Class KurtosisRTIndicator

All Implemented Interfaces:
Resettable, RTIndicator

public class KurtosisRTIndicator extends AbstractWindowSeriesRTIndicator
Rolling Excess Kurtosis with O(1) per-tick updates.

Measures the "tailedness" of the distribution of values in the rolling window. A normal distribution has excess kurtosis of 0; positive (leptokurtic) indicates heavy tails; negative (platykurtic) indicates light tails. Critical for risk management — heavy tails mean more extreme events.

Uses the sample excess kurtosis formula:

  K = [n(n+1) / ((n-1)(n-2)(n-3))] * Σ((x_i - x̄)/s)⁴ - 3(n-1)² / ((n-2)(n-3))
computed from four power-sum accumulators (sum, sumSq, sumCub, sumQuad).

Returns 0 when n < 4 or when standard deviation is zero.