On This Page
Key Takeaways
- Tail risk concerns extreme outcomes and the losses a position produces in those scenarios.
- Value at Risk is the loss threshold for one horizon: at 99%, the model gives a worse loss a 1% chance.
- Expected shortfall is the average loss once that threshold breaks, so it shows how deep the tail runs.
- The worked normal model gives one-day 99% VaR of $27,916 and expected shortfall of $31,983.
- A matched protective put sets the floor in advance: 100 shares at $100 with a 90-strike put at $2 lose at most $1,200 at expiration before fees.
What is tail risk?
Tail risk is exposure to extreme outcomes at the far ends of a return distribution, such as a sudden market drop far larger than a typical day’s move. For a trader, the key calculation connects each extreme outcome with the entire position’s profit or loss, not just the percentage move in the market.
Use the contract quantities, entry premiums and payoff structure. This turns a percentage move into a concrete dollar result.
How is tail risk measured with VaR and expected shortfall?
Value at Risk is the loss threshold for one horizon: at 99%, the model gives a worse loss a 1% chance. Expected shortfall is the average loss in that worst 1%, so it measures how deep the tail runs.
The worked position is $1 million. Its one-day return model has mean zero and normal standard deviation of 1.2%, giving a dollar standard deviation of $12,000. The 99% VaR is $27,916, and the expected shortfall for the worst 1% is $31,983.
A separate −4.8% return scenario produces a $48,000 loss, far past both model numbers. VaR marks where the tail starts, expected shortfall gives its average depth, and the scenario prices one specific extreme move.
Formula
VaR(c) = c-quantile of loss L; ES(c) = [1 / (1 − c)] × integral from c to 1 of VaR(u) du.- c is the confidence level and 1 − c is the share of outcomes in the loss tail: 1% at 99%.
- For a continuous loss distribution, expected shortfall equals average loss beyond the VaR threshold.
- In the normal model, expected shortfall = standard deviation × φ(z) / (1 − c), where φ(z) is the normal density at the VaR quantile z.
- Use the same position, horizon and loss distribution for both measures.
Worked Example
Model the one-day simple return on a $1,000,000 equity-index position as normal, with mean zero and standard deviation 1.2%. Compare the model’s 99% loss quantile and expected shortfall with a separate −4.8% return scenario.
- Dollar standard deviation = 1.2% × $1,000,000 = $12,000
- 99% normal quantile z = 2.326348
- 99% VaR = $12,000 × 2.326348 = $27,916
- 99% expected shortfall = $12,000 × φ(2.326348) / 0.01 = $31,983
- Separate −4.8% return scenario = $48,000 loss
The model’s 99% threshold is $27,916, and its average loss within the worst 1% is $31,983. The separate −4.8% scenario loses $48,000 and lands past both model numbers. Run that scenario on the full position before you size it.
Tail risk hedging with a protective put
Consider 100 shares bought at $100 and one matching 90-strike protective put bought at $2 per share. The combined entry cost is $10,200.
At expiration, stock plus put is worth at least $9,000. At stock $80, the shares are worth $8,000 and the put pays $1,000. At stock $60, the shares are worth $6,000 and the put pays $3,000. Both outcomes produce the same $1,200 loss before fees.
The strike, expiration, underlying and quantity define that matched payoff. Compare the entry premium and the floor across the strikes in your plan.
How to size a position for tail risk
Check volatility skew and options flow in the expiration you plan to trade, then evaluate the full position across the price scenarios that matter in the options profit calculator.
For a defined-loss structure, calculate the loss per unit and trade the whole-number quantity that fits the dollar allocation. Recalculate the complete payoff when changing a leg or rolling an expiration.
Limitations of tail risk measures
- Size against the tail beyond VaR. The $27,916 threshold marks where the worst 1% begins, and the average loss inside that 1% is the $31,983 expected shortfall.
- Stress the position past the model. Both measures come from the assumed distribution, horizon and position: the worked normal model puts one-day 99% VaR at $27,916, yet the separate −4.8% scenario loses $48,000, a move that a normal model treats as extremely rare.
- Run explicit scenarios such as the −4.8% day beside any historical tail estimate. Extreme days are rare, so an estimate built from history rests on few observations.
- Mark the hedge to market before expiration. The protective put floor holds at expiration for the matched share quantity; before then, the position’s value follows the put’s market price, and fees come on top of the $1,200 maximum loss.
Bottom line
Tail risk is the exposure a position carries to rare, extreme moves, and it is best stated in dollars for the whole position. Use VaR for the threshold, expected shortfall for the average loss beyond it, explicit scenarios for moves a model calls unlikely, and a defined-loss structure such as a protective put when the floor must be known in advance.
Clear answers
Frequently Asked Questions
Sources and Further Reading
- CFA Institute: Measuring and Managing Market Risk (Refresher Reading)
- CFA Institute Digest: Volatility versus Tail Risk, Which One Is Compensated in Equity Funds
- Fidelity: What Is Hedging
How entries are sourced, calculated and updated: editorial methodology.