Thesis
momentum-code applies a straightforward rule: each scheduled run scans a 24-name universe of large-cap US equities and buys the top positive price movers, capped per position. No shorting, no leverage, no sector tilts — just a clean cross-sectional momentum filter on names that institutional money already knows well (AAPL, NVDA, JPM, XOM, and twenty others).
The appeal is the simplicity. Momentum is one of the most replicated factors in academic finance, and anchoring it to a blue-chip universe limits blow-up risk from single-name implosions.
Backtest Snapshot
Over a 451-day backtest window, the strategy returned 19.76% on a starting equity of roughly $10,000, compounding at a 10.6% CAGR. The Sharpe ratio came in at 0.65 — respectable but not exceptional — and the maximum drawdown reached 20.48%, which is meaningful for a strategy running only five trades.
Turnover was high at 92.74%, a natural consequence of a signal that rotates into whatever moved most recently. Fees totalled a flat $5, and there were no FX costs given the all-USD universe.
Walk-Forward Validation
Four sequential folds were run from August 2024 through May 2026. All four finished in the green — a consistency point worth noting:
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +15.38% | 2.47 | 6.12% |
| 2 | Jan 2025 – Jul 2025 | +3.27% | 0.46 | 16.55% |
| 3 | Jul 2025 – Dec 2025 | +12.18% | 2.31 | 5.21% |
| 4 | Dec 2025 – May 2026 | +13.71% | 1.93 | 6.32% |
Fold 2 is the outlier: during the volatile early-2025 tape, the strategy earned only 3.27% with a Sharpe of 0.46 and a 16.55% drawdown. Folds 1, 3, and 4 were considerably cleaner, each producing double-digit gains and Sharpes above 1.9.
The out-of-sample composite return is 13.71% at a Sharpe of 1.93 — better than the full-window number, which is an encouraging sign that performance didn't degrade on unseen data.
However, the strategy failed its formal validation gate. The Probabilistic Sharpe Ratio (PSR) sits at 0.811 and the Deflated Sharpe Ratio (DSR) at 0.338, tested across six parameter trials. The DSR in particular penalises for the number of configurations tried before landing on this one — a standard guard against data-mining. Until DSR clears a threshold closer to 0.5+, the validation system correctly withholds a passing grade.
Live Activity — A Quiet July
Since the last executed trades on June 1 (XOM at $148.69, NVDA at $224.20) and May 31 (BAC, HON, MSFT), the strategy has run daily but rejected every candidate signal. Six consecutive scheduled runs from June 30 through July 7 returned zero executions. Portfolio value has drifted between $9,401 and $9,663 on $608.79 cash, implying modest mark-to-market gains from the held positions rather than new rotation.
The rejection streak isn't necessarily alarming — a momentum filter simply found no names meeting its entry threshold in a range-bound or mixed tape. But five trades in total since inception is a thin sample on which to judge live execution quality.
Strengths and Risks
Strengths: Four-for-four positive folds, a clean blue-chip universe, and an OOS Sharpe of 1.93 are genuine positives. The strategy doesn't overfit on exotic tickers or micro-cap noise.
Risks: The 20.48% max drawdown is steep for a large-cap momentum approach. Fold 2's underperformance shows real sensitivity to choppy, mean-reverting markets. Most importantly, the DSR of 0.338 means the risk-adjusted edge hasn't been statistically established beyond the number of trials run. Promoting this strategy to real capital before DSR improves would be premature.
momentum-code is one to watch — the cross-fold consistency is real — but it needs a wider trade sample and a cleaner validation pass before it earns full confidence.