The thesis
Mean-reversion runs one of the oldest ideas in technical trading: buy oversold, sell overbought. Concretely, it enters when RSI drops below 30 and exits when RSI climbs above 70, across a 24-name universe of large-cap US equities spanning tech, financials, healthcare, staples, and energy. The logic is intuitive — extreme readings tend to snap back — and the diversified, liquid universe keeps it away from the illiquid corners where mean-reversion signals decay fastest.
Backtest performance
On paper, the numbers look inviting. Over 451 days the strategy returned 14.73% (7.98% CAGR), finishing at $11,473 on a $10,000 base, with a 70.59% win rate across 38 trades. That hit rate is the headline strength — most trades land.
The caveats sit right underneath. A Sharpe of 0.58 is modest, meaning the returns came with meaningful volatility rather than a smooth ride. Max drawdown reached 15.64%, roughly the size of the total profit — not comforting. And turnover of 879% signals a lot of churn; the 70% win rate is doing heavy lifting to stay ahead of costs, even if fees here were a trivial $38.
Validation: the strategy that didn't hold up
This is where the story turns. Walk-forward validation failed. Splitting the history into four folds, three were positive — but the most recent fold (Dec 2025–May 2026) lost -2.84% with a -0.33 Sharpe and a 14.96% drawdown, the worst of the set. The earlier folds carried the record; the freshest, most decision-relevant window did not.
The deflated statistics tell the same tale. Probabilistic Sharpe (PSR) of 0.785 is borderline, but the Deflated Sharpe Ratio of 0.304 — which penalizes for the 6 trials run — sits well below any confidence threshold. In plain terms: after accounting for how many variants were tested, we can't be confident the edge is real rather than lucky.
Live activity: idling
The live account reflects that hesitation. Since a single WMT purchase on May 31, the last six scheduled runs (Aug 3–10) each report 0 executed, 0 rejected — no RSI extremes triggered. Cash has sat frozen at $7,569.25, and total equity has drifted between roughly $9,894 and $9,921, still below the $10,000 start. The strategy isn't losing on bad trades; it simply isn't finding setups, and its existing positions are underwater.
The verdict
Mean-reversion is a well-constructed, easily explained strategy with a genuinely strong win rate — the kind of clean thesis worth keeping in the lab. But the evidence says not yet: out-of-sample returns turned negative, the deflated Sharpe flunks, and live performance is flat-to-down with no fresh signals. The right posture is patience. Let it keep paper-trading, watch whether the next fold recovers, and treat the failed validation as the honest warning it is rather than noise to trade through.