Industrial output and stock price revisited: an application of the multivariate indirect causality model

This paper presents an analysis of the empirical relationship between stock returns, industrial production, money supply, inflation and interest rates across five countries-Canada, France, Japan, Taiwan and the USA. Specifically, we estimate a five-variable vector autoregression model in order to answer the question: does industrial production predict stock returns directly or indirectly (i.e. does industrial production help predict a variable that itself predicts stock returns)? The key result is that there is no direct and significant statistical relationship in any of the five countries, but there is strong evidence of an indirect relation in Taiwan (via money supply) and another indirect relation in the USA (via interest rate). This indirect causality is verified by examining the relative predictability of stock returns both with and without the additional information. Predictability increases when the indirect relationship is exploited. Copyright Blackwell Publishing Ltd and The Victoria University of Manchester, 2004.