This paper develops a structural model of the costs and beliefs required to rationalize household direct stock ownership. In the model, households believe they can learn information about individual stock returns through costly research. The model provides a novel explanation for many empirical features of household portfolios. Further, the model identifies the distributions of both household research costs and household beliefs about the predictability of individual stock returns. Identification depends only on households’ wealth and portfolio choices. Parameter estimates suggest that most households have modest beliefs about the benefits of individual stock research, although a minority must expect extraordinary returns.
direct stock ownership household beliefs research costs under-diversification