Portfolio and risk · Managing a portfolio
Tail-risk protection: insuring a portfolio against a crash
How crash insurance with far out-of-the-money puts works, what a tail-risk fund did from 2017 to 2026, and how it compares with holding cash or bonds.
On this page
Tail-risk protection is insurance against a severe market decline. The "tail" is the far end of the distribution of returns, where the rare, very large losses sit. Fat tails: why extreme market moves are more common than models say shows that such losses happen far more often than a bell curve predicts. A tail hedge is a position designed to lose a little in most periods and gain a great deal in exactly those periods, so that the portfolio as a whole falls less.
The idea is attractive and the instruments exist. The difficulty is the price. Insurance that pays in a crash is something many investors want, so it is rarely cheap, and it must be paid for through all the years in which no crash comes. This lesson explains how the protection works and measures a real tail-risk fund against simpler alternatives.
The idea
Ordinary insurance works like this: pay a small premium every year, and collect a large sum if the house burns down. Most years the premium is simply lost. Nobody regards that as a bad investment, because the purpose is not profit.
A tail hedge applies the same structure to a portfolio. A put option with a strike price 20% or 30% below the market costs very little, because the market rarely falls that far before the option expires. If it does fall that far and quickly, the option's value can multiply many times. A few percent of a portfolio in such options can offset a meaningful part of the loss on the rest.
The figure shows the two faces of the trade. The tail-risk fund alone is a line that slopes steadily down, interrupted by short jumps. The jumps are what it is for. The slope is what they cost.
How it works
A typical tail-risk fund holds most of its money in Treasury bonds and spends a small, regular amount on put options on a stock index, with strikes below the market. Three things determine the result.
| Feature | Effect |
|---|---|
| How far below the market | Further is cheaper, and pays only in a deeper fall |
| Time to expiry | Longer costs more and decays more slowly; must be rolled before expiry |
| Speed of the decline | A fast fall raises implied volatility and the puts' value; a slow one lets them expire |
The third row is the one most often overlooked. An option is a bet on a fall by a date. A 30% decline spread over a year can pass beneath a series of three-month puts without any of them paying much, because each is replaced at a new, lower strike before the market reaches the old one.
A worked example: nine years of a tail-risk fund
The data: monthly total returns of SPY, TAIL and AGG, June 2017 to October 2026 (113 months). Mixes are rebalanced monthly; cash earns the Treasury bill rate.
| Portfolio | Return a year | Volatility | Deepest fall | Oct.–Dec. 2018 | Feb.–Mar. 2020 | 2022 | $100,000 became |
|---|---|---|---|---|---|---|---|
| 100% SPY | 14.7% | 15.7% | −23.9% | −13.5% | −19.4% | −18.2% | $362,635 |
| 100% TAIL | −7.4% | 11.2% | −53.0% | +14.8% | +20.3% | −13.1% | $48,667 |
| 90% SPY / 10% TAIL | 12.6% | 13.4% | −21.9% | −10.8% | −15.8% | −17.2% | $304,885 |
| 90% SPY / 10% cash | 13.5% | 14.2% | −21.6% | −12.2% | −17.6% | −16.1% | $330,168 |
| 60% SPY / 40% AGG | 9.5% | 10.5% | −20.0% | −7.5% | −11.5% | −15.8% | $234,515 |
In the two fast declines it worked. In late 2018 the fund gained 14.8% while stocks lost 13.5%, and in early 2020 it gained 20.3% while stocks lost 19.4%. The 90/10 mix with the fund lost 1.4 and 1.8 points less than the 90/10 mix with cash.
In the slow decline it did not. Through 2022 the fund lost 13.1%. Stocks fell gradually over nine months, so the puts paid little, and rising interest rates cut the value of the Treasury bonds that make up most of the fund. The mix with cash lost less.
Over the whole period, cash did better. Ten percent in the tail fund cost 2.1 points of return a year against all stocks; ten percent in cash cost 1.2. The deepest falls were almost the same, 21.9% and 21.6%. The insurance paid out twice and still cost more than it returned.
