Financial institutions have traditionally used VaR and expected shortfall to analyze risk. But do these standard techniques actually work, and, if so, which one provides more accurate and reliable risk forecasts? Jon Danielsson and Chen Zhou offer a comprehensive breakdown.