Empiryczne modele równowagi ogólnej: zagadnienia numeryczne estymacji bayesowskiej
Empirical General Equilibrium Models: Numerical Issues of the Bayesian Estimation

Summary/Abstract: The paper presents numerical issues related to the estimation of dynamic general equilibrium models with Bayesian techniques. As an illustrative example, a simple New-Keynesian model with nominal rigidities in price and wage settings is taken from the literature. Estimating structural parameters requires the application of numerical methods – optimization algorithms for maximum likelihood and Monte Carlo techniques for the Bayesian estimation of the Markov Chain. A broad literature that treats the estimation of dynamic macroeconomic models rarely presents any aspects of the numerical side of applications and its potential impact on results and inference. The paper concentrates on the issue of convergence criteria of the Markov Chain Monte Carlo techniques and the methods of monitoring parameter stability.

• Details
• Contents