Question

23) Which of the following statements is not true about historical simulation approach using weighing of observations? Recent observations are given more weight than more distant observations in the past In order to find VaR at a given confidence level, weights are summed up starting from the worst outcome until the required percentile of the distribution is reached It incorporates volatility updating procedures such as EWMA The method reflects better current volatilities and macroeconomic conditions as compared to the basic approach 24) The model building approach as a method of defining market risk VaR: Is based on Modigliani and Miller theory. is based on Markowitz's portfolio theory Is based on Cooke ratio regarding the capital requirements Is based on the nonlinear products variances and covariances 25) Suppose that we have bought $50 million of Google shares. The daily volatility of the Google stock is 2%. Changes in the value of the stock are normally distributed. The one- day 99% market risk VaR using the model-building approach is (hint: N(-2.33) = 0.01): $2.33 million $1 million $2.66 million Depends on yearly volatility of the price of a Google share

Answer 1

23). In traditional historical simulations, all observations are equally weighted so option (I) is incorrect.

24). The model building approach for computing VaR is based on Markowitz portfolio theory. Option (II) is correct.

25). 1-day 99% VaR = daily volatility*amount invested*2.33

= 2%*50*2.33 = 2.33 milion (Option I)