2010-11: An Empirical Investigation of Consumption CAPMs in the Australian Market (Working paper)

Abstract

We test conditional (C)CAPMs in the Australian equity market. The conditional variables used are Lettau and Ludvigson's (2001a, 2001b) consumption-wealth ratio, Campbell and Cochrane's (1999) surplus consumption ratio and Santos and Veronesi's labor income to consumption ratio (2006). We test the cross-sectional implications of these variables using the Fama-French 25 size and book-to-market portfolios and the Australian industry portfolios. The Fama-MacBeth (1973) cross-sectional regressions on the 25 size/BM portfolios show that the conditional model using some scaling variable can perform better than the unconditional models. However, these conditional models cannot outperform the Fama-French three-factor model. The conditional CCAPM with the labour income to consumption ratio as the scaling factor can match more closely to the Fama- French three-factor model. We also find consumption growth is non-contemporaneously related to portfolio returns as the consumption risk premium is positive and statistically significant using consumption growth two-quarter ahead. The asset pricing models perform poorly for the 19 industry portfolios.