Market-neutral residual momentum — forward paper

Buy the stocks with the strongest beta-adjusted (residual) 9-month momentum, short the weakest, rebalanced monthly, and sit out (hold flat) in months the market is in a deep drawdown or high-volatility regime — the forward-looking switch that de-risks the momentum-crash months. The aim is a few percent a year uncorrelated to the market (and to a short-premium options book), not a high return. This page tracks it live, honestly, as the record accrues.

Current positions

Ratio: dollar-neutral — 25 long / 25 short, ~100% gross exposure each side, equal-weight (4.0% per long name). Long minus short = market-neutral by construction; realised beta is tracked above.

Ratio: dollar-neutral — 97 long / 96 short, ~100% gross exposure each side, equal-weight (1.0% per long name). Long minus short = market-neutral by construction; realised beta is tracked above.

Full decile book (~100 names per side — faithful to the backtest)

Tradeable book (top/bottom 25 per side — what an individual would run)

The frozen model

Survivorship-free US common stocks, point-in-time top-1000 by dollar volume, $5 price floor. Signal: 9-1 residual momentum (return minus trailing-12-month beta × SPY), monthly decile long-short, equal-weight. Risk-off switch (flat that month): SPY in a ≥10% trailing-12-month drawdown or trailing-6-month SPY monthly volatility above the frozen in-sample threshold (4.60%). Friction 40 bps round-trip per leg per rebalance. Short-leg borrow, tiered: 1%/yr on liquid ≥$10 names, 10%/yr on <$10 names (momentum losers skew hard-to-borrow); the long leg pays no borrow. Any change to this model is a new slug.

In-sample context (≤2022, exploration — not a forward claim): the same construction ran +7.3%/yr at Sharpe 0.58, max-DD 22%, realised beta −0.14. Modest, and out-of-sample expect less — which is exactly why it is run forward in paper before any capital or sizing decision. Lineage: discovery D-005/D-006 → exploration lab/explore/stock-momentum-xsec → this tracker.