Class FairValueEngine

java.lang.Object
com.quantfinlib.pricing.FairValueEngine
public final class FairValueEngine extends Object
Latency-adjusted fair value for rapidly updating order books. Maintains the size-weighted microprice plus a short-window mid drift estimate, so a consumer that is latencyNanos behind the market can project the "true mid" at the moment its order would actually arrive.

Zero allocation after construction (fixed ring of samples); single writer.

  • Constructor Summary

    Constructors
    Constructor
    Description
    FairValueEngine()
    256-sample ring over a 500 ms drift window.
    FairValueEngine(int capacity, long windowNanos)
     

  • Method Summary

    Modifier and Type
    Method
    Description
    double
    driftPerSecond()
    Estimated mid drift in price units per second over the lookback window.
    double
    latencyAdjustedFair(long latencyNanos)
    Fair price projected latencyNanos into the future: microprice plus drift over the latency horizon.
    double
    microprice()
    Latest microprice (NaN before the first quote).
    static double
    microprice(double bid, double ask, double bidSize, double askSize)
    Size-weighted microprice: I*ask + (1-I)*bid, I = bidSize/(bidSize+askSize).
    void
    onQuote(double bid, double ask, double bidSize, double askSize, long timestampNanos)
    Feed a top-of-book update.

    Methods inherited from class java.lang.Object

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

  • Constructor Details

    • FairValueEngine

      public FairValueEngine(int capacity, long windowNanos)
      Parameters:
      capacity - ring capacity (rounded up to a power of two)
      windowNanos - lookback window for the drift estimate

    • FairValueEngine

      public FairValueEngine()
      256-sample ring over a 500 ms drift window.

  • Method Details

    • onQuote

      public void onQuote(double bid, double ask, double bidSize, double askSize, long timestampNanos)
      Feed a top-of-book update.

    • microprice

      public static double microprice(double bid, double ask, double bidSize, double askSize)
      Size-weighted microprice: I*ask + (1-I)*bid, I = bidSize/(bidSize+askSize).

    • microprice

      public double microprice()
      Latest microprice (NaN before the first quote).

    • driftPerSecond

      public double driftPerSecond()
      Estimated mid drift in price units per second over the lookback window.

    • latencyAdjustedFair

      public double latencyAdjustedFair(long latencyNanos)
      Fair price projected latencyNanos into the future: microprice plus drift over the latency horizon.