Bayesian Estimation of
Nonlinear Dynamic General Equilibrium (DGE) Models

The aim of the project is to develop a toolkit and a graphical user interface for the specification, solution, estimation and simulation of nonlinear DGE models.

Click for a larger view

Mailing List

There is a moderated mailing list for news and discussion about JBendge. Please subscribe if interested:


Current Status

JBendge is constantly under development and collaborators are highly welcomed.

  • project version 2.0 beta, July 19, 2008, (requires Java 6) download

  • solves for linear and nonlinear optimal policies from nonlinear dynamic first order conditions with rational expectations

  • estimates linearized or nonlinear models

  • estimation and solution is parallelized to run on multicore CPUs

  • Smolyak sparse grid methods are used to avoid the curse of dimensionality for function approximation, numerical integration of expectations and filters

  • various nonlinear filters implemented

  • interactive parallel Metropolis-Hastings algorithm for Bayesian posterior estimation

  • platform independent