Omnibus test for covariate effects in conditional copula models

Conditional copulas describe the conditional dependence and influence that covariates have on structure between two (or more) variables. Of interest is to test null hypothesis a specific effect. This paper proposes an omnibus for testing of specified effect covariates. The statistic designed having power against many alternatives, can be used variety covariate effects (no effects, linear partial etc.). A special case problem do not affect structure. In this semiparametric framework marginal distribution functions are estimated using nonparametric kernel techniques parametric model maximum likelihood estimation. We establish asymptotic under hypothesis, evaluate finite-sample performance via simulation study, which also includes comparisons with alternative tests. real data analysis illustrates practical use test.