 50% of their G-SIB surcharge as a leverage ratio buffer (e.g., 1% additional leverage ratio for a bank with a 2% G-SIB surcharge) on top of the 3% minimum.
Explanation: G-SIB leverage ratio buffer: leverage ratio ≥ 3% + 50%×G-SIB buffer%. For a 2% G-SIB bank: leverage ratio ≥ 3% + 1% = 4% minimum. This prevents G-SIBs from using high leverage to amplify systemic risk. A, C, D are incorrect.
Question 141
Under IFRS 9, the '12-month ECL' (Stage 1) represents:
Select an option first.
Correct answer: B — The expected credit loss attributable to default events that are possible within the next 12 months, not the total loss if default occurs within 12 months.
Explanation: 12-month ECL = PD(12-month) × LGD × EAD. It is the ECL for defaults that could occur in the next 12 months — not the full lifetime ECL. The PD is the 12-month conditional PD, applied to the full LGD × EAD if default occurs at any point in that period. A is lifetime ECL (Stage 2/3); C is incurred loss (old IAS 39); D is stress loss.
Question 142
A bank's 'Internal Capital Adequacy Assessment Process' (ICAAP) should cover which risks?
Select an option first.
Correct answer: B — All material risks including those not covered by Pillar 1: credit risk, market risk, operational risk, IRRBB, concentration risk, pension risk, reputational risk, strategic risk, and liquidity risk.
Explanation: ICAAP scope: comprehensive — covers all material risks, whether or not they are captured by Pillar 1. Pillar 2 risks specifically include: IRRBB (interest rate risk in the banking book), credit concentration, liquidity risk, pension deficit, reputational, and strategic risk. A, C, D are too narrow.
Question 143
The 'Net Stable Funding Ratio' (NSFR) addresses which risk category?
Select an option first.
Correct answer: B — Structural liquidity risk — ensuring banks fund long-term illiquid assets with stable, longer-term funding sources rather than short-term wholesale funding.
Explanation: NSFR (Basel III): Available Stable Funding / Required Stable Funding ≥ 100%. It prevents the maturity mismatch that caused the 2008 liquidity crisis (funding long-term illiquid assets with short-term wholesale CP). A, C, D are different risk categories.
Question 144
In credit risk governance, the 'credit review function' (independent credit review / loan review) serves to:
Select an option first.
Correct answer: B — Independently assess the credit quality of loans, validate rating accuracy, identify emerging portfolio problems, and provide objective feedback to senior management and the board.
Explanation: Independent credit review: a third line of defence function (or strong second line) that: (1) reviews loans post-origination to confirm accurate risk ratings; (2) identifies deterioration not yet reflected in formal ratings; (3) assesses adequacy of provisions; (4) reports findings to the board/audit committee. A is the approval function; C and D are business/pricing functions.
Question 145
A bank's 'credit concentration risk' is best managed through:
Select an option first.
Correct answer: B — A combination of single-name limits, sector limits, geographic limits, and regular portfolio-level concentration reporting to the risk committee.
Explanation: Concentration risk management: (1) single-name limits (Basel large exposure ≤25% Tier 1); (2) sector limits (e.g., max 20% of portfolio in real estate); (3) geographic limits; (4) portfolio analytics identifying hidden concentrations (e.g., multiple companies exposed to the same commodity). A raises capital but doesn't reduce concentration; C and D are too narrow or unrelated.
Question 146
The 'Special Assets' or 'Workout' team at a bank deals with:
Select an option first.
Correct answer: B — Managing distressed and defaulted credit exposures — working to maximise recovery through restructuring, enforcement of collateral, or sale of the distressed debt.
Explanation: Special Assets/Workout: takes over credit management of loans transferred from the regular credit team when they become non-performing. Responsibilities: (1) assess recovery options; (2) negotiate restructuring; (3) enforce security; (4) manage insolvency/bankruptcy. Goal = maximise NPV of recovery. A, C, D are different functions.
Question 147
The 'economic capital' framework enables a bank to:
Select an option first.
Correct answer: B — Allocate capital to business lines based on risk-adjusted returns, enabling comparison of risk-adjusted profitability (RAROC) across different credit activities with very different risk profiles.
Explanation: Economic capital utility: different business lines (e.g., mortgage lending vs. leveraged finance) have very different risk profiles and capital consumption. RAROC = Risk-adjusted income / Economic capital standardises profitability comparison. Enables capital allocation to highest RAROC activities. A is wrong (regulatory capital is a regulatory minimum, not replaceable); C and D are incorrect.
Question 148
Under the 'Standardised Approach for Counterparty Credit Risk' (SA-CCR), the exposure at default for a derivative is calculated as:
Select an option first.
Correct answer: B — Replacement cost (current MTM or zero if negative) plus potential future exposure (PFE), where PFE is calculated using supervisory factor tables that depend on asset class and maturity.
Explanation: SA-CCR: EAD = α × (RC + PFE), where α = 1.4, RC = max(MTM, 0) after netting and collateral, PFE = supervisory aggregate addon based on asset class (IR, FX, credit, equity, commodity) × notional × supervisory factor. A ignores PFE; C uses arbitrary multiplier; D is VaR-based (different approach).
Question 149
A bank's 'credit risk dashboard' for portfolio monitoring should include:
Select an option first.
Correct answer: B — Key portfolio metrics including NPL ratio, provision coverage, rating migration analysis, concentration metrics, sector breakdowns, and vintage performance.
Explanation: Effective credit risk dashboard: comprehensive view covering (1) portfolio quality (NPL, watchlist, rating distribution); (2) coverage (provisions/NPLs); (3) concentration (top 10 exposures, sector/geography); (4) performance trends (vintage analysis, rating migrations); (5) capital adequacy. A, C, D are too narrow.
Question 150
The 'BCBS 239' principles for risk data aggregation and risk reporting require banks to:
Select an option first.
Correct answer: B — Ensure risk data is accurate, complete, timely, and adaptable — capable of being rapidly aggregated in normal conditions and stress periods to support informed decision-making.
Explanation: BCBS 239 (January 2013): 14 principles for risk data aggregation and reporting. Key requirements: (1) data accuracy and integrity; (2) completeness across all risk types; (3) timeliness (near-real-time in stress); (4) adaptability to management requests; (5) group-level aggregation capability. Post-2008, many banks could not aggregate risk data quickly enough. A, C, D are incorrect.
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.