A joint test for serial correlation and heteroscedasticity in fixed-T panel regression models with interactive effects

摘要

In this paper, a joint test is proposed for serial correlation and heteroscedasticity in fixed-T panel regression models with interactive effects. If the idiosyncratic errors are serially uncorrelated and homoscedastic, the proposed test can be shown to be asymptotically chi-square distributed under some mild conditions. A small Monte Carlo simulation experiment is carried out for illustrations. (C) 2020 Elsevier B.V. All rights reserved.