FRM Part II · FRM Exam Part II · Backtesting VaR
A bank's trading desk reports 12 exceptions in 250 days against a 99% one-day VaR. A review finds that 8 of them occurred on days when traders had entered positions after the VaR run and before close. Which cause of exceptions does this best describe?
The exceptions are best explained by intraday trading. VaR is based on the previous close portfolio, so positions added after the VaR run alter actual P&L relative to what the model measured. This is a position-change cause, not a flaw in distribution, mapping or confidence level.
- AIntraday trading that changed the portfolio from the one used for VaRCorrect
- BPoor choice of confidence level
- CModel misspecification of the return distribution
- DRisk-factor mapping error
Explanation
VaR is computed on end-of-day positions. When positions change during the day, actual P&L reflects a different portfolio from the one measured, so exceptions can arise even if the model is sound. The other causes would not be tied specifically to positions entered after the VaR run.
Did you get it right without looking?
One question tells you little. A timed set on Backtesting VaR shows your real accuracy, how long you take and where you lose marks.
More Backtesting VaR questions
- A bank backtests a 95% one-day VaR over 500 days and observes 40 exceptions. The Kupiec likelihood ratio statistic is LR = -2 ln[(0.95)^460 …
- A bank uses a 99% one-day VaR model and tests it with a two-sided binomial test at the 5% significance level over a long sample. The model p…
- A risk manager backtests a 99% one-day VaR model over 250 trading days. Under the assumption that the model is correctly calibrated and exce…
- A risk analyst evaluates a bank's 99% VaR model with 250 observations. Assume the model is correct, so exceptions are binomial with p = 1%. …
- A bank's 99% one-day VaR model produced 4 exceptions in 250 days, all in the same two-week period following a market shock. Which conclusion…
- A risk manager backtests a 99% one-day VaR model and rejects the model whenever the number of exceptions exceeds a chosen cutoff. In this co…