Exploring Investment Optimization and “Greenwashing” from ESG Disclosure: A Dual Examination of Investor Perception

Abstract

This study investigates the intricate correlation between environmental, social, and governance (ESG) information and the financial performance of companies, with a focus on the fundamental nature of ESG and its influence on the choices made by investors. This study examines available literature and data analysis to uncover how disclosing ESG information impacts investment optimization. Additionally, it clarifies the relationship between greenwashing and the advancement of green financial products. The study employs the XGBoost ensemble learning method, using non-financial features of ESG combined with financial features to construct a prediction model, achieving a prediction accuracy rate of 71.26%. Furthermore, applying this model aims to further utilize it in stock selection and constructing a stock pool. By analyzing the financial performance of companies predicted by the model, we will select potential high-performance stocks to build an investment portfolio. Then, we use the Markowitz portfolio theory to optimize the weight combination of stocks in the pool to maximize expected returns and minimize risk. After backtesting the investment portfolio using the closing prices in 2021, its annualized return was a positive 5.76%, significantly higher than the benchmark portfolio. Additionally, this study provides theoretical references and practical guidance for insight and addressing the potential large-scale greenwashing behavior under the trend of increasing ESG information disclosure in the future.