本篇論文旨在研究臺灣中小型企業的逾期率預測模型，我們的實證模型指出：ROA、資產週轉率、GDP成長率以及通貨膨脹率等上升時，會降低逾期率的機率；相對的，融資率的增加則會提高逾期的機率。上述的證實結果不受到固定效果模型與隨機效果模型，及當期模型與前期模型的影響，即結果具有頑強性。

This paper studies the default predicted model of Taiwan small and medium industry. Our empirical model shows that the rise of ROA, asset turnover ratio, GDP growth rate and inflation rate can reduce the default probability. By contrast, the increase of loan rate increases the default probability. Results are rather robust even when we consider the fixed effect and random effect of panel model.