Can I annualize Expected Shortfall with square-root-of-time the same way I annualize volatility?

Question

AcadiFi · Accepted Answer

Expected Shortfall can only be scaled cleanly under strong distributional assumptions. The square-root-of-time rule comes from volatility logic under independent, identically distributed returns with stable variance. Tail measures are more fragile because autocorrelation, fat tails, and changing regimes distort how bad outcomes accumulate over longer horizons.

Suppose Meridian Pension Overlay computes a monthly ES from five years of history. If the return process were close to normal and independent through time, a scaling shortcut might be used as a rough approximation. But if the portfolio contains options, illiquid credit, or clustered volatility, the scaled annual ES may be seriously misleading because the tail is not behaving like a simple volatility multiple.

```mermaid
flowchart TD
  A[Monthly ES] --> B{Strong iid assumption?}
  B -->|Yes| C[Approximate scaling may be acceptable]
  B -->|No| D[Re-estimate at target horizon]
  D --> E[Use simulation or scenario aggregation]
```

The safer FRM answer is to prefer direct horizon estimation or simulation when tail behavior matters. The more nonlinear the portfolio, the less comfortable you should feel with naive time scaling. Scaling can be a convenience, but it is not a universal law for expected shortfall in real portfolios or exams. Practice more in our FRM question bank.

Can I annualize Expected Shortfall with square-root-of-time the same way I annualize volatility?

Master Part I with our FRM Course

Related Questions

Practice Questions