The fund's monthly record explains why.
| Measure (TAIL, 113 months) | Value |
|---|---|
| Total return | −51.3% |
| Months with a gain | 38 of 113 |
| Best month | +11.4% (March 2020) |
| Worst month | −8.9% |
| Correlation with SPY | −0.70 |
| Months SPY fell; months of those in which TAIL rose | 36; 23 |
| Average TAIL return in the 11 months SPY fell more than 5% | +5.0% |
| Average SPY return in those 11 months | −7.8% |
A correlation of −0.70 is unusually negative, and the fund did rise when stocks fell hard. It gained about two-thirds of a point for each point stocks lost in the worst months. The problem is the other 100 or so months.
By calendar year the pattern is plain: the fund gained 2.9% in 2018, 6.9% in 2020 and 5.5% in 2025, and lost between 9.6% and 14.3% in each of 2019, 2021, 2022, 2023 and 2024.
Using it
- Compare with the plain alternatives first. Holding 10% less in stocks protects in every kind of decline, fast or slow, and costs only the return forgone. A tail hedge has to beat that. Over 2017 to 2026 this one did not.
- Know what it buys. The real benefit of a tail hedge is cash at the bottom of a crash, when everything else is down. That cash is useful only if it is used: rebalanced into stocks at low prices. A hedge that is held and never harvested gives its gains back as the market recovers.
- Size it as a cost. Advocates typically suggest spending a small, fixed share of the portfolio each year, on the understanding that it will usually be lost.
- Consider the behavioural case. Some investors can hold more stocks, and hold them through a crash, only because they know a hedge is in place. If protection that loses money on average prevents selling at the bottom, it may still pay for itself. That is a judgement about the investor, not about the hedge.
- Watch the price of insurance. Option prices rise with implied volatility. After a fall, when demand for protection is highest, the same put can cost several times what it did a month before. Volatility and the VIX: historical vs implied volatility explained explains the gauge.
Where it fails or misleads
- Insurance sellers are paid on average. Research by Ilmanen and others finds that buying index puts has had a negative expected return over long periods: buyers overpay, on average, for crash protection. A tail hedge is expected to cost money.
- Slow bear markets slip underneath. 2022 here; much of 2000 to 2002 for longer-dated history. The hedge is for speed as much as for depth.
- The other holdings in a tail fund carry their own risk. A fund that keeps its collateral in intermediate or long Treasuries is also a bond fund, which hurt in 2022.
- Timing the hedge defeats it. Buying protection after a scare means paying the highest price for it, and dropping it after years of losses means being uninsured when the event arrives. It works only as a permanent position.
- One fund over nine years is a small sample. Other tail strategies are built differently, and a period containing a crash on the scale of 1987 or 2008 could look very different.
- Naming the tail in advance is hard. A hedge built for a stock crash does not cover an inflation shock or a bond-market rout.
Compared with its neighbours
| Approach | Cost in normal years | In a sudden crash | In a slow bear market |
|---|---|---|---|
| Hold fewer stocks | Return forgone | Proportionally smaller loss | Proportionally smaller loss |
| High-quality bonds | Lower expected return | Often gain | Depends on interest rates |
| Protective put (near the market) | High premium | Floor under the loss | Floor, if renewed in time |
| Tail hedge (far puts) | Small, steady loss | Large gain | Little or none |
| Stop-loss orders | None until triggered | May sell far below the stop | Sells on the way down |
Hedging a portfolio: four ways to reduce risk and what each costs compares the first three in detail; Protective put: insuring a stock position with a put option covers the option itself.
How Finabase does it
Finabase does not model tail hedges, price a hedging programme or show a portfolio's expected loss in a crash.
The pieces can be looked at directly. The comparison view overlays SPY and a tail-risk fund as percent changes over ranges up to two years, which shows the opposite movement and the steady decline. The option chain for an index fund lists puts at every strike and expiry with their prices and implied volatility, so the current cost of a put 20% below the market can be read off and compared with one 5% below. The Indices page shows the VIX, the market's gauge of what index options cost. The paper-trading account can hold a small position in a tail fund next to a stock fund to watch how the two behave together.
