Exploring the Factor Zoo With a Machine-Learning Portfolio

60 Pages Posted: 9 Aug 2018 Last revised: 2 Feb 2021

See all articles by Halis Sak

Halis Sak

Hong Kong University of Science & Technology (HKUST) - Department of Finance

Tao Huang

Beijing Normal University-Hong Kong Baptist University United International College

Michael Chng

Xian Jiaotong Liverpool University; Department of Finance, Deakin University

Date Written: January 18, 2021

Abstract

Over the years, top journals have published hundreds of characteristics to explain stock return, but many have lost significance. What fundamentally affects the time-varying significance of characteristics? We combine machine-learning (ML) and portfolio analysis to uncover patterns in significant characteristics. We train ML models on 106 characteristics to predict stock returns. From out-of-sample ML portfolio analysis, we reverse-engineer important characteristics that ML models uncover, which are unobservable. The ML portfolio’s dominant characteristics rotate between proxies for investor arbitrage constraint and firm financial constraint. We show that the credit cycle could fundamentally explain cross-sectional stock return over time.

Keywords: Machine Learning, Characteristics, Return Predictability, Portfolio Evaluation

JEL Classification: G12, G32

Suggested Citation

Sak, Halis and Huang, Tao and Chng, Michael, Exploring the Factor Zoo With a Machine-Learning Portfolio (January 18, 2021). WRDS Research Paper, Available at SSRN: https://ssrn.com/abstract=3202277 or http://dx.doi.org/10.2139/ssrn.3202277

Halis Sak (Contact Author)

Hong Kong University of Science & Technology (HKUST) - Department of Finance ( email )

Clear Water Bay, Kowloon
Hong Kong

Tao Huang

Beijing Normal University-Hong Kong Baptist University United International College ( email )

2000 Jintong Road
Zhuhai, Guangdong 519087
China

Michael Chng

Xian Jiaotong Liverpool University ( email )

BB520, 111 Ren Ai Lu
Higher Education Zone, SIP
Suzhou, Jiangsu 215123
China

Department of Finance, Deakin University ( email )

75 Pigdons Road
Victoria, Victoria 3216
Australia

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
576
Abstract Views
2,628
rank
70,250
PlumX Metrics