Machine Learning versus Economic Restrictions: Evidence from Stock Return Predictability

72 Pages Posted: 18 Sep 2019 Last revised: 6 Apr 2020

See all articles by Doron Avramov

Doron Avramov

Interdisciplinary Center (IDC) Herzliyah

Si Cheng

Chinese University of Hong Kong - Department of Finance

Lior Metzker

Hebrew University of Jerusalem

Date Written: April 5, 2020

Abstract

This paper shows that machine learning methods often fail to clear standard economic restrictions. Investments based on deep learning signals extract profitability from difficult-to-arbitrage stocks and during alleviated limits-to-arbitrage market states. Value-weighting returns and excluding microcaps or distressed stocks considerably attenuate profitability. Performance further deteriorates in the presence of trading costs due to high turnover or extreme positions in the tangency portfolio implied by the pricing kernel. Despite their opaque nature, machine learning methods identify mispriced stocks consistent with most anomalies. Beyond economic restrictions, deep learning signals are profitable in long positions, remain viable in recent years, and command low downside risk.

Keywords: Machine Learning, Return Prediction, Neural Networks, Financial Distress, Fintech

JEL Classification: G10, G11, G12, G14

Suggested Citation

Avramov, Doron and Cheng, Si and Metzker, Lior, Machine Learning versus Economic Restrictions: Evidence from Stock Return Predictability (April 5, 2020). Available at SSRN: https://ssrn.com/abstract=3450322 or http://dx.doi.org/10.2139/ssrn.3450322

Doron Avramov

Interdisciplinary Center (IDC) Herzliyah ( email )

P.O. Box 167
Herzliya, 46150
Israel

Si Cheng (Contact Author)

Chinese University of Hong Kong - Department of Finance ( email )

12/F, Cheng Yu Tung Building
No.12, Chak Cheung Street
Shatin, N.T.
Hong Kong

HOME PAGE: http://www.bschool.cuhk.edu.hk/staff/cheng-si/

Lior Metzker

Hebrew University of Jerusalem ( email )

Mount Scopus
Jerusalem, IL Jerusalem 91905
Israel

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
739
Abstract Views
2,679
rank
36,732
PlumX Metrics