摘要：In this paper,we propose a discrete time model to measure the default spread for bank loans. The model provides a closed-form solution for the short- and medium-term default spread,which we assume to be dependent on the default probabilities,the losses given default,the risk grades transition probabilities,seen in a Markov chain,the prime rate and the economic cycle phases. The model is tested with real data provided by a bank,and allows one to conclude that the actual spread is,on the one hand,insufficient to cover the whole credit risk for the low-risk clients and,on the other hand, excessive for the high-risk clients. We believe that this study may contribute to improving the pricing for bank loans.
