The thesis
Bollinger-reversion is a textbook mean-reversion play: buy when price closes below the lower Bollinger band, sell when it pushes above the upper band. It runs live across a 24-name universe of large-cap US equities — mega-cap tech (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), staples (PG, KO, WMT, COST), and healthcare (JNJ, UNH, PFE, ABBV). The logic bets that stretched moves snap back. It's simple, interpretable, and easy to reason about — which is exactly why we should hold it to a high evidentiary bar.
Backtest performance
Over 451 days the strategy returned 17.55% (final equity $11,755 on a $10k base), a 9.46% CAGR, with a 63.89% win rate across 76 trades. Those are healthy hit-rate numbers. The catch is a Sharpe of 0.66 and a max drawdown of 20.57% — you're taking meaningful volatility for that return. Turnover clocks in at a hefty 1,720%, so the strategy trades actively; fees totalled $76, modest here but a drag that scales with churn.
Validation: where it stumbles
Walk-forward validation did not pass, and the reason is instructive. All four out-of-sample folds were positive — a genuine strength, suggesting the edge isn't a single lucky window. But look at the trajectory:
| Fold | Return | Sharpe |
|---|---|---|
| 1 (Aug'24–Jan'25) | 7.67% | 1.78 |
| 2 (Jan–Jul'25) | 1.42% | 0.25 |
| 3 (Jul–Dec'25) | 0.77% | 0.20 |
| 4 (Dec'25–May'26) | 0.40% | 0.14 |
The edge is monotonically decaying. Most of the headline return came from the earliest fold; the most recent out-of-sample stretch returned a near-flat 0.4% at a Sharpe of 0.14. The probabilistic Sharpe ratio (0.814) looks reassuring, but the deflated Sharpe ratio of 0.342 — which discounts for the 6 trials run — signals the performance may not survive multiple-testing scrutiny. That combination is the classic fingerprint of a strategy that fit the past better than it forecasts the future.
Recent live activity
The live account tells a consistent story. The last executed trades were on 2026-08-07 (buy 1 CAT at $843.72, sell 25 DIS at $105.15). Since then, every scheduled run from 2026-08-13 through 2026-08-20 executed zero trades, rejecting one or two candidates each time. Cash has sat flat at $2,752.28 while total equity drifted from ~$10,036 down to $9,623 — the strategy is holding, not acting, and slipping modestly.
Verdict
The strengths are real: four-for-four positive folds and a 64% win rate argue there's a genuine mean-reversion signal here. But the decaying out-of-sample returns, sub-0.35 deflated Sharpe, 20%+ drawdown, and a live book that has stopped transacting all point the same way. This is a candidate for paper-trading observation, not capital — until the edge shows it can hold up in the most recent regime, bollinger-reversion is a strategy to watch, not trust.