Frequently asked questions
What is tail risk?
The risk of a rare, very large loss: an outcome in the far end, or tail, of the distribution of returns. Stock markets produce such losses more often than standard models assume.
Do tail-risk funds make money?
Not on their own over typical periods. They are designed to lose slowly and gain sharply in a crash. The one examined here lost about half its value over nine years that included two sharp declines.
Is a tail hedge better than holding cash?
Over June 2017 to October 2026, no: a 90/10 mix with cash earned more and fell about as far. A tail hedge gains more than cash in a sudden crash and costs more the rest of the time.
Why not buy protection only when a crash looks likely?
Because crashes are not reliably predictable, and protection becomes expensive precisely when many people think one is likely.
How much of a portfolio do investors put into tail hedges?
Those who use them generally commit a small percentage, treated as an annual expense. The rest of the portfolio is often held more aggressively than it would be without the hedge.
Summary
Tail-risk protection trades a small, regular loss for a large gain in a crash. A tail-risk fund from 2017 to 2026 did what it promised in two fast declines, gaining 14.8% and 20.3% while stocks fell, and still lost half its value overall. A portfolio with 10% in it earned less than one with 10% in cash and fell about as far, and in the slow decline of 2022 the hedge lost money too. It is insurance against speed. For most portfolios the first defence remains holding an amount of stock whose worst loss can be lived with.
Check your understanding
4 questions on this lesson. Each answer comes with the reason, and nothing is scored.
What is the basic payoff pattern of a tail hedge?
- Large steady gains with rare small losses
- Small steady losses with rare large gains
- The same return as the stock market with less volatility
Show the answer
B. Small steady losses with rare large gains It is structured like insurance: a premium lost in most periods, and a large payoff in the rare periods of severe decline.
The tail-risk fund gained in the fast declines of late 2018 and early 2020 and lost 13.1% in 2022. Why?
- Options pay for a fall by a date, so a gradual decline lets them expire with little value, and the fund’s bonds also fell as rates rose
- The fund stopped buying options in 2022
- Stocks rose in 2022
Show the answer
A. Options pay for a fall by a date, so a gradual decline lets them expire with little value, and the fund’s bonds also fell as rates rose A slow decline passes beneath a series of short-dated puts that are each rolled to lower strikes. The Treasury collateral also lost value when interest rates rose.
Over 2017 to 2026, how did 90% SPY with 10% in the tail fund compare with 90% SPY and 10% cash?
- The tail mix earned more and fell far less
- The cash mix earned more and its deepest fall was about the same
- They were identical
Show the answer
B. The cash mix earned more and its deepest fall was about the same The cash mix returned 13.5% a year with a deepest fall of 21.6%; the tail mix 12.6% with 21.9%. The simple alternative is the benchmark any hedge must beat.
When is the gain from a tail hedge actually captured?
- When the hedge’s gains are used to buy other assets at depressed prices
- Automatically, whatever the investor does
- Only when the hedge is sold before the crash
Show the answer
A. When the hedge’s gains are used to buy other assets at depressed prices The hedge supplies cash at the low. If it is simply held, its value falls back as markets recover and the benefit is lost.
Sources and further reading
- Nassim Nicholas Taleb, The Black Swan: The Impact of the Highly Improbable, 2nd ed. (Random House, 2010)
- Antti Ilmanen, “Do Financial Markets Reward Buying or Selling Insurance and Lottery Tickets?”, Financial Analysts Journal, vol. 68, no. 5 (2012)
- Roni Israelov, “Pathetic Protection: The Elusive Benefits of Protective Puts,” Journal of Alternative Investments, vol. 21, no. 3 (2019)
- Vineer Bhansali, Tail Risk Hedging: Creating Robust Portfolios for Volatile Markets (McGraw-Hill, 2014)
- Finabase chart data — monthly bars and dividends of SPY, TAIL and AGG, June 2017 to October 2026; 3-Month Treasury Bill rate (TB3MS) via FRED; statistics computed by Finabase
