Financial Risk Manager Part 2

Get started today

Ultimate access to all questions.

Explanation:

Since the test is being conducted at the two-tailed 95% level, the cutoff value for the test will be 1.96. To test whether the model is accurate, the following hypothesis is tested:

\frac{(x - pT)}{\sqrt{p(1 - p)T}} > z = 1.96

\frac{x - 0.02 \times 252}{\sqrt{0.02 \times 0.98 \times 252}} = 1.96

\Rightarrow x > 9.39596

In this case, the acceptable number of exceptions allowed for Black to conclude that the model is correctly calibrated is 9.

Explanation:

Since the test is being conducted at the two-tailed 95% level, the cutoff value for the test will be 1.96. To test whether the model is accurate, the following hypothesis is tested:

\frac{(x - pT)}{\sqrt{p(1 - p)T}} > z = 1.96

\frac{x - 0.02 \times 252}{\sqrt{0.02 \times 0.98 \times 252}} = 1.96

\Rightarrow x > 9.39596

In this case, the acceptable number of exceptions allowed for Black to conclude that the model is correctly calibrated is 9.

Comments (0)

No comments yet.

Q.2637 Jason Black, a risk analyst at a large multinational bank, is backtesting the VaR model of the bank. The model being tested is a daily, 98% VaR model. If the backtest is conducted for one year at a two-tailed 95% test confidence level, and assuming that a year has 252 trading days, what is the least acceptable number of daily losses that will lead Black to conclude that the model is calibrated correctly?

Exam-Like

Community

UAnonymous

Last updated: July 5, 2026 at 07:18