本研究採用混合面板數據迴歸，系統性檢驗臺灣上市企業之自願性氣候風險揭露、董事會性別多樣性與信用風險之間的關聯性。實證結果顯示，在全樣本分析中，法說會氣候揭露不確定性與企業違約距離呈現顯著的負向關係，顯示當企業管理階層對氣候政策或轉型路徑的表述愈趨模糊時，外部債權人與資本市場將感知到更高的資訊不對稱與營運不確定性，進而侵蝕企業的信用安全邊際並縮短違約距離。而作為內部治理指標的女性董事比例，則與企業違約距離具備實質的正向關係。產業異質性分析進一步證實，在半導體產業中展現出更高的敏感度，其女性董事比例與信用風險之正向關聯遠高於整體市場，且統計顯著性顯著提升至1%水準；與整體市場不同的是，管理階層主動揭露氣候負面衝擊能為企業帶來「誠信溢價」，與違約距離呈顯著正向關係。方法論對比則證實，企業違約距離相較於預期違約頻率更能精準且靈敏地捕捉新興非財務資訊與信用風險之對應關係。本研究成功整合環境資訊與內部治理特徵，不僅豐富了氣候財務學之實證文獻，亦為企業信用風險提供了具現實價值的評估面向。;This study employs a pooled panel regression model to systematically examine the relationships among voluntary climate risk disclosure， board gender diversity， and credit risk measured by distance to default and expected default frequency of listed companies in Taiwan. The empirical results indicate that in the full-sample analysis， climate disclosure uncertainty in earnings calls exhibits a significantly negative relationship with a firm′s distance to default， demonstrating that when corporate management’s articulations regarding climate policies or transition pathways become increasingly ambiguous， external creditors and capital markets perceive higher information asymmetry and operational uncertainty， which subsequently erodes the firm′s credit safety margin and shortens its distance to default. Conversely， the proportion of female directors， serving as an internal governance indicator， possesses a substantially positive relationship with distance to default. Industry heterogeneity analysis further confirms that the semiconductor industry exhibits higher sensitivity within this framework; the positive association between the proportion of female directors and credit risk is far stronger than that of the overall market， with the statistical significance rising prominently to the 1% level. Distinct from the overall market， active disclosure of negative climate impacts by management in the semiconductor industry yields a premium for firms， demonstrating a significantly positive relationship with distance to default. Methodological comparisons reveal that distance to default， compared to expected default frequency captures the correspondence between emerging non-financial information and credit risk more precisely and sensitively. By successfully integrating environmental information and internal governance characteristics， this study not only enriches the empirical literature in climate finance but also provides a practically valuable perspective for evaluating corporate credit risk.
