No. 473 - Revisiting the implications of heterogeneity in financial market participation for the C-CAPM

Recent studies have explored the possibility that accounting for limited participation in financial markets, and in the stock market in particular, might explain the empirical inconsistency of the Consumption-based Capital Asset Pricing Model (C-CAPM). The reasoning is that if non-shareholders’ consumption growth co-varies less than shareholders’ with share returns, it is misleading to include their expenditure in the consumption measure used to test the model. This paper reviews the implications of household portfolio heterogeneity for various well-known characterizations of the empirical failure of the model, such as the inconsistency of consumption-based asset pricing factors with Hansen and Jagannathan bounds, the equity premium puzzle and the rejection of the overidentifying restrictions to the model. Specifically, it provides a unified framework of analysis, based on the US Consumer Expenditure Survey, to assess the extent to which the empirical inconsistency of the C-CAPM can be attributed to the use of aggregate data that do not allow us to account for limited participation in asset markets. The evidence supports the view that accounting for portfolio heterogeneity improves the empirical performance of the model and helps rationalize some of the puzzling findings, but by itself is not enough to reconcile the theory with the empirical evidence.

Published in 2004 in: Review of Finance, v. 8, 3, pp. 445-480