The thesis
Momentum SMA20 Rider is a mechanical trend-follower running live over a 24-name large-cap universe (AAPL, MSFT, NVDA, JPM, XOM and peers). It ranks candidates by how far they trade above their SMA50, buys only names whose price is also holding above the SMA20 — a confirmation filter — and rides the move. Exits are rule-bound: an RSI14 above 75 flags overbought froth, and a close below SMA20 signals the trend structure has broken. When the book is full, capital rotates out of laggards into higher-ranked names. There is no discretion here — it is a clean "cut losers, let winners run" design.
Recent activity
The strategy fires once per scheduled weekday-evening run. Over the past week it kept a steady cadence: it sold AAPL (5 @ $339.88) and MSFT (3 @ $526.14), and bought V (4 @ $403.62) and WMT (16 @ $118.23). Earlier rotations trimmed DIS, NVDA and CAT and added XOM and ABBV. Notably, roughly half of recent runs logged a rejected order alongside the executed one (e.g. 2026-10-20: 1 executed, 1 rejected), which hints that cash or sizing constraints are regularly blocking the second-choice trade. Live equity has hovered in a tight band around $9,700–$9,750 all week — essentially flat, not the fireworks the backtest implies.
Backtest and validation
On paper the record is strong: +81% total return over 1,233 days, a 12.89% CAGR, Sharpe 0.98 and a 20.9% max drawdown. The win rate is only 37%, which is textbook momentum — a minority of large winners carrying a majority of small losers. Validation is where the strategy earns real credibility. Across four walk-forward folds, three were positive, the most recent out-of-sample segment returned 43.5% at a 2.18 Sharpe, and the probabilistic Sharpe ratio sits at a confident 0.986.
Strengths and risks
The honest caveats live in two numbers. The deflated Sharpe ratio — which penalises for the 24 trials run during development — falls to 0.58, well below the raw figure; some of the headline performance is selection luck, not pure edge. And fold 3 (2024-02-28 to 2025-05-17) returned −0.28% with the deepest 23% drawdown: in a choppy, trendless regime the model simply spins its wheels and pays fees. With turnover near 27,000% and a flat $1-per-trade fee across 1,427 trades, cost drag is a permanent headwind.
The verdict: a well-built, properly validated momentum engine whose edge is genuine but conditional. It needs trends to exist, and the current flat live equity is a timely reminder that it does not manufacture them.