Autori
Hirk, Rainer
Hornik, Kurt
Vana, Laura

Titolo
Multivariate ordinal regression models: an analysis ofcorporate credit ratings
Periodico
Statistical methods & applications : Journal of the Italian Statistical Society
Anno: 2019 - Volume: 28 - Fascicolo: 3 - Pagina iniziale: 507 - Pagina finale: 539

Correlated ordinal data typically arises from multiple measurements on a collectionof subjects. Motivated by an application in credit risk, where multiple credit ratingagencies assess the creditworthiness of a firm on an ordinal scale, we consider mul-tivariate ordinal regression models with a latent variable specification and correlatederror terms. Two different link functions are employed, by assuming a multivariatenormal and a multivariate logistic distribution for the latent variables underlying theordinal outcomes. Composite likelihood methods, more specifically the pairwise andtripletwiselikelihoodapproach,areappliedforestimatingthemodelparameters.Usingsimulated data sets with varying number of subjects, we investigate the performance ofthe pairwise likelihood estimates and find them to be robust for both link functions andreasonable sample size. The empirical application consists of an analysis of corporatecredit ratings from the big three credit rating agencies (Standard & Poor’s, Moody’sand Fitch). Firm-level and stock price data for publicly traded US firms as well as anunbalanced panel of issuer credit ratings are collected and analyzed to illustrate theproposed framework.



SICI: 1618-2510(2019)28:3<507:MORMAA>2.0.ZU;2-H

Esportazione dati in Refworks (solo per utenti abilitati)

Biblioteche ACNP che possiedono il periodico