Quantitative Risk Analysis Flashcards
7 cards from real CRA practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 7 Quantitative Risk Analysis flashcards as text
In the Poisson process model for operational risk frequency, what does the parameter λ (lambda) represent?
Answer: The average number of loss events per unit time
Lambda is the rate parameter of the Poisson distribution, representing the mean number of events occurring per time period.
When a quantitative risk model consistently underestimates losses during back-testing, the most likely cause is:
Answer: The model's assumptions do not capture tail risk or regime changes adequately
Systematic underestimation during back-testing usually indicates the model fails to capture fat-tailed distributions or shifts in risk regime.
Which of the following best describes 'risk aggregation' in enterprise-wide quantitative risk management?
Answer: Combining individual risk exposures across business units accounting for diversification and correlations
Risk aggregation combines exposures across the enterprise while accounting for correlation and diversification benefits to derive a total risk picture.
In quantitative credit risk modeling, the term 'asset correlation' in the Vasicek model determines:
Answer: The degree to which borrowers' asset values move together, affecting portfolio default correlation
Asset correlation in the Vasicek/ASRF model drives how strongly individual default events are linked through common systematic risk factors.
A risk manager wants to estimate the 99.9% VaR for a portfolio but has only 500 data points. The most appropriate approach is to:
Answer: Fit a parametric extreme value distribution to augment the tail estimate
With limited data, parametric methods like EVT provide more reliable tail estimates than relying on sparse empirical observations at extreme quantiles.
The 'butterfly effect' concept is most relevant to which quantitative risk challenge?
Answer: Non-linear systems where small input changes produce disproportionately large outcomes
The butterfly effect describes sensitive dependence on initial conditions in non-linear systems, which makes long-run quantitative prediction extremely difficult.
In multi-period risk modeling, which of the following correctly describes the 'square root of time' rule for scaling VaR?
Answer: Multiply 1-day VaR by the square root of the number of holding days assuming i.i.d. returns
Under the assumption of independent, identically distributed returns, the T-day VaR equals the 1-day VaR multiplied by √T.