Chart patterns and candlestick setups
Does breakout and retest work?
Rejected Pre-registered and tested, July 2026
The setups the rule throws away outperformed the ones it keeps.
What the pattern claims
One of the most widely taught entries there is. Price breaks above a level it has failed at before, on heavy volume. Rather than chasing it, you wait: price comes back down to test that old level, the level holds as support, and then you buy. The claim is that the retest confirms the breakout was real and gives you a better price.
How we tested it
Formalised from a chart the owner supplied. A close above the 60-session high on at least 1.5 times average volume, in an uptrend; then within ten sessions a retest coming within 2% of that level and holding it without closing 3% below. 1,425 occurrences across 760 S&P 600 small-cap companies, 2021–2026, corrected for companies that later disappeared.
Every study here was pre-registered: the exact rule, the pass and fail thresholds and the data window were written down and cryptographically fingerprinted before the test was run. That makes it impossible to move the goalposts after seeing the answer.
Results assume you buy at the next morning's open, not at the closing price that triggered the signal, and they include companies that were later delisted or went bust. Returns are measured against what the rest of the market did on the same days, so a rule that made money only because it fired on days everything rose scores zero here.
What happened
The full pattern returned −3.48% over 60 days, and was beaten by all five of its control groups — including simply buying the breakout and skipping the retest entirely.
But the real result is a comparison inside the setup itself. We split every breakout-and-retest into the ones where support held and the ones where it broke. Those two groups together account for every occurrence exactly, so there is no selection effect between them.
Setups where support held returned −0.89%. Setups where support broke returned +0.08%. The condition the pattern is named for picked the worse half.
What it means
Two things carry beyond this study.
First, waiting for confirmation has now cost money in four out of four tests we have run on it. Here, buying the breakout with no wait beat waiting by about one percentage point.
Second, a warning about how convincing the raw numbers looked. Measured plainly, the pattern returned −3.48% with a t-statistic of −6.11, which is the kind of number that would normally settle an argument. Once we accounted for the fact that these setups cluster on the same calendar days, it became −0.89% with a t-statistic of −1.88. A t of −6 that was really −1.9. Signals conditioned on volume spikes fire together, and if you do not correct for that you will believe almost anything.
A rejection is not "this never works"
It means: tested on this universe, over this period, against a bar written down in advance, it did not clear the bar. Where a test lacked the power to decide either way we record it as undecided rather than rejected. The full research ledger has all 192 experiments and what the failures have in common.
Other patterns we tested
- Does the bull flag pattern work? The pole is the only live part of the pattern. Every flag condition made the result worse.
- Does MACD divergence work? The loose version changed the outcome by less than one hundredth of a percent. The strict version widened both tails.
- Does buying the dip after a spike work? The first pattern in this family to lose money significantly rather than merely do nothing.
- Does a big RSI drop mark a buying opportunity? The family's only statistically significant result, and it vanished completely once we accounted for when it fired.
- Do three shrinking red candles signal a bottom? The only chart pattern that beat all its own control groups — and it still could not clear the error bars.
- Does a tight base after a sharp drop work? Removing the base conditions improved the result. The base was the problem.