DIME MCMC: A Swiss Army Knife for Bayesian Inference

53 Pages Posted: 20 Oct 2022 Last revised: 19 Aug 2023

Date Written: October 17, 2022

Abstract

This paper proposes a Differential-Independence Mixture Ensemble (DIME) sampler for the Bayesian estimation of economic models. DIME allows the efficient estimation of models that may be computationally expensive to evaluate with challenging, multimodal, high-dimensional posterior distributions and ex-ante unknown properties. It combines the advantages of gradient-free global multi-start optimizers with the properties of Monte Carlo Markov chains to quickly explore the typical set. DIME is used to estimate a two-asset heterogeneous agent New Keynesian (``HANK'') model, for the first time including the households' preference parameters. The results suggest that household heterogeneity plays a less prominent role in explaining the empirical macroeconomic dynamics.

Keywords: Bayesian Estimation, Monte Carlo Methods, Heterogeneous Agents, Global Optimization JEL: C11, C13, C15, E10

JEL Classification: C11, C13, C15, E10

Suggested Citation

Boehl, Gregor, DIME MCMC: A Swiss Army Knife for Bayesian Inference (October 17, 2022). Available at SSRN: https://ssrn.com/abstract=4250395 or http://dx.doi.org/10.2139/ssrn.4250395

Gregor Boehl (Contact Author)

University of Bonn ( email )

Adenauerallee 24-42
Bonn, D-53113
Germany

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
48
Abstract Views
372
PlumX Metrics