Systematic Tail Hedge Overlay

Overview

Every other volatility strategy on this list is, in some form, selling insurance. This one buys it. And it is here for a reason: if you run any of the short-volatility strategies, this is the thing that lets you survive the day they all go wrong at once.

The idea is deliberately simple. Take a small, fixed percentage of your portfolio every year, call it 1 percent, and spend it on far out of the money index puts. In a normal year, those puts expire worthless and you are down 1 percent. In the year the market falls 35 percent, those puts, bought when they were nearly free, are suddenly worth many multiples of what you paid, because a deep out of the money option's value explodes non-linearly in a crash.

That non-linearity is called convexity, and it is the entire product. You are not trying to offset your losses one for one. You are trying to own something whose payoff accelerates precisely when everything else is collapsing.

Why bother, when the market goes up over time?

This is the right question, and the answer is about compounding rather than about averages.

Losses do not hurt symmetrically. A portfolio that falls 50 percent needs a 100 percent gain to get back to even. A portfolio that avoids the worst of a crash, and has cash to buy at the bottom, compounds from a much higher base. The tail hedge is not primarily about feeling safe. It is about protecting the compounding.

The second reason is capacity. A hedged investor can hold more risk in the rest of the book. If your tail is protected, you do not have to hold cash you did not want to hold, and you do not have to sell equities at the bottom because your risk limits forced you to. The hedge buys you the ability to stay invested.

Strategy logic

• Budget: Fix it as a percentage of the portfolio, annually. This is the discipline that makes the strategy work. It is a cost of doing business, like an insurance premium, and it is not a trade you are trying to win.
• Purchase schedule: Buy a slice every month. Never try to time when protection is cheap, because it is expensive exactly when you want it most, and if you wait for a good entry you will not have a hedge on when the shock arrives.
• Strike: Far out of the money, typically around a 5 delta put. Use delta rather than a fixed percentage so the strike automatically moves further out when volatility is high and closer when it is low.
• Tenor: Two to three months. Very short-dated puts are cheap but decay to nothing before a slow-developing crisis matures. Long-dated puts have less convexity per dollar.
• Monetisation: This is the rule most people never write down, and it is the one that determines whether the strategy actually works.

The monetisation problem

A tail hedge that is never sold is a tail hedge that never paid you anything.

In March 2020, plenty of people owned puts that went up thirty or fifty times. Many of them held all the way through, watched the market recover, and their options expired worthless anyway. They correctly predicted the crash, correctly owned the hedge, and made nothing.

So write the rule in advance:

• Trigger: Sell when the hedge reaches some multiple of its cost, or when VIX crosses a threshold, or when the index breaches a drawdown level. Any defensible rule beats no rule.
• Redeploy: The proceeds go straight back into equities at the lower price. This is not optional, it is the actual mechanism by which the strategy adds return. The hedge payoff is not the point. The point is that the payoff lets you buy the dip with real money at the moment when everyone else is selling.

Without a monetisation and redeployment rule, you have bought an expensive lottery ticket and forgotten to cash it.

Where it works and where it hurts

It works in a crash, obviously. It works in 1987, in 2008, in March 2020, and in any sharp, fast dislocation where volatility explodes and skew steepens at the same time.

It hurts everywhere else, and it hurts constantly. Expect the hedge leg to lose money in something like eight or nine years out of ten. Expect flat, quiet, grinding bull markets to make you look and feel foolish. Expect to be asked, repeatedly, why you are paying for insurance against a fire that never comes.

The bleed is not a bug. The bleed is the premium. But knowing that intellectually does not make it easy, and the historical failure rate of tail hedging programs is overwhelmingly behavioural rather than analytical. People cancel the policy in year four and the fire is in year five.

It also works poorly against a slow grind down. A market that drifts lower by 25 percent over eighteen months without a volatility spike will not pay your far out of the money puts much at all. This hedge is calibrated for the fast crash, not the long bear.

Backtest design checklist

• Never backtest the hedge in isolation. Standalone, it shows a horrible return and a horrible Sharpe ratio and looks like the worst strategy ever devised. It is only meaningful as an overlay on a portfolio.
• Report the combined compound growth rate, not just the drawdown reduction. The real question is whether protecting the compounding beat the cost of the premium over the full cycle, and honest answers to this are closer than the marketing suggests.
• Model the monetisation rule explicitly, and test how sensitive results are to it. They are extremely sensitive. This is where most of the variance in outcomes lives.
• Use real option prices with real spreads. Far out of the money options have terrible bid-ask spreads, and you are crossing that spread every single month.
• Test the slow-bear scenario, not just the fast crash, so you see the case where the hedge does not save you.

Common failure modes

• Cancelling the hedge during the calm. The single most common failure, and it is almost universal.
• No monetisation rule. Correctly owning the hedge into a crash and then never selling it.
• Trying to time the purchase. Waiting for cheap volatility means being unhedged when it matters.
• Over-hedging. Spending 4 percent a year instead of 1 percent turns a sensible insurance policy into a drag so large that it destroys the compounding it was meant to protect.
• Expecting it to pay in every downturn. It is built for the violent one. It will disappoint you in a slow one.

Our notes and suggestions

We think this belongs in the toolkit of anyone running short-volatility strategies, and honestly of anyone running a levered or concentrated equity book. If you are selling puts, selling variance, or shorting VIX futures, you are implicitly short the crash. A tail hedge is how you convert an unsurvivable exposure into a survivable one, and paying for it out of the premium you are collecting is a natural and coherent way to fund it.

Be honest about the trade-off. Over a long enough horizon, a pure buy-and-hold investor with the discipline to never sell may well beat a tail-hedged one, because the premium compounds against you. The hedge earns its keep for people who are levered, who have liabilities, who have investors that redeem, or who simply know they will not have the nerve to hold through a 50 percent drawdown. That last group is much larger than the number of people who admit to being in it.

What would change our mind: a structural fall in the cost of far out of the money protection, which would make the hedge cheap enough to be an obvious buy, or evidence that our own book is no longer implicitly short volatility.