Opening range breakout on SPY, 59 sessions of 5-minute bars, every trade counted after costs
Who this applies to: a 30-minute opening range breakout traded on the SPY exchange-traded fund during regular US hours, from 28 August to 20 November 2026 (59 full sessions). It is one rule on one instrument over about three months: a small sample, not a verdict on breakout trading in general. Not financial advice.
The rule (calculated on 5-minute bars)
- Opening range: the high and low of the 9:30 to 10:00 New York time bars.
- After 10:00, the first bar that trades above the range high opens a long at the range high; the first that trades below the range low opens a short at the range low. One trade a day at most.
- Stop: the opposite side of the range. Otherwise exit at the close of the 15:55 bar.
- Costs: a round trip of 2 cents per share (spread), and a second run at 4 cents per share (spread plus 1 cent slippage on each side).
What we found
- The rule traded on 58 of the 59 sessions: 28 longs and 30 shorts; 27 trades hit the stop.
- 24 of the 58 trades made money, a win rate of 41%.
- Before costs, the 58 trades made 12.83 dollars per share in total, an average of 3.02 basis points per trade with a standard deviation of 50.59 basis points.
- After 4 cents per share of costs, the total fell to 10.51 dollars per share and the average to 2.50 basis points per trade.
- The t-statistic of the average after costs was 0.38: the result cannot be told apart from zero.
- By our arithmetic, with the same average (2.50 basis points) and spread of results (50.59 basis points), it would take about 1,638 trades for the average to reach two standard errors from zero, about six and a half years of one trade per trading day.
How the numbers work
| Figure | Value | Where it comes from |
|---|---|---|
| Sessions | 59 | calculated from the 5-minute bars |
| Trades | 58 | calculated |
| Winning trades | 24 (41%) | calculated |
| Total, gross | 12.83 dollars per share | calculated |
| Total after 4 cents per trade | 10.51 dollars per share | calculated |
| Average after costs | 2.50 basis points per trade | calculated |
| Standard deviation | 50.59 basis points | calculated |
| t-statistic | 0.38 | calculated: mean / (sd / square root of 58) |
| Trades for t = 2 | about 1,638 | calculated: (2 x 50.59 / 2.50) squared |
What we are unsure of
- Fills are assumed exactly at the range boundary; a real stop or breakout order often fills worse when price jumps through the level, so the result is optimistic.
- When a bar both breaks out and crosses the stop, the order within the bar is unknown; we take the breakout and check the stop from the next bar.
- Five-minute bars hide moves inside the bar; tick data would be more exact.
- Three months is one market regime; the same rule can look very different over years.
Data
- SPY 5-minute bars, regular hours, Yahoo Finance chart data downloaded 2026-11-21; each trade (date, side, points per share, entry price, exit reason) is in the data file beside this page.
- Background: the opening range breakout idea is described by Toby Crabel, Day Trading with Short Term Price Patterns and Opening Range Breakout (1990).