Technological bias at the exchange rate market

• Author: Svitlana Galeshchuk
• Date: 01 April 2017
• DOI: http://doi.org/10.1002/isaf.1408
• Published date: 01 April 2017

SPECIAL ISSUE ARTICLE

Technological bias at the exchange rate market

Svitlana Galeshchuk

1,2

1

Laboratoire d'Informatique de Grenoble,

Université Grenoble Alpes, Grenoble, France

2

Faculty of Accounting and Audit, Ternopil

National Economic University, Ukraine

Correspondence

Svitlana Galeshchuk, Laboratoire

d'Informatique de Grenoble, Université

Grenoble Alpes, Grenoble, France.

Email: svitlana.galeshchuk@univ‐grenoble‐

alpes.fr

Summary

Prediction of exchange rates has been a topic for debate in economic literature since the late

1980s. The recent development of machine learning techniques has spurred a plethora of studies

that further improves the prediction models for currency markets. This high‐tech progress may

create challenges for market efficiency along with information asymmetry and irrationality of

decision‐making. This technological bias emerges from the fact that recent innovative approaches

have been used to solve trading tasks and to find the best trading strategies. This paper

demonstrates that traders can leverage technological bias for financial market forecasting. Those

traders who adapt faster to the changes in market innovations will get excess returns. To support

this hypothesis we compare the performance of deep learning methods, shallow neural networks

with baseline prediction methods and a random walk model using daily closing rate between

three currency pairs: Euro and US Dollar (EUR/USD), British Pound and US Dollar (GBP/USD),

and US Dollar and JapaneseYen (USD/JPY). The results demonstrate that deep learning achieves

higher accuracy than alternate methods. The shallow neural network outperforms the random

walk model, but cannot surpass ARIMA accuracy significantly. The paper discusses possible out-

comes of the technological shift for financial market development and accounting conforming

also to adaptive market hypothesis.

KEYWORDS

deep network, prediction for exchange rates, technological bias, tradingstrategy

1|INTRODUCTION

Recent developments of the new financial instruments and the data

mining techniques together with machine learning methods pose a

question as to whether these innovations will change the market's

environment, so that classical ‘tried and tested’economic theories will

not be applicable anymore. Evolution of the trading strategies

demands evolution of the economic theory to understand market

changes and adjust macroeconomic policy.

The efficient market hypothesis (EMH) is one of the most

well‐known financial market theories of the financial market. It

suggests the financial market cannot be predicted accurately enough

for consistent excess returns. Investors possess all information in the

market and exploit it. Available information is reflected in the current

prices. The fact that the profitability of the most popular approaches

of technical analysis decreased significantly after the mid‐1990s calls

into question the efficiency of some of these tools and is used by the

researchers as a proof of the EMH.

We believe that rapid development of heuristic computational

tools, enhancement of computers and computer equipment, and

growth of electronic trading make room for a new form of efficient

market challenge –technological bias.

If we consider the same pace of technological progress of trading

infrastructure and computational tools along with software in the

coming years, the traders who are able to adapt to these technological

changes will get more profitable trading solutions than those who

require more time to adjust. Assuming that technology is considered

as ‘scientific knowledge used in practical ways in industry’(Oxford

Dictionary), we define technological bias at the market as a bias arising

from enhancing trading tools and/or trading environment based on the

development of information technology. Technological bias has not

motivated enough discussions among academics, despite its growing

significance. It has worried other interested groups, mainly market

traders and state decision‐makers.

Our contribution is in providing evidence against the EMH, while

displaying technological bias. Hence, the goal of the article is to

investigate the upcoming technological shift in the foreign‐exchange

market with demonstration of cutting‐edge deep learning tools

performance in the prediction of foreign‐exchange rates in comparison

with the baseline time‐series methods and random walk model.

DOI: 10.1002/isaf.1408

80 Copyright © 2017 John Wiley & Sons, Ltd. Intell Sys Acc Fin Mgmt. 2017;24:80–86.wileyonlinelibrary.com/journal/isaf