👍The Equity Premium and the One Percent

In the fall of 2013, Kieran told me that some detrended version of the Piketty-Saez top 1% income share series predicts subsequent stock market returns. I thought that was interesting, so we started to work on this project. The idea is that when there are agents with heterogeneous preferences (say, discount rate or risk aversion), the patient and/or risk-tolerant type will invest more in stocks and accumulate wealth, but their wealth also drives up stock prices and predicts subsequent lower returns. We confirmed this mechanism both theoretically using a general equilibrium model and empirically using instrumental variable regressions using estate tax data.

In those days we were not as good researchers as we are now, so it took time to improve the model and the empirical analysis. One cool thing we learned along the way is the uniqueness of equilibrium with arbitrary homothetic preferences and collinear endowments, due to Eisenberg (1961). This result is crucial for deriving unambigous comparative statics. Due to our immaturity, we had a hard time publishing this paper, getting rejected by most of the top 5 journals. Eventually we submitted the paper to RFS in March 2017 and got an R&R, but at each round one of the referees (an empirical researcher) came up with new complaints and the editor threatened to reject the paper, which was such a pain. The paper was accepted in July 2019 after the fourth round.