Distributional HANK models reshape central bank DSGE toolkits
Christophe Cahn, Patrick Fève and Julien Matheron trace the evolution of DSGE models in central banking in a Banque de France paper. The authors conclude that incorporating heterogeneous agents and sequence-space methods modernizes structural policy analysis across central banks.
Microfoundations meet household heterogeneity
The paper examines how dynamic stochastic general equilibrium models evolved into operational central-bank tools following the Smets-Wouters benchmark.
While medium-scale New Keynesian frameworks successfully combined nominal rigidities, real frictions, and Bayesian estimation, the 2008 financial crisis exposed critical gaps.
Central banks subsequently integrated labor-market slack, open-economy linkages, and financial balance-sheet frictions to model unconventional policy interventions.
The frontier has now shifted toward heterogeneous-agent New Keynesian architectures, which embed uninsurable income risk and borrowing constraints to evaluate how policy affects diverse households.
Sequence spaces break numerical limits
Computational advances have transformed the operational deployment of heterogeneous-agent models.
Methodologies such as sequence-space Jacobians and projection-plus-perturbation algorithms allow researchers to solve high-dimensional distributions in seconds.
This speed enables Bayesian estimation on standard macroeconomic time series alongside microeconomic survey data, making HANK models feasible for quarterly forecasting rounds and historical decompositions.
Guardrails against narrative drift
The paper rightly frames DSGE models as essential guardrails for counterfactual policy rather than literal truth.
Yet adding complex distributional layers risks compounding identification traps without guaranteeing better forecasts.
Central banks need diversified model portfolios rather than a single all-encompassing engine.