A multi-factor, credit migration model for sovereign and corporate debts

This paper develops a multi-factor, Markov chain model for rating migrations and credit spreads that is applicable to both sovereign and corporate debts. The modelʹs central feature is to allow transition matrices to be time-varying and driven by rating specific latent variables which encompass economic factors like the business cycle. There are three main contributions. First, the model incorporates well-documented empirical properties of transition matrices such as their dependence on business/credit cycles, and it also allows for inter-rating variations in credit quality changes. Second, instead of focusing solely on empirical modeling of rating transitions, the paper also shows how the empirical model can be implemented for actual valuations. Third, the estimation and calibration procedures are easy to follow and implement.