medium · FRM Part 1 Foundations of Risk Management

Which deficiency would a correct-count backtest most plausibly fail to detect, and why does it matter most?

1. Clustering of the exceptions in a short window, which signals the model fails to capture volatility dynamics and conditional coverage, leaving the firm exposed to runs of breaches even though the unconditional frequency looks correct.
2. Systematic overstatement of VaR on calm trading days, which a count test misses entirely because holding excess capital above the true risk level is never penalized and the firm would simply carry a needlessly prudent buffer.
3. An error embedded in the mean return assumption used to build the model, which a count test misses because VaR at the 99% confidence level is driven overwhelmingly by the mean rather than by the shape of the underlying distribution's tail.
4. Use of a shorter 250-day historical estimation window instead of a longer 500-day window, which a count test misses because window length mainly affects the smoothness of the VaR series over time, not the actual number of exceptions recorded.

Sign up free to see the explanation and track your rank →