Which of the following methods of estimating VaR most likely assumes that the return distributions for the risk factors in the portfolio are normal?