The Thesis
The mean-reversion strategy is one of the oldest systematic ideas in equities: prices that have moved too far, too fast tend to snap back. Headmars implements this with a clean RSI trigger — buy when RSI falls below 30 (oversold), sell when it climbs above 70 (overbought). The universe spans 24 large-cap U.S. names across tech, financials, healthcare, consumer staples, and energy, giving the strategy broad exposure while staying within liquid, well-covered names.
Backtest Performance
Over 451 days of backtest history, the strategy turned $10,000 into $11,473 — a 14.73% total return at a 7.98% CAGR. The win rate of 70.6% across 38 trades is genuinely strong; most trend-following systems would envy that hit rate. Maximum drawdown came in at 15.64%, which is meaningful but not alarming for an equity strategy.
The Sharpe ratio of 0.58 is the first place to pause. It is positive, but comfortably below the 1.0 threshold most risk managers use as a baseline. The strategy earns its return, but the ride is bumpier than the headline numbers suggest.
Turnover of 879% over the period is worth noting — this strategy trades actively, and while per-trade fees are modest at $1 each, high turnover amplifies the impact of any friction that scales with volume.
Walk-Forward Validation: A Mixed Picture
Headmars runs a four-fold walk-forward validation, and this is where the picture gets more nuanced.
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +2.06% | 0.57 | 4.24% |
| 2 | Jan 2025 – Jul 2025 | +11.10% | 1.32 | 10.53% |
| 3 | Jul 2025 – Dec 2025 | +2.21% | 0.46 | 8.80% |
| 4 | Dec 2025 – May 2026 | −2.84% | −0.33 | 14.96% |
Three of four folds are positive, and Fold 2 is genuinely impressive at a 1.32 Sharpe. But Fold 4 — the most recent, and the closest proxy for live conditions — is the worst: a 2.84% loss, a negative Sharpe, and a drawdown of nearly 15%. The validation system correctly flagged this: status is failed.
The Probabilistic Sharpe Ratio (PSR) of 0.785 says there is a 78.5% chance the true Sharpe exceeds zero — encouraging, but not decisive. The Deflated Sharpe Ratio (DSR) of 0.304 is more sobering; after adjusting for the number of trials tested (6), the probability that the strategy's edge is real drops considerably.
Live Activity
The strategy deployed on May 31, 2026, with a $10,000 paper allocation. It opened one position immediately: 21 shares of WMT at $115.75, spending roughly $2,431 and leaving $7,569 in cash. The two subsequent daily runs (June 1 and 2) found no new signals — no buys, no sells. Current portfolio value sits at $9,950, a modest paper loss consistent with the recent drawdown environment the validation data already warned about.
Verdict
Mean-reversion is a credible starting point: the win rate is high, the logic is transparent, and the backtest return is real. But the validation failure is not a bureaucratic technicality — it reflects genuine deterioration in the strategy's most recent live-like period. RSI-based reversion can underperform in trending or momentum-driven markets, and the late-2025 fold suggests that may be exactly the environment the strategy walked into.
This strategy is worth running in paper mode and monitoring closely. Before any capital promotion, the key question is whether the recent fold represents a regime shift or ordinary variance. A longer out-of-sample window and a stronger DSR would both need to improve before this one earns full confidence.