This paper suggests a solution to the puzzling finding documented in Moskowitz and Vissing-Jorgensen (2002) that the return to an index of private equity is equal to the return to the CRSP index of public equity even though investment in private firms is substantially riskier. It presents an occupational choice model where human and financial capital can be jointly invested in a private business as an alternative to working for pay and investing in the stock market. Entrepreneurs learn about their firm’s unobservable average return over time. Only firms that pay a high average return survive but the existence of young firms that learn about their, possibly low, productivities drives down the return to the aggregate stock of private equity. Agents are willing to make the initial investment at a negative premium on the stock to buy the option of enjoying the more favorable self-employment earnings process. I calibrate the model and show that this mechanism can explain the empirical fact mentioned above. The learning model in the paper obtains this result by using a self-employment earnings process that is consistent with the results in Hamilton (2000) both in terms of average excess earnings over the ones for paid-employees and in terms of a tenure profile of entrepreneurial earnings that starts below and then overtakes the one for workers after a few years.
