Home / Current Issue / Paper 1719944
A Comparative Study of Deep Learning Models for ESG Index Volatility Prediction
Subject area: Management and Commerce · Area of research: Finance (ESG)
DOI: https://doi.org/10.64388/IREV10I1-1719944
Abstract
The inclusion of Environmental, Social, and Governance (ESG) factors has achieved recognition in investment strategies, influencing market dynamics and investor behavior. However, predicting the volatility of ESG indices presents additional complexities due to heightened sensitivity to various external macroeconomic and geopolitical factors. This study develops an structured computational framework by utilizing deep learning architectures namely Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Convolutional Neural Network (CNN) to enhance ESG index volatility prediction. The models were rigorously evaluated using standard performance metrics and statistical validation through Welch’s t-tests to ensure robustness and reliability of outcomes. Experimental results indicate that the GRU model achieved superior performance compared to both LSTM and CNN, while achieving superior accuracy and interpretability. These results provide meaningful insights for both investors and researchers looking for data-driven strategies for ESG market forecasting.
Keywords
ESG Investing, ESG Indices, Deep Learning, Volatility Prediction, Machine Learning
How to cite this paper
@article{1719944,
author = {Priya Rani},
title = {A Comparative Study of Deep Learning Models for ESG Index Volatility Prediction},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {10},
number = {1},
pages = {2317-2329},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1719944.pdf},
abstract = {The inclusion of Environmental, Social, and Governance (ESG) factors has achieved recognition in investment strategies, influencing market dynamics and investor behavior. However, predicting the volatility of ESG indices presents additional complexities due to heightened sensitivity to various external macroeconomic and geopolitical factors. This study develops an structured computational framework by utilizing deep learning architectures namely Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Convolutional Neural Network (CNN) to enhance ESG index volatility prediction. The models were rigorously evaluated using standard performance metrics and statistical validation through Welch’s t-tests to ensure robustness and reliability of outcomes. Experimental results indicate that the GRU model achieved superior performance compared to both LSTM and CNN, while achieving superior accuracy and interpretability. These results provide meaningful insights for both investors and researchers looking for data-driven strategies for ESG market forecasting.},
keywords = {ESG Investing, ESG Indices, Deep Learning, Volatility Prediction, Machine Learning},
month = {July},
doi = {https://doi.org/10.64388/IREV10I1-1719944}
}