信用評等：期間與產業差異分析

Study Period and the Industry Effect of Credit Rating

本文旨在分析產業差異與期間更替對企業長期債務信用評等之影響。我們以美國 S&P 500企業為對象，針對不同產業別，採用三個以5年為一期的次期間，分別萃取適當的財務變數並進行羅吉特（logit）迴歸分析，然後根據這些財務變數進行個別產業與期間模式的區別能力分析。研究結果發現，相同產業在不同的時期，以及不同產業在相同的基間，其決定信用風險的財務變數存有相當程度之差異。相較於過去研究採用全體企業來做為評等模式建構之基礎，以個別產業建立評等模式之區別能力明顯提升。此外，本研究亦發現影響信用評等之重要財務變數，會隨著市場景氣循環以及企業結構的改變而有所不同，凸顯了信用平等的產業及期間差異化特性。

The paper examines the study period and industry effect of credit rating. We analyze different industry groups of the S&P 500 companies across a sample period of 15 years key financial variables are selected using logit regression and multiple discriminant analysis. We find that for either different sub-periods in the same industry or different industries during the same study period, determinants of credit rating can vary significantly. Result also show that industry-specific rating models perform better than that based on all firms in the market. Also of interest is the finding the key rating determinants do change with business cycles and with stages of industrial development, providing evidence that credit ratings are industry specific and are study period in natures.