KILLED · the deadest thing we ever tested
Hypothesis
Price touching the lower Bollinger Band (20, 2σ) is "oversold" and mean-reverts to the mid-band — the entry taught in virtually every beginner course.
Math — the band
Bands are a rolling z-score envelope:
$$ BB_{lower} = \mu_{20} - 2\sigma_{20}, \qquad \mu_{20}=\tfrac{1}{20}\sum_{i=1}^{20}C_{t-i} $$
The taught trade: long the touch, exit at $\mu_{20}$.
Method
Long on close back above the band after a touch, exit at mid-band, −3% failsafe, 12h cap. 30 most-liquid USDT-M perps, 84 days of 5m bars, 0.20% round-trip friction. n = 11,938 trades.
Results
| Trades | 11,938 |
| Win rate | 40% |
| Mean net per trade | −0.20% |
| Profit factor | 0.48 |
| t-statistic | −19.9 |
Why it failed
- A 2σ touch on a trending asset is not "oversold" — it is the trend working. The band is a volatility envelope, not a value signal.
- The gross edge is a few basis points at best; 20 bps of friction per round trip turns a coin-flip into a wood-chipper.
- t = −19.9 across twelve thousand trades. This is not "needs better filters". This is dead at any confidence level you like.
The single most-taught entry in retail trading is, on modern crypto perps, a fee-donation machine. PF 0.48 on n=11,938. No filter we tried rescues it.
Mean-reversion signals must beat the friction floor, not zero. On 5m crypto, 0.20% round trip is a wall that band-touch reversion never clears.