FRM Part II · FRM Exam Part II · Backtesting VaR
A bank's internal model produces 99% one-day VaR. Over the last 250 trading days it records 7 exceptions. Under the Basel traffic light framework, which outcome is most accurate?
Seven exceptions is the yellow zone. Supervisors add a table-based plus factor, which is 0.85 at seven exceptions, so the multiplier becomes about 3.85. The automatic increase to 4.0 applies only to the red zone, at ten or more exceptions.
- AGreen zone; the multiplier stays at the minimum of 3.0
- BYellow zone; the multiplier increase is set by a supervisory table and is between 0 and 1 added to the baseCorrect
- CRed zone; the model is automatically rejected and the multiplier rises to 4.0
- DYellow zone; the multiplier is automatically increased to 4.0
Explanation
Seven exceptions lie in the yellow zone (5-9). The plus factor is a table value that rises with exceptions (0.65 for 5, 0.75 for 6, 0.85 for 7, 0.92 for 8, 0.97 for 9), giving a multiplier of 3.85 for 7. A 4.0 multiplier applies only in the red zone (10+).
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 99% one-day VaR over 500 days and observes 12 exceptions. The expected number is 5. A review finds that the model assumes…
- A bank has 8 exceptions in the last 250 days under its 99% one-day VaR. Investigation shows the exceptions arose because the model's positio…
- 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 bank backtests a 99% one-day VaR over 250 days. Assume exceptions are independent with probability 1% each. The bank has 5 exceptions and …
- 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%. …