The thesis
The Momentum SMA20 Rider is a trend-following strategy with a clean, testable logic. It ranks a 24-name large-cap universe — the usual megacaps plus staples, financials, and energy (AAPL, MSFT, NVDA, JPM, KO, XOM, CVX, and so on) — by distance above the 50-day moving average, then buys the leaders only when price is also holding above its SMA20. That double confirmation is the whole idea: not just "it's gone up," but "it's gone up and the shorter-term structure agrees." Exits are equally mechanical — cut on RSI14 above 75 (overbought) or a close back below SMA20 (structure broken) — with capital rotating into higher-ranked names when the book is full.
Backtest and validation
On paper the numbers are attractive: an 81% total return over 1,233 trading days, a 12.89% CAGR, and a final equity of $18,100 from a $10k base. Crucially, the strategy passed walk-forward validation — 3 of 4 folds positive, an out-of-sample return of 43.48%, and a probabilistic Sharpe ratio (PSR) of 0.986, which is genuinely strong.
But read the folds, not just the headline. Fold 4 (mid-2025 to mid-2026) did the heavy lifting: +43.48% at a 2.18 Sharpe with only a 5.38% drawdown. Fold 3, by contrast, was essentially flat (-0.28%) while suffering a 23% drawdown — a full-sized loss of capital for no gain. The deflated Sharpe ratio (DSR) of 0.583, which discounts for the 24 configurations trialled, is the more sober read: real edge, but a fraction of the raw figure. The full-sample Sharpe of 0.98 says the same thing — this is a below-1.0 risk profile, not a smooth compounder.
Recent activity
The live paper book has been quiet and slightly heavy. Recent scheduled runs mostly executed zero trades, and total equity has drifted from ~$10,095 (Sep 3) to ~$9,780 (Sep 10) — a modest give-back. When it does trade, the whipsaw shows: KO was bought at $89.75 on Aug 28, sold at $88.05 on Sep 1, then re-bought at $87.82 on Sep 9. MSFT round-tripped more happily, bought at $483.90 and sold at $491.56.
The honest risks
Two things demand attention. First, the 37% win rate — this strategy is wrong on nearly two trades out of three, and depends entirely on winners running far enough to pay for the losers. Second, turnover of ~27,000% across 1,427 trades. That churn is a structural drag; even at a modest $1 per trade the fee load is non-trivial on a small book, and the KO round-trips are exactly the kind of noise-driven flip-flopping momentum systems are prone to.
Verdict
A validated, honestly-tested momentum strategy with a real but modest edge. The out-of-sample strength is encouraging; the flat fold, low hit rate, and heavy turnover are the price of admission. Worth riding — with eyes open to the drawdowns.