← All articles Research / Opening Gaps

Gap Trading Strategy: What 225 Opening Gap Tests on NASDAQ Showed

Most gap trading advice starts from the overnight gap and asks whether it fills. Tomas tested the assumption instead. In BreakoutOS he built 225 long continuation gap strategies on E-mini NASDAQ across three different gap definitions, then ran the fade and short-side versions. The familiar overnight gap came last. Here's how the tests were built, what came out of them and how to run the same test on your own platform.

Quick answer

1. Of three gap definitions tested for long continuation on E-mini NASDAQ, the usual overnight gap was the weakest. The aftermarket gap was the strongest.

2. The best-looking setting entered at the aftermarket gap open plus 4.6 times ATR(20). The neighbouring multipliers, 4.4 and 4.8, worked well too.

3. The short side was harder. Short continuation didn't produce a usable candidate, and the one promising short fade still needs more testing. All results are hypothetical backtests.

Go deeper on gaps

The Gap Breakout Playbook shows how to find the hidden gaps in markets that barely close, with 11 plays and three model strategies. See The Gap Breakout Playbook.

What Is an Opening Gap When the Market Never Closes?

A gap is the move between one session's close and a later session's open. If the later open is higher, it's a gap up. If it's lower, it's a gap down.

E-mini NASDAQ trades almost around the clock, so your chart may show bars right through the gap. It still works as a measurement. What matters is the relationship between the two prices you pick, and that depends on how you split the day into sessions: premarket, the regular session and the aftermarket.

Session times are the first thing to check. Some charts show exchange time and some show your local time. Define your sessions once, the same way every time, and test which one behaves best rather than assuming the familiar overnight window wins.

Gap Continuation vs Gap Fade

There are two opposite ways to trade a gap, on each side of the market. They're different ideas, so each one needs its own test.

ApproachThe gapThe idea
Long continuation (gap and go)Gap upPrice keeps rising
Long fade (gap fill)Gap downPrice recovers into the gap
Short fadeGap upPrice reverses lower
Short continuationGap downPrice keeps falling

The Rules in Plain Words

In this method a gap doesn't trigger a trade on its own. The gap open becomes the point of initiation, the reference level a breakout is measured from. Price then has to travel a set distance beyond it before the strategy enters.

  • Reference level: the open after the gap, using the session definition you're testing.
  • Long entry: a stop order at the reference level plus a multiple of ATR (Average True Range). For a short, the reference level minus the multiple.
  • Exit: end of day. These are day trades.
  • No stop loss or filter yet. The first job is to find out whether the raw idea has an edge. Risk controls come next.

Measuring the distance in ATR means the entry level widens when the market is volatile and tightens when it's calm. It also gives you something to test: how much extra movement to demand after the gap, instead of buying every qualifying open.

How the 225 Variations Were Built

Tomas tested long continuation first, on E-mini NASDAQ hourly data, varying three things at once.

What changedValues testedCount
Gap definitionPremarket (midnight to 9 a.m.), main session, aftermarket3
ATR period5, 20, 403
ATR multiplier0.2 to 5.0 in steps of 0.225
Total3 × 3 × 25225

The main-session definition is deliberately unusual: it treats the move between the premarket close and the aftermarket open as the gap. The point was to see whether a different way of dividing the day finds better behaviour than the one traders usually talk about.

Which Gap Worked Best?

Gap definitionResult in this test (long continuation)
Conventional overnight gap (midnight to the open)Weakest. Almost no useful continuation results
Main session treated as the gapPerformed well
Aftermarket gapStrongest

Finding good-looking results in a batch of 225 is the easy part. To keep it honest, the candidates were ranked on their in-sample Robustness Suite scores first, and only then checked on the out-of-sample data they'd never seen. Many of the top in-sample candidates held up on the unseen data. One aftermarket candidate scored about 80% on the Robustness Suite. That's a robustness score. It isn't a win rate or a chance of future success.

The best-looking setting was the aftermarket gap with ATR(20) times 4.6. What makes it worth a closer look is that 4.4 and 4.8 also worked well. One good setting on its own is often luck. A good neighbourhood is more likely to be real.

Keep the scope in mind. This was long continuation on one market. It doesn't show that aftermarket gaps win in every market or direction, and it still needs a stop loss and possibly a filter before it's a finished strategy.

The Fade and Short-Side Tests

Tomas then ran the same kind of experiment on the other three approaches.

