The use of domestic and world market indexes in the estimation of time-varying betas

This paper generates time-varying estimates of Australian industry betas relative to an Australian market index and a world market index using the Kalman filter approach. As a means of comparison, these conditional estimated betas are used to forecast each industry's return in-sample. The forecast error metrics suggest that the estimates of conditional risk relative to the domestic market index are preferred to estimates generated using the world market index, irrespective of the industry concerned. While not to suggest time-varying betas estimated relative to a domestic index are universally superior, these results suggest that they are preferable in certain circumstances.