This study analyzes the determinants of rating changes and the variables' marginal effects on rating change probabilities. Based on the results, it presents transition matrices by computing transition probabilities. Furthermore, this study analyzes subsamples of the data set, conditional on the business cycle and the economic strength of a country, by using interaction effects. The Author of this study thereby verifies whether or how the transition matrices change by including interaction effects. He applies a latent variable approach, using an ordered probit model, to calculate the effects of different variables on the probabilities of rating changes.