Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective

Dome Lohpetch, David Corne

Research output: Contribution to conferenceOther

Abstract

Genetic programming (GP) is increasingly investigated in finance and economics. One area of study is its use to discover effective rules for technical trading in the context of a portfolio of equities (or an index). Early work in this area used GP to find rules that were profitable, but were nevertheless outperformed by the simple "buy and hold" (B&H) strategy. Attempts since then tend to report similar findings, except for a handful of cases where GP methods have been found to outperform B&H. Recent work has clarified that robust outperformance of B&H depends on, mainly, the adoption of a relatively infrequent trading strategy (e.g. monthly), as well as a range of factors that amount to sound engineering of the GP grammar and the validation strategy. Here we add a comprehensive study of multiobjective approaches to this investigation, and find that multiobjective strategies provide even more robustness in outperforming B&H, even in the context of more frequent (e.g. weekly) trading decisions.

Original languageEnglish
Pages192-199
DOIs
Publication statusPublished - 2011
EventIEEE Congress on Evolutionary Computation (CEC) - New Orleans, LA, United States
Duration: 5 Jun 20118 Jun 2011

Conference

ConferenceIEEE Congress on Evolutionary Computation (CEC)
CountryUnited States
CityNew Orleans, LA
Period5/06/118/06/11

Fingerprint

Genetic programming
Finance
Robustness
Grammar
Equity
Technical trading
Trading strategies
Economics
Factors

Keywords

  • financial trading
  • genetic programming
  • multiobjective algorithms

Cite this

Lohpetch, D., & Corne, D. (2011). Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective. 192-199. IEEE Congress on Evolutionary Computation (CEC), New Orleans, LA, United States. https://doi.org/10.1109/CEC.2011.5949618
Lohpetch, Dome ; Corne, David. / Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective. IEEE Congress on Evolutionary Computation (CEC), New Orleans, LA, United States.
@conference{c9ab32a0c382499d8ae967e5bff025ce,
title = "Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective",
abstract = "Genetic programming (GP) is increasingly investigated in finance and economics. One area of study is its use to discover effective rules for technical trading in the context of a portfolio of equities (or an index). Early work in this area used GP to find rules that were profitable, but were nevertheless outperformed by the simple {"}buy and hold{"} (B&H) strategy. Attempts since then tend to report similar findings, except for a handful of cases where GP methods have been found to outperform B&H. Recent work has clarified that robust outperformance of B&H depends on, mainly, the adoption of a relatively infrequent trading strategy (e.g. monthly), as well as a range of factors that amount to sound engineering of the GP grammar and the validation strategy. Here we add a comprehensive study of multiobjective approaches to this investigation, and find that multiobjective strategies provide even more robustness in outperforming B&H, even in the context of more frequent (e.g. weekly) trading decisions.",
keywords = "financial trading , genetic programming , multiobjective algorithms",
author = "Dome Lohpetch and David Corne",
year = "2011",
doi = "10.1109/CEC.2011.5949618",
language = "English",
pages = "192--199",
note = "IEEE Congress on Evolutionary Computation (CEC) ; Conference date: 05-06-2011 Through 08-06-2011",

}

Lohpetch, D & Corne, D 2011, 'Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective' IEEE Congress on Evolutionary Computation (CEC), New Orleans, LA, United States, 5/06/11 - 8/06/11, pp. 192-199. https://doi.org/10.1109/CEC.2011.5949618

Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective. / Lohpetch, Dome; Corne, David.

2011. 192-199 IEEE Congress on Evolutionary Computation (CEC), New Orleans, LA, United States.

Research output: Contribution to conferenceOther

TY - CONF

T1 - Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective

AU - Lohpetch, Dome

AU - Corne, David

PY - 2011

Y1 - 2011

N2 - Genetic programming (GP) is increasingly investigated in finance and economics. One area of study is its use to discover effective rules for technical trading in the context of a portfolio of equities (or an index). Early work in this area used GP to find rules that were profitable, but were nevertheless outperformed by the simple "buy and hold" (B&H) strategy. Attempts since then tend to report similar findings, except for a handful of cases where GP methods have been found to outperform B&H. Recent work has clarified that robust outperformance of B&H depends on, mainly, the adoption of a relatively infrequent trading strategy (e.g. monthly), as well as a range of factors that amount to sound engineering of the GP grammar and the validation strategy. Here we add a comprehensive study of multiobjective approaches to this investigation, and find that multiobjective strategies provide even more robustness in outperforming B&H, even in the context of more frequent (e.g. weekly) trading decisions.

AB - Genetic programming (GP) is increasingly investigated in finance and economics. One area of study is its use to discover effective rules for technical trading in the context of a portfolio of equities (or an index). Early work in this area used GP to find rules that were profitable, but were nevertheless outperformed by the simple "buy and hold" (B&H) strategy. Attempts since then tend to report similar findings, except for a handful of cases where GP methods have been found to outperform B&H. Recent work has clarified that robust outperformance of B&H depends on, mainly, the adoption of a relatively infrequent trading strategy (e.g. monthly), as well as a range of factors that amount to sound engineering of the GP grammar and the validation strategy. Here we add a comprehensive study of multiobjective approaches to this investigation, and find that multiobjective strategies provide even more robustness in outperforming B&H, even in the context of more frequent (e.g. weekly) trading decisions.

KW - financial trading

KW - genetic programming

KW - multiobjective algorithms

U2 - 10.1109/CEC.2011.5949618

DO - 10.1109/CEC.2011.5949618

M3 - Other

SP - 192

EP - 199

ER -

Lohpetch D, Corne D. Multiobjective algorithms for financial trading: Multiobjective out-trades single-objective. 2011. IEEE Congress on Evolutionary Computation (CEC), New Orleans, LA, United States. https://doi.org/10.1109/CEC.2011.5949618