On this page
Expected Tail Loss: Sizing the Severity of Crashes
**Expected tail loss** is the average loss a portfolio suffers in its worst outcomes, the slice of the distribution beyond the value at risk cutoff. It is the same measure that regulators chose to replace value at risk for trading book capital, because it captures the depth of a crash rather than just its starting point.
Key Takeaways
- Expected tail loss is the average of all losses worse than the value at risk threshold.
- It goes by several names: expected shortfall, conditional value at risk, and average value at risk.
- The common mistake is assuming normal returns, which understates the fat-tailed losses ETL is built to capture.
- Basel rules replaced value at risk with expected shortfall for market risk capital after the 2008 crisis.
Key Takeaways
- Expected tail loss is the average of all losses worse than the value at risk threshold.
- It goes by several names: expected shortfall, conditional value at risk, and average value at risk.
- The common mistake is assuming normal returns, which understates the fat-tailed losses ETL is built to capture.
- Basel rules replaced value at risk with expected shortfall for market risk capital after the 2008 crisis.
What It Is
Expected tail loss, often written ETL, answers the question value at risk leaves open. Value at risk tells you the threshold loss you would not expect to breach at, say, 97.5 percent confidence. It is silent on how bad things get once that threshold is crossed.
Expected tail loss fills the gap by averaging every loss in the tail beyond the threshold. It carries several interchangeable names, including expected shortfall, conditional value at risk, average value at risk, and mean shortfall. They describe the same underlying quantity.
The Intuition
A burglar alarm that only tells you someone crossed the fence, but not whether they took a bicycle or emptied the house, is half a tool. Value at risk is that alarm. It marks the line of the bad zone without describing the damage inside it.
Expected tail loss describes the damage. By averaging the worst outcomes, it reflects whether the tail is shallow or catastrophic. This matters because two portfolios with identical value at risk can hide very different tail severity, and the deep tail is what wipes out capital. After the 2008 crisis exposed how badly value at risk understated extreme losses, the Basel framework moved trading book capital onto expected shortfall for exactly this reason.
How Expected Tail Loss Works
Expected tail loss is the conditional average of losses beyond the value at risk level. If alpha is the tail probability, the calculation averages all outcomes in that worst alpha slice:
ETL_alpha = average of losses worse than VaR at the (1 - alpha) confidence level
Equivalently, it is the expected loss conditional on exceeding the threshold:
ETL = E[ Loss | Loss > VaR ]
In practice with simulated or historical returns, you rank outcomes from worst to best, isolate the worst alpha fraction, and average them. Because it integrates the whole tail, ETL reflects the shape of extreme losses, not a single quantile.
Expected tail loss is a coherent risk measure. It satisfies subadditivity, so the risk of a combined portfolio never exceeds the sum of its parts. Value at risk can break this rule and punish diversification, which is one technical reason regulators preferred expected shortfall.
Worked Example
A trading desk simulates 10,000 daily profit-and-loss outcomes and wants the 97.5 percent expected tail loss, the standard confidence level under the Basel trading book rules. The worst 2.5 percent of outcomes is the worst 250 scenarios.
The 97.5 percent value at risk is the loss at the 250th-worst scenario, which works out to 2,000,000 dollars. That is the threshold. To get the expected tail loss, average the losses across all 250 of those worst scenarios.
Suppose those 250 losses sum to 750,000,000 dollars.
ETL = 750,000,000 / 250 = 3,000,000 dollars
The expected tail loss is 3,000,000 dollars, 50 percent larger than the 2,000,000 dollar value at risk. Capital sized to the value at risk alone would leave the desk short by 1,000,000 dollars per average tail day. Sizing to expected tail loss reserves against the severity of the crash, not merely its entry point.
Common Mistakes
- Assuming normal returns. A normal distribution has thin tails and understates extreme losses. Real markets are fat-tailed, so a normal-based ETL is too low.
- Treating it as identical to VaR. Value at risk is the threshold; expected tail loss is the average beyond it. ETL is always at least as large.
- Using too small a tail sample. The tail holds the fewest observations. With few scenarios, the ETL estimate is noisy and often understated.
- Confusing the many names. Expected shortfall, conditional value at risk, and average value at risk are the same measure. Mixing them up causes needless confusion.
- Omitting the confidence level. A 95 percent ETL differs from a 97.5 percent ETL. Always state the level so the figure is comparable.
Frequently Asked Questions
What is expected tail loss in simple terms? Expected tail loss is the average size of a portfolio's worst losses, looking only at outcomes beyond the value at risk threshold. It tells you how deep losses go once a bad event occurs.
How does expected tail loss affect investment decisions? It sizes capital and reserves against severe events that value at risk underweights. Because it captures tail depth, it guides hedging and position limits better than a single threshold figure.
What is a real-world example of expected tail loss? A trading desk with a 97.5 percent value at risk of 2,000,000 dollars finds its worst 250 scenarios average a 3,000,000 dollar loss. That expected tail loss is 50 percent larger than the threshold.
How can investors use expected tail loss effectively? Estimate it with a fat-tailed model across many scenarios, state the confidence level, and reserve against the ETL rather than the value at risk. This protects against the deep losses that destroy capital.
How is expected tail loss different from value at risk? Value at risk reports a single threshold loss and ignores anything worse. Expected tail loss averages all the losses beyond that threshold, capturing the severity of the tail rather than just its edge.
Sources
- AnalystPrep. "Describe Extensions of VaR." https://analystprep.com/study-notes/cfa-level-2/describe-extensions-of-var/
- AnalystPrep. "Fundamental Review of the Trading Book (FRTB)." https://analystprep.com/study-notes/frm/part-2/operational-and-integrated-risk-management/fundamental-review-of-the-trading-book-frtb/
- CFA Institute. "Measuring and Managing Market Risk." https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2026/measuring-managing-market-risk
- Corporate Finance Institute. "Value at Risk (VaR)." https://corporatefinanceinstitute.com/resources/career-map/sell-side/risk-management/value-at-risk-var/
Disclaimer
This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.