Forecasting stock market returns by combining sum-of-the-parts and ensemble empirical mode decomposition

ABSTRACT

In this article, we combine the sum-of-the-parts (SOP) method with Ensemble Empirical Mode Decomposition (EEMD) to forecast stock market returns. We obtain very significant stock return predictability both in statistical and economic terms. Interestingly, the strongest performance is achieved by the extended SOPEEMD method to forecast stock market returns when the price-earnings multiple growth is forecasted using the dividend yield as predictor ($R_{oos}^2$of 21.25%) with monthly data and the book-to-market ratio as predictor achieves $R_{oos}^2$ of 20.05% with monthly data. The highest monthly CER gains for the extended SOPEEMD method are for book-to-market ratio reach 14.11%. Furthermore, the evidence based on robust check supports the feasibility of our forecasting strategy.

KEYWORDS:

• Predictability
• stock returns
• asset allocation
• ensemble empirical mode decomposition

JEL CLASSIFICATION:

• G11
• G12
• G14

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ http://www.hec.unil.ch/agoyal/.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China granted [71771030, 11301041,71671018].