Opening Range Breakout: What the 1,637% Paper Really Says

Someone posted the chart in your trading group: a 5-minute opening range breakout that turned $25,000 into about $435,000 over eight years, while buying and holding the S&P 500 turned the same money into about $75,000. It comes from a real

Someone posted the chart in your trading group: a 5-minute opening range breakout that turned $25,000 into about $435,000 over eight years, while buying and holding the S&P 500 turned the same money into about $75,000. It comes from a real paper, and the numbers on the chart are real. What the screenshot leaves out is the same paper's first result, where the identical strategy made 3.2% a year. The famous opening range breakout result is mostly a stock-selection result, and the parts that decide whether you could trade it live are the parts the tables never price.

The rules fit in five lines, and one of them is a stop order

An opening range is the high and low of the first few minutes after the US stock market opens at 9:30 a.m. Eastern. A breakout is price trading beyond that range. Zarattini, Barbon, and Aziz tested the 5-minute version on roughly 7,000 US stocks from 2016 through 2023, with delisted companies kept in the data so dead stocks don't quietly vanish from the results.

Their rules:

• Only stocks priced above $5, averaging at least 1,000,000 shares a day, with a 14-day Average True Range (ATR, the average size of a recent daily price range) above $0.50.
• If the first 5-minute candle closes up, place a buy stop at its high. If it closes down, place a sell stop at its low. A doji (open equals close) gets no order.
• The stop loss sits 10% of the stock's ATR from the entry. Anything still open at the 4:00 p.m. close is sold.
• Each position is sized so a stop-out costs 1% of the account, with leverage capped at 4x.
• The only cost modeled is a $0.0035 per share commission, the entry tier at Interactive Brokers at the end of 2023.

A buy stop is an order that only goes live once price touches your level, so what it actually fills at is its own question. And a stop loss at 10% of ATR is strikingly tight next to the 1.5x to 3x multiples most guides recycle.

The plain version made 29% in eight years

Table 1 is the one that rarely gets screenshotted. Run the rules on every stock that clears the liquidity screen and $25,000 grows by 29%, roughly $7,500. The Sharpe ratio (return per unit of volatility, higher is better) comes in at 0.48 against 0.78 for the S&P 500.

It did stay quiet. Maximum drawdown was 13% against 34% for the index, and the worst single day was -0.8% against -10.9%. Quiet, though, is not what the screenshot advertises.

One volume filter turned 29% into 1,637%

Relative volume compares how much traded in today's first five minutes with the average first-five-minute volume of the previous 14 days. Above 100% means busier than normal. The paper's second strategy keeps only stocks at 100% or higher and then trades the 20 highest readings each day.

Result: 41.6% a year, a Sharpe of 2.81, and $25,000 becoming roughly $435,000. The paper's Figure 4 shows why, measuring the average trade net of commission in R, where one R is the amount you lose if the stop is hit.

A breakout tells you price is moving. Relative volume tells you whether anyone else showed up to move it.

— The Tradoki desk note

The lesson isn't "breakouts work." The same breakout was worth almost nothing on ordinary stocks and a lot on unusual ones. That is a scanner study wearing a breakout costume, and the scanner is the part you'd have to run every single morning at 9:35, the moment the first five minutes close.

The 5-minute window was the best of four, and the authors don't know why

The paper also ran 15-, 30-, and 60-minute opening ranges, all with the volume filter:

Opening range	Total return	Annual return	Sharpe
5 minutes	1,637%	41.6%	2.81
15 minutes	272%	17.4%	1.43
30 minutes	21%	2.3%	0.21
60 minutes	39%	4.1%	0.40
S&P 500	198%	14.2%	0.78

The 30-minute version made less per year than the unfiltered 5-minute baseline. The authors say the reason the shortest window wins is unclear and needs more research.

When four windows get tried and the shortest wins by a mile, every variant you tried raises the number of trades you need before the result means anything.

The best stocks still lost about four trades in five

Among the 25 best-performing stocks on the 5-minute version, the paper's own table lists win ratios between 17% and 24%. The top name, DDD, earned 385R at a 21% win ratio.

