استعمال بعض طرائق بيز الحصين لتقدير انموذج GARCH (1,1)‎

Abstract EN

In this paper, invulnerable Bayesian capabilities were found to estimate a model of self-regression models conditioned by heterogeneity of variance, which is the generalized GARCH (1.1) model when the distribution of errors follows the normal distribution through three invulnerable biometric methods for estimating which are the (Y ̂BM Bayes) method and the (BM) method.

Bayes Shrinkage and BM Bayes Shrinkage, which combines the Bayes method with three restrictive fortified methods, namely the BM.

Huber method, the two proposed methods (BM.

Hample) and (BM.

Tukey), and then apply what was mentioned in the theoretical side on a time series consisting of (1254) views of the daily selling prices of Basra oil for the period (2/1/2008 - 31/12/2012) through the application of the BM Bayes Shrinkage method, which was the best when applied to the estimated parameters values By means of (MLE) where it achieved the lowest value of the comparison standard (MSE).