Cash-flow or return predictability at long horizons? The case of earnings yield

Paulo Fraga Martins Maio*, Danielle Xu

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

5 Citations (Scopus)


We examine the predictive ability of the aggregate earnings yield for both market returns and earnings growth by estimating variance decompositions at multiple horizons. Based on weighted long-horizon regressions, we find that most of the variation in the earnings yield is due to return predictability, with earnings growth predictability assuming a minor role. However, by using implied estimates from a first-order restricted VAR, we find an opposite predictability mix. The inconsistency in results stems from a misspecification of the restricted VAR. Using an unrestricted first-order VAR estimated by OLS, or alternatively, estimating the restricted VAR by the Projection Minimum Distance method, produces long-run variance decompositions that are substantially more similar to the decomposition obtained under the direct method. Hence, earnings yield is not fundamentally different from the dividend yield. These results suggest that the practice of analyzing long-run return and cash-flow predictability from a restricted VAR can be quite misleading.
Original languageEnglish
Peer-reviewed scientific journalJournal of Empirical Finance
Pages (from-to)172-192
Number of pages21
Publication statusPublished - 12.2020
MoE publication typeA1 Journal article - refereed


  • 512 Business and Management
  • predictability of stock returns
  • earnings-growth predictability
  • weighted long-horizon regressions
  • earnings yield
  • VAR implied predictability
  • present-value model
  • dividend yield
  • Projection Minimum Distance method


Dive into the research topics of 'Cash-flow or return predictability at long horizons? The case of earnings yield'. Together they form a unique fingerprint.

Cite this