The thesis
momentum-code is deliberately simple: on each scheduled run it ranks a universe of 24 US large caps — the usual mega-cap suspects across tech, financials, healthcare, staples and energy — and buys the strongest positive movers, with a cap per position so no single name can dominate the book. It is long-only, rules-based, and makes no attempt to time exits with anything clever. The appeal is legibility: you can read the entire thesis in one sentence.
Recent activity
The last six scheduled runs (16–22 Oct) are a flat line. Every run logged 0 executed against 1–3 rejected orders, with cash pinned at $608.79 and total equity drifting in a tight $10,374–$10,473 band. The last real fills were months ago, in late May and early June: entries in MSFT ($450.24), HON ($237.86), BAC ($51.60), NVDA ($224.20) and XOM ($148.69). Since then the book has been effectively static — fully invested with almost no dry powder, which is the likely reason rebalance attempts keep bouncing off the cash constraint rather than filling.
Backtest and validation
Over 451 days the backtest returned 19.76% (10.6% CAGR) to a final equity of $11,976, on just 5 trades and 92.74% turnover, with trivial fees ($5 total). The headline is flattered by context, though: that return comes with a modest 0.65 Sharpe and a steep 20.48% max drawdown — a poor ratio of reward to pain.
Walk-forward cross-validation looks more encouraging on the surface. All four folds were positive (15.4%, 3.3%, 12.2%, 13.7%), and the most recent out-of-sample fold posted a strong 1.93 Sharpe on a contained 6.3% drawdown. Yet the gate still reads passed: false. The reason is in the deflated numbers: with six trials behind the search, the Probabilistic Sharpe (0.81) survives, but the Deflated Sharpe (0.34) does not clear the bar. Penalise for how many variants were tried, and the edge stops being statistically convincing.
Strengths and risks
Strengths: consistency across every validation fold, a transparent one-line rule, a low trade count and negligible costs. Risks: a 20%+ drawdown against a sub-1.0 full-sample Sharpe is uncomfortable; the reported 0% win rate across 5 trades reflects that these are still-open entries with no closed winners yet, not a run of losses; and concentration in a handful of movers plus near-zero cash leaves the strategy no room to act on fresh signals.
Bottom line
momentum-code is a clean, honest momentum rule with respectable headline returns and encouraging fold-by-fold consistency. But the deflated statistics say the apparent edge may be a product of the search, and in live operation it has spent the past week unable to trade at all. Right now it is less an active momentum engine than a buy-and-hold position waiting on cash — worth keeping on the watchlist, not yet worth leaning on.