Thesis
momentum-code runs a simple, disciplined playbook: each session it screens a 24-stock universe for the day's top positive movers, buys the leaders, and caps each position to limit concentration. The universe is deliberately liquid and blue-chip — mega-cap tech (AAPL, MSFT, GOOGL, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE, ABBV), consumer staples and discretionary (PG, KO, WMT, COST, MCD, NKE, HD), energy (XOM, CVX), and industrials (CAT, HON, DIS). The logic is clean: capture near-term price momentum in names with sufficient liquidity to absorb the position.
Backtest Summary
Over 451 days, the strategy grew a $10,000 notional stake to $11,976 — a 19.76% total return, or roughly 10.6% annualised. Transaction costs were negligible: $5 in commissions, no FX drag, and minimal slippage from concentrating in five positions. The Sharpe of 0.65 is moderate, and the 20.48% maximum drawdown is meaningful relative to those returns, pointing to a risk-adjusted profile that has headroom to improve.
One figure worth contextualising: the backtest records a win rate of 0 across five trades. With positions likely still open at the end of the test window, this almost certainly reflects an accounting artifact — unrealised P&L not yet tallied as wins — rather than a clean sweep of losses. Realised P&L should be watched as current holdings close.
Walk-Forward Validation
The cross-validation story is more nuanced than the headline return suggests. All four out-of-sample folds finished in the green:
| 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 that demands attention: a 16.55% drawdown dwarfs every other period's, and the 0.46 Sharpe is a sharp step down from the flanking folds. Early-to-mid 2025 was a volatile, momentum-punishing regime — the kind of market where this strategy shows its key vulnerability.
Despite all folds being positive, validation did not pass. The Deflated Sharpe Ratio (DSR) of 0.338 is the sticking point. With six optimisation trials in the search history, the DSR adjusts the observed Sharpe downward to account for selection bias — and that adjusted figure does not clear the threshold. The Probabilistic Sharpe Ratio of 0.811 is more encouraging, but the DSR gate exists for precisely this reason: distinguishing a genuine edge from a plausible artefact of parameter search. The strategy needs more independent test periods, not more parameter tuning.
Recent Activity
The strategy has been largely dormant through late June and early July. Six consecutive scheduled runs (26 June – 3 July) executed zero trades, logging ten total signal rejections. Portfolio value drifted from roughly $9,457 to $9,648 over that stretch — mark-to-market appreciation from existing holdings, with no new entries.
The last executed trades date to late May and early June: buys in XOM ($148.69 × 13 shares), NVDA ($224.20 × 8), BAC ($51.60 × 38), HON ($237.86 × 8), and MSFT ($450.24 × 4). A quiet signal queue in a rangebound tape is expected behaviour for a momentum strategy — it should only fire when movers clear the threshold, not chase noise.
Outlook
momentum-code has a coherent thesis and a directionally consistent track record. Its strengths are consistency across folds and low operational overhead. Its risks are equally clear: fold 2's drawdown reveals genuine exposure to choppy, mean-reverting markets, and the DSR flags that observed performance has not yet earned statistical credibility across enough independent periods. Live observation status is appropriate; an auto-deploy gate upgrade would require the DSR to improve organically through continued out-of-sample performance — something time and a favourable regime can provide, but parameter tinkering cannot.