Thesis
momentum-code runs a deliberately legible idea: buy the top positive movers across a 24-name large-cap universe (AAPL, MSFT, NVDA, JPM, XOM and peers), capped per position. It is a classic trend-following premise — let the market nominate leadership and size into it — with position caps as the only real risk control. Simplicity is a feature here: there is little room for the kind of parameter sprawl that quietly overfits.
Recent activity
The last six scheduled runs (8–15 October 2026) tell a flat story: zero trades executed and a steady drip of rejections — three on both 8 and 13 October, two on most other days. Cash has been frozen at $608.79 for the entire week while total equity drifted between roughly $10,411 and $10,591. The last orders that actually filled were buys back on 31 May–1 June 2026: XOM, NVDA, BAC, HON and MSFT. In short, the book has been fully invested and effectively dormant for over four months; the nightly runs want to add exposure but lack the cash to clear the per-position cap, so the orders bounce.
Backtest and validation
The backtest spans 451 days and finishes at $11,976.21 — a 19.76% total return, 10.6% CAGR, on just five trades and $5 of fees. Sharpe is a modest 0.65 and max drawdown a meaningful 20.48%. Walk-forward validation looks stronger on the surface: all four folds were positive (15.38%, 3.27%, 12.18%, 13.71%), out-of-sample return was 13.71% with an OOS Sharpe of 1.93, and probabilistic Sharpe reads a healthy 0.811.
Yet the verdict is a fail. The tell is the deflated Sharpe ratio of 0.338, computed across six trials — once you penalise for the number of configurations tried, the edge shrinks below the bar. The full-period drawdown of 20.48% also exceeds every single fold's worst (16.55%), suggesting the strategy's pain concentrates in stretches the folds don't isolate.
Strengths and risks
Strengths are real: the logic is transparent, turnover is low (92.74%, roughly one rotation), costs are negligible, and every fold made money. That consistency deserves respect.
The risks are equally clear. The win rate of 0 simply reflects that none of the five open positions have been realised — there is no closed-trade record to lean on yet. The failed deflated-Sharpe gate warns the headline return may be partly a product of selection. And operationally, a week of zero fills with cash stuck at $608.79 means the thesis isn't being expressed at all right now.
Verdict: a clean idea with a believable walk-forward, held back by a failed robustness gate and a cash-starved, stalled live book. I'd want the sizing/capital problem fixed and at least one full round-trip realised before trusting the numbers.