What explains the variance of prices and returns? Time-series vs. cross-section

This paper studies the relative importance of discount rates and cash flows with a focus on the differences between time-series and cross-sectional variance tests. I show that the following holds for the market, different types of portfolios, and individual stocks: (a) changes in expected returns drive the majority of the time-series volatility in price ratios and unexpected returns, and (b) differences in expected cash flows generate most of the cross-sectional variance in valuations and unexpected returns. Contrary to previous results in the literature, I conclude that individual stocks or portfolios look similar to the market. These findings are robust to short- and long-run regressions and hold when using dividends or (clean surplus accounting) earnings as cash flows. Finally, I present a simple present-value model with latent expected returns and dividend growth rates that explains most of these results.