Why does a price sometimes skip a price entirely?

A gap is an empty band on the chart: yesterday's range and today's range don't overlap. It looks like the price jumped. What actually happened is more mundane and more useful to understand.

Gaps are made of closed markets

Exchanges close. Information keeps arriving — earnings after the bell, overnight news, a competitor's announcement in another time zone. When the market reopens, an opening auction matches the executable interest that accumulated and clears at whatever single price balances it.

That clearing price can sit far from yesterday's close, and the daily record holds nothing in between. The gap isn't a leap; it's a hole in the price record, and the hole is in the record rather than in trading.

A US-listed stock also trades in a pre-market session and an after-hours session, and an earnings release published after the bell is repriced in the after-hours session that evening. Those prints appear in neither the daily bar nor the regular session's five-minute bars. Apple reported after the close on 30 July 2026; its five-minute record ends at the closing print and resumes at the next morning's open, with nothing between them:

The band on the chart is where the record has no bars. Whether a resting stop can trigger in those sessions is a broker setting; the trade-off is a thin book and a wide spread against a fill hours earlier than the open.

This explains something people find surprising: instruments that trade close to continuously barely gap at all. Major FX pairs run around the clock on weekdays and still pause over the weekend, so that is where their gaps are; spot crypto does not pause even then, and gaps least of all. Gaps are largely an artifact of trading hours, not a property of prices.

The names traders use

You'll encounter a vocabulary here, and it's worth knowing while remembering that several of these labels can only be applied after the fact:

• Common gap — small, often in thin trading, no identifiable cause.
• News or earnings gap — a specific release repriced the security overnight. By far the most frequent large gap.
• Breakaway gap — a gap that takes the price out of a long-established range.
• Exhaustion gap — a gap late in an extended move. You can only classify one as "exhaustion" once you know the move ended, which makes it a description of history, not a category you can apply live.

A worked example

Illustrative and rounded: a stock closes Thursday at $88.20. Earnings are released after the close. Friday's opening auction clears at $79.40, and the session's high never exceeds $81.

Be careful about where the empty band actually is. Friday traded between $79.40 and $81, so those prices did print; the untraded band runs from $81 up to Thursday's low, and that is the gap on the chart — the shaded strip below. What the headline figure measures is something adjacent and easier to compute: the close-to-open move, (88.20 − 79.40) ÷ 88.20 ≈ −10%, which is the market's overnight repricing delivered in one auction match rather than gradually. Quote whichever you mean, and do not call one the other.

"Gaps always fill" — the folk claim

You will hear that gaps are eventually "filled" — that the price returns to trade through the empty band. Many do. But examine the claim properly:

• It's untestable without a time limit. Over an unlimited horizon, a volatile security will revisit almost any nearby price eventually. "Always fills, given enough time" is close to a statement about volatility, not about gaps.
• It survives on survivorship. Filled gaps are visually obvious on a chart — you can see the price come back through. Unfilled gaps from years ago sit far away, off the screen, unremembered.
• Fixing a window changes everything. Ask instead: what fraction of gaps larger than 3% are filled within 20 sessions? That question has an answer you can count. The vague version doesn't.

One data trap

Not every gap is real. Dividends and splits create artificial gaps in unadjusted price data — a stock that goes ex-dividend appears to drop overnight, and a 4-for-1 split looks like a catastrophic collapse. Adjusted price series remove these. Before you interpret any gap, confirm which series you're looking at — this is the adjusted-price lesson from Markets Foundations doing real work.

In the data

Both artificial gaps come with a calendar. Every split is a dated ratio of new shares for old, and every dividend has an ex-dividend date, the first session the buyer no longer gets it, which is when the price steps down, not the later payment date. Here are Apple's splits:

An adjusted price history has both kinds of step removed. So a gap that shows in the traded closes and not in the adjusted ones was never a market event.

Try it now

1. From the schematic above, write down both numbers and label them differently: the width of the untraded band, and the close-to-open move. They are not the same quantity and most reporting conflates them.
2. Now measure a real one. In the five years below, find a session whose open sits outside the previous session's range, and compute both numbers. One such pair of sessions is below, 30 and 31 July 2026, the morning after an earnings report, so you can do the arithmetic on exact prices before hunting for your own on the chart. Then track how many sessions passed before price traded back through the band — or note that it hasn't yet.

3. Failure hunt: find an "obvious" gap that is not a market event at all. Take the 2020 split from the list in the section above; below are the sessions either side of it, traded and adjusted. A gap that is present in the traded close and absent from the adjusted close never happened to anybody. Ex-dividend dates make the same kind of false gap, only much smaller.