The Role of Heterogeneity in Asset Pricing : The Effect of a Clustering Approach

In this paper we use a novel clustering approach to study the role of heterogeneity in asset pricing. We present evidence that the equity premium is consistent with a stochastic discount factor calculated as the average of the household clusters' intertemporal marginal rates of substitution in the 1984-2002 period. The result is driven by the skewness of the cluster-based cross-sectional distribution of consumption growth, but cannot be explained by the cross-sectional variance and mean alone. We find that nine clusters are sufficient to explain an equity premium with relative risk aversion coefficient equal to six. The result is robust to various averaging schemes of cluster-based consumption growth used to construct the stochastic discount factor. Lastly, the analysis reveals that standard approximation schemes of the stochastic discount factor using individual household data produce unreliable results, implying a negative stochastic discount factor