Using a principal component analysis for multi-currencies-trading in the foreign exchange market

Subtitle:

Article type: Research Article

Authors: Byun, Hyun Woo^a | Baek, Seungho^b | Ahn, Jae Joon^c | Oh, Kyong Joo^{a; *} | Kim, Tae Yoon^d

Affiliations: [a] Department of Information and Industrial Engineering, Yonsei University, Seoul, Korea | [b] Booth School of Business, University of Chicago, Chicago, IL, USA | [c] Department of Information and Statistics, Yonsei University, Wonju, Seoul, Korea | [d] Department of Statistics, Keimyung University, Daegu, Korea

Correspondence: [*] Corresponding author: Kyong Joo Oh, Department of Information and Industrial Engineering, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 120-749, Korea. Tel.: +82 2 2123 5720; Fax: +82 2 364 7807; E-mail:[email protected]

Abstract: This study proposes a multi-currencies trading algorithm that applies a stock-trading algorithm to the foreign exchange (FX) market. Our algorithm applies a principal component analysis and artificial neural networks to produce an induced classifier from the FX market. Our algorithm yields reasonable profits. In addition, we discuss a basic procedure that the currency-trading algorithm must follow.

Keywords: Multi-currencies-trading algorithm, principal component analysis, induced classifier, artificial neural networks

DOI: 10.3233/IDA-150738

Journal: Intelligent Data Analysis, vol. 19, no. 3, pp. 683-697, 2015