Thesis
bollinger-reversion is a classic mean-reversion play across a 24-name large-cap universe (AAPL, MSFT, NVDA, JPM, XOM, and peers). The rule is simple: buy when price closes below the lower Bollinger band, sell when it pushes above the upper band. The bet is that short-term dislocations in liquid, well-covered stocks revert rather than trend.
Backtest Performance
On paper, the strategy looks appealing. Over 451 days it returned 17.55% (final equity $11,755, ~9.46% CAGR) on 76 trades with a 63.89% win rate. But the headline masks two concerns. The Sharpe of 0.66 is modest, and a max drawdown of 20.57% roughly matches the total return — you endured about as much peak-to-trough pain as you ultimately earned. Turnover of 1,721% is also high; on a real book, slippage and financing would erode the edge that $76 in modeled fees does not capture.
Validation: The Warning Sign
This is where the story turns. Walk-forward validation failed. All four folds were positive, which sounds reassuring, but the trajectory is a steady decay:
- 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
The out-of-sample return collapses to 0.4% with an OOS Sharpe of 0.14. Most of the full-sample gain was earned in a single early window. The deflated Sharpe ratio of 0.342 — well short of a confident threshold — tells us that across 6 trials, the observed edge is not statistically distinguishable from luck. A PSR of 0.814 is the more optimistic read, but the DSR is the one that accounts for selection.
Recent Live Activity
The live account echoes the caution. The last executed trades were in late May and early June 2026 — buys in DIS, GOOGL, PG, WMT, and COST, plus a PG sell. Since then, six scheduled runs (July 8–15) executed zero trades, with a handful of rejections. Cash sits pinned at $1,001.98 and total equity has drifted between $9,710 and $9,882 — below the implied $10,000 start. The bands simply aren't triggering, and what little the book holds is slightly underwater.
Verdict
The strategy's logic is sound and its win rate is genuinely good. But the evidence points to an edge that was strongest early and has faded — exactly the profile validation is designed to flag. Until out-of-sample Sharpe recovers, treat bollinger-reversion as a candidate for re-tuning or retirement, not fresh capital.