Resumo

ESPINOSA ACUNA, Oscar Andrés  e  VACA GONZALEZ, Paola Andrea. FITTING THE CLASSICAL AND BAYESIAN GARCH MODELS WITH STUDENT-T INNOVATIONS TO THE COLCAP INDEX. rev. econ. Caribe [online]. 2017, n.19, pp.1-32. ISSN 2011-2106.

In this article, it is fitted two Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models to the Colcap financial index. One of them is analyzed using a Classical (or Frecuentist) procedure, whose parameters were estimated by maximum likelihood, and the other one is estimated via a Bayesian approach using the Metropolis-Hastings algorithm. Both models were estimated with Student-t innovation. By means of different information criteria, the estimated models by the bayesian and classic approaches are evaluated.

JEL CODES: K49

Palavras-chave : Conditional Heteroskedasticity Models; Metropolis-Hastings Algorithm; COLCAP Index.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )