影響台灣地區商業銀行逾放比之因素

Abstract

本研究以1999年至2004年選取台灣43家本國商業銀行為研究對象，利用縱橫面資料模型（Panel Data Model）之隨機及固定效果模型，探討影響逾放比之原因後，再與以逾放認定較嚴格之廣義逾放比相較其結果是否一致；另利用統計方法檢驗金融控股公司與非金融控股公司之子銀行、新銀行與舊銀行及公營與非公營銀行間分別在逾放比與廣義逾放比方面是否存在顯著差異。其實證結果發現如下： 1.影響逾放比因素有經濟成長率、銀行對消費者放款比例、實質資產規模排序值及資本適足率等級別、銀行對中小企業放款比例及銀行設立年限。 2.在應變數為廣義逾放比時，經濟成長率、銀行對中小企業放款比例、銀行對消費者放款比例、實質資產規模排序值及資本適足率等級別此五項因素與以逾放比為應變數所得之結果相同。 3.在逾放比方面，金融控股公司與非金融控股公司之子銀行、新舊銀行間存在顯著差異，且金融控股公司之子銀行平均逾放比低於非金融控股公司之子銀行，新銀行平均逾放比低於舊銀行；在廣義逾放比方面，金融控股公司與非金融控股公司之子銀行存在顯著差異，且金融控股公司之子銀行平均廣義逾放比低於非金融控股公司之子銀行。

This paper studies 43 banks from Taiwan during the 1999-2004 period. The panel data regression and Wilcoxon rank sum test are used to analyze the factors influencing non-performing loan (NPL) ratios of Taiwan’s commercial banks. We compare the NPL ratios between banks controlled and not controlled by financial-holding companies, held and not held by the government, and established before and after 1990. Our major findings are as follows: (1) The effects of economic growth rate, loan at consumer ratio, real asset size ranking, and capital adequacy ratio on NPL ratio are significantly negative. (2) The effects of loan ratio to small- and medium-sized enterprises and bank age on NPL ratios are significantly positive. (3) Other things being equal, control by financial holding company significantly helps reduce NPL ratios. (4) Relaxing the definition of NPL to general NPL will not change our major findings.