This document discusses applying "square root of time" scaling to calculate value at risk (VaR) over longer time horizons when direct calculation is not possible due to insufficient historical data. It explains that regulators may request VaR measures over periods like 2 weeks, but with only 4 years of data there would be too few samples to directly calculate a 99% VaR. Therefore, 1-day VaR is computed and multiplied by a scaling factor based on the square root of time. While theoretically incorrect due to non-constant volatility, this approach is examined to determine if it provides an acceptable approximation, especially for tail distributions like 95% or 99% VaR. MATLAB code is presented to automate the calculations and visually examine the