Strategy Thesis
Bollinger-reversion is a classic mean-reversion play: buy when price closes below the lower Bollinger band and sell when it climbs back above the upper band. The universe is deliberately blue-chip — 24 large-caps spanning tech (AAPL, MSFT, NVDA, GOOGL), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV), consumer staples (PG, KO, WMT, COST), and a handful of cyclicals (CAT, HON, DIS, NKE). The idea: established names with mean-reverting price behaviour, enough liquidity to fill cleanly, and low idiosyncratic blow-up risk.
Backtest Snapshot
Across 451 days and 76 trades, the strategy returned 17.55% (9.46% CAGR) with a win rate of 63.89% — better than a coin flip by a meaningful margin. The Sharpe of 0.66 is modest but not embarrassing for a strategy running at this cadence.
The pain point is drawdown. A 20.57% max drawdown is steep relative to the return, and turnover of 1,720% means the strategy is churning the portfolio aggressively. With $1 per-trade fees already baked in (76 total, $76 in fees), slippage on real size could erode the edge further.
Walk-Forward Validation: A Clear Decay Pattern
This is where the story gets complicated. The four-fold walk-forward validation tells a consistent, troubling 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% |
All four folds are positive — which sounds encouraging — but the trajectory is unmistakable: returns and risk-adjusted performance deteriorate monotonically from fold to fold. The out-of-sample Sharpe lands at 0.14 and the OOS return at 0.40%, barely above zero. The Probabilistic Sharpe Ratio of 0.814 is reasonable, but the Deflated Sharpe Ratio of 0.342 — which adjusts for the number of trials (6) — reflects the multiple-testing penalty and pulls confidence down sharply. Validation status: failed.
Recent Live Activity
The live paper-trading run corroborates the walk-forward picture. The last executed trades were in early June — a DIS buy at $99.18, a round-trip on PG (bought at $140.03, sold at $148.40 nine days later), and simultaneous entries in GOOGL, WMT, and COST at end of May. Since then, six consecutive daily runs across July 1–8 have produced zero executions, with 1–2 signals rejected per session. Portfolio equity has ranged from $9,706 to $9,893 during this period, reflecting mark-to-market drift on existing positions rather than active trading.
The signal drought likely means the current market is not pushing these blue-chips to band extremes — consistent with a lower-volatility or trending regime where mean-reversion setups simply don't materialize.
Strengths and Risks
Strengths: High win rate, intuitive and auditable logic, a diversified large-cap universe that limits single-stock blowup risk, and all four OOS folds finishing in the green.
Risks: Monotonically decaying performance across folds is a red flag — the strategy may have been partially fit to the 2024 volatility environment. Max drawdown at 20.57% is uncomfortably large relative to the cumulative gain. The DSR of 0.342 indicates the edge is fragile after correcting for multiple trials. And the high turnover creates real friction that backtests are unlikely to fully capture.
Outlook
Bollinger-reversion is worth watching but not yet worth trusting with live capital. The next productive step is regime-filtering: does performance hold when the broader market is in a mean-reverting rather than trending state? A volatility gate or VIX-conditional entry filter could help it concentrate trades in environments where the edge is genuinely present.