Chart patterns and candlestick setups
Does a big RSI drop mark a buying opportunity?
Rejected Pre-registered and tested, July 2026
The family's only statistically significant result, and it vanished completely once we accounted for when it fired.
What the pattern claims
A large red candle combined with the Relative Strength Index collapsing by 12 or more points in a single day. The idea is capitulation: a shock violent enough to have flushed out the sellers, with a bounce to follow.
There is a real thought underneath it. An RSI point-change is automatically scaled to how volatile the stock has recently been, because of how the indicator is built. So the same −7% day registers as a shock on a normally quiet company and barely registers on one that has been falling hard for weeks. The question became: does scaling a shock to the company's own recent character pick a better set of shocks than a plain percentage does?
How we tested it
6,444 occurrences on S&P 500 companies, buying at the next morning's open.
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
No, and slightly worse. We built a control that used a plain one-day percentage drop, calibrated purely to fire the same number of times — a −5.10% threshold matched the RSI rule's selectivity to within 1%, and it was set on counts alone, never on returns. The plain percentage returned −0.56% at 60 days. The RSI-scaled version returned −0.65%.
The full signal returned −1.05% at 60 days. The deeper threshold we had pre-registered as the stronger version — a 16-point collapse — was worse at −1.46%. A dose-response running backwards.
What it means
This study produced the first genuinely significant number in eight consecutive tests: a t-statistic of −3.19. It looked like we had finally found something, even if that something was a signal to sell.
Then we measured it against what the rest of the market did on the same days. The effect became +0.10%, with a t-statistic of 0.19, and a median of exactly zero. It disappeared entirely.
The whole of that convincing negative was one fact: this pattern fires on days the market is falling. It was a market-timing observation wearing a stock-selection costume — the third one we had found in a row, and the sharpest. If a shock-based signal shows you a large t-statistic, check when it fires before you believe any of it.
The transferable part: volatility-scaling a price shock adds nothing over just using the raw percentage. There is no need to re-dress this idea in a different oscillator.
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 breakout and retest work? The setups the rule throws away outperformed the ones it keeps.
- Does buying the dip after a spike work? The first pattern in this family to lose money significantly rather than merely do nothing.
- 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.