Discovering stock trading preferences by self-organizing maps and decision trees

Chih Fong Tsai, Yuah Chiao Lin, Yi Ting Wang

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

Stock trading activities are always very popular in many countries. Generally, investors with various backgrounds have different preferences over the stocks they trade. In literature, a number of studies examine the institutions' holding preferences for certain stock characteristics when choosing the security portfolio. However, very few studies investigate the stock trading preferences of individual investors. In this paper, we focus on two factors which affect the portfolio choices of investors, which are stock characteristics and investor features. In particular, a self-organizing map (SOM) is used to group a certain number of clusters based on a chosen dataset. Then, the decision tree model is used to extract useful rules from the clusters which contain the most trading records in the sample. We find that if the investors are females, less wealthy, and make stock trades with lower frequencies, they will be more careful and conservative. On the other hand, if the investors are males, having a high level of wealth, and make stock trades very often, they tend to choose stocks with high EPS, high market-to-book, and high prices.

Original languageEnglish
Pages (from-to)603-611
Number of pages9
JournalInternational Journal on Artificial Intelligence Tools
Volume18
Issue number4
DOIs
StatePublished - Aug 2009

Keywords

  • Decision trees
  • Self-organizing maps
  • Stock characteristics
  • Trading preferences

Fingerprint

Dive into the research topics of 'Discovering stock trading preferences by self-organizing maps and decision trees'. Together they form a unique fingerprint.

Cite this