Conditional Value at Risk, also called Expected Shortfall, answers the question VaR ignores: when you do breach the VaR threshold, how bad is it on average? CVaR is the mean loss across all the outcomes worse than VaR, so it measures the depth of the tail rather than just its edge.
Because it accounts for extreme losses, CVaR is generally the more conservative and better-behaved risk measure for fat-tailed assets like crypto. A strategy with the same VaR as another but a much larger CVaR is hiding nastier tail events.
Formula
CVaR_α = average loss given that the loss exceeds VaR_α
Example
Two strategies share a 95% VaR of 3%, but one has a CVaR of 4% and the other 9% — the second has far heavier tail risk.
How Noon Barbari uses Conditional Value at Risk (CVaR)
Every concept here is implemented in the platform. Open the relevant docs or tool to see it in action.
See risk metrics in noonbarbari →Related terms
- Statistics
Value at Risk (VaR)
A loss threshold you won't exceed at a given confidence level over a period.
- Risk
Maximum drawdown
The deepest peak-to-trough decline observed across the entire equity curve.
- Statistics
Standard deviation
Square root of variance — the most common volatility measure in trading.
- Backtesting
Stress Test
Replaying a strategy through historically hard regimes and adversarial shocks.