Heterogeneity shocks explain under 5 percent of macro dynamics
Researchers at the Central Bank of Russia have developed a new DSGE model incorporating agent, regional, and sectoral heterogeneity. Published in July 2026, the working paper estimates that heterogeneity shocks account for a minimal share of macroeconomic fluctuations.
Mapping regional and sectoral ties
Andrei Shulgin and Anna Novak from the Volga-Vyatka Main Branch of the Bank of Russia introduce the MARS dynamic stochastic general equilibrium model.
Designed for regional macroeconomic analysis supporting Board of Directors meetings, the framework integrates three types of heterogeneity: household access to financial markets, geographical differences across regions, and sectoral variations in export exposure and pricing.
Estimated using Bayesian methods on data from 2014 to 2025, the model bridges national monetary policy aggregates with regional business cycles in Russia, the Volga-Vyatka macroregion, and foreign economies.
The approach accounts for interregional spill-overs through goods and labour markets.
The minor footprint of regional shocks
While multi-agent and regional models are increasingly common in central bank research, combining agent, household, and sectoral dimensions in a single framework is novel.
The authors apply a conservative approach, viewing heterogeneity not as a primary economic problem to solve, but as an informational tool to improve macroeconomic identification and conditional forecasting.
Spill-overs operate via regional goods competition, interregional labour markets, and national policy variables like the key rate.
Sobering limits of regional divergence
The study delivers a sophisticated tool for regional analysis, yet its core empirical finding is sobering.
If heterogeneity shocks drive under five percent of aggregate dynamics, the practical payoff for monetary policy remains limited.
Nevertheless, the framework provides a rigorous foundation for regional forecasting.