At a 20% win rate, the arithmetic is unforgiving. Ten losses in a row has a probability of 0.8 to the tenth power, about 11%, in any given ten-trade stretch. That is roughly one stretch in nine.

That streak is the real price of the 1,637%. Most traders don't quit a strategy over a drawdown percentage. They quit on trade number eleven, which is why position sizing decides whether a streak like that is survivable.

The tables price commission and nothing else

The paper models a commission of $0.0035 per share. I searched the full text of the February 2024 version and the word slippage never appears. Slippage is the gap between the price you expected and the price you got.

The entry here is a stop order, fired in the first minutes of the session, in stocks that are unusually busy by design. That is where fills drift. Here is an illustration with hypothetical numbers, not measured ones. In the paper's own example a stock has a $5 ATR, so R is $0.50, and each cent of slippage costs 0.02R. Four cents in total, in and out, erases the 0.08R average for the above-100% bucket. On a stock at the paper's ATR floor of $0.50, the entire stop is five cents wide.

Two more limits. The window is a single eight-year stretch in which the S&P 500 nearly tripled. And the account starts at $25,000, the old pattern day trader minimum, with leverage up to 4x, while the rule that replaced it in 2026 watches intraday margin instead of counting trades.

Independent tests are thin, and the closest one is a different market

The nearest public reproduction of the stock version I found, a QuantConnect research post, covers only 2016, the first year of the paper's own window.

An independent researcher's 2026 arXiv preprint tested opening range breakouts on Micro E-mini Nasdaq-100 futures across 947 trading days from 2021 to 2025. It used walk-forward validation (fit on the past, test on the unseen stretch after it) and charged 2.0 points, $4.00 per micro contract, round trip for spread, fees, and slippage. No breakout variant passed. The shortest holds lost money after costs, at -0.82 points per trade long and -3.45 short. The best, a long held 75 minutes, made +2.82 points per trade with a t-statistic of 0.88, well under the 2.0 the study required, and it was unstable from year to year.

That is not a refutation. It is a different market, a six-bar opening range, no relative-volume filter, a different entry, and a single-author preprint. What it shows is what a two-point toll does to a small intraday edge.

So the honest status of the 1,637% is a published result, net of commission only, with no independent post-2023 test of the filtered version that I could find. A trader who wanted to work with the idea would run the breakout with their own broker's fills, in a demo account first, and log slippage on every trade before trusting any average-R figure from a paper.

● FAQ

What is the 5-minute opening range breakout strategy?

It marks the high and low of the first five minutes after the 9:30 a.m. Eastern open, then places a buy stop above the high if that first candle closed up, or a sell stop below the low if it closed down. In the Zarattini, Barbon, and Aziz paper, the stop loss sits 10% of the stock's 14-day ATR from entry, and any position still open is closed at 4:00 p.m.

Does the opening range breakout strategy work?

In one published backtest, the unfiltered 5-minute version on US stocks returned 29% over 2016 to 2023, or 3.2% a year, while the version limited to high relative-volume stocks returned 1,637%. The cost model was commission only, and I found no independent post-2023 test of the filtered version. A separate independent study on Micro Nasdaq-100 futures found no opening range breakout variant that passed its significance tests after costs.

What is a stock in play?

A stock trading with unusual volume on a given day, usually because of a catalyst such as an earnings report, an FDA decision, or a merger. The paper measures it with relative volume: the first five minutes' volume today divided by the average of the previous 14 days. It trades only the 20 stocks with the highest readings, each at 100% or more.

What is the win rate of an opening range breakout?

The paper does not headline one. Its table of the 25 best-performing stocks on the 5-minute version shows per-stock win ratios of 17% to 24%, which is a hindsight-selected best case, not an average. Many small losses paid for by a few large winners is the shape a stop at 10% of ATR produces.

Why does slippage matter so much for an opening range breakout?

The entry is a stop order triggered in the first minutes of the session, in stocks that are unusually active, and the stop itself is tight. In the paper's own example the stop is 50 cents wide, so each cent of slippage costs 0.02R. The paper's average trade above 100% relative volume earned 0.08R, which four cents of total slippage would erase.