The thesis
Momentum SMA20 Rider plays a clean, well-worn idea: names trading furthest above their SMA50, with price still holding above the SMA20, are in confirmed uptrends worth riding. The strategy ranks a 24-name large-cap universe — the usual megacap suspects (AAPL, MSFT, NVDA, GOOGL) alongside financials, staples, healthcare and energy — by distance above the SMA50, then buys the strongest uptrends. Exits are rule-based and unsentimental: cut when RSI14 pushes above 75 (overbought) or when price closes back under the SMA20 (trend structure broken), and rotate capital into higher-ranked names when the book is full.
Backtest and validation
Over 1,233 trading days the backtest grew equity to $18,100.23, an 81% total return, or roughly 12.89% CAGR. Risk-adjusted, that's a Sharpe of 0.98 against a maximum drawdown of 20.91% — respectable, not spectacular.
The validation work is the more reassuring part. Walk-forward testing across four folds passed, with three of four folds positive. Out-of-sample return came in at 43.48% with an out-of-sample Sharpe of 2.18, and the Probabilistic Sharpe Ratio sits at 0.986 across 24 trials, with a Deflated Sharpe of 0.583. In plain terms: the edge survives being tested on data it wasn't tuned on, and the deflated figure accounts for the multiple trials run.
The folds tell an honest story, though. Fold 2 (Dec 2022–Feb 2024) was the star at +31.84% (Sharpe 1.59), and Fold 4 carried the out-of-sample strength at +43.48% (Sharpe 2.18, a tidy 5.38% drawdown). But Fold 3 (Feb 2024–May 2025) was essentially flat at −0.28% with a 23% drawdown — a reminder that momentum stalls badly in choppy, trendless regimes.
The risks worth naming
Two numbers deserve a hard look. The win rate is just 37.13% across 1,427 trades — this is a strategy that is wrong more often than it is right and relies entirely on letting winners run while cutting losers short. That works until it doesn't, and it demands discipline the rules thankfully enforce. Second, turnover is a staggering 26,971%. That churn means fees are a structural drag ($1,427 in the backtest), and real-world slippage could erode the paper edge more than the model assumes.
Recent live activity
The live book has been cautious. The 2026-10-01 run executed two sells — trimming AAPL (5 @ $329.83) and ABBV (8 @ $260.50) — lifting cash to $4,286.27 against a total of $9,663.71. The preceding week was quiet: four straight runs with zero executions, the book holding steady around $9.7k. Earlier moves added MSFT and a sizeable PFE position (70 @ $28.42) on 2026-09-24.
Verdict
A genuinely validated momentum engine with sensible exits — but one whose low hit-rate and extreme turnover mean it lives or dies on regime and execution quality. Worth riding; worth watching the fees.