Skip to content

FRM Part II · FRM Exam Part II · Beyond Exceedance-Based Backtesting of Value-at-Risk Models

An analyst compares two VaR models using average quantile scores over 500 days and finds Model X has a lower (better) average score than Model Y by a small margin. Before concluding that X is superior, what is the most appropriate next step?

The analyst should test whether the difference in scores is statistically significant, for example with a Diebold-Mariano type test on the daily score differentials. A slightly lower average could be sampling noise, so it does not by itself prove one model is better.

  1. AApply a statistical test on the score differential, such as a Diebold-Mariano type test, to see if the difference is significantCorrect
  2. BChoose Model X because any lower average score proves superiority
  3. CChoose the model with the higher average VaR
  4. DDiscard both models because scores cannot be compared across models

Explanation

A small difference in average scores may arise from sampling noise. A test on the series of daily score differences, such as Diebold-Mariano, assesses whether the gap is statistically significant. Lower average alone does not establish superiority, and a higher VaR is not inherently better.

Did you get it right without looking?

One question tells you little. A timed set on Beyond Exceedance-Based Backtesting of Value-at-Risk Models shows your real accuracy, how long you take and where you lose marks.

Practise Beyond Exceedance-Based Backtesting of Value-at-Risk ModelsFree diagnostic mock

More Beyond Exceedance-Based Backtesting of Value-at-Risk Models questions