Thesis
momentum-code runs a simple idea: buy the top positive movers in a 24-name large-cap universe, capped per position. The universe spans mega-cap tech (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV) and staples/industrials (PG, KO, WMT, COST, CAT, HON). It is a classic cross-sectional momentum tilt with no shorting — concentration is the whole bet.
Recent Activity
The strategy deployed its capital in a burst on 29–30 October, buying MA (3 @ $636.60), HD (6 @ $328.57), NKE (52 @ $38.38), XOM (10 @ $186.99) and WMT (16 @ $123.53). That drew cash down to $177.08.
Since then it has done nothing. Across six scheduled runs (2–9 November), every single one reports 0 executed, 1–2 rejected, with cash frozen at $177.08. In practice momentum-code is fully invested and out of dry powder — new momentum signals keep firing, but there is no cash to act on them, so they bounce. Total equity has drifted sideways in a tight band, from about $9,876 to $10,186, finishing the window near $10,012. Flat, in other words.
Backtest & Validation
The historical numbers look attractive on the surface: +19.76% total return over 451 days, a 10.6% CAGR, final equity of $11,976, negligible fees ($5) and zero FX cost. The walk-forward split is the real strength — all four folds were positive (+15.38%, +3.27%, +12.18%, +13.71%), with out-of-sample return of 13.71% and a healthy OOS Sharpe of 1.93. Individual folds even hit Sharpes of 2.47 and 2.31.
And yet validation did not pass. Two things explain the red flag. First, the full-sample Sharpe is only 0.65 — far below the glossy per-fold figures, which means a drawdown straddling fold boundaries dragged the aggregate down. The 20.48% max drawdown exceeds any single fold's worst (16.55%), confirming that pattern. Second, the deflated Sharpe ratio is just 0.338 once you adjust for the six trials run; the probabilistic Sharpe (0.811) is decent, but DSR is the honest, multiple-testing-aware number, and it is weak.
The Balance
Strengths: a coherent, cheap-to-run thesis, consistently positive out-of-sample folds, and low frictional cost. Risks: the sample is tiny — 5 trades and a reported win rate of 0 leave almost no statistical ground to stand on; the headline Sharpe and >20% drawdown are uncomfortable for a long-only book; and right now the live strategy is structurally stuck, unable to rotate because it holds no cash buffer.
momentum-code earns a cautious watch, not conviction. The validation failure is doing its job — treat the backtest as a hypothesis, not a promise, and watch whether it can free up cash to actually trade its signal.