← 論文一覧に戻る

ESG-DIVERGE:再現可能なESG格付不一致と頑健性分析

ESG-DIVERGE: Reproducible ESG Rating Disagreement and Robustness Analysis (原題)

Stephen Kamran

Zenodo (CERN European Organization for Nuclear Research)ジャーナル2026-11-11#ESG経営インパクト: 資金調達対象セクター: finance
DOI: 10.5281/zenodo.24144950
原典: https://doi.org/10.5281/zenodo.24144950

🤖 gxceed AI 要約

日本語

ESG-DIVERGEは、ESG格付プロバイダー間の不一致を再現可能に分析するPythonベースの研究ソフトウェアである。中国上場企業377社・1,323社年の観測値を用い、格付不一致の持続性や企業財務特性との関連をパネル回帰で検討する。商用格付データは含まず、集計結果と再現手順を公開する点に特徴がある。

English

ESG-DIVERGE is a Python-based reproducibility framework for analyzing cross-provider ESG rating disagreement. Using 1,323 firm-year observations on 377 Chinese listed firms (2019–2024), it examines rating divergence persistence and links to firm financial characteristics via panel regressions. It ships workflows and aggregate results, not proprietary rating data.

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

格付機関間の不一致は、SSBJ・有報でのESG情報開示や投資家の格付利用における解釈の幅を示す。日本企業が複数格付をどう扱うかを考える上で再現可能な分析手法として参考になる。

In the global GX context

Rating disagreement is central to ISSB/CSRD-era debates on how investors consume ESG ratings and how disclosure quality affects divergence. The reproducible workflow offers a template for transparency-oriented rating research, though it is China-focused and non-causal.

👥 読者別の含意

🔬研究者:格付不一致の再現可能な計量ワークフローと限界の整理を参考にできる。

🏢実務担当者:複数格付のばらつきを自社評価の文脈でどう解釈するかの示唆を得られる。

🏛政策担当者:格付データの権利・再現性・透明性に関する論点を把握できる。

📄 Abstract(原文)

ESG-DIVERGE: Reproducible ESG Rating Disagreement and Robustness Analysis ESG-DIVERGE is an independently developed Python-based research software and reproducibility framework for examining cross-provider Environmental, Social, and Governance (ESG) rating disagreement, associated firm financial characteristics, and methodological robustness. The research workflow evaluates ESG disagreement through harmonised rating measures, descriptive statistics, persistence analysis, panel regression specifications, and exploratory lagged associations. Research Coverage The accompanying analysis examines 1,323 cleaned firm-year observations representing 377 Chinese listed company identifiers during 2019–2024, using secondary academic replication data attributed to MSCI and FTSE Russell. Software and Research Outputs The archive contains: A Python/Jupyter research notebook with stored execution outputs removed. Statistical analysis and robustness-testing workflows. Aggregate descriptive and regression results. Statistical verification documentation. Methodological limitations and data provenance records. Reproducibility instructions and file-integrity documentation. Data Sources and Attribution The underlying ESG-related observations originate from the publicly accessible academic replication dataset: Guo, W., and Du, Y. (2026). ESG Rating Disagreement and Bank Loans. Mendeley Data, Version 1. https://doi.org/10.17632/krnfm86389.1 The original dataset authors are acknowledged for their academic replication materials. MSCI and FTSE Russell are acknowledged as the identified sources of the underlying ESG rating information. Data Rights and Access Restrictions This software archive excludes original company-level MSCI/FTSE rating datasets and proprietary firm-level analytical observations. The included analytical workflow requires users to obtain necessary input data independently through authorised sources. The public availability of the original academic repository does not establish unrestricted redistribution rights over third-party commercial ESG information. Aggregate research outputs also remain subject to applicable rights review. Scientific Scope and Limitations The software supports exploratory statistical analysis and reproducibility. Its computational verification does not constitute independent authentication of original commercial ratings, confirmation of historical point-in-time information availability, or external empirical validation. Reported statistical associations should not be interpreted as causal relationships or evidence of individual corporate greenwashing. Researcher Stephen KamranMS Banking & Financial EconomicsGovernment College University Lahore, PakistanORCID: https://orcid.org/0009-0000-3063-5611 Associated Research Notebook https://www.kaggle.com/code/stephenkamran/esg-diverge-reproducible-esg-rating-disagreement Version: 0.3.0Year: 2026

🔗 Provenance — このレコードを発見したソース

🔔 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。