The RATS Software Forum

Attached is a procedure for testing a set of series for multivariate ARCH effects. The null is that the series are mean zero, not serially correlated and with a fixed covariance matrix. It performs an LM by regressing the crossproducts of the series (that is u(i,t) x u(j,t) for all combinations of i and j) on a constant and its lag(s) and testing the coefficients on the lags. The number of degrees of freedom is

(n(n+1)/2)^2 x the number of lags

since it is including all crossproducts on all crossproducts.



@MVARCHTest( options ) start end
# list of series

(start and end default to the common range of the series)

Options

LAGS=number of ARCH lags to test [1]
[PRINT]/NOPRINT

Variables Defined

%CDSTAT	Test statistic
%SIGNIF	Significance level of the test

Is this based on any paper, i.e. are there any references for this? How to cite? For the MVQStat, for instance, you refer to Hosking (1981).

It's a fairly obvious extension of the LM test to multivariate ARCH. Based upon the description, it sounds like

R. Scott Hacker & Abdulnasser Hatemi-J, 2005. "A test for multivariate ARCH effects," Applied Economics Letters, Taylor and Francis Journals, vol. 12(7), 411-417

is the same thing, though that's behind a pay wall, so I can't be sure.

TomDoan wrote:It's a fairly obvious extension of the LM test to multivariate ARCH. Based upon the description, it sounds like

R. Scott Hacker & Abdulnasser Hatemi-J, 2005. "A test for multivariate ARCH effects," Applied Economics Letters, Taylor and Francis Journals, vol. 12(7), 411-417

is the same thing, though that's behind a pay wall, so I can't be sure.

I think the article you cited comes very close. Thank you, Tom!