The thesis
"Momentum SMA20 Rider" is a textbook trend-follower dressed in moving averages. It ranks a 24-name large-cap universe — the usual megacap-tech-plus-defensives roster (AAPL, MSFT, NVDA, JPM, KO, XOM, and friends) — by how far each trades above its SMA50, then buys the leaders provided price is also holding above the SMA20. The exit discipline is where the character shows: positions are cut when RSI14 pushes past 75 (overbought) or when price closes back below the SMA20 (trend structure broken), with capital rotating into stronger ranks once the book is full.
It's a coherent, well-worn idea: ride confirmed uptrends, refuse to argue with a broken trend, and don't overstay into froth.
The backtest numbers
On paper, the strategy delivers. Over 1,233 trading days it turned a notional book into $18,100 of final equity — an 81% total return, or roughly 12.89% CAGR, at a Sharpe of 0.98. Maximum drawdown was a stomachable 20.91%.
Two caveats leap off the tape. First, the win rate is only 37.13% across 1,427 trades. That's not a flaw so much as a signature — momentum systems live by cutting losers small and letting a minority of winners run. But it means the equity curve depends on a handful of trades doing the heavy lifting, and it will feel unpleasant to hold. Second, turnover is a staggering 26,971%. That churn racked up $1,427 in fees (about a dollar a trade), a real drag that any live deployment must respect.
Does the validation hold up?
This is the encouraging part. The strategy passed walk-forward validation across four folds, three of them positive, with a healthy out-of-sample return of 43.48% and an out-of-sample Sharpe of 2.18. The Probabilistic Sharpe Ratio of 0.986 is high, and the Deflated Sharpe Ratio of 0.583 — after accounting for 24 trials — stays comfortably positive, suggesting the edge isn't purely a data-mining artefact.
The folds tell an honest story, though. Fold 2 (+31.84%) and fold 4 (+43.48%) were excellent; fold 4 in particular paired big returns with a tiny 5.38% drawdown. But fold 3 was essentially flat at -0.28% with the worst drawdown of the set (23.02%). Trend-following pays out lumpily, and this strategy will have long, frustrating stretches that go nowhere.
Live activity
Live, the book is quiet and nearly fully invested — cash sits at $337.64. The last two weeks show mostly zero-execution scheduled runs, with occasional single trades: a KO position bought at 87.82 on Sept 9 after being sold at 88.045 on Sept 1 and originally bought at 89.75 — a small round-trip that hints at whipsaw near the SMA20 line. Recent rotation trimmed MSFT, V, JPM, and BAC while adding CVX and JNJ.
Notably, total portfolio value stood at $9,842 on Sept 11 — modestly below a $10,000 starting line. The simulated edge is real and validated; the live proof is still being written.
Verdict
A well-constructed, validation-backed momentum system with a defensible edge. Watch the turnover, brace for the low hit rate, and judge it over quarters, not weeks.