Browse the library

Trading · Trading styles

Gap trading: price gap types, gap fills and what the data shows

What price gaps are, the four classic gap types, how often gaps fill based on two years of data for 10 large US stocks and ETFs, and the risks of trading them.

On this page
  1. Why gaps happen
  2. Measuring a gap
  3. The four classic gap types
  4. What the data shows: how often do gaps fill?
  5. A real example: NVIDIA, January 27, 2025
  6. Risks of trading gaps
  7. Gap trading compared with related approaches
  8. How Finabase helps
  9. Frequently asked questions
  10. Summary

A price gap is a jump between one bar and the next with no trading in between. On a daily chart, a stock that closed at $100 and opened the next morning at $104 has gapped up 4%. Gaps happen because news and orders keep arriving when the regular session is closed: earnings after the bell, economic data before the open, events overseas, and trading in the pre-market and after-hours sessions.

Gap trading is the study of what prices tend to do after such a jump. Some traders look for gaps to "fill," meaning the price returns to the prior close. Others look for gaps that keep going in the same direction. Both ideas are popular, and both are easy to overstate. This guide covers the mechanics and the classic gap types, then tests the most common claim, that "gaps always fill," on two years of real data. It is an educational outline, not a trading method.

Why gaps happen

US stocks trade most heavily in the regular session, 9:30 a.m. to 4:00 p.m. Eastern Time. Outside it, information keeps arriving and some trading continues in thinner extended-hours sessions. When the regular session reopens, the opening auction sets a price that reflects everything that happened overnight. If that price is far from the prior close, the chart shows a gap.

Typical causes, roughly from most to least common for an individual stock:

  • Earnings and guidance, usually released before the open or after the close. See Trading earnings season: implied moves, IV crush and earnings gaps.
  • Company news: mergers, product or drug-trial results, management changes, legal decisions.
  • Analyst upgrades and downgrades published before the open.
  • Macro news: inflation and employment data at 8:30 a.m. ET, central-bank decisions, and moves in overseas markets and index futures overnight.
  • Corporate actions such as a stock going ex-dividend, which mechanically lowers the opening price by roughly the dividend.

A halt during the session creates the same effect inside the day: the stock reopens at a new price with nothing in between.

Measuring a gap

Some traders define gaps from the prior day's high or low instead of its close. A full gap up opens above the prior high, so the two bars do not overlap at all; a partial gap up opens above the prior close but inside the prior range. On a candlestick chart (see How to read candlestick charts: a guide with real data) a full gap shows as visible empty space between candles. The statistics below use the simpler close-to-open definition.

The four classic gap types

The classic descriptions come from Robert Edwards and John Magee's 1948 book on chart patterns. They describe the context in which a gap appears rather than its size.

Type Where it appears Classic interpretation
Common gap Inside a trading range, on ordinary volume Noise; often fills quickly
Breakaway gap Out of a range or base, often on news and heavy volume Start of a new move; often does not fill soon
Runaway gap In the middle of a strong trend Continuation; sometimes called a measuring gap
Exhaustion gap Late in a long, accelerating move Last burst before a reversal; tends to fill

The weakness of this scheme is that you usually know which type a gap was only afterwards. A gap that looked like a breakaway can turn out to be exhaustion a few days later. Treat the labels as vocabulary for describing charts, not as a forecast.

What the data shows: how often do gaps fill?

To test the claim that gaps fill, we measured every daily opening gap for ten large, liquid US securities: SPY, QQQ, AAPL, MSFT, NVDA, AMZN, META, TSLA, JPM and XOM. The sample covers daily bars from October 31, 2024 to October 29, 2026, 500 sessions each and 5,000 symbol-days in total. For each day we computed the gap from the prior close to the open and recorded whether the price traded back to the prior close during the same session.

Same-day gap fill rate by gap size, 10 large US stocks and ETFs, Oct 31, 2024 – Oct 29, 2026 Up 0.5–1% 55% (n=708) Down 0.5–1% 57% (n=533) Up 1–2% 37% (n=462) Down 1–2% 36% (n=416) Up 2–4% 28% (n=179) Down 2–4% 25% (n=150) Up 4%+ 19% (n=47) Down 4%+ 14% (n=51) Daily bars, 5,000 symbol-days. Filled = traded back to the prior close the same session. Finabase chart data.
Same-day gap fill rates for 10 large US stocks and ETFs, daily bars October 31, 2024 to October 29, 2026. n = number of gaps in each bucket. Finabase analysis of Finabase chart data.
Gap size Gap-ups Filled same day Gap-downs Filled same day
0.5–1% 708 54.8% 533 56.8%
1–2% 462 36.8% 416 36.3%
2–4% 179 27.9% 150 24.7%
4% or more 47 19.1% 51 13.7%

Three findings stand out.

  1. Gaps were common. About half of all sessions (2,546 of 5,000) opened at least 0.5% away from the prior close.
  2. The bigger the gap, the less likely a same-day fill. Small gaps filled a little more than half the time, which is close to what random intraday movement would produce. Gaps of 4% or more filled the same day less than one time in five.
  3. Eventually, most did fill, but not all. For gaps of 2% or more, we also checked the following 20 sessions (excluding gaps too recent to follow that long). About 27% of gap-ups filled the same day, 57% within the next five sessions and 73% within 20. For gap-downs the figures were about 22%, 61% and 83%.

