The thesis
Mean-reversion is one of the oldest ideas in systematic trading: prices that stretch too far from their recent average tend to snap back. This strategy implements the textbook version of that bet — buy when RSI falls below 30 (oversold), sell when RSI climbs above 70 (overbought). It runs across a 24-name universe of large-cap US equities spanning tech, financials, healthcare, staples, and energy — the kind of liquid, well-behaved names where clean RSI signals are most plausible.
Backtest: the good news
On paper, the numbers are attractive. 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. Seven of every ten trades closed green — exactly the profile you'd expect from a strategy that harvests small, frequent snap-backs.
But two figures temper the enthusiasm. The Sharpe ratio is a modest 0.58, meaning returns came with meaningful volatility, and the max drawdown reached 15.64% — larger than the annual return itself. Turnover of 879% also flags a busy, fee-sensitive approach; here fees were only $38, but at scale that churn matters.
Validation: the warning
This is where the story turns cautious. The walk-forward validation gate failed. Across four folds, three were positive — but the most recent fold (Dec 2025 → May 2026) lost 2.84% with a negative Sharpe of -0.33 and a 14.96% drawdown. The out-of-sample results echo that: OOS return of -2.84% and OOS Sharpe of -0.33.
The deflated statistics are more encouraging than the raw ones suggest a strategy this simple should worry about overfitting, and with only 6 trials the multiple-testing penalty is light — PSR sits at 0.785 and the Deflated Sharpe Ratio at 0.304. Read together, the signal is real but thin: it worked in calmer regimes (fold 2 returned 11.1% at a 1.32 Sharpe) and stumbled in the most recent one.
Live activity
The live paper account has gone quiet. The last six scheduled runs (July 27 through Aug 3) each executed zero trades — unsurprising, since RSI thresholds only fire at extremes, and nothing in the universe has been decisively oversold or overbought. The account total has drifted between roughly $9,901 and $9,971, sitting at $9,893.95 on Aug 3 with $7,569.25 in cash. The last actual fill was a WMT buy of 21 shares at $115.75 back on May 31.
Verdict
Mean-reversion is a coherent, interpretable strategy with a genuinely strong hit rate — but the failed validation and negative most-recent fold are honest red flags. It deserves to keep running as a live paper experiment, not fresh capital, until it proves it can navigate the regime that broke fold 4.