  • Long fade (225 variations): encouraging results on unseen data, with substantial drawdowns and the risk controls still to be worked out.
  • Short fade (225 variations): several candidates looked good in-sample and then fell away on the out-of-sample data.
  • Short continuation: struggled even in-sample. No viable candidate came out of this run.
  • A Thursday-only short fade: filtering the best aftermarket short fade to Thursdays gave 114 trades, a 57.9% win rate and a net profit to drawdown ratio of 9.19. That filter was chosen after looking at the results, the sample is small and it may be overfitted, so it needs testing on other indexes before anyone trades it.

Finally, he combined the long continuation, the long fade and the short fade into one portfolio, then tested weekday filters to bring the drawdown down. On their own, the long continuation strategy ran at almost 8:1 net profit to drawdown and the fade at about 4:1. That portfolio is the example in How to Pass a Prop Firm Challenge: Start With a Diversified Portfolio. All of these figures are hypothetical backtests.

How to Test a Gap Strategy Yourself

  1. Define the sessions once and use them the same way for every test.
  2. Use the gap open as the reference level. The entry only comes once price travels beyond it.
  3. Test several ATR periods and a wide range of breakout distances.
  4. Pick candidates on in-sample results, then check them once on data they never saw.
  5. Make sure neighbouring settings work too.
  6. Add a stop loss and any filters last, and test them the same way.

Test gap strategies in BreakoutOS

Answer four quick questions and see which BreakoutOS tools fit how you trade, and what to do next.

Get Your Free Roadmap  →

Run This Test on Your Platform

Gap strategies live or die on session settings, so match them on whichever platform you trade before you compare results:

  • TradingView: match the chart's session (regular or extended) and timezone to the sessions you defined.
  • NinjaTrader: check the trading hours template on the chart and the strategy.
  • TradeStation: check the symbol's session settings and whether times show in exchange or local time.
  • MultiCharts: check the symbol's session settings and timezone.
  • MetaTrader 5: your broker's server time often differs from exchange time, so convert your session times.
  • AmiBroker: check the database's intraday session settings.
  • Python: make sure your timestamps are in exchange time before you define the sessions.

Where BreakoutOS Fits

Every test in this article was built in BreakoutOS with its opening-gap presets, without code. You pick the market and the gap template, set the ranges to test, and it builds every variation, ranks them on in-sample robustness and lets you check the winners on unseen data. Portfolio Builder then shows how several gap strategies work together.

An E-mini NASDAQ opening gap strategy is also the one we used to check the TradingView export: 389 of 393 trades matched exactly in TradingView's Strategy Tester, in hypothetical backtests with no commission or slippage. See the BreakoutOS for TradingView page for the full comparison, and How to Backtest a Strategy in TradingView for the settings to check on your side.

Research results are hypothetical backtests and do not guarantee future trading performance.

Ready-made gap strategies

The Gap Breakout Playbook shows how to find the hidden gaps in markets that never close, with three complete strategies and code for seven platforms. See The Gap Breakout Playbook.

The Gap Breakout Playbook - find the hidden gaps in markets that never close, with three complete strategies and code for seven platforms
The Gap Breakout Playbook. See what's inside →

Frequently Asked Questions

A gap trading strategy trades the move between one session's close and the next session's open. In the breakout version tested here, the open after the gap becomes a reference level, and the strategy enters only if price travels a set distance (a multiple of ATR) beyond it, then exits at the end of the day.
A gap fill (or fade) strategy bets that price moves back into the gap, against its direction. A gap and go (or continuation) strategy bets that price keeps moving the way it gapped. They're opposite ideas, so they need separate tests, and a gap definition that works for one may not work for the other.
In Tomas's test of 225 long continuation strategies on E-mini NASDAQ, the aftermarket gap was strongest, treating the main session as the gap also worked well, and the conventional midnight-to-open gap was weakest. That's one market and one direction in hypothetical backtests, so test it on your own market before relying on it.
Yes, as a measurement. Even when bars run through the night, the relationship between a chosen session's close and a later open can still define a useful reference level. The key is defining the sessions consistently and checking whether your chart shows exchange time or local time.
Yes. The tests here started without one to check whether the raw idea had an edge, but a finished strategy needs a protective stop. Size it sensibly: Tomas's rule is a stop at least 1.5 to 2 times the biggest 1-minute bar, so the backtest can resolve it.
Tomas Nesnidal

About the Author

Tomas Nesnidal, known to the systematic trading community as Mr. Breakouts, is a breakout trading specialist, hedge fund co-founder, and creator of BreakoutOS. He has managed institutional portfolios using breakout strategies for over 15 years, trading from 65+ countries. He is the author of The Breakout Trading Revolution and co-founder of Breakout Trading Academy.