Page 3 - MFB22侯國弘

Abstract

Stock selection strategies are a topic that has been arousing great

interest and heated discussion among investors. The popular

environment, social and governance (ESG) indicators and big data

predictions are two types of such strategies. Since the concept of

sustainable operations from an ESG perspective which was introduced

by a United Nations report in 2005, many governments and enterprises

around the world have been committed to pursuing the business goal of

ESG sustainable development. Meanwhile, there are myriad ways in

which big data predictions are used for stock selection. The ‘volume of

internet posts’ is one of the big data prediction-based stock selection

methods. This study mainly uses ESG indicators and the volume of

internet posts for individual stocks to formulate portfolios, and

compares the performance of these portfolios with those of Yuanta

Taiwan Top 50 (0050.TW) exchange traded funds (ETF) and other ESG

related ETFs. We first use Taiwan Stock Exchange (TWSE) of Taiwan

Corporate Governance Assessment Indicator and Google search volume

index to select the top 15, top 20, top 25, top 30, top 35 and top 40

companies. Then we use the stocks of the selected companies to

construct the portfolios. Further, Markowitz’s Modern Portfolio Theory

was adopted to generate the efficient frontier of the portfolios.

According to the empirical research results, the rate of return and

Sharpe ratio of portfolios selected based on a combination of ESG

indicators and the volume of internet posts over the three-year period

perform better than those of Yuanta Taiwan Top 50 ETF (0050.TW) and

other ESG related ETFs. On this ground, this study argues that

ii