Dissecting Market Expectations in the Cross-Section of Book-to-Market Ratios
28 Pages Posted: 16 Sep 2019 Last revised: 15 Nov 2019
Date Written: November 14, 2019
I find no evidence that partial least squares based on disaggregated book-to-market ratios produces a model of market premiums with persistently positive out-of-sample R2, as originally documented for market returns. This is consistent with time variation in predictability, for example, and does not necessarily invalidate the method. The two main drivers of the original performance are: (i) The sample period and (ii) the use of market returns as forecasting targets. Two other drivers are using, as regressors, (iii) the book-to-market ratios of the specific portfolios double-sorted by size and book-to-market (iv) divided by their standard deviations.
Keywords: predictability, out-of-sample, equity premium, disaggregated
JEL Classification: G11, G12, G14
Suggested Citation: Suggested Citation