Barberis (2000) demonstrates that estimation risk plays an important role in long-run asset allocation. After incorporating estimation risk, he finds that predictability in returns makes investors allocate substantially more to stocks in the long run than in the short run. We show that, when the nonnegative equity premium (NEP) condition is imposed on predictive regressions, investors actually allocate roughly the same or less to stocks in the long run. The reason is that the NEP reduces mean reversion in returns by providing informative prior beliefs. A long-horizon investor who ignores reduction in mean reversion may overallocate to stocks by a sizeable amount.
Please see the link below for a full text:
https://doi.org/10.1016/j.pacfin.2020.101397
Yugui Zhang, Jie Zhu, Xiaoneng Zhu, Investing for the long run when expected equity premium is nonnegative, Pacific-Basin Finance Journal, Volume 63, October 2020, 101397.
