Does the Opening Range Breakout Strategy Work? 6 Tested
Does the opening range breakout strategy work? In our lab, no. We coded six popular versions exactly as published and tested them on SPY, ES and NQ futures, and gold, with realistic trading costs. None of them beat random entries taken at the same time of day.
That's the verdict in one line: opening range breakouts, as commonly published, did not beat random entries at the same time of day on SPY, ES, NQ or gold. One version came close enough to be interesting, and we'll show why it still didn't pass. If you've seen ORB pitched as a simple daily edge, our test doesn't support that.
Where the opening range breakout comes from
The idea is simple. Mark the high and low of the first few minutes of the session. If price breaks above that range, go long. If it breaks below, go short. Put a stop on the other side or a set distance away, and take profit at a multiple of the risk.
The name usually traces back to trader Toby Crabel. His 1990 book Day Trading With Short Term Price Patterns and Opening Range Breakout was published by Traders Press, according to Wikipedia's entry on Crabel. Since then it's become one of the most widely taught day trading setups.
There's serious research nearby, too. In Market intraday momentum (Journal of Financial Economics, 2018), Lei Gao, Yufeng Han, Sophia Zhengzi Li and Guofu Zhou found that the S&P 500 ETF's first half hour return, measured from the prior close, helped predict its last half hour return from 1993 to 2013. Closer to ORB itself, Carlo Zarattini and Andrew Aziz studied the strategy on QQQ from 2016 to 2023 in Can Day Trading Really Be Profitable?. A follow up with Andrea Barbon, A Profitable Day Trading Strategy for the U.S. Equity Market, reported a Sharpe ratio of 2.81 for a 5 minute ORB on the top 20 "stocks in play". The Sharpe ratio measures return per unit of risk.
Those studies test specific designs, like limiting trades to stocks with unusual news driven activity. They aren't the rules we tested, so our result doesn't overturn them. What it does tell you about is the versions retail traders actually get handed.
How we tested six ORB strategies
All six rules came from public templates on roboquant.dev's marketplace, a strategy template site: M30 ORB + M5 Breakout, NY ORB Retest, Giga ORB, Flash Trigger ORB, Crazy ORB and Scarface ORB. Each was coded exactly as published.
We used SPY (5 minute data from 2018-06 to 2026-10, with some cells covering 2023-08 to 2026-06), ES and NQ futures (about two years, 2024-10 to 2026-09), and gold. Each rule ran on the timeframes it specifies, plus one step either side. We call each market and timeframe combination a "cell".
Costs came in two models. The conservative one uses measured spread plus commission. The cheap one uses half the spread but only counts a limit target as filled if price trades one tick through it.
The key check is a random-entry control: the same trade, same side, same stop and target, taken at random times in the same time of day slot, drawn 500 times. It shows what luck alone produces. Because we tested many cells, one lucky fluke is expected somewhere, so we applied a Bonferroni correction. That raised the bar to roughly p < 0.004. In plain terms, p is roughly how often random entries did as well as the strategy, so 0.004 means about 4 in 1,000.
We also ran a look-ahead check on every cell, deliberately poisoning future data to confirm the code wasn't peeking at prices it couldn't have known yet. To pass, a rule needed a net profit after costs, a win over the random control at the corrected bar in some cell, and the same sign across markets and timeframes.
Opening range breakout strategy results, rule by rule
A few terms first. A basis point (bp) is one hundredth of one percent. The t-statistic is how sure we are that the average trade isn't really zero; about 2 or higher is the usual bar. Profit factor is gross winnings divided by gross losses, so anything under 1 loses money.
- M30 ORB + M5 Breakout: range from 9:30 to 10:00, entry on the first close outside it with high relative volume, 3 ATR stop, 1.5R target. Gross results ran from -2.6 to +2.7 bps a trade in the index cells, with no t above 1.2. Best control p was 0.042 (gold). On the template's own showcase cell (NQ 5 minute) we got a profit factor of 0.95; the template page claims 1.22.
- NY ORB Retest: a limit order back at the 15 minute range edge. ES and NQ made +2.7 to +5.5 bps gross with a 58 to 64% win rate, but best net t was 1.3 and best control p 0.052. SPY 15 and 30 minute cells, and every 30 minute cell, were negative.
- Giga ORB: 9:30 to 9:45 range with a candle shape and bias filter, 1R target. Results ran from -4.0 to +5.8 bps a trade, with no t above 1.5 in either direction, and best control p was 0.12. 2026 year to date was negative in every cell. Showcase NQ 15 minute: 49% win rate and profit factor 1.04, against a claimed 55.9% and 1.22.
