Default Prediction Model for SME’s : Evidence from Indonesian Market Using Financial Ratios

Banks face adverse selection problem in granting loans due to asymmetric information about their debtor's financial condition which in turn causes bank to be exposed to credit risk. The Basel II accord requires banks to develop quantitative credit risk model as a base for determining their capital adequacy level; besides, regarding the fact that the bank exposed to higher credit risk for SMEs loan, the Advanced Internal Rating Based (A-IRB) approach in the Basel II accord suggests to separate between corporates lending exposure and SMEs as a result of SMEs informational opacity. In this paper, we develop a default prediction model for SMEs using a logistic regression model on a sample of 1,845 SMEs over the period of 2005-2007. This model serves both as an instrument in determining capital cushion and a screening tool for credit granting to SMEs. We propose a so called three-step forward procedure for parameters estimation to overcome the multicollinearity problem in the Altman & Sabato's model (2007). Based on the model, we find that some financial ratios i.e., current ratio, interest coverage ratio, and average collection period are significant as default predictors of SMEs. In addition, we also test a default prediction using probit model that is consistent with the result of the logit model. Finally, our logit-default prediction model has hit ratios of 87.81% for in-sample data and 89.54% for out-of-sample data, while the probit model has hit ratios of 87.98% for in-sample data and 89.78% for out-of-sample data. These results show that our models have higher accuracy than that of Altman & Sabato (2007)