Thesis
Bollinger-reversion is a classic mean-reversion play: buy when price dips below the lower Bollinger band, sell when it pushes above the upper band. It runs against a 24-name large-cap universe spanning tech (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV) and staples/industrials (PG, KO, WMT, COST, CAT, XOM). This is a hand-written, built-in strategy — the reviewer approved it at risk 0 back on 2026-05-31 — so there is no black-box authorship to second-guess here, only the edge itself.
Recent activity
Trading has been quiet. The 2026-10-29 scheduled run executed two buys — CAT (3 shares @ $647.16) and GOOGL (8 shares @ $283.35) — lifting the book to a total of $9,957.81. The following day's run (2026-10-30) executed nothing and the total slipped to $9,859.39, leaving the strategy fractionally underwater against its $10k Season 2 reset. Notably, it is carrying a large cash balance ($5,749.53, roughly 58% of the book), consistent with a strategy that only deploys when bands are breached and otherwise waits.
Backtest and validation
On paper the full backtest looks appealing: 17.55% total return over 451 days (9.46% CAGR), a 63.89% win rate across 76 trades, and final equity of $11,755. Fees were a negligible $76 and FX cost zero. But two numbers temper the enthusiasm — a Sharpe of just 0.66 and a 20.57% max drawdown. For a strategy that trades infrequently, swallowing a one-fifth drawdown for sub-1.0 risk-adjusted returns is a meaningful ask.
The walk-forward validation is where the thesis wobbles. Headline flags: validation failed, despite all four folds being positive. The giveaway is the trend:
| Fold | Window | Return | Sharpe |
|---|---|---|---|
| 1 | 2024-08 → 2025-01 | 7.67% | 1.78 |
| 2 | 2025-01 → 2025-07 | 1.42% | 0.25 |
| 3 | 2025-07 → 2025-12 | 0.77% | 0.20 |
| 4 | 2025-12 → 2026-05 | 0.40% | 0.14 |
That is monotonic decay. The edge that looked real in late 2024 has all but evaporated by mid-2026. The out-of-sample figures confirm it — OOS return of 0.40% and OOS Sharpe of 0.14. A PSR of 0.814 is encouraging, but the deflated Sharpe ratio (DSR) of 0.342, adjusted for 6 trials, says the apparent skill does not clear the multiple-testing bar.
Verdict
The strengths are genuine: a clean, interpretable rule, a high win rate, and every fold in the black. But the risks dominate. Returns compress fold-over-fold, out-of-sample performance is essentially flat, and the deflated statistics flag probable overfitting. A 63.89% win rate paired with a 0.14 OOS Sharpe suggests the losers are large relative to the frequent small wins. Bollinger-reversion earns its place as a live benchmark, but on this evidence it is a strategy to watch sceptically — not to size up.