Working Paper: CEPR ID: DP17391
Authors: Hanno Kase; Leonardo Melosi; Matthias Rottner
Abstract: We leverage recent advancements in machine learning to develop an integrated method to solve globally and estimate models featuring agent heterogeneity, nonlinear constraints, and aggregate uncertainty. Using simulated data, we show that the proposed method accurately estimates the parameters of a nonlinear Heterogeneous Agent New Keynesian (HANK) model with a zero lower bound (ZLB) constraint. We further apply our method to estimate this HANK model using U.S. data. In the estimated model, the interaction between the ZLB constraint and idiosyncratic income risks emerges as a key source of aggregate output volatility.
Keywords: Machine Learning; Neural Networks; Bayesian Estimation; Global Solution; Heterogeneous Agents; Nonlinearities; Aggregate Uncertainty; HANK Model; Zero Lower Bound
JEL Codes: C11; C45; D31; E32; E52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| neural networks (C45) | improved estimation efficiency (C51) |
| parameters as pseudostate variables (C29) | comprehensive mapping of model’s equilibrium law of motion (C62) |
| neural network's scalability (C45) | ability to estimate complex models (C51) |
| higher idiosyncratic risk (D81) | increased frequency of encountering the ZLB (E31) |
| estimation method (C51) | recovery of the true data-generating process (C51) |
| speed and efficiency of neural network method (C45) | comparison to traditional methods (C52) |