--- title: Chapter 5 Implementing risk forecasts layout: default ---

Implementing risk forecasts

  1. Outline one advantage and two disadvantages of using Historical Simulation for forecasting risk.

  2. Derive VaR for continuously compounded returns.

  3. Describe the so called burn time in an EWMA model.

  4. Outline two disadvantages of obtaining VaR forecasts using parametric methods such as GARCH.

  5. What is the recommended sample size for HS?

  6. Suppose the historical annual volatility and mean of returns on the Japanese TOPIX index are 20% and 2%, respectively, while the annual historical Japanese inflation is 0.1%. Further suppose that returns on the Japanese stock market are IID normally distributed. The probability level for all VaR calculations is 1%. Consider an investor who has 1,000 invested in the index.

    1. Based on the information given above and reasonable assumptions, would you expect the long run real investment performance to be less than 2%, 2%, 3% or more a year?

    2. Based on the information given above, what is the one month VaR?

    3. Based on the information given above as well as your answer in the question above, what is the 1,000 year VaR?

    4. What issues does your answer to the last question raise about the concept of VaR?