Short strangles, measured

What a short strangle is

You sell an out-of-the-money put below the market and an out-of-the-money call above it, on the same underlying and the same expiry. You collect two premiums and keep both if the price finishes between the strikes. If it falls through the put strike or rises through the call strike, you lose the move beyond that strike, less the premium.

The put side behaves like a cash-secured put. The call side is a naked call, and a naked call has no cap: a stock can rise further than it can fall. That is the whole trade in two lines, and most arguments about strangles are arguments about how often the uncapped side matters.

Strangle vs straddle

A straddle sells the put and the call at the same strike, at the money. It collects the most premium, and one leg is in the money as soon as the price moves at all, so the trade keeps a profit only while the move at expiry stays smaller than the premium collected. It is a bet on how far the price ends from today. A strangle moves both strikes out, collects less, and wins on a range instead of a point.

Buying either one is a long volatility bet. In our replays, buying the at-the-money straddle 30 to 45 days out lost on every ticker we tried, about 8% of its cost a trade on QQQ and about 25% on SPY. That loss is the premium sellers collect: options on these funds were priced for more movement than showed up. It does not follow that selling them is free, because the seller also carries the tail.

What the history says

We replayed a 25-delta put plus a 25-delta call, 30 to 45 days to expiry, held to expiry, on about 100 evenly spaced entry days per ticker from April 2022 to 2026. Short legs were filled at the real bid, with fees, and settled on the closing price. Collateral is booked as the put strike, the cash a broker would hold for the put alone.

On QQQ the strangle won 72% of the time and earned about +0.2% a trade on that collateral. The worst trade lost about 21% of it. On SPY it won 80% of the time, earned about +0.9% a trade, and the worst trade lost about 11%. Entered below the 200-day moving average, the QQQ strangle won only 54% of the time and lost about 1.6% a trade.

On a 3x semiconductor fund the strangle won 62% of the time and lost money on average, and its worst trade lost more than three times the cash set aside for the put. A put cannot lose more than its strike, so that loss came from the call side in a rally. That is the uncapped leg doing what it does on a fund built to move three times its index.

Why the call side carries the risk

Index funds have drifted up over most of the last decade, and calls are priced cheaper than equally distant puts because investors pay up for crash protection. So the call leg collects less than the put leg for the same distance, and the price is more likely to drift toward it. Over ten years of SPY and QQQ, selling the call side as a spread lost about 7% a trade on the dollars at risk, which our call credit spread guide covers.

The strangle hides this because the put premium is usually large enough to carry it. Look at the two legs separately before you decide the combined number is an edge. In our samples most of the strangle's return was the put, and most of its worst days were the call.

Capital, sizing and the defined-risk version

A strangle on margin ties up far less than the put strike, which is why its return on margin looks large. The margin is not the risk. The risk is a gap through either strike, and on a strangle there are two directions for it to come from. Size on the loss you would take on a gap, not on the margin the broker asks for.

Buying a wing beyond each short strike turns the strangle into an iron condor: less premium, a known worst case, and far less capital per position, which lets a small account spread its risk over more expiries instead of one large bet. If you only want the side that has held up, the cash-secured put or the put credit spread keeps the put premium and drops the uncapped leg.