On Jarque-Bera normality and cusum parameter change tests for BCTT-GARCH models

In this paper, we study the Jarque-Bera (JB) and cusum tests for the normality of innovations and parameter change in BCTT-GARCH models. In order to demonstrate the validity of JB normality and cusum parameter change tests, we derive their limiting null distributions under mild conditions.

► The Box-Cox transformed threshold GARCH models were considered for leverage effect of financial time series. ► We studied two different tests for the normality of innovations and parameter change. ► Asymptotic distributions of two tests are derived for their validation. ► Asymptotic distributions of two tests based on residuals are identical to the one based on true innovations.