The Thesis
Mean-reversion is one of the oldest ideas in systematic trading, and this strategy runs a textbook version of it: buy oversold names when RSI drops below 30, sell overbought names when RSI climbs above 70. It operates over a 24-stock universe of US large caps — the megacap tech complex (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV), and staples/industrials (PG, KO, WMT, COST, CAT, XOM). The logic assumes short-term price extremes in liquid, well-covered names tend to snap back toward a mean.
Backtest Performance
On paper, the numbers are attractive. Over 451 days the strategy returned 14.73% (a 7.98% CAGR), finishing at an equity of $11,473 from a $10,000 base. The headline strength is the 70.59% win rate across 38 trades — mean-reversion systems often trade this way, taking many small, high-probability wins.
The caveats sit right alongside. A Sharpe of 0.58 is modest for that return, meaning the ride was bumpier than the total suggests, and the 15.64% max drawdown is real pain. Turnover of 879% is high — this is an active strategy, and while fees came in at only $38, that churn is a friction and a fragility signal.
Validation: The Honest Verdict
This is where the story turns cautious. Walk-forward validation did not pass. Three of four folds were positive, but the most recent fold (Dec 2025 – May 2026) returned -2.84% with a -0.33 Sharpe and a 14.96% drawdown — the out-of-sample window is negative. The earlier folds carried the full-sample result, particularly fold 2's excellent 11.1% / 1.32 Sharpe run.
The deflation metrics tell the same story with more nuance. The Probabilistic Sharpe Ratio of 0.785 is reasonably encouraging, but the Deflated Sharpe Ratio of 0.304 — which penalises for the 6 trials run — is weak. In plain terms: once you account for how many variants were tested, the edge is not statistically convincing, and the strategy's performance is front-loaded into favourable regimes.
Recent Activity
Live, the strategy has gone quiet. The last executed trade was a 21-share WMT buy at $115.75 on 31 May 2026. Every scheduled run since late August has logged 0 executed, 0 rejected — no RSI extremes have triggered. Cash sits parked at $7,569.25, and total equity has drifted between roughly $9,724 and $9,852, meaning the live book is below its starting mark and largely idle.
Balance of Evidence
The strengths are genuine: a clear, interpretable thesis, a high win rate, and a positive full-period return. The risks are equally genuine: failed validation, a negative most-recent fold, a soft deflated Sharpe, and months of live inactivity that leave capital sitting on the sidelines. This is a strategy worth watching, not yet a strategy worth trusting with conviction — the market simply hasn't handed it the extremes it needs, and validation suggests its recent edge may be thinner than the backtest implies.