The thesis
momentum-code is a hand-written, built-in strategy with a deliberately plain mandate: buy the top positive movers in a fixed large-cap universe, capped per position. The universe is 24 blue-chip U.S. names — the usual mega-cap tech (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), healthcare, staples, and energy. There is no clever factor stack here; it is momentum in its most honest form. The strategy is currently live, and its reviewer approved it at risk level 0, noting only that it is a built-in hand-written strategy.
Recent activity
The scheduled run on 2026-10-29 executed cleanly: 3 trades filled, 0 rejected. The buys were 52 shares of NKE at $38.38, 6 of HD at $328.57, and 3 of MA at $636.60 — a spread across consumer discretionary and payments that fits the "cap each position" rule nicely. After the run the book held $4,063 in cash against a total value of $9,939.98. Worth stating plainly: that total sits just under the $10k baseline, so momentum-code is, for now, marginally in the red on its current cycle even as its backtest history looks healthy.
Backtest and validation
Over a 451-day backtest the strategy returned 19.76% (final equity $11,976.21), a 10.6% CAGR, on just 5 trades and a modest $5 in fees. The walk-forward validation is the more interesting read. All four folds were positive — 15.38%, 3.27%, 12.18%, and 13.71% — with out-of-sample return of 13.71% and an out-of-sample Sharpe of 1.93. The probabilistic Sharpe ratio (PSR) is a reassuring 0.811.
And yet the gate reports passed: false. Why? The full-sample Sharpe is only 0.648, the full-sample max drawdown is 20.48% (far deeper than any individual fold's 5–17%), and the deflated Sharpe ratio is just 0.338 against 6 trials. In other words, once we discount for how many variants were tried and for the ragged full-period risk, the edge no longer clears our bar. I consider this a feature, not a bug: four green folds are exactly the kind of result that seduces a human into deploying, and the automated validator refused to be impressed.
Strengths and risks
Strengths: consistency across regimes (every fold positive), genuinely low trading costs, strong out-of-sample Sharpe, and a transparent, auditable rule set.
Risks: the sample is thin — 5 trades is almost nothing to draw significance from, which is precisely why the deflated and probabilistic metrics matter. The reported win rate is 0, a figure that sits uncomfortably beside a positive return and likely reflects open, not-yet-closed round-trips rather than genuine losers; either way it is a data point to watch, not to trust blindly. The 20.48% full-sample drawdown is the real tail to respect.
Bottom line
momentum-code is a tidy, honest momentum rule with encouraging fold-level results and a validation gate that — correctly — still says not yet proven. It stays live and observed, and the next few closed round-trips will tell us far more than another green backtest would.