Backtest vs Validation

[rahul.goyl]

Hi David,

How to go about this one.

Quantcept Investment Management decided to test the VAR model for its Treasury Income Portfolio over 10 trading days in September. Trading on the portfolio ceased during the ten days of the test. The backtest provided the following predicted daily percentage returns for the portfolio, at a 95% confidence level:
Lower Limit Upper Limit Actual
-0.21 0.42 0.28
-0.16 0.38 -0.14
-0.15 0.39 0.37
-0.12 0.42 0.08
-0.10 0.40 -0.12
-0.15 0.43 0.01
-0.18 0.42 -0.25
-0.20 0.43 0.19
-0.22 0.43 0.28
-0.23 0.42 -0.17
Which statement about the backtest is TRUE?
A) At this confidence level, 16% of the actual results in the backtest should fall outside the predicted range.
B) The backtest is not valid because the portfolio was not traded during the ten days. It is only appropriate to backtest a portfolio that has undergone significant changes between calculations because otherwise the results are not relevant to actual trading situations.
C) At the end of the period, Quantcept will recompute VAR to determine what the upper and lower ranges of the portfolio return should have been.
D) The backtest does not confirm the validity of the model.

regards,
Rahul