Two winners on camera, 141,027 trades in the test
An opening-range fade from a popular trading video, tested exactly as presented across nine years and every stock in our test universe. The video shows two winning trades. The test shows a coin flip.
In our last post we said that most strategies taught in trading videos show no edge once you test them. Here is one of them, in full.
It is a scalp on the opening range, and it is a good example precisely because it sounds right. It has a setup, a filter, an entry signal, a stop and a target. The video walks through two trades, and both win.
The strategy, as presented
The rules, exactly as the video gives them:
- Box the first fifteen minutes. The high and low of the first fifteen-minute bar of the day mark the box.
- Only trade a “liquidity candle”. The box has to be big enough: its range must be at least a quarter of the stock’s average daily range over the last fourteen days.
- Wait for a reversal outside the box. Until 11:00, on five-minute bars, look for a reversal candle outside the box that points back against the colour of the opening bar — a hammer, or an engulfing candle.
- Stop and target. The stop goes just beyond the reversal candle. The target is the far side of the box.
That is the whole method. It fits on an index card, which is what makes it testable.
Two places where a video can cheat without meaning to
Before any number, the rules had to be pinned down, because a demonstration on a chart can quietly use information a trader would not have had at the time.
When do you know a candle is engulfing? Only once it has closed. On a chart after the fact it looks as if you could have been in the trade as soon as price touched the previous candle’s extreme. Live, you could not. So we tested both readings: a loose one that fills on the touch, and a strict one that waits for the candle to close as a true engulfing and enters at the next bar’s open.
When do you enter on a hammer? Also only once it has closed — so at the next bar’s open.
Neither point is dishonest. Both are the kind of thing that turns a coin flip into a winning demonstration, and neither is visible unless you write the rules down as code.
How we tested it
Nine years of five-minute bars, January 2017 to August 2026, every stock in our test universe. The fourteen-day range was computed only from what was known on that day — never with hindsight. The rules were followed to the letter: every setup taken, none skipped because the last three lost.
One thing to be clear about. This test ran on price bars, not through our matching engine. That matters, because bars flatter a strategy: they assume you get filled at the price the bar shows. If a strategy has no edge even on bars, the matching engine will not rescue it.
What came out
A basis point is a hundredth of a percent. Here is the average result per trade, before any cost of trading:
| Version | Trades | Per trade, before costs | Hit target / stop / closed at day end |
|---|---|---|---|
| Loose entry, far target | 217,303 | +0.8 basis points | 18% / 73% / 9% |
| Strict entry, far target | 141,027 | +1.2 basis points | 18% / 73% / 9% |
| Strict entry, near target | 141,027 | −1.2 basis points | 33% / 57% / 9% |
Around one basis point per trade, either way, is less than it costs to trade. After a realistic cost, all three lose. The best single signal on its own — the hammer, without the engulfing entries — barely breaks even.
And it is not a matter of the wrong years. The first half of the period and the second half look the same: near zero.
Why it is a coin flip
The two targets make the reason visible.
With the far target, a winning trade pays about four and a half times what a losing trade costs — but the far side of the box is reached only 18% of the time.
With a nearer target, back into the box, you win more often — 33% of the time — but a winner pays only 0.7 times the risk.
Move the target and the win rate moves to compensate. What does not move is the expectation, which stays at about zero. You are betting on the direction of the open, from the other side, and the market does not pay for that.
The filter that filters nothing
The “liquidity candle” is presented as the quality check that separates good days from bad ones. It is met on about 90% of days. A filter that lets nine days in ten through is not selecting anything. It makes the method feel selective without selecting.
Two winners on camera
The video shows two trades, and both win. The test has 141,027 trades.
Any rule that trades often enough produces winning trades — thousands of them. Picking two for a video is not dishonest; it is what a demonstration is. But two winners out of 141,027 trades tell you nothing about the rule. They tell you which trades made it into the video. That is survivorship bias in its simplest form.
What to take from it
Write the rules down. Resolve every point where the video is vague, in the direction a live trader would actually face. Run them across every stock, not the ones on screen. Then count.
This one came back a coin flip, as most do. If the presenter has an edge, it is not in these rules.
Nine years of five-minute bars, January 2017 to August 2026. Backtest figures of a published strategy, not a track record: past performance does not indicate future results.