Multiple-State Models for the Analysis of Time-to-Event Data: Application to Corporate Rating

• Main Authors: Chih-wei Yang, 楊織蔚
• Other Authors: Pao-sheng Shen
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/6h3nh2

碩士 === 東海大學 === 統計學系 === 103 === A multi-state model (MSM) is a model for a continuous time stochastic process allowing individuals to transit among a nite number of states. The MSM of disease have been of use in the past in designing and evaluating cancer screening programs. Recently MSM are also used to examines the determinants of dierent states of nancial distress. In this article, using MSM, we investigate key factors that aect corporate rating transitions. The transition intensities are modeled using Cox Markov models (CMM)/Cox semi-Markov models (CSMM). Based on a data set of 123 rms in Taiwan over the period 1995 to 2010, we investigate key factors that aect the entry of four states of corporate rating: low risk, moderate risk, high risk and default.