A Novel Method For Stock Selection Using Weight Threshold

JIITA, vol.5 no.1, p.401-411, 2021, DOI:

A Novel Method For Stock Selection Using Weight Threshold

Thien Nguyen 1), Trang Nguyen 2)
1) Natural Language Processing and Knowledge Discovery Laboratory, Faculty of Information Technology, Ton Duc Thang University, Ho Chi Minh City, Vietnam
2) Faculty of Natural Sciences, Novosibirsk State University, Novosibirsk, Russia

Abstract: This paper studies the portfolio with the optimal in-sample Sharpe ratio, constructed from a large stock pool. Basing on the portfolio, this paper proposes to select only stocks which weights surpass a certain weight threshold in order to avoid over-fitting. The weights of the selected stocks will be normalized to 100%. The new portfolio has better out-of-sample Sharpe ratio than out-of-sample Sharpe ratio of the original portfolio, as the same time, has significantly smaller size, which is more manageable for investors.
The first part of the paper explains the goal of our work. The second part reviews related works on building an investment portfolio. The third section presents the proposed method. The fourth part describes the data in this study. The fifth and sixth parts of the paper discusses experimental results and conclusion, respectively.
With regard to the focus of this conference, an experiment was performed on 6-year historic price data of stocks listed on the Ho Chi Minh City Stock Exchange (HSX), the largest stock market in Vietnam.
Experimental results show that the proposed method of building a portfolio from the HSX market obtains the better out-of-sample Sharpe ratio, reducing the size of portfolio.

KeywordS: Portfolio construction, Sharpe ratio, Ho Chi Minh City Stock Exchange