This dissertation consists of three parts. The first chapter presents an analysis of the structural difference between a make-whole callable and a traditional callable bond. Based on the analysis, we construct a reduced-form model for the make-whole callable bond. The second chapter empirically investigates validation of our model with the extended Kalman filter. In this chapter, we show not only that our model is valid for the sequence of the make-whole callable bond behavior, but also that our model outperforms the model from Jarrow et al. (2010).The third chapter examines the association between issuer's debt structure and yield spreads. Specifically, we investigate whether or not an investor requires compensation for liquidity risk. Diamond (1991) introduces liquidity risk as the risk of a borrower being forced into inefficient liquidation when refinancing is not available. According to Diamond's argument, the firm holding the larger proportion of short-term debt to its debt structure is more vulnerable to the unforeseen negative event. Consequently, it will increase firm's risk. Through our tests in this chapter, we find that for investment grade bonds, the results consistently show that the fraction of debt maturing in one or two year is positively related to the yield spreads.