We investigate wealth returns on an administrative panel containing the disaggregated balance sheets of Swedish residents. The expected return on household net wealth is strongly persistent, determined primarily by systematic risk, and increasing in net worth, exceeding the risk-free rate by the size of the equity premium for households in the top 0.01 percent. Idiosyncratic risk is transitory but generates substantial long-term dispersion in returns in top brackets. Systematic and idiosyncratic risk both drive the cross-sectional distribution of the geometric average return over a generation. Furthermore, wealth returns explain most of the historical increase in top wealth shares.

Boissel and Matray (2022) find that investment increased after 2013 in French firms facing higher dividend taxes. We identify an alteration in the code plotting the event study of the effect of this reform on investment. Using identical data and removing this alteration, we find differential pre-trends between treated and control firms. We also establish that the controls referred to as “size growth,” used in all the difference-in-difference specifications, effectively are controls for lagged investment, i.e., the main outcome variable. Removing such controls attenuates differential pre-trends but leaves no clear event study evidence of a positive effect of dividend taxation on investment.