Gap Risk: The Overnight Move No Buffer Absorbs

The assumption hiding inside every buffer

A short put with a strike 8% below the current price feels like it has 8% of room. That framing carries an assumption so ordinary it is rarely stated: that if the underlying travels those 8%, it will do so through the intervening prices, in market hours, while you are able to watch.

Most of the time this holds. Stocks grind. They fall 2% on Tuesday and 3% on Thursday and you have time to think.

Gaps are the cases where it does not hold. The stock closes at $100 and opens at $84. There was no $92, no $88, no sequence of worsening prints. The buffer was not consumed; it was skipped. Everything a seller might have done at $95 was unavailable, because $95 never traded.

This matters more for put sellers than for almost anyone else, because the position is short an option whose value explodes precisely in the scenario where you cannot act.

Where gaps come from

Roughly two-thirds of the trading day happens when you cannot trade. A stock trades for six and a half hours and then carries risk for seventeen and a half, plus weekends. Information does not respect that schedule.

Earnings. The largest and most predictable source. Companies report outside market hours specifically so the information can be absorbed before continuous trading resumes, which is exactly the mechanism that produces a gap. A stock that routinely moves 8% on earnings is telling you, in advance, that a single overnight is capable of consuming a buffer built for ordinary weeks.

Regulatory and clinical events. Trial results, approvals, rejections, investigations. These are binary and they are announced when they are ready.

Deal news. An acquisition gaps a stock up, which is harmless to a put seller. A collapsed deal gaps it down, which is not.

Guidance and pre-announcements. A company that revises its outlook eleven days before the scheduled report produces the earnings gap without the earnings date, which means it produces it on a day nobody had marked.

Overnight macro. Rate decisions, geopolitical events, a foreign session that reprices an entire sector before the US open. These gap everything at once, which is the case that also defeats diversification.

Why the usual protections do not apply

The instinct is to reach for a stop. It does not work here, and the reason is worth being precise about.

A stop-loss order is an instruction to sell once a price is touched. In a gap, that price is never touched — the market opens below it. The order becomes a market order at the open, and it fills at whatever the opening auction produces. A stop placed at $95 on a stock that opens at $84 executes at $84 or worse. It did not limit the loss; it converted an open position into a closed one at the bottom of the move.

Nor does monitoring help. There is no vigilance that reacts to a price that did not print. The move is complete before you can respond, and this is true whether you are asleep, watching the pre-market, or sitting at the terminal at 09:29.

The only protections that function are the ones chosen before the position exists: which underlyings, how far from the money, over what dates, and at what size.

What the buffer is really telling you

A safety buffer, honestly read, is a statement about ordinary conditions. It says: for the underlying to reach this strike through normal trading, it has to fall by this much, and here is how often it has done that historically.

It is not a barrier. Nothing about a strike sitting 8% below the market prevents the stock trading at $84. The buffer measures distance, not protection, and the difference between those two ideas is the entire subject of this article.

Which is why a buffer derived from how an underlying has actually behaved is worth more than one derived from a pricing model. The model that produces delta assumes prices move continuously and returns are normally distributed. Both assumptions are approximately true on ordinary days and false on exactly the days that matter. A distribution built from observed behaviour has the gaps already in it, because they happened.

The earnings case, which is the whole problem in miniature

Selling a put that spans an earnings announcement is the clearest version of this trade-off, and it is not automatically a mistake.

The premium is elevated ahead of earnings because the market prices the expected move. You are being paid more, and you are being paid more for a specific, identifiable reason. That is a fair exchange rather than a trap.

What makes it dangerous is misreading which risk you have taken. The elevated premium compensates for the move the market expects. It does not compensate for the move nobody expects, and the distribution of earnings reactions has a long left tail — the occasional 25% single-day decline on a guidance cut that no implied move contemplated.

So the question is not whether to sell through earnings. It is whether the strike sits outside what this underlying has actually done through prior announcements, rather than outside what the implied move suggests. Those two distances are frequently different, and the second is systematically the more generous of the two.

Why the premium is the wrong thing to weigh it against

A gap cannot be managed after the fact, so everything that determines the outcome is fixed at entry. That makes the relevant comparison awkward, because the two quantities involved are on wildly different scales.

The premium on a short put is measured in cents per share. A gap is measured in tens of percent of the underlying. Against a 25% overnight decline the credit is a rounding error, and the arithmetic in why a 90% win rate can still lose money shows what happens to expectancy when the rare outcome is deep enough — the win rate barely moves, and the ledger moves a lot.

Two further properties of gap risk are worth naming, because both are easy to misread. Tails are not uniform across underlyings: a utility and a small-cap biotech can quote similar premium and carry entirely different left-tail behaviour, so premium is not a proxy for gap exposure. And several names exposed to the same catalyst behave as one exposure rather than several, which is the subject of sector concentration.

How StratosIQ handles it

Gap exposure is what the safety floor exists to price. ShieldIQ measures each strike against how the underlying has actually behaved rather than against the distance a pricing model implies — for Strike, against observed earnings reactions; for Patrol, against multi-day return distributions. Both are built from moves that happened, so the gaps are in the data rather than assumed away.

The result is reported as a status: Fortified, Secure, Tight or Exposed. A contract that fails the safety floor does not publish, whatever its premium looks like. That is a structural gate rather than a factor a strong showing elsewhere can offset.

The two engines take opposite positions on earnings deliberately. Strike works relative to the announcement, requiring the contract to expire after it, and prices the event as the thing being paid for. Patrol excludes contracts whose expiration spans an announcement, so the steady-state book is not carrying event risk it was not compensated for. Inverting the earnings condition is the reason there are two products rather than one.

The full published record, contract by contract and including every loss, is on the performance page.

Worth reading next

Assignment covers what happens after a gap takes a strike, and why the outcome belongs in your records as a loss. Short puts in a market-wide drawdown looks at the case where every position gaps on the same morning.