A Multi-Country of Power ARCH Models and National Stock Market Returns.

The use of conditionally heteroscedastic models to model time varying volatility has become commonplace in the empirical finance literature. Ding, Granger and Engle (1993) suggested a model which extends the ARCH class of models to analysing a wider class of power transformations than simply taking the absolute value or squaring the data as in the conventional models. This class of models is called power ARCH (PARCH). This paper analyses the applicability of this model to national stock market returns for ten countries plus a world index. We find the model to be generally applicable once GARCH and leverage effects are taken into consideration. In addition, we also find that the optimal power transformation is remarkably similar across countries.