Modelling and trading the London, New York and Frankfurt stock exchanges with a new gene expression programming trader tool
Author | Andreas Karathanasopoulos |
DOI | http://doi.org/10.1002/isaf.1401 |
Date | 01 January 2017 |
Published date | 01 January 2017 |
RESEARCH ARTICLE
Modelling and trading the London, New York and Frankfurt
stock exchanges with a new gene expression programming
trader tool
Andreas Karathanasopoulos
Olayan School of Business, American
University of Beirut, Beirut, Lebanon
Correspondence
Andreas Karathanasopoulos, Olayan School of
Business, American University of Beirut,
Beirut, Lebanon.
Email: andreas.kara@hotmail.com
Summary
The scope of this manuscript is to present a new short‐term financial forecasting and trading tool:
the Gene Expression Programming (GEP) Trader Tool. It is based on the gene expression pro-
gramming algorithm. This algorithm is based on a genetic programming approach, and provides
supreme statistical and trading performance when used for modelling and trading financial time
series. The GEP Trader Tool is offered through a user‐friendly standalone Java interface. This
paper applies the GEP Trader Tool to the task of forecasting and trading the future contracts
of FTSE100, DAX30 and S&P500 daily closing prices from 2000 to 2015. It is the first time that
gene expression programming has been used in such massive datasets. The model's performance
is benchmarked against linear and nonlinear models such as random walk model, a moving‐
average convergence divergence model, an autoregressive moving average model, a genetic pro-
gramming algorithm, a multilayer perceptron neural network, a recurrent neural network a higher
order neural network. To gauge the accuracy of all models, both statistical and trading perfor-
mances are measured. Experimental results indicate that the proposed approach outperforms
all the others in the in‐sample and out‐of‐sample periods by producing superior empirical results.
Furthermore, the trading performances are improved further when trading strategies are imposed
on each of the models.
KEYWORDS
gene expression,genetic algorithm, trading, transaction
1|INTRODUCTION
Many technical trading algorithms have been introduced in the last
decade, with artificial neural networks being the dominant machine‐
learning technique for financial forecasting applications (Zhang,
2012). However, these techniques present certain drawbacks, such
as overfitting issues, data snooping bias, and a curse of dimensionality,
as have been mentioned by Karathanasopoulos, Sermpinis, Laws, and
Dunis (2014). Despite the vast number of different financial forecast-
ing algorithms which are continuously being published, only a few of
them have been combined to construct efficient trading strategies,
and even fewer have provided user‐friendly graphical user interfaces.
Related to the new tool presented in this research paper it is important
to show some description characteristics of the main partof the model.
Algorithms based on gene expression programming (GEP) are domain‐
independent problem‐solving techniques that run in various
environments. GEP can be categorized in the forecasting bracket
known in the finance world as ‘evolutionary algorithms’. These envi-
ronments have been structured in a manner which approximates prob-
lems in order to produce forecasts with a high degree of accuracy. The
basis for this type of problem‐solving technique derives from the
Darwinian principle of reproduction and survival of the fittest. Addi-
tionally, they can be considered similar to the biological genetic opera-
tions of crossover recombination and mutation. These algorithms have
been successfully applied to many real‐world problems. Some recent
research (Yaghouby, Ayatollahi, Bahramali, Yaghouby, & Alavi, 2010;
Alavi & Gandomi, 2011; Divsalar, Firouzabadi, Sadeghi, Behrooz, &
Alavi, 2011; Divsalar, Roodsaz, Vahdatinia, Norouzzadeh, & Behrooz,
2012; Gandomi & Alavi, 2011; Gandomi, Alavi, Mirzahosseini, &
Moghadas, 2011; Gandomi, Yang, Talatahari, & Alavi, 2013; Yang,
Gandomi, Talatahari, & Alavi, 2012; Zargari, Zabihi, Alavi, & Gandomi,
2012) proves that GEP outperforms all the genetic algorithms due to
Received: 7 February 2016 Revised: 5 August 2016 Accepted: 6 October 2016
DOI 10.1002/isaf.1401
Intell Sys Acc Fin Mgmt. 2017;24:3–11. Copyright © 2016 John Wiley & Sons, Ltd.wileyonlinelibrary.com/journal/isaf 3
To continue reading
Request your trial