The thesis
momentum-code runs one of the oldest ideas in systematic trading: buy the top positive movers in the universe, capped per position. It works across a concentrated set of 24 US large-caps — mega-cap tech (AAPL, MSFT, NVDA, GOOGL), financials (JPM, BAC, V, MA), healthcare, staples, and energy. The per-position cap is the strategy's main risk control, and the universe's liquidity keeps execution assumptions honest.
Backtest performance
Over a 451-day backtest, the strategy returned 19.76%, ending on $11,976 of equity for a 10.6% CAGR. Risk-adjusted, the picture is more sober: a Sharpe of 0.65 against a max drawdown of 20.48%. Turnover of 92.74% is moderate, and fees were negligible ($5 total), so costs are not eroding the edge.
One figure demands a caveat: a reported win rate of 0% across just 5 trades. Taken literally that contradicts the positive return, so it is almost certainly an artifact of how open, unrealized positions are counted rather than evidence the book never made money. Still, five trades is a thin sample — treat the headline return as suggestive, not conclusive.
Walk-forward validation
This is where the strategy earns real credibility. Across 4 out-of-sample folds, all 4 were positive (15.38%, 3.27%, 12.18%, 13.71%), with fold Sharpes of 2.47, 0.46, 2.31, and 1.93. The most recent OOS window returned 13.71% at a 1.93 Sharpe with a shallow 6.32% drawdown. Consistency across regimes — including the weak-but-positive early-2025 fold — is the kind of robustness that survives a live deployment.
Yet the validation gate reports passed: false. The reason is in the deflated metrics: a PSR of 0.811 is respectable, but the Deflated Sharpe Ratio sits at just 0.338 after accounting for 6 trials. In plain terms, once you penalize for how many variants were tested, the evidence isn't strong enough to rule out luck. The gate is doing its job — flagging that four good folds on a five-trade engine is a small foundation.
Live activity
The live sleeve tells a more frustrating story. The last executed trades were a cluster of buys on 31 May–1 June 2026 (XOM, NVDA, BAC, HON, MSFT). Since then, every scheduled run from 22–29 July executed 0 trades and rejected 2–3 candidates each. Cash is pinned at $608.79 and total equity has drifted sideways around $10,080–$10,130 — roughly flat on a $10,000 base.
The bottom line
momentum-code has a defensible edge and unusually consistent walk-forward results, but two yellow flags keep it from a clean bill of health: a failed deflated-Sharpe gate rooted in trial multiplicity, and a live book that is fully invested and rejecting new candidates rather than compounding. Watch whether rejected trades reflect prudent risk caps or a stalled allocation engine.