I document a new stylized fact: the higher the share of institutional ownership in a stock, the more its valuation is driven by expected-return variation rather than by changes in dividend-growth expectations. As general equilibrium outcomes, expected returns inevitably depend on investors’ properties and the circumstances they are under. Time-variation in the volatility of the pricing kernel of institutional investors acting as marginal investors in stocks with high institutional ownership translates into time-varying expected returns in those stocks. In my model, imperfect sharing of time-varying risk generates cross-sectional differences in return predictability depending on ownership. My findings help explain the weak return predictability of small and value stocks and the postwar predictability reversal.

We introduce Implied Volatility Duration (IVD) as a new measure for the timing of uncertainty resolution, with a high IVD corresponding to late resolution. Portfolio sorts on a large cross-section of stocks indicate that investors demand on average about seven percent return per year as a compensation for a late resolution of uncertainty. In a general equilibrium model, we show that `late' stocks can only have higher expected returns than `early' stocks if the investor exhibits a preference for early resolution of uncertainty. Our empirical analysis thus provides a purely market-based assessment of the timing preferences of the marginal investor.


I test the hypothesis that the differences in return predictability between the value and growth portfolios are indeed due to value stocks having shorter cash flow duration. I find that duration acts as amplification for the change in dividend yields that is caused by discount rate variation. However, differences in return predictability across book-to-market sorted portfolios go beyond the effect of duration. For comparable cash flow duration, discount rate variation explains about 40% more of the dividend yield variation for growth stocks as opposed to value stocks. This is consistent with recent research suggesting that the exposure to value-specific risks is not driven by duration.