Work

Modelos autorregresivos para la varianza condicionada heteroscedastica (ARCH)

Year: 1994
Type: preprint
Abstract: En el presente trabajo se introduce al lector en los modelos autorregresivos para la varianza condicionada heterocedastica, incidiendo en los problemas que plantean los esquemas mas sencillo y sugiriendo diversas soluciones. Se describe el concepto, las hipotesis y los modelos que explican la varianza condicionada en el tiempo desde diversas perspectivas del analisis estadistico: relacion lineal o no lineal entre las variables y metodos de estimacion de los parametros. Finalmente, se discuten diversos contrastes que permiten escoger entre diversas especificaciones alternativas. In this paper the Autoregressive Conditional Heteroskedasticity (ARCH) models are introduced to the reader. The problems implied by their different parameterizations are pointed out and the corresponding solutions are suggested. Several statistical perspectives are used to explain the concept, the assumptions and the models that explain the temporal behaviour of the conditional variances. The paper exposed lineal and non-lineal relationships and several methods of estimation of the parameters of the models. Finally, some tests that allow to choose between alternative specifications are described.(This abstract was borrowed from another version of this item.)
Language: Spanish
Subfield: Finance
Open Access status: green
OpenAlex ID: W3121128629
Cites:
Cited by: 2
Related to: 10