ESG聲量對於ESG基金中個股評等的影響

Abstract

本研究旨在探討聲量變數對於企業之ESG（環境、社會與公司治理）評等之影響，聚焦於國內市場中具代表性的ESG基金，其所涵蓋之上市上櫃公司標的總檔數共255檔，並以2017年初至2024年底作為樣本期間，主要聲量變數包括新聞報導數（News）、Google搜尋熱度（SVI）、機構投資人持股（MGR_Hold）與重大ESG法規（Regulation），其中，新聞報導資料更是透過網路爬蟲技術蒐集自各大新聞平台，並輔以人工方式進行篩選與檢查，以確保資料之準確性與相關性，並納入企業規模、財務槓桿、獲利能力與公司年齡等控制變數。研究運用固定與隨機效果模型，分別檢驗各基金樣本與整體合併樣本的迴歸結果，並檢視聲量因素於不同情境下之解釋力與穩定性。實證結果顯示，企業在市場中所獲得的聲量，特別是媒體報導頻率，與其ESG 評等之間存在穩定且正向的關聯，顯示資訊曝光程度在評等機制中扮演關鍵角色。聲量與ESG評等間的關係亦可能隱含企業外部形象與內部永續作為的一致性差異，部分企業的高聲量未必完全對應其實質永續表現，凸顯聲量變數在檢視ESG評等可靠性及企業永續承諾落實度上的潛在意涵。相比之下，其他聲量來源如搜尋熱度與機構持股則顯現出樣本依賴性與效果上的不對稱性，僅在特定基金或模型中展現統計顯著性，顯示其影響力可能受限於產業特性、投資人偏好。This study aims to examine the impact of attention-related variables on corporate ESG (Environmental, Social, and Governance) ratings, focusing on representative ESG funds in the domestic market. The dataset comprises a total of 255 constituent companies covered by these funds, with the sample period spanning from early 2017 to the end of 2024. The primary attention variables include the number of news reports (News), Google search intensity (SVI), institutional ownership (MGR_Hold), and major ESG-related regulations (Regulation). Notably, news report data were collected through web scraping techniques from major news platforms and subsequently filtered and verified manually to ensure the accuracy and relevance of the information. Control variables include firm size, financial leverage, profitability, and firm age. The study employs both fixed effects and random effects models to examine regression results for individual fund samples as well as the combined full sample, assessing the explanatory power and stability of attention variables under different scenarios.The empirical results show that firms’ market-level attention, particularly the frequency of media coverage, exhibits a stable and positive relationship with ESG ratings, suggesting that information exposure plays a critical role in the rating mechanism. The relationship between attention and ESG ratings may also reflect potential discrepancies between firms’ external image and their actual sustainability practices, where a high level of attention does not necessarily correspond to substantive ESG performance. This underscores the potential of attention variables in evaluating the reliability of ESG ratings and the extent to which firms fulfill their sustainability commitments. By contrast, other attention sources such as search intensity and institutional ownership demonstrate sample dependency and asymmetric effects, showing statistical significance only in certain funds or models. These findings suggest that the influence of these variables may be shaped by industry characteristics and investor preferences.