Advantages of ESG Indexes Compared to Traditional Ones in Predicting Stock Prices

Abstract

This paper investigates the predictability of traditional and ESG indices in the Ukrainian stock market, examining potential differences between models. The study tests two hypotheses: (H1) ESG indices exhibit lower predictability than traditional indices, and (H2) different forecasting models should be applied to ESG and conventional indices. Various forecasting models, including AR, ARIMA, ARCH, GARCH, TGARCH, Logit, and Probit regressions, are employed, along with additional tests, using daily data from the Ukrainian stock market (UX, PFTS, and WIG indices) spanning 2015–2022. The findings confirm both hypotheses for the case of returns, indicating less predictability for ESG indices and the need for distinct models. For volatility, there is limited evidence supporting Hypothesis 1, but Hypothesis 2 is confirmed. Possible factors explaining the results include higher information transparency, liquidity, and trading activity in ESG indices. The research has implications for academics and practitioners, emphasizing the importance of employing different models for forecasting ESG and traditional indices. It also highlights the preference for traditional indices in trading and speculative activities. The study suggests that a shift toward ESG indices represents a move toward more efficient markets.