Protective puts and tail hedges, measured

What a protective put is

You own shares, or an index fund, and buy a put option below the current price. If the price falls through the strike, the put gains what the shares lose below it, so your loss is capped at the distance to the strike plus what you paid. If nothing happens, the put expires and the premium is gone. It is insurance with a deductible and a monthly bill.

A tail hedge is the same idea pushed further out: a cheap put far below the market, around 10 delta, meant to pay only in a crash. It costs little each month and is supposed to pay a lot rarely. The question is whether the rare payouts cover the monthly bills.

What the history says

We replayed buying a roughly 10-delta put 30 to 45 days out, at the real ask, held to expiry, on about 100 entry days per ticker from April 2022 to 2026. That window includes the 2022 bear market. On QQQ the put paid off 5% of the time and lost about 63% of its cost per trade on average. On SPY it paid off 8% of the time and lost about 35% a trade. On a 3x semiconductor fund it paid off 9% of the time and lost about 57%.

The puts did pay in 2022, which is the point of owning them. Across the whole window the payouts did not cover the premiums. That is the same variance risk premium that put sellers collect, seen from the buyer's side: index options were priced for more movement than showed up.

Buying it when volatility looks cheap made it worse

The usual advice is to buy protection when it is cheap, when implied volatility is low relative to its own history. We tested that directly. On QQQ, puts bought when implied volatility sat more than one standard deviation below its 60-day average never paid off in the sample and lost essentially their whole cost, and the same held on QQQ and SPY when implied volatility was below recent realized volatility. Buying when the VIX was unusually low lost a little less than buying every month, and still lost. Only the 3x Nasdaq fund showed a positive cheap-volatility cell, on 24 trades, which is too few to lean on.

Cheap protection tends to be cheap because the market is calm, and calm stretches tend to continue for a while before they end. Timing the purchase on cheapness bought more of the expiring months and few of the paying ones.

Cheaper ways to limit a drawdown

If the goal is to keep a bad year survivable rather than to profit from a crash, the measured levers were simpler. Holding less of the risky asset lowers the drawdown one for one and costs no premium. For option sellers, the width of a put credit spread is a protective put bought at the same time as the sale, and it caps the loss on every trade at a known number.

In our ten-year spread replays, the fraction of the account put at risk decided the drawdown odds far more than any hedge did: at about 10% of equity per spread a 20% drawdown within a year did not occur, and at 30% it occurred in about 38% of years. Our drawdown lesson has the full table.

When a protective put still makes sense

Insurance is worth paying for when the loss it prevents is one you cannot afford, not when it is expected to make money. A concentrated position you cannot sell for tax or employment reasons, a known event you cannot sit out, or a short window where a drawdown would force a sale are all reasons to pay the premium knowingly. Paying it every month as a standing strategy, on our sample, was a steady cost.