The thesis
Bollinger-reversion is a textbook mean-reversion play: buy when price falls below the lower Bollinger band, sell when it pushes above the upper band. It trades a 24-name large-cap universe spanning tech (AAPL, MSFT, NVDA, GOOGL), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV) and consumer staples (PG, KO, WMT, COST). The premise is that liquid blue chips over-extend and snap back — a bet on noise, not trend.
Backtest performance
On paper, the numbers are inviting. Over 451 days the strategy returned 17.55% (9.46% CAGR), finishing at $11,755 on a $10k base with a 63.89% win rate across 76 trades. But the quality metrics temper the enthusiasm: a Sharpe of just 0.66 and a 20.57% max drawdown mean returns came with real volatility. Turnover of 1,721% signals a hyperactive book, though at $76 total fees the frictional cost stayed modest.
The validation problem
This is where the story turns. Walk-forward validation failed the gate, and the fold-by-fold breakdown shows why. All four folds were positive — but the edge is front-loaded and fading:
- Fold 1 (Aug 2024–Jan 2025): +7.67%, Sharpe 1.78
- Fold 2: +1.42%, Sharpe 0.25
- Fold 3: +0.77%, Sharpe 0.20
- Fold 4 (Dec 2025–May 2026): +0.40%, Sharpe 0.14
That is monotonic decay toward zero. The headline 17.55% is dominated by one early, favourable regime. Out-of-sample return was 0.40% at a 0.14 Sharpe — statistically indistinguishable from doing nothing. The Deflated Sharpe Ratio of 0.342, adjusted for the 6 trials run, confirms the in-sample Sharpe doesn't clear the bar once you account for selection. A PSR of 0.814 looks healthier, but DSR is the stricter, more honest read here.
Live activity: quietly stalled
The live book reflects that fading signal. The last executed trades were on 7 August 2026 (bought 1 CAT, rotated DIS). Every scheduled run since — 7 September through 14 September — reports 0 executed, 2–3 rejected. With $2,752 cash idle and portfolio value drifting between roughly $9,570 and $9,764, the strategy is sitting below its $10k starting mark while it waits for band-touch entries that aren't triggering.
Verdict
The strengths are genuine: a high win rate, a coherent thesis, and consistently positive folds. But the risks dominate. The out-of-sample edge has eroded to near-nothing, the failed validation gate is the correct call, and live performance is underwater and inactive. Bollinger-reversion looks like a strategy whose backtest flattered an early regime that markets have since arbitraged away. It belongs in observation, not conviction.