Template-Type: ReDIF-Paper 1.0
Title: An Endogenous Gridpoint Method for Distributional Dynamics
Author-Name: Christian Bayer
Author-Email: christian.bayer@uni-bonn.de
Author-Name: Ralph Luetticke
Author-Email: ralph.luetticke@uni-tuebingen.de
Author-Name: Maximilian Weiss
Author-Email: maximilian.weiss@uni-tuebingen.de
Author-Name: Yannik Winkelmann
Author-Email: yannik.winkelmann@uni-tuebingen.de

Classification-JEL: C46, C63, E32
Keywords: Numerical Methods, Distributions, Heterogeneous Agent Models, Linearization

Abstract: The “histogram method” (Young, 2010), while the standard approach for analyzing
distributional dynamics in heterogeneous agent models, is linear in optimal policies.
We introduce a novel method that captures nonlinearities of distributional
dynamics. This method solves the distributional dynamics by interpolation instead
of integration, which is made possible by making the grid endogenous. It retains
the tractability and speed of the histogram method, while increasing numerical efficiency
even in the steady state and producing significant economic differences in
scenarios with aggregate risk. We document this by studying aggregate investment
risk with a third-order solution using perturbation techniques.

Note: 
Length: 21
Creation-Date: 2024-05
Revision-Date: 
File-URL: https://www.crctr224.de/research/discussion-papers/archive/dp548
File-Format: application/pdf
Handle: RePEc:bon:boncrc:CRCTR224_2024_548