Skip to content

Strategies

Every strategy, every caveat.

Lifecycle states are read live from the strategy registry. Only PAPER_ACTIVE strategies receive paper capital, and only a named human can promote one.

10 strategies shown

  • Cross-Sectional Momentum

    EXPERIMENTAL

    Ranks stocks by 12-month return skipping the latest month.

    Data
    Daily adjusted bars
    Horizon
    ~1 month

    Evidence: Out-of-sample Sharpe not significantly different from an equal-weight universe (90% CI includes zero).

  • Residual Momentum

    EXPERIMENTAL

    Momentum with the market's contribution removed.

    Data
    Daily adjusted bars
    Horizon
    ~1 month

    Evidence: Under validation; experimental.

  • Quality & Value

    EXPERIMENTAL

    Cash-flow and earnings yield, with profitability as a quality screen.

    Data
    SEC XBRL filings
    Horizon
    ~3 months

    Evidence: Highest point estimate among our pods, but its edge over an equal-weight universe is not statistically significant.

  • Trend Following

    EXPERIMENTAL

    Holds an asset only while its own trend is positive.

    Data
    Daily bars, ETFs
    Horizon
    ~1 month

    Evidence: Experimental; does not beat an equal-weight universe in our tests.

  • Mean Reversion

    EXPERIMENTAL

    Buys short, sharp pullbacks inside long-term uptrends.

    Data
    Daily bars
    Horizon
    ~1 week

    Evidence: Fails our cost-stress test; kept experimental.

  • Earnings Intelligence

    EXPERIMENTAL

    Post-earnings drift after large standardized surprises.

    Data
    SEC 10-Q/10-K
    Horizon
    ~2 months

    Evidence: Experimental; not significant vs equal weight.

  • Insider Buying

    EXPERIMENTAL

    Clusters of open-market insider purchases.

    Data
    SEC Form 4
    Horizon
    ~3 months

    Evidence: Underperforms an equal-weight universe in our out-of-sample test.

  • ML Stock Ranking

    Research only

    Ridge and gradient-boosted ranking of 15 point-in-time features.

    Data
    Prices + SEC fundamentals
    Horizon
    ~1 month

    Evidence: Adds no value over simple baselines in our test — reported as a failure, not promoted.

  • Regime Overlay

    Research only

    Scales exposure down when trend, volatility or breadth deteriorate.

    Data
    Daily bars
    Horizon
    Daily

    Evidence: Reduces drawdowns; the risk-adjusted gain is not significant and the return cost is. Research only.

  • Multi-Factor

    EXPERIMENTAL

    Combines momentum, value and the research council.

    Data
    All of the above
    Horizon
    ~1 month

    Evidence: Experimental; combined pods roughly match an equal-weight universe.