Thesis
Mean-reversion runs one of the oldest ideas in technical trading: buy what looks oversold, sell what looks overbought. Concretely, it enters when RSI drops below 30 and exits when RSI climbs above 70, applied across a 24-name universe of US large caps spanning tech (AAPL, MSFT, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE), and staples (PG, KO, WMT, COST). It's a bet that short-term price extremes in liquid mega-caps tend to snap back.
Recent Activity
The live book has been quiet. Of the last six scheduled runs (Sep 21–28), only two executed a trade, and the rest passed with zero fills — unsurprising for an RSI strategy that only acts at extremes. Recent executed buys include MCD (10 shares @ $238.58), HD (8 @ $299.12), and BAC (42 @ $58.21); no sells appear in the recent tape.
Worth flagging: after the Sep 21 and 24 buys, cash fell from $2,720.79 to $329.71, so the book is now close to fully deployed. Total equity has also drifted down over the window — from $9,841.20 on Sep 22 to $9,619.30 on Sep 28 — a modest but consistent slide as positions sit and wait for reversion.
Backtest Performance
Over 451 days the strategy returned 14.73% (final equity $11,473.17, ~7.98% CAGR), with a 70.59% win rate across 38 trades. Costs were negligible: $38 in fees, no FX. The headline win rate is genuinely attractive, and the drivers make sense — mean-reversion tends to win often with small gains and lose occasionally with larger ones.
The caveats sit in the risk stats. A Sharpe of 0.58 is mediocre, and a 15.64% max drawdown against a ~14.73% total return means you endured roughly as much peak-to-trough pain as you ultimately earned. Turnover of 879% also signals a strategy that trades a lot to extract that return.
Validation: The Red Flag
This is where enthusiasm should cool. The walk-forward validation failed. Across four folds, three were positive (+2.06%, +11.10%, +2.21%) but the most recent — Dec 2025 to May 2026 — lost 2.84% with a negative Sharpe of −0.33 and a 14.96% drawdown. That fold is also the out-of-sample readout: OOS return −2.84%, OOS Sharpe −0.33.
The multiple-testing statistics reinforce the concern. Against 6 trials, the deflated Sharpe ratio (DSR) is just 0.304 — well below the confidence you'd want that the edge is real rather than the best of several attempts. The probabilistic Sharpe (0.785) is more forgiving but not decisive.
Bottom Line
Mean-reversion has a coherent thesis, a high hit rate, and low trading costs — real strengths. But the shape of its returns (deep drawdowns relative to gains, high turnover) and, more importantly, a failed cross-validation with a negative most-recent fold argue for caution. The 70% win rate is doing a lot of narrative work that the risk-adjusted and out-of-sample numbers don't support. It stays live, but on a short leash: I'd watch whether the current near-fully-invested book can reverse the recent equity slide before adding conviction.