Machine Learning Portfolios with Equal Risk Contributions

30 Pages Posted: 8 Aug 2019 Last revised: 9 Dec 2019

See all articles by Alexandre Rubesam

Alexandre Rubesam

IESEG School of Management; French National Center for Scientific Research (CNRS) - Lille Economie & Management (LEM) UMR 9221

Date Written: November 6, 2019

Abstract

We use machine learning methods to forecast individual stock returns in the Brazilian stock market, using a unique data set including technical and fundamental predictors. We find that portfolios formed on the highest quintile of predicted returns significantly outperform market benchmarks. However, portfolios formed on the lowest quintile of predicted returns earn positive returns and have high volatilities, making traditional long-short strategies unnatractive. To resolve this problem, we propose an equal risk contribution (ERC) ensemble approach to build a portfolio combining long-short portfolios obtained with various machine learning methods such that (i) the risk contributions of all individual long-short portfolios are equal, and (ii) the aggregate risk contribution of all long positions equals that of all short positions. The ERC ensemble portfolio outperforms, on an after cost, risk-adjusted basis, all individual machine learning long-short portfolios, as well as equally-weighted ensembles of these portfolios.

Keywords: machine learning, forecasting, return prediction, risk parity, equal risk contribution

JEL Classification: C53, G11, G15

Suggested Citation

Rubesam, Alexandre, Machine Learning Portfolios with Equal Risk Contributions (November 6, 2019). Available at SSRN: https://ssrn.com/abstract=3432760 or http://dx.doi.org/10.2139/ssrn.3432760

Alexandre Rubesam (Contact Author)

IESEG School of Management ( email )

Socle de la Grande Arche
1 Parvis de la Defense
Puteaux, Paris 92800
France

French National Center for Scientific Research (CNRS) - Lille Economie & Management (LEM) UMR 9221 ( email )

Lille
France

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
303
Abstract Views
1,306
rank
109,704
PlumX Metrics