Strategy Thesis
Bollinger-reversion is a classic mean-reversion approach applied to a 24-stock large-cap universe spanning tech, financials, healthcare, consumer staples, energy, and industrials. The rule is deliberately simple: buy when price closes below the lower Bollinger band, sell when it closes above the upper band. The bet is that blue-chip equities tend to snap back toward their rolling mean after short-term dislocations — a historically well-documented phenomenon in liquid, institutionally-covered names.
Backtest Headline Numbers
Over the 451-day backtest period, the strategy logged a 17.55% total return (9.46% CAGR) on a starting equity of $10,000, finishing at $11,755.25. The win rate of 63.89% across 76 trades is solid — nearly two in three trades closed profitable. Turnover was high at 1,720.89%, consistent with a short-term tactical strategy cycling through positions frequently.
The Sharpe ratio of 0.66 is acceptable but not impressive. The 20.57% max drawdown is the more concerning figure: it implies the strategy can erase roughly a fifth of capital before recovering, which is a meaningful psychological and capital-management hurdle for any real-money deployment.
Walk-Forward Validation: A Clear Warning Sign
The four-fold walk-forward analysis tells a more cautious story.
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +7.67% | 1.78 | 7.06% |
| 2 | Jan 2025 – Jul 2025 | +1.42% | 0.25 | 20.56% |
| 3 | Jul 2025 – Dec 2025 | +0.77% | 0.20 | 6.66% |
| 4 | Dec 2025 – May 2026 | +0.40% | 0.14 | 11.19% |
The pattern is stark: returns decay fold-by-fold, from a strong 7.67% Sharpe-1.78 performance in the first period down to a near-flat 0.40% in the most recent window. The out-of-sample return is just 0.40% with a Sharpe of 0.14 — barely above noise.
The validation gate correctly flagged this: passed: false. The Probabilistic Sharpe Ratio (PSR) of 0.814 is borderline, but the Deflated Sharpe Ratio (DSR) of 0.342 — which adjusts for the number of trials tested (6) — suggests the in-sample edge does not survive multiple-comparison correction. In plain terms: the strategy may have been fitted to a regime that has since faded.
Recent Live Activity
Despite the validation flag, the strategy continues running in paper-trade mode. Activity over the past week has been light:
- June 9: Sold 17 shares of PG at $148.40, netting a clean gain on the June 1 entry at $140.03 — a textbook band-to-band reversion in roughly eight days.
- June 1: Entered GOOGL (6 shares at $377.90), PG (17 shares at $140.03), WMT (21 shares at $115.75), and COST (2 shares at $956.32) on apparent oversold signals.
- June 2–8: Four consecutive daily runs produced zero or rejected trades, reflecting either no signals or insufficient cash/position sizing room.
Current portfolio value sits at $10,109.08 with $3,506.27 in cash following the PG exit — a modest recovery from the $9,870 trough logged on June 2.
Assessment
Strengths: Transparent, rules-based logic; reasonable win rate; performs well in mean-reverting regimes (fold 1 is genuinely strong).
Risks: Performance degrades monotonically across folds, suggesting either regime dependency or mild overfitting during parameter selection. The 20.57% max drawdown is disproportionate to the Sharpe on offer. The DSR failing with just 6 trials is a red flag — a larger parameter search would likely push it lower.
This strategy earns its place as a live paper-trade monitor, but the validation gate exists for a reason. Promotion to real capital should wait for evidence that fold-4 conditions represent a temporary regime shift, not a permanent edge decay.