132.9 points behind buy and hold after costs; 3 of 4 checks run · 2 passed.
Bollinger squeeze breakout · 20 of the largest US companies
It returned +15.2% in total over the test, after costs; buying and holding over the same period returned +148.2%, 132.9 points more. The checks test whether the result holds up (on the unseen years, nearby settings and shuffled trades); they don't compare it with buying and holding. It kept making money on the years it wasn't tuned on (+1.3%). The nearby-settings check wasn't run (it isn't run on a group of stocks), so this doesn't show whether the result depends on the exact settings. With the trades in a worse order, the worst fall was much deeper. 210 trades, enough to judge.
The rules: A quiet spell followed by a breakout: enters when the price closes above the upper Bollinger band (20, 2) after the previous day's ATR was under 2% of the price; leaves when the price closes below the middle band. The quiet spell is measured by ATR, a stand-in for a narrow band.
StratLab ran these rules through its own backtest on every stock in the group, after costs, and through four checks: unseen years, nearby settings, a bad-luck drawdown and enough trades. The page shows the return beside buy and hold and the result of each check. A verdict describes the past, not what comes next, and nothing here is investment advice.