The consumption–wealth ratio, real estate wealth, and the Japanese stock market

Abstract

The first contribution of this paper, in following the works of Lettau and Ludvigson (2001a,b), is to construct a Japanese consumption–wealth ratio data series and to examine whether it explains Japanese stock market data. We find that the consumption–wealth ratio does predict future stock returns, but the evidence is weaker than that from US data, and the source of predictability is limited to observations after the collapse of the asset bubble at the beginning of the 1990s. The consumption–wealth ratio also helps to explain cross-sectional Japanese stock returns. The second contribution of the paper is that we propose new consumption–wealth ratio that more explicitly deal with household real estate wealth utilizing Japanese aggregate-level data. Such “real estate augmented” consumption–wealth ratio perform better than the consumption–wealth ratio calculated with only financial wealth data. While the scaled factor model proposed by Lettau and Ludvigson performs relatively well with Japanese data, the book-to-market related anomaly pointed out by Chan et al. (1991) and Jagannathan et al. (1998) remains strong.