« Back to Results

Developments in Macroeconomics and Finance

Paper Session

Sunday, Jan. 8, 2023 10:15 AM - 12:15 PM (CST)

Hilton Riverside
Hosted By: Society for Nonlinear Dynamics and Econometrics
  • Chair: Tatevik Sekhposyan, Texas A&M University

Estimating HANK Models with Neural Networks

Hanno Kase
University of Minnesota
Leonardo Melosi
Federal Reserve Bank of Chicago
Matthias Rottner
Deutsche Bundesbank


We leverage recent developments in machine learning to develop methods to estimate large and complex nonlinear macroeconomic models, e.g. HANK models. Our method relies on neural networks because of their appealing feature that even models with hundreds of state variables can be solved. While likelihood estimation requires the repeated solving of the model, something that is infeasible for highly complex models, we overcome this problem by exploiting the scalability of neural networks. Including the parameters of the model as pseudo-state variables in the neural network, we solve this extended neural network and apply it directly in the estimation. As a proof of concept, we demonstrate with a tractable RANK model augmented with a zero lower bound that our approach coincides with an estimation based on conventional methods for nonlinear models. To show the full potential of our approach, we then estimate a quantitative HANK model that features nonlinearities on an individual (borrowing limit) and aggregate level (zero lower bound) using simulated data.

The Unattractiveness of Indeterminate Dynamic Equilibria

Julian Ashwin
London Business School
Paul Beaudry
University of British Columbia and Bank of Canada
Martin Ellison
University of Oxford


Macroeconomic forces that generate multiple equilibria often support locally-indeterminate dynamic equilibria in which a continuum of perfect foresight paths converge towards the same steady-state. The set of rational expectations equilibria (REE) in such environments can be very large, although the relevance of many of them has been questioned on the basis that they may not be learnable. In this paper, we document the existence of a learnable REE in such situations. However, we show that the dynamics of this learnable REE do not resemble perturbations around any of the convergent perfect foresight paths. Instead, the learnable REE treats the locally-indeterminate steady-state as unstable, in contrast to it resembling a stable attractor under perfect foresight.

Talking Over Time: Dynamic Central Bank Communication

Laura Gati
European Central Bank


This paper studies the optimal dynamic communication strategy of central banks using a Bayesian persuasion game framework. In a dynamic environment, financial market participants and the general public have misaligned interests because the present and future have different relevance in their optimization problems, leading to a novel tradeoff for the monetary authority. Compared to the static benchmark, I show that the central bank’s optimal dynamic communication policy should put a higher weight on talking about the present state than the future. In addition, the central bank should strategically send more noisy signals than in the static benchmark.

A Bewley Model of Asset Pricing

Dong-Hyun Ahn
Seoul National University
Hwagyun Kim
Texas A&M University
Eunseong Ma
Louisiana State University


Capital adjustment costs and stochastic capital depreciation exist as additional frictions. Our model quantitatively accounts for economic data, including consumption and income distributions, and excels at producing the key properties of asset prices. Limited risk-sharing with costly capital adjustment lies at the center of explaining the cross-section of stock returns. Our model supports empirical asset pricing models employing firm-specific investment and profitability as factors. Relaxing the assumption of an incomplete market weakens or reverses quantitative results.

Ed Herbst
Federal Reserve Board
Thomas Lubik
Federal Reserve Bank of Richmond
Tatjana Dahlhaus
Bank of Canada
Jack Favilukis
University of British Columbia
JEL Classifications
  • E1 - General Aggregative Models
  • G1 - Asset Markets and Pricing