The thesis
momentum-code runs a simple, legible idea: buy the top positive movers across a 24-name large-cap universe — AAPL, MSFT, NVDA, JPM, XOM and peers — with a cap on each position so no single name dominates. It is a trend-following sleeve, not a stock-picker: it rents whatever is already working and bounds concentration risk by construction. The strategy is flagged live.
Recent activity
The last executed fills are old. The five most recent trades are all buys dated 31 May–1 June 2026 — XOM, NVDA, BAC, HON and MSFT — and nothing has traded since. Every scheduled run from 30 September through 7 October reports 0 executed against 1–3 rejected orders. Across that window cash is pinned at $608.79 while total equity drifts in a narrow band, from $10,405 up to $10,557 and back to $10,459.
The read is unflattering but clear: the book is effectively fully invested, the per-position caps and thin cash leave no room to act on new signals, and the engine is proposing trades it cannot fund. The strategy is not broken so much as stuck — holding its June positions and logging daily rejections.
Backtest and validation
Over 451 days the backtest returns +19.76% (final equity $11,976), a 10.6% CAGR, on just five trades with $5 in fees and 92.7% turnover. The headline blemish is a 20.48% max drawdown — larger than the drawdown of any individual walk-forward fold (6.1%, 16.6%, 5.2%, 6.3%), a reminder that stitched-together segments can mask a deeper peak-to-trough.
The walk-forward evidence is genuinely encouraging: all four folds are positive (+15.4%, +3.3%, +12.2%, +13.7%), the out-of-sample tail returns +13.7% at a 1.93 Sharpe, and the probabilistic Sharpe ratio sits at a healthy 0.81.
And yet validation did not pass. The tell is the deflated Sharpe ratio of 0.34 against 6 trials. Once the Sharpe is haircut for the number of configurations tried, the evidence is no longer strong enough to reject luck — a sensible, overfitting-aware gate doing exactly its job.
Strengths and risks
Strengths: a transparent thesis, consistently positive out-of-sample folds, a strong PSR, and negligible trading costs. There is a real signal here.
Risks: the reported win rate is 0% across 5 trades — consistent with a book of still-open buys and no closed round-trips, so treat it as not yet earned rather than a loss record. The deep aggregate drawdown, the failed deflated-Sharpe gate on only six trials, and the operational reality of a cash-starved book throwing daily rejections all argue for caution.
Verdict: a promising trend sleeve whose statistics are not yet robust enough to trust unsupervised, and whose live plumbing — capital allocation and order sizing — needs attention before the thesis can express itself again.