In this contribution we propose to estimate the probability of financial default of companies and the correlated rating classes, using efficiently the informa- tion contained in different databases. In this respect, we propose a novel approach, based on the recursive usage of Bayes theorem, that can be very helpful in inte- grating default estimates obtained from different sets of covariates. Our approach is ordinal: on one hand, the default response variable is categorized in nine rating classes that are ordered according to the average default probability. On the other hand, covariates induce partitioning of companies into the rating classes, according to their measurement levels, when categorical, or to their quantiles, when they are continuous. The application of our proposal to credit risk databases shows that it performs quite efficiently, allowing to obtain, in a natural way, additional measures such as the default value at risk and the lift index.

Ordinal models for financial evaluation

CERCHIELLO, PAOLA;GIUDICI, PAOLO STEFANO
2012-01-01

Abstract

In this contribution we propose to estimate the probability of financial default of companies and the correlated rating classes, using efficiently the informa- tion contained in different databases. In this respect, we propose a novel approach, based on the recursive usage of Bayes theorem, that can be very helpful in inte- grating default estimates obtained from different sets of covariates. Our approach is ordinal: on one hand, the default response variable is categorized in nine rating classes that are ordered according to the average default probability. On the other hand, covariates induce partitioning of companies into the rating classes, according to their measurement levels, when categorical, or to their quantiles, when they are continuous. The application of our proposal to credit risk databases shows that it performs quite efficiently, allowing to obtain, in a natural way, additional measures such as the default value at risk and the lift index.
2012
9788861298828
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11571/570253
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact