GX & Decarbonization Research

M. Rao, Falguni Nayak

This study analyzes the short-term relationship between ESG performance and stock prices of 45 Indian BSE-listed companies across nine industries in the post-BRSR period (FY2023-24 and FY2024-25). Using descriptive statistics, ANOVA, regres…

Sasmita Mishra, Zefree Lazarus Mayaluri, C. Liew +2

Proposes a hybrid deep learning model combining ESG scores and news sentiment for asset return forecasting. Uses TFT, SVR residual correction, and gated fusion to quantify ESG-sentiment interactions, finding regime dependence. A latency-opt…

Zilong Yang

This paper examines how negative ESG news affect short-run trading in China's A-share market (2015-2023). Using a BERT classifier, it identifies 4,873 events and finds abnormal turnover +2.84pp (28.3% relative increase), volume +35%, and id…

Guangxing Hu, Bin Li

This study examines the impact of China's Intelligent Manufacturing Demonstration Projects (IMDPs, 2015-2018) on corporate ESG ratings using a quasi-experimental design with propensity score matching and difference-in-differences. IMDP part…

Soon-Ae Choi

This study employs a systematic review to analyze how AI/ML technologies can mitigate the structural dilemma of high initial investment costs and declining financial performance due to strengthened ESG regulations in the Korean cruise indus…

Kahyun Lee

This paper proposes a data-driven predictive framework that uses machine learning to map financial disclosures to ESG ratings, demonstrating practical applicability in linking disclosure content to scores.

Giuseppe Pernagallo, F. Quatraro, Eleonora Rubichi

This paper proposes a text-based circularity index using mandatory non-financial disclosures from large Italian companies. By computing cosine similarity between FTSE MIB sustainability reports (2017-2022) and a circular economy vocabulary …

Benjamin Josef Schüßler, Jakob Prange

This study extends a sentence-level dataset of German ESG reports with crowdsourced readability annotations and evaluates various readability scoring methods. It finds that while LLM prompting can distinguish clear from hard-to-read sentenc…

Tamás Zelles, Bernadett Domokos, Sándor Remsei

This paper examines how combining narrative techniques with machine learning can enhance decision-making, particularly in accounting and sustainability reporting. It finds that narrative-driven reporting with expert interpretation improves …

Huipeng Yang, Yihang Yu, Yuhan Wu +2

This study constructs a climate risk index from annual report text mining of Chinese A-share firms (2016-2024) and examines its impact on the cost of debt financing. Using regression and mediation analysis, it identifies two channels: reduc…

Hafiz Muhammad Naveed, Huaping Sun, Rabia Akram +2

Using an interpretable MNN-DeepSHAP framework on data from 13 countries (2001-2025), this study reveals structural drivers of CO2 intensity and ecological footprint. It finds that improving renewable energy efficiency (lower levelized cost …

Tanzina Hossain, Mahfuja Malik, K. S. M. Tozammel Hossain +1

This study predicts ESG scores for U.S. manufacturing and service sectors using ten machine learning algorithms. Results show ESG scores are more predictable in manufacturing than services, with XGBoost best for manufacturing and Random For…

Hussam Musa, Zdenka Musová, Frederik Rech +1

This study applies XGBoost and SHAP analysis to predict ESG ratings of 974 Slovak manufacturing firms using multi-year financial data. Leverage, liquidity, debt-servicing capacity, firm age, and tax-related indicators emerged as important p…

Edward I. Altman, Francesco Baldi, Claudia D'Ippolito +1

This study develops a novel ESG rating model called Z-ESG, applying Altman's Z-score logic to environmental, social, and governance indicators. Using discriminant analysis and logistic regression on 325 European listed firms, it produces ES…

M T H Nguyen, Anbinh Vuong, Vanha Tran +1

This paper proposes a method to automate the assessment of sustainability reports aligned with GRI standards using large language models (LLMs) and knowledge graph-based retrieval-augmented generation (RAG). A case study of Vietnamese banks…

Petr Procházka, Емил Велинов, Paul Lacourbe

This paper reviews AI tools that enable sustainability governance in multi-tier supply chains. It categorizes existing tools, identifies gaps, and helps low-resource users (SMEs, lower-tier suppliers) access solutions. The evaluation covers…

Berra Öz, Ali Hakan Işık

This paper analyzes the use of large language models (LLMs) in sustainability reporting, focusing on the performance comparison of retrieval-augmented generation (RAG) and low-rank adaptation (LoRA) techniques. It evaluates their effectiven…

Anonymous Author(s)

This replication study validates a method for refining prompts by learning from their own mistakes in analyzing sustainability reports. It demonstrates how iterative improvement based on incorrect outputs enhances prompt quality for ESG tex…

Anonymous Author(s)

This paper proposes and validates a method to refine prompts for analyzing sustainability reports, where the language model learns from its own mistakes. A case study demonstrates the effectiveness of this approach.

Md. Qamruzzaman, Abdulrahman Alomair, Mohammed Alomair

This paper examines how FinTech adoption, intellectual capital, ESG disclosure, and dividend policy affect firm value in financial institutions in an emerging economy (Bangladesh). Using econometric methods including deep neural networks, i…