We also checked whether the day's close was beyond the open in the gap's direction, a rough test of "gap and go." For most sizes it was close to a coin flip, about 48–51%. Gap-ups of 4% or more closed above their open on only 15 of 47 days (32%).

A real example: NVIDIA, January 27, 2025

NVIDIA (NVDA), daily, Dec 2, 2024 – Mar 14, 2025: a −12.5% gap, filled 15 sessions later $120 $140 Prior close $142.62 (Jan 24) Jan 27 open $124.80 (-12.5%), high $128.40 Jan 2025 Feb Mar
NVIDIA (NVDA), daily bars, December 2, 2024 to March 14, 2025. The dashed line is the January 24 close. Finabase chart data.

On Friday, January 24, 2025, NVIDIA closed at $142.62. Over the weekend, reports about a low-cost artificial-intelligence model from the Chinese start-up DeepSeek raised doubts about future demand for AI chips. On Monday, January 27, NVIDIA opened at $124.80, a gap of −12.5%. It never got back near the prior close that day: the high was $128.40, and it closed at $118.42, down 17.0% from Friday.

By the classic labels this looked like a breakaway gap on news and heavy volume. Yet the gap filled 15 sessions later, when the stock traded at $143.44 on February 18. Then the stock fell again, closing at $106.98 on March 10. Volume on the gap day was about 819 million shares, roughly four times its 20-day average. A trader who bought the gap expecting a quick fill sat through a further 6% drop on day one; one who shorted it expecting a breakaway saw the whole move reverse within three weeks. The same gap supported both stories at different times.

Risks of trading gaps

  • Stops do not protect against gaps. A stop order becomes a market order when triggered, and if the stock opens beyond the stop it fills near the open. Investor.gov's bulletin on stop orders makes this point directly. See Stop-loss placement: volatility, structure and time stops.
  • The open is the most volatile part of the day. Spreads are wider and prices jump around while the market digests overnight news. Early fills can be poor.
  • Extended-hours prices can mislead. Pre-market quotes come from thin trading and can differ substantially from the official open.
  • Halts. News-driven gaps are the ones most likely to come with a trading halt or volatility pause.
  • Survivorship in examples. Charts shared online tend to show the gaps that filled neatly. The data above shows how many did not, at least not quickly.
Approach Core idea Main risk
Gap fill (fade) Trade against the gap, toward the prior close Large gaps often keep going
Gap and go Trade in the direction of the gap Close was near a coin flip vs the open in our sample
Earnings drift Hold in the direction of an earnings surprise for weeks Evidence weaker in recent years
Opening range breakout Wait for the first 15–30 minutes, trade a break of that range False breaks in choppy markets

How Finabase helps

Finabase does not place trades; it shows where gaps come from. The chart draws daily or intraday candles so gaps are easy to spot and measure against the prior close. The trending page lists the most active stocks in the regular session, pre-market and after hours, which is where the next day's gaps usually start, and the earnings calendar shows which companies report on each day, the most predictable source of large gaps. Each symbol page also shows the previous close and the day's open, high and low, the four numbers a gap is measured with.

Frequently asked questions

Do gaps always fill?

No. In our sample of 10 large US stocks and ETFs over two years, most small gaps filled the same day, but gaps of 4% or more filled the same day less than 20% of the time. Many larger gaps filled within a few weeks, but some took months, and 9 of the 98 gaps of 4% or more had still not filled by the end of the sample.

What is the difference between a full gap and a partial gap?

A full gap opens beyond the prior day's entire range, above its high or below its low. A partial gap opens beyond the prior close but still inside the prior day's range.

What causes a stock to gap up or down?

News released while the regular session is closed: earnings, company announcements, analyst rating changes, economic data and overnight moves in futures and foreign markets. The opening auction then prices all of it at once.

Which gap types are most important?

Classic chart analysis names four: common, breakaway, runaway and exhaustion. They describe where a gap appears in a trend, but they are usually only clear after the fact.

Summary

A gap is the market repricing news that arrived while the regular session was closed. The classic gap types give a vocabulary for what a gap might mean, but not a forecast. Our two-year test on ten large US stocks and ETFs found that small gaps filled the same day about half the time, large gaps rarely did, and most gaps of 2% or more filled within 20 sessions, sometimes after a large move the other way. Gaps are as much a risk to manage as a pattern to trade, because no stop order can stop a price from opening beyond it.

Sources and further reading

  1. Edwards and Magee — Technical Analysis of Stock Trends (1948; later editions)
  2. SEC, Investor.gov — Investor Bulletin: Risks of extended-hours trading
  3. SEC, Investor.gov — Investor Bulletin: Stop, stop-limit and trailing stop orders
  4. Finabase chart data — daily bars for SPY, QQQ, AAPL, MSFT, NVDA, AMZN, META, TSLA, JPM and XOM, October 31, 2024 to October 29, 2026 (Finabase analysis)