gmm_test: GMM Orthogonality Test

In tstests: Time Series Goodness of Fit and Forecast Evaluation Tests

GMM Orthogonality Test

Description

The GMM orthogonality test of Hansen (1982).

Usage

gmm_test(x, lags = 1, skewness = 0, kurtosis = 3, ...)

Arguments

x	a series representing the standardized residuals of some estimated model.
lags	the lags for the co-moment test.
skewness	the skewness of the estimated model residuals.
kurtosis	the kurtosis of the estimated model residuals.
...	not currently used.

Details

For parametric models estimated with a particular distribution, the skewness and kurtosis should flow from the distributional model. See for example dskewness and dkurtosis.

Value

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

References

\insertRef

Hansen1982tstests

Examples

library(tsgarch)
library(tsdistributions)
library(data.table)
library(xts)
data("spy")
spyr <- na.omit(diff(log(spy)))
spec <- garch_modelspec(spyr, model = "egarch", order = c(2,1), constant = TRUE,
distribution = "jsu")
mod <- estimate(spec)
skewness <- dskewness("jsu", skew = coef(mod)["skew"], shape = coef(mod)["shape"])
# kurtosis is dkurtosis is the excess over the Normal (3) so we add back 3
# since the test takes the actual not excess kurtosis.
kurtosis <- dkurtosis("jsu", skew = coef(mod)["skew"], shape = coef(mod)["shape"]) + 3
test <- gmm_test(residuals(mod, standardize = TRUE), lags = 2, skewness = skewness,
kurtosis = kurtosis)
print(test, collapse = TRUE, include.decision = TRUE)

tstests documentation built on Oct. 30, 2024, 9:28 a.m.