Thesis
momentum-code runs a simple, transparent idea: buy the top positive movers across a 24-name large-cap universe — spanning tech (AAPL, MSFT, NVDA), financials (JPM, BAC, V, MA), healthcare, staples, and energy — with a cap on each position. It is a trend-following, chase-the-strength approach, and it is currently flagged live.
Backtest Performance
Over 451 days, the strategy returned 19.76% (CAGR 10.6%), lifting a notional $10,000 to $11,976. That headline is respectable, but the risk profile deserves scrutiny. The Sharpe ratio is a modest 0.65, and the maximum drawdown reached 20.48% — roughly the size of the entire return. In other words, an investor would have needed to stomach a peak-to-trough decline about as large as the full-period gain. Turnover of 92.74% is high for a five-trade sample, and the reported win rate of 0% reflects that no round-trip positions have been closed yet; every executed trade remains an open long, so there are no realized winners to count.
Validation
Here the picture is genuinely encouraging on one axis and cautionary on another. Across four walk-forward folds, all four were positive (15.38%, 3.27%, 12.18%, 13.71%), and the out-of-sample fold delivered a 13.71% return at a 1.93 Sharpe — notably better than the in-sample 0.65. The probabilistic Sharpe ratio (PSR) sits at a healthy 0.811.
And yet the validation gate reports passed: false. The culprit is the deflated Sharpe ratio (DSR 0.338), which discounts performance for the 6 trials run during development. With multiple-testing baked in, the strategy has not cleared the bar for statistical confidence. That is the honest read: consistent folds, but not enough edge to rule out luck.
Recent Activity — A Warning Sign
The live trail is the most concerning part. The last executed trades were the June 1 buys of XOM and NVDA, following an end-of-May tranche (BAC, HON, MSFT). Since then, every scheduled run from August 7 through August 14 has produced 0 executed, 1–2 rejected. Cash has been pinned at $608.79 the entire time, with total equity drifting between roughly $10,720 and $10,830.
The likely mechanism: the portfolio is fully invested, and the per-position cap plus thin remaining cash means new momentum signals cannot be funded, so they are rejected. The strategy is effectively frozen in its June allocation.
Verdict
momentum-code has a clean thesis and encouraging fold consistency, but three flags temper enthusiasm: a drawdown as large as its return, a DSR that fails the validation gate, and — most pressing — an execution engine that has done nothing but reject trades for over two months. The idea may have merit; the live deployment needs attention before that can be proven.