The thesis
momentum-code is about as unadorned as a trend strategy gets: buy the top positive movers in its universe, capped per position. The universe is 24 US large caps — mega-cap tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), staples (PG, KO, WMT, COST) and energy (XOM, CVX) — so the strategy is fishing in a deep, liquid pond rather than chasing small-cap noise. The per-position cap is the only real risk control baked into the rules, and it does the obvious job of stopping a single hot name from dominating the book.
Recent activity
The live book opened on 29 October with three executions — MA, HD and NKE — followed by WMT and XOM on the 30th. Since then it has gone quiet: the runs on 2 and 3 November executed zero trades and rejected two apiece. With cash sitting at just $177.08, the strategy is effectively fully invested, and the rejections look like the position cap and thin cash doing exactly what they should — refusing to over-concentrate. The trade-off is inertia: once deployed, the book barely rotates.
Performance in the live window is unremarkable and slightly negative. Total equity has drifted from $9,939.98 on day one to $9,875.75 on 3 November, a touch under the implied $10k start. That is noise over five sessions, not signal, but it is worth stating plainly rather than dressing up.
Backtest and validation
The backtest is genuinely encouraging on the surface: 19.76% total return over 451 days, a 10.6% CAGR, a 20.48% max drawdown, and almost nothing lost to friction ($5 in fees, zero FX). Turnover of 92.74% confirms this is a low-churn, hold-the-winners approach. The Sharpe of 0.65, however, is modest — the return came with real volatility.
The walk-forward validation is the more interesting story. All four folds were positive (15.4%, 3.3%, 12.2%, 13.7%), with out-of-sample return of 13.71% and a strong OOS Sharpe of 1.93. On consistency alone, that is a clean sheet. And yet the gate failed. The reason is in the deflated Sharpe ratio of 0.338: after adjusting the 0.81 probabilistic Sharpe for the six trials behind this candidate, the confidence that the edge is real — not a product of search — collapses. This is the validation layer doing its job, resisting a result that looks good precisely because several variants were tried.
Strengths and risks
Strengths: a simple, auditable rule set; four-for-four positive folds; negligible trading costs; and a position cap that visibly restrains concentration.
Risks: a reported win rate of 0 against only 5 trades is almost certainly a metrics artifact given the positive folds, but it flags how little the sample size supports. The headline 0.65 Sharpe and 20% drawdown are unglamorous, the DSR says the backtest edge may not survive multiple-testing scrutiny, and the fully-invested, rejection-heavy live behaviour means the book can go stale. momentum-code earns its live slot on consistency, but the deflated Sharpe is a warning worth heeding, not waving away.