Nonlinear Dependence and Conditional Heteroscedasticity : A Notes from Malaysian Daily Stock Returns

This study tests whether the behaviour of daily stock returns for the sample of three banks and the composite index in the Malaysian market are nonlinear dependence. Using three nonlinear testing procedures, the study suggests nonlinearity in the return series for all cases. The cause for the nonlinear dependence appears to be conditional heteroscedasticity. We then test for the model adequacy using the GARCH (1,1) model and find that the model does not fit well to the data generating process of the return series except for the composite index. The results lead to the suggestion that, GARCH (p,q) model or more complex threshold model can possibly explain the microeconomic data better