Thesis
momentum-code runs a straightforward rule: at each scheduled trigger, scan the 24-stock large-cap universe — names like NVDA, MSFT, XOM, and JPM — and buy the top positive movers, subject to a per-position cap. No discretion, no macro overlay. The edge, if it exists, is purely price-based momentum within a curated blue-chip pool.
Backtest Overview
Over 451 days of simulation the strategy compounded to +19.76% total return, implying a 10.6% CAGR. Final paper equity sits at $11,976 on a notional $10,000 start. Turnover is high at 92.74%, meaning the book nearly fully rotates each period — that's consistent with a pure momentum design but worth watching in any live context where friction matters. Fees for the full backtest period totalled just $5, reflecting the small number of executed trades (5).
The headline Sharpe of 0.65 is modest, and the 20.48% max drawdown is the most notable risk figure. For a strategy selecting only from mega-cap, liquid names, a one-in-five paper loss from peak is not trivial — it suggests momentum crowding can flush quickly when market leadership rotates.
Walk-Forward Folds
The more interesting story is in the cross-validation detail. Four sequential out-of-sample folds all closed positive:
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +15.38% | 2.47 | 6.12% |
| 2 | Jan 2025 – Jul 2025 | +3.27% | 0.46 | 16.55% |
| 3 | Jul 2025 – Dec 2025 | +12.18% | 2.31 | 5.21% |
| 4 | Dec 2025 – May 2026 | +13.71% | 1.93 | 6.32% |
Fold 2 is the outlier — a thin 3.27% return with a 16.55% drawdown and a near-flat Sharpe. That window (early 2025) was visibly hostile to momentum strategies as leadership rotated. Folds 1, 3, and 4 look far cleaner, with Sharpes above 1.9 and drawdowns under 7%. The out-of-sample return across the final fold is 13.71% with a Sharpe of 1.93, which is the strongest individual reading in the set.
Validation Status
Despite four positive folds, momentum-code has not passed platform validation. The key constraint is the Deflated Sharpe Ratio (DSR) of 0.338 — the platform's primary overfitting-adjusted metric, which accounts for the number of trials run (6 in this case). A DSR below roughly 0.5 suggests a material probability that the observed Sharpe is noise rather than signal. The Probabilistic Sharpe Ratio (PSR) of 0.811 is more encouraging but doesn't override the DSR gate.
This is the right call. Six strategy trials with a 0.65 full-period Sharpe leaves meaningful uncertainty about whether the edge is real or an artifact of the parameter search.
Recent Live Activity
The last executed trades were placed in late May and early June 2026 — buys in XOM, NVDA, BAC, HON, and MSFT. Since then, six consecutive daily runs have registered zero executions, with a total of nine rejected signals over that window. Cash has held steady at $608.79 while total equity has ranged from $9,401 to $9,537, tracking the broader market.
The rejection pattern suggests no ticker in the universe has cleared the momentum threshold on a day the strategy also has room to add. That's not necessarily alarming — momentum strategies naturally go quiet when the market lacks clear directional leaders.
Risks to Watch
- Fold 2 regime risk: a repeat of a choppy, rotation-heavy environment could compress returns significantly.
- DSR overhang: with only 5 executed trades and 6 parameter trials, the statistical base is thin. More live history is needed before the DSR moves meaningfully.
- High turnover cost: in a real-money context, 92.74% annual turnover on a larger book would erode the thin fee buffer quickly.
Bottom Line
momentum-code shows genuine directional consistency — four-for-four positive folds is not nothing — but the strategy is still in the "prove it" phase. The DSR gate is doing exactly what it should: holding a plausible-but-unproven idea to a higher standard before it earns autonomous capital.