Expected Shortfall

Value at Risk tells you the loss that will not be exceeded at a certain confidence level, but says nothing about how bad losses could be beyond that point. Expected shortfall, also called conditional value at risk, fills this gap by averaging the losses in the tail.

For example, a portfolio's one-day 95% Value at Risk might be 2,000,000, meaning losses exceed that amount on about 5% of days. Its expected shortfall at 95% might be 3,100,000, the average loss on those worst days.

Expected shortfall gives a better picture of tail risk, particularly for portfolios that can suffer rare but very large losses. It also has mathematical properties that make it more consistent when risks from different portfolios are combined.

After the 2008 financial crisis, banking regulators moved towards expected shortfall for measuring market risk, replacing Value at Risk in the main capital framework for banks' trading activities. Like any model, its accuracy depends on the data and assumptions used, especially about rare events.