BICORRTEST Procedure

@BiCorrTest computes the "C" and "H" portmanteau statistics for autocorrelation and third order dependence, from Hinich(1996). The "H" test is the more commonly used. The null is pure white noise against an alternative of third order dependence (which would be fairly deep non-linear dependence—GARCH is a form of second order dependence).

@BiCorrTest(options) x start end

Parameters

x	series to analyze
start, end	range of x to use. By default, the defined range of x.

Options

C=power of T for number of lags [.4]

Determines the number of lags in the formula as a function of the number of observations.

[PRINT]/NOPRINT

TITLE=title for report ["Bicorrelation Test, Series xxxx"]

Variables Defined

%%BICORRH	H statistic (REAL)
%%BICORRHSIGNIF	Significance of H (REAL)
%%BICORRC	C statistic (REAL)
%%BICORRCSIGNIF	Significance of C
%%BICORRLAGS	Number of lags used