Benefits of using Bayesian estimation for macromodels of Ukraine: the case of application to bivariate VAR model
Loading...
Date
2014
Authors
Stelmashenko, Yaroslava
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Ukrainian econometricians often face a shortage o f observations necessary for providing precise answers to complex macroeconomic questions. Recent studies ha\’e shown that the Bayesian Estimation approach
can solve this problem as it is partially based on nonsample information. In this paper the theoretical analysis and practical application o f using the Bayesian Estimation is presented. A bivariate VAR(2) model has been build to estimate quarterly GDP growth and CPIfor Ukraine using Gibbs sampling and a Minnesota prior. The empirical results show robust correlation betM’een the estimate and actual quarterly GDP and CPI figures, indicating the ability o f the Bayesian Estimation to provide a high level o f accuracy in macromodels o f Ukraine.
Description
Keywords
the Inverse Wishart Distribution, bivariate VAR, Bayesian Estimation, Gibbs sampling, Minnesota prior, random walk, двомірна векторна авторегресійна модель, байєсівське оцінювання, вибірка Гіббса, Міннесота prior, випадкове блукання, зворотний розподіл Уішарта
Citation
Stelmashenko, la. Benefits of using Bayesian estimation for macromodels of Ukraine: the case of application to bivariate VAR model / la. Stelmashenko // Наукові записки НаУКМА. - 2014. - Т. 159 : Економічні науки. - С. 81-84.