Q5 (20 points) - Back testing your VaR calculated from Question 4 We will test whether...

Transcribed Image Text:

Q5 (20 points) - Back testing your VaR calculated from Question 4 We will test whether your 1-day VaR calculated from Question 4 is a good risk measure. Use the VaR calculated using the actual volatility of your portfolio returns (i.e., from (iii) in Q4). We will back test your 1-day VaR using the security prices from January 1999 to December 2001. The data is located in the worksheet labelled "DataProblem5 and6". (a) (5 points) Assuming that your portfolio holding is the same as in Question 4, calculate daily returns of your portfolio from 1999 - 2001. Plot the results. (b) (7 points) Report the number of times that 1-day 99% VaR of your portfolio is violated during 1999 - 2001. Note that the violation occurs when the portfolio return falls greater than what your VaR measure suggests. Calculate the violation rate of your 1-day VaR measure. Show your work. If your VaR measure is good, what do you expect the violation rate to be? (c) (5 points) Using the statistical method introduced in class (Binomial distribution), test whether your 1-day 99% VaR measure is an acceptable model. Show your work. Q5 (20 points) - Back testing your VaR calculated from Question 4 We will test whether your 1-day VaR calculated from Question 4 is a good risk measure. Use the VaR calculated using the actual volatility of your portfolio returns (i.e., from (iii) in Q4). We will back test your 1-day VaR using the security prices from January 1999 to December 2001. The data is located in the worksheet labelled "DataProblem5 and6". (a) (5 points) Assuming that your portfolio holding is the same as in Question 4, calculate daily returns of your portfolio from 1999 - 2001. Plot the results. (b) (7 points) Report the number of times that 1-day 99% VaR of your portfolio is violated during 1999 - 2001. Note that the violation occurs when the portfolio return falls greater than what your VaR measure suggests. Calculate the violation rate of your 1-day VaR measure. Show your work. If your VaR measure is good, what do you expect the violation rate to be? (c) (5 points) Using the statistical method introduced in class (Binomial distribution), test whether your 1-day 99% VaR measure is an acceptable model. Show your work.

Answer rating: 100% (QA)

a To calculate daily returns of your portfolio from 19992001 you would typically use the formula Daily Return Portfolio Value Today Portfolio Value Ye... View the full answer