Opening Gap: A Gap in the Chart, Not Trading
An opening gap is a session opening at a price outside the previous session's range, leaving untraded space on the chart. Nothing failed to trade in that space because the market was closed, which is why a gap is a feature of the chart rather than of the market.
How it works
A market closes at one price and opens at another. The band between the previous close and the new open contains no bars, because no bars could form while the exchange was shut.
That is the single most useful sentence about gaps and the one most often skipped. Nothing failed to trade. There was no trading to fail. Opinions changed while the market was closed, and the first price available in the morning reflects the new consensus.
Which immediately disposes of the idea that a gap is an “imbalance to be corrected.” The imbalance concept describes price moving too fast during trading. A gap describes the market being shut. Different events, frequently conflated.
The two kinds, and telling them apart
A gap on real news is a repricing. Earnings, a regulatory decision, an acquisition — the world changed and the old price is simply wrong now. There is no reason for price to return to a level that reflected the previous state of knowledge.
A gap on no news is thin trading. A small number of participants moved price in a session with almost no volume, and the ordinary population disagrees when it arrives. That gap has a much better claim to closing, because nothing changed except who was awake.
Size sorts them better than anything else on the chart. Large gaps almost always mean something happened; small ones are usually noise. Pooling both into one “gap fill rate” statistic — which is what most published figures do — produces a number that describes neither.
And the clearest evidence for what causes gaps is which markets have them. Crypto trades continuously and gaps rarely. Forex gaps mainly over the weekend. Stocks and futures gap regularly, because they close every day. The gap is a property of the session structure, not of the instrument’s character.
In practice: what a gap does to everything else on the chart
Every indicator computed from closes sees a gap as a large move, because it is one. The relative strength index spikes, the moving average convergence divergence indicator’s histogram prints its biggest bar, momentum reaches an extreme — all from a price change in which nobody transacted.
Which means indicator readings immediately after a gap are describing a closed market. That is not a malfunction; it is what those formulas do with the data they are given.
Volume at the open is the best available separator. Heavy participation supporting the new price suggests the repricing has been ratified; light participation suggests the overnight move has not met the wider market yet.
The bid-ask spread is at its widest in the opening minutes, which is exactly when gap trades happen. So the real cost of a gap trade is higher than the round-trip figure quoted elsewhere on this site, and the difference is largest on the most volatile opens.
Aggregate to weekly bars and most gaps vanish inside them. A gap exists relative to the previous bar, and longer bars have wider ranges to absorb it.
A stop order does not work across a gap. It becomes a market order when price trades through the level, and if the market opens beyond it, the fill is at the open — not at the stop. Position size calculated from stop distance is therefore an estimate that a gap can invalidate entirely.
That is the single most important practical consequence on this page. Overnight risk is not the risk you sized for.
Each trade costs 2% of a typical bar’s range in round-trip costs on this history, and at the open with wide spreads that is a floor rather than an estimate.
And the order book has never seen those prices in this context. There is no resting supply inside a gap, no memory, and nothing waiting to pull price back.
What an opening gap is not
It is not an inefficiency. The market was closed. There was no opportunity to trade those prices and nobody missed one.
It is not a fair value gap. That forms during trading, when price moves too fast for one candle’s range to overlap the next but one. This forms when the market is shut.
It is not required to fill. The gap fill page covers why that claim needs both a definition and a deadline before it means anything.
And it is not tradeable with a normal risk calculation. The stop cannot protect the position across the event that created the gap.
When it fails
In a range every gap fills, because price is oscillating and revisits everything. Anyone learning gap fading during a range learns that it works almost always, then meets a trending market with a method calibrated on the wrong regime.
The second failure is fading the news gap. The gaps offering the largest apparent opportunity are the ones on the biggest news, which are the least likely to reverse and the most likely to keep going.
A third is the overnight position sized on a daytime stop. The stop is a price; the gap is a jump past it. The loss is whatever the open decides.
A fourth is trading the first minute. The widest spreads of the day, the least stable prices, and the highest chance of a fill materially away from the quote you saw.
And a fifth is quoting a fill-rate statistic you have not computed on your own instrument. Every instrument has a different rate, every definition of “filled” gives a different answer, and this site cannot supply the number either — its shared history is generated continuously and contains exactly zero gaps, which is recorded in the measurements file rather than papered over.
The original data
The shared 576-bar history behind every chart on this site is continuous by construction: each bar opens
where the previous one closed, at base resolution and at every aggregation of it, so it contains 0 gaps in 576 bars. That is recorded in research/series-measurements.json, along with the note that no
gap-fill rate can be computed from this data and no page here should quote one.
Publishing a zero is more useful than publishing a borrowed figure. Every other page on this site quotes numbers measured on that history; this one cannot, and saying so is the honest alternative to repeating someone else’s fill rate from a different market. The specification for computing your own is short: your instrument, gaps above a stated size, full fill rather than partial, within a stated window, split by whether news was scheduled. That is an afternoon of data work and it settles a question you would otherwise take on faith for years.
Related
Gap trading covers the strategies built around these. Gap fill is the claim that needs a number. And premarket and after-hours is the session where the gap is actually created.
Gaps were the thing that finally made me stop trading stocks intraday with overnight positions. Not because the gaps went against me more often than not, but because the size of the outcome had nothing to do with the size of the risk I thought I had taken.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.