- Flash Trigger ORB: the most consistent. Positive in every ES and NQ cell, +1.6 to +3.1 bps net, but net t only 1.1 to 1.7, with best control p 0.062.
- Crazy ORB: premarket range with a tiny 0.35 ATR target. Win rate of 72 to 77%, but the average trade was about zero gross and negative after costs in 10 of 12 cells. Best control p was 0.18.
- Scarface ORB: break and retest with a rejection wick, 2R target. It lost gross on SPY and NQ, was negative net in 7 of 8 cells, and its control p was 0.36 or worse.
Why a 77% win rate can still lose
Crazy ORB is the textbook case. ATR, or average true range, is the typical size of a bar. A target of 0.35 ATR sits very close to the entry, while the stop is much further away. Price touches a nearby target far more often than a distant stop, so a high win rate comes from the geometry and doesn't show any edge. When your target is about a third of your stop, you need to win roughly three trades in four just to break even before costs. That's why a 72 to 77% win rate still left the average trade around zero.
The template that told the truth
Flash Trigger deserves credit. Its showcase cell reproduced closely: we measured a 58% win rate and profit factor 1.21, against the template page's claimed 61.6% and 1.36. So the template is honest. The problem is that the edge is too small to tell apart from luck on two years of data. A control p of 0.062 is a long way from 0.004. We've left it running as a forward paper shadow, which means simulated trades with no money involved.
There's a wider lesson here. When a showcase result does reproduce, it's usually down to market drift or tiny-target geometry rather than a real edge.
A coding bug that flatters ORB backtests
Our look-ahead check caught the same bug three times in this batch. The code skipped trades whose entry bar gapped past the stop. Those trades are instant losers, so leaving them out made results look better than they were. If you backtest your own ORB, check how your code handles that case.
These six were part of a batch of 25 public templates from the same site, all tested the same way. All 25 failed. In this lab, the opening range breakout family has never produced a survivor. Across about 290 strategy ideas tested so far, only 2 are clean survivors.
What this test can't tell you
ES and NQ had only about two years of data, which is short for detecting a small edge. We covered SPY, ES, NQ and gold, not individual stocks, so the stocks in play approach from the research above went untested. Your costs may differ from both of our models. We tested the rules as published, not every possible tweak. Flash Trigger's forward paper results are still pending. And past results don't guarantee future ones.
Frequently asked questions
Is the opening range breakout profitable?
Not in the six popular versions we tested. After realistic costs, none beat random entries taken at the same time of day on SPY, ES, NQ futures or gold. The most consistent, Flash Trigger ORB, made a small net profit on ES and NQ, but that profit was too small to tell apart from luck.
Why do ORB strategies show such high win rates?
Often the geometry produces them, not an edge. A small target paired with a wide stop gets hit far more often, so win rates of 72 to 77% come almost automatically. In our test of Crazy ORB, the average trade was still about zero before costs and negative after costs in 10 of 12 cells.
What timeframe is best for an opening range breakout?
In our test, no timeframe rescued it. We ran each rule on the timeframes it specified, plus one step either side. Some cells were positive, but none beat the random-entry control at our corrected bar, and NY ORB Retest was negative in every 30 minute cell we tested.
Does academic research support the opening range breakout?
Some does, for specific designs. Zarattini, Barbon and Aziz reported strong results for a 5 minute ORB limited to stocks in play, and Gao and coauthors found the first half hour helped predict the last. Those aren't the retail templates we tested, which didn't beat random entries on SPY, ES, NQ or gold.
Want a strategy put through the same tests? Email info@synergytrading.net with the subject "Verdict request". It's a waitlist: there's no price and no promised date, and nothing is sold yet. Synergy Trading is not a registered broker-dealer or investment adviser, and nothing here is a recommendation to buy or sell.
Educational research only, not investment advice. Past results do not guarantee future results.
Sources
- Wikipedia: Toby Crabel
- Gao, Han, Li and Zhou, Market intraday momentum (EconPapers)
- Zarattini and Aziz, Can Day Trading Really Be Profitable? (Concretum Group)
- Zarattini, Barbon and Aziz, A Profitable Day Trading Strategy for the U.S. Equity Market (Concretum Group)
- roboquant.dev strategy marketplace
- roboquant.dev: The Flash Trigger ORB template
Want your strategy tested?
We run the same test on any rule set: realistic costs, a random-entry control and a look-ahead check. Email us the rules to join the waitlist. Nothing is for sale yet.
Request a verdict