Class AlphaReport

java.lang.Object
com.quantfinlib.alpha.AlphaReport
public final class AlphaReport extends Object
Alpha reporting — the diagnostics that explain a factor's P&L rather than just totalling it:
  • Alpha decay — mean IC as a function of the forward horizon. A signal predictive at 1 day but dead at 5 needs fast, expensive trading; the halfLife estimate says how long the edge survives, which bounds the viable rebalance cadence.
  • Factor attribution — OLS of portfolio returns on factor return streams: how much of the "alpha" is just repackaged momentum/value beta, and what residual (true) alpha remains.
  • Curves and ratios — cumulative return, drawdown series, and the full ratio set (Sharpe, Sortino, Calmar, CAGR, max drawdown) reused verbatim from backtest.PerformanceAnalytics, so alpha reports and strategy backtests can never disagree on a definition.
  • Rolling metrics — the windowed Sharpe that shows whether performance is steady or one lucky year.

  • Nested Class Summary

    Nested Classes
    Modifier and Type
    Class
    Description
    static final record 
    AlphaReport.Attribution
    OLS attribution: per-bar residual alpha, factor betas, and fit quality.
    static final record 
    AlphaReport.Decay
    IC per horizon plus the interpolated half-life of the shortest-horizon IC.

  • Method Summary

    Modifier and Type
    Method
    Description
    static AlphaReport.Attribution
    attribute(double[] portfolioReturns, double[][] factorReturns, List<String> factorNames)
    Regresses portfolio returns on factor return streams (with an intercept) via the normal equations: r_p = α + Σ βᵢ·fᵢ + ε.
    static AlphaReport.Decay
    decayProfile(AlphaContext ctx, AlphaFactor factor, int startIndex, int[] horizons)
    Evaluates the factor's mean IC at each horizon.
    static double[]
    drawdownCurve(double[] equity)
    Drawdown series: fraction below the running peak (0 at new highs).
    static double[]
    returnsOf(double[] equity)
    Per-bar simple returns of an equity curve — the input to attribution/rolling.
    static double[]
    rollingSharpe(double[] returns, int window, int periodsPerYear)
    Rolling annualized Sharpe over a trailing window of per-bar returns; NaN until the window fills.
    static PerformanceMetrics
    summarize(double[] equity, int periodsPerYear)
    The full ratio set on an equity curve — Sharpe, Sortino, Calmar, CAGR, max drawdown — computed by the same engine the backtesters use, so definitions never fork between research and backtest reports.

    Methods inherited from class java.lang.Object

    clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

  • Method Details

    • decayProfile

      public static AlphaReport.Decay decayProfile(AlphaContext ctx, AlphaFactor factor, int startIndex, int[] horizons)
      Evaluates the factor's mean IC at each horizon. The half-life is the first horizon (linearly interpolated) where the IC falls below half of its shortest-horizon value; +∞ when it never does within the tested range — the honest answer for slow signals.
      Parameters:
      horizons - ascending forward horizons in bars

    • attribute

      public static AlphaReport.Attribution attribute(double[] portfolioReturns, double[][] factorReturns, List<String> factorNames)
      Regresses portfolio returns on factor return streams (with an intercept) via the normal equations: r_p = α + Σ βᵢ·fᵢ + ε. The intercept is the residual alpha — what survives after the known factors take their share. Keep the factor count far below the bar count; the normal equations of collinear factors are a data problem, not a solver problem.
      Parameters:
      portfolioReturns - per-bar portfolio returns
      factorReturns - one per-bar return stream per factor, aligned

    • returnsOf

      public static double[] returnsOf(double[] equity)
      Per-bar simple returns of an equity curve — the input to attribution/rolling.

    • drawdownCurve

      public static double[] drawdownCurve(double[] equity)
      Drawdown series: fraction below the running peak (0 at new highs). Guards peak > 0 exactly like risk.RiskMetrics.maxDrawdown so min(drawdownCurve) and the headline max-drawdown metric can never disagree on a curve that touches zero.

    • rollingSharpe

      public static double[] rollingSharpe(double[] returns, int window, int periodsPerYear)
      Rolling annualized Sharpe over a trailing window of per-bar returns; NaN until the window fills. The steadiness plot: a flat positive line is a strategy, a single spike is an anecdote.

      Uses the SAMPLE standard deviation (n−1) — the same definition as risk.RiskMetrics.sharpeRatio behind summarize(double[], int) — so a full-sample rolling window reproduces the headline Sharpe exactly rather than differing by √(n/(n−1)).

    • summarize

      public static PerformanceMetrics summarize(double[] equity, int periodsPerYear)
      The full ratio set on an equity curve — Sharpe, Sortino, Calmar, CAGR, max drawdown — computed by the same engine the backtesters use, so definitions never fork between research and backtest reports.