escc_test: Engle-Sheppard Constant Correlation Test

View source: R/tests.R

escc_testR Documentation

Engle-Sheppard Constant Correlation Test

Description

The constant correlation test of Engle and Sheppard (2001).

Usage

escc_test(x, order = c(1, 1), lags = 1, constant = FALSE, ...)

Arguments

x

an xts matrix of stationary data.

order

the GARCH model order.

lags

the number of lags to test for.

constant

whether to include a constant (mean) in the GARCH model estimation.

...

none.

Details

The whitened residuals from a constant correlation GARCH model are used to test the hypothesis that they are IID using an artificial regression of the outer products of tese residuals on a constant and their lagged outer products. The null hypothesis of constant correlation can then be tested using a Wald-type test on the constant and coefficients of the lagged regressors (which should be zero under the null), with the test statistic asymptotically distributed as Chi-squared with (lags + 1) degrees of freedom. In the returned table, the parameter “J” represents the test statistic.

Value

An object of class “tstest.escc” which has a print and as_flextable method.

Note

The test makes use of parallel functionality for estimating the univariate GARCH models and the user can control the number of threads used using plan.

References

\insertRef

Engle2001tsmarch


tsmarch documentation built on Nov. 2, 2026, 9:06 a.m.