Blockchain-based dual ESG index for corporate sustainability assessment

Abstract

Investors are increasingly relying on Environmental, Social, and Governance (ESG) indexes to obtain a third-party assessment of corporate sustainability performance. However, due to the inflexible ESG assessment framework, obscure assessment process, and subjective weight preference, the current ESG ratings mainly focus on the large listed companies, and the rating results are questioned by the ESG stakeholders. This thesis proposes a novel Dual ESG Index (DESGI) using blockchain technology to provide a flexible, transparent, and objective ESG rating methodological solution. Precisely, this study consists of four modules. Module I aims to shed fresh light on proposing the dual ESG index (DESGI) by borrowing the rationale and concepts from the academic credit system. First, the theoretical DESGI model is developed to measure the depth and width of corporate sustainability performance. Second, the architecture of the blockchain-based DESGI platform is designed to bring transparency into the various ESG rating services. Third, a practical roadmap for ESG blockchain design, development, application, and validation is provided. Based on the blockchain-based DESGI platform, three ESG rating services are established as the following modules. Module Ⅱ formulates a rule-based DESGI service using smart contract to provide a flexible and reliable ESG rating. This service is designed for the scenario where ESG raters already have ESG weight preferences. First, the framework of the rule-based DESGI using blockchain technology is proposed. Second, ESG GPA (grade point average) and ESG credit measure the depth and width of corporate sustainability performance, respectively. Third, the case study explores the feasibility of the rule-based DESGI services using smart contract and crypto token and the sensitivity analysis validates the DESGI stability. Module Ⅲ provides a robust ESG rating service using SMAA-2 (Stochastic Multi-criteria Acceptability Analysis). This service is designed for the scenario where ESG raters face uncertainties in ESG weight preference. On the one hand, SMAA-2 provides the advantages of robustness and no subjective weight preferences to assess the ESG data. On the other hand, blockchain provides a transparent ledger to store and share ESG data among listed companies, investors, and stakeholders alike. Results show that the proposed blockchain-based ESG assessment approach using SMAA-2 can robustly analyze the companies' ESG performances and benchmark the sustainability level in the peer industry. Module Ⅳ investigates a data-driven DESGI rating service using machine learning techniques to provide an objective ESG rating service. This service is designed for the scenario where the ESG raters own massive ESG historical data. First, we collect and preprocess the real ESG data from 100 textile and apparel companies from 2011 to 2020. Second, the ESG course credits are calculated using combined Criteria Importance Through Intercriteria Correlation (CRITIC) and Maximum Entropy Bootstrap (Meboot) methods due to their advantages of objectiveness and robustness. Third, we analyze the ESG dataset using five types of machine-learning techniques. The results show that the random forest regression has the optimal ESG prediction, which provides opportunities to foresee the corporate sustainability performance and help the listed companies make optimal ESG strategies.