Thesis
momentum-code runs a simple, legible idea: buy the top positive movers across a 24-name large-cap universe (think AAPL, MSFT, NVDA, JPM, XOM, UNH), capped per position so no single winner dominates the book. No regime filters, no shorting — just rotate into strength and size it sensibly. The appeal is that you can explain it in one sentence, which matters when a strategy has to be trusted with real capital.
Recent activity
The book executed its opening trades at the end of May and start of June 2026 — buys in MSFT, HON, BAC, then NVDA and XOM — and has been quiet since. Every scheduled run from 20–27 October reports 0 executed, 1–2 rejected, with cash pinned at $608.79 and total equity drifting between roughly $10,375 and $10,533 on a $10,000 base. So the strategy is modestly in the green live, but it is effectively fully invested and its recent signals are being rejected rather than filled — likely bumping against per-position caps or available cash. Worth watching: a momentum book that cannot rotate is just a static long portfolio.
Backtest
Over 451 days the backtest returned 19.76% (final equity $11,976), a 10.6% CAGR, with a Sharpe of 0.65 and a max drawdown of 20.48%. Turnover was 92.74% and total fees just $5 — cheap, but on only 5 trades. That is a very thin sample: the headline return rests on a handful of positions, so the Sharpe and drawdown figures carry wide error bars. The reported 0% win rate is best read as an artefact of open, unrealised positions rather than a genuine loss record.
Validation
This is where it gets interesting. Across 4 walk-forward folds, all 4 were positive — 15.4%, 3.3%, 12.2% and 13.7% — and the most recent out-of-sample fold posted a 1.93 Sharpe on a 13.71% return with a shallow 6.3% drawdown. The Probabilistic Sharpe Ratio of 0.811 is encouraging. Yet the gate reads passed: false, driven by a Deflated Sharpe Ratio of 0.338: once you haircut for the 6 trials run during search, the edge no longer clears the multiple-testing bar.
Verdict
The honest read is split. Consistency across folds and a strong, low-drawdown OOS window are real positives, and the thesis is robust and transparent. Against that: the full-sample Sharpe is pedestrian, drawdowns approach 20%, the trade count is too small to be confident, and the formal validation gate fails on deflation. Add the week of rejected orders and the case is "promising but unproven." I would keep it live at small size, fix the order-rejection issue so it can actually rotate, and demand more trades before trusting the fold numbers.