Determinants of the Directional Changes in the Australian Stock Market Return

Abstract

This paper predicts the directional changes in excess stock market return using macroeconomic variables. To minimize risks, an accurate forecasting of directional changes in excess return is required. To predict the directional changes this study first estimated the binary dependent dynamic probit model with the macroeconomic variables such as real GDP, CPI, interest rate, inflation rate, money supply (M3), unit labor cost index, commodity price index, oil price, US stock market index (S&P 500) and exchange rate (Trade weighted index), consumer confidence, business confidence, inflation differential between Australia and US, interest rate differentials between Australia and US, term structure of interest rates (10 year bond minus 90 day bills), dwelling approvals. First the dynamic probit model was estimated with growth rate of macroeconomic variables. Then dynamic probit model was estimated with the volatility (volatility estimates of variables were constructed by taking the absolute value of deviations from the four quarterly (annual) mean) of the macroeconomic variables. Finally the predicting performance of the dynamic probit model with macroeconomic variables and dynamic probit model with the volatility of the macroeconomic variables were compared using the quadratic probability score (QPS).