A VaR method assuming returns are normally distributed: VaR = z × σ × value × √t, where z is the standard-normal quantile for the confidence level. It is simple but understates risk for fat-tailed or option-heavy (nonlinear) portfolios.