Not yet listed. Not yet listed. Over the test window the family lost after costs, and the quality bar needs a positive result over at least twenty trades.
The first fifteen minutes of a trading day are where overnight news gets priced. Orders that queued since yesterday's close hit the book, the price finds a level, and by half past nine most stocks have drawn a short range: a high and a low that mark out what the market thinks the news was worth. When a stock later pushes through the top of that range with more volume than usual, something is still being bought, and the breakout trader buys it too.
It is one of the oldest day-trading ideas there is, and the reason it has survived is not that it works often. It is that when it works, it works by more than it costs when it fails.
The rule, and the asymmetry it depends on
From 09:35 the strategy watches for a five-minute close above the highest point of the previous three bars, with volume above its recent average, while the stock is above its short-term average price. All three, or no trade. A stop sits one to one and a half percent below the entry and a target two to three percent above it, and that ratio is the whole design: the target is twice the stop, so the strategy can lose two trades for every one it wins and still break even before charges.
That 'before charges' is doing a lot of work, and the test shows how much.
What the test found
It lost after costs. With the target twice the stop, the strategy breaks even before charges if it wins one trade in three, and needs to win somewhat more often than that to pay them. Over the window it did not. Brokerage and taxes are charged on every trade, winners and losers alike, and a design that is a wash on paper becomes a slow leak on a statement. The strategy is not wrong about breakouts. It is a business with margins too thin to pay its overheads.
Two losses for every win, with the win twice the size, is break-even on a whiteboard. On a broker's statement it is a loss, by exactly the charges.
The bug that made it look better than it was
An earlier version of this strategy tested cleanly over a year of five-minute prices and placed no trades at all. It compared the close with the highest high of the last three bars, including the current one, and a bar's high is never below its own close, so the condition could not be true. Zero trades, zero losses, and a result that said nothing. The rule now looks at the three bars before the current one, which is what 'the opening range' means, and the test went from empty to honest. An empty test is the most dangerous kind, because it looks like a pass.
Who it might suit, one day
Someone who can sit through weeks of small losses waiting for the morning that pays, which is a specific temperament and not a common one. And, if it is ever listed, only on instruments that move enough in a morning for the target to be worth the charges. That is a design question we have not answered, and the family stays unlisted until we have.
If you take one thing away
The opening range breakout bets that a stock clearing its early high will keep going, and relies on wins twice the size of its many losses. In our test the charges ate the difference, and it stays unlisted.
Think a stricter filter or wider stop fixes opening range breakout? Open the rule sheet, change the numbers, and run your own test.
Customize Opening range breakout