LLM-Assisted and Rule-Based Assessment of ESG Disclosure Quality and Its Association with External ESG Ratings: Exploratory Evidence from S&P 500 Energy Firms
LLM支援とルールベースによるESG開示品質評価と外部ESG格付との関連性：S&P500エネルギー企業からの探索的証拠 (AI 翻訳)
H. Jung, Shaopeng Che, Haein Lee
🤖 gxceed AI 要約
日本語
本研究は、S&P500エネルギー企業のサステナビリティ報告書を対象に、LLMを用いた構造化コンテンツ分析によりESG開示品質指標（QER、TAR、RIS）を構築し、外部ESG格付（S&P Global ESGスコア）との関連性を検証した。個別指標では弱い相関だったが、複合指標ではより明確な正の相関が確認された。開示品質と格付の乖離パターンも観察され、開示品質の多面性が示唆された。
English
This study constructs ESG disclosure quality indicators using LLM-assisted content analysis of sustainability reports from S&P 500 Energy firms. It finds limited positive associations between individual indicators and S&P Global ESG Scores, but a stronger relationship with a composite index. Disclosure-rating divergence patterns suggest multidimensionality in disclosure quality.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本のSSBJ開示基準においても開示品質の評価が重要視されており、本手法は日本語サステナビリティ報告書への応用可能性を示唆する。特に、エネルギーセクターの日本企業が自社の開示品質を客観的に評価する際の参考となる。
In the global GX context
This paper explores how AI can assess ESG disclosure quality, relevant for global ISSB implementation and rating agency methodologies. The findings underscore the need to consider multiple quality dimensions, which can inform disclosure standard-setters and investors.
👥 読者別の含意
🔬研究者:Provides a novel methodology for constructing quantitative ESG disclosure quality indicators using LLMs, and offers exploratory evidence on their relationship with external ESG ratings.
🏢実務担当者:Highlights that composite disclosure quality metrics may align better with ESG ratings than single indicators, guiding firms on how to improve their sustainability reports.
🏛政策担当者:The observed disclosure-rating divergence suggests that disclosure quality assessments should be multidimensional, informing standards like ISSB on how to evaluate and reward quality.
📄 Abstract（原文）
Environmental, social, and governance (ESG) disclosure has become an important source of information for external stakeholders. As sustainability reporting has expanded, distinguishing disclosure quantity from disclosure quality has become increasingly important. This study examines how sustainability disclosure quality is associated with external ESG evaluation outcomes among Standard & Poor’s (S&P) 500 Energy Sector firms. ESG-related claims were identified and classified from sustainability reports using large language model (LLM)-assisted structured content analysis. Based on the resulting corpus, three disclosure quality indicators were constructed: the Quantitative Evidence Ratio (QER), the Target Accountability Ratio (TAR), and the Reporting Infrastructure Score (RIS). These indicators were integrated into a composite Disclosure Quality Index (DQI) and examined in relation to S&P Global ESG Scores using Spearman’s rank correlation. The results indicated limited positive associations for the individual indicators, whereas the composite DQI showed a more pronounced positive relationship. Disclosure–rating divergence patterns were also observed, indicating that relatively favorable disclosure quality positions do not consistently correspond to higher ESG Score rankings. Overall, the findings suggested that sustainability disclosure quality may be multidimensional and that its association with external ESG evaluation outcomes became more apparent when disclosure characteristics were considered in combination. However, because the analysis is restricted to a small sample of S&P 500 Energy Sector firms, the findings should be interpreted as exploratory sector-specific evidence with limited generalizability.
🔗 Provenance — このレコードを発見したソース
- semanticscholar https://doi.org/10.3390/su18136882first seen 2026-07-13 06:43:40
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。