Thesis
momentum-code runs a simple, legible idea: buy the top positive movers across a 24-name large-cap universe — the usual mega-cap tech, financials, staples and energy — with a cap per position to avoid concentration. There is no regime filter or hedge here; it is pure long momentum, and its behaviour should be read in that light.
Backtest and validation
On a 451-day backtest the strategy returned 19.76% (final equity $11,976, CAGR 10.6%), against a 20.48% max drawdown and a Sharpe of 0.65. Turnover was modest at 92.74% and fees were negligible ($5 total). The headline reward-to-risk is unremarkable: roughly one unit of drawdown for each unit of return.
The walk-forward work is where it looks better. All four folds were positive (+15.38%, +3.27%, +12.18%, +13.71%), with fold Sharpes of 2.47, 0.46, 2.31 and 1.93, and the most recent out-of-sample fold returning 13.71% at a 1.93 Sharpe. Consistency across non-overlapping windows is genuinely encouraging.
And yet the validation gate reports passed: false. The reason is in the risk-adjusted, multiple-testing-aware numbers: the probabilistic Sharpe ratio is a respectable 0.81, but the deflated Sharpe ratio is only 0.34 after penalising for 6 trials. In plain terms, once you account for how many variants were tried before landing on this one, the edge is no longer statistically convincing. That is the gate doing its job, not a bug.
A caveat on win rate
The reported win rate of 0% across 5 trades should not be taken at face value. The five executed trades (XOM, NVDA, BAC, HON, MSFT, bought late May–early June) are all still open positions, so there are no realised wins to count. With a positive total return, the zero here reflects unclosed trades, not five losers.
Recent activity — the real concern
Live behaviour is the flashing light. Across six consecutive scheduled runs (2026-07-31 through 2026-08-07), the strategy executed zero trades and rejected between one and three orders each day. Cash has been pinned at $608.79 the entire week — the book is effectively fully invested, and every fresh momentum signal is being turned away for want of buying power. Portfolio value drifted from $10,372 up to $10,734 over that stretch, so the held positions are carrying it, but the strategy is no longer doing anything.
Verdict
The positives are real: four-for-four positive folds and a healthy PSR. The risks are equally real: a 20% drawdown, a full-sample Sharpe of 0.65, a deflated Sharpe that fails the gate, and a live book that has been inert for a week. The prudent read is that momentum-code has a plausible but unproven edge that the validation framework is right to flag — and that its cash constraint needs fixing before the thesis can even be tested going forward.