Shock Propagation and Macroeconomic Fluctuations
New research reveals why granular firm shocks explain less than 10% of real-world economic volatility.
Economists Antoine Mandel and Vipin P. Veetil have published a significant theoretical paper, 'Shock Propagation and Macroeconomic Fluctuations,' challenging core assumptions in macroeconomic modeling and AI-driven economic forecasting. The research investigates how idiosyncratic shocks at individual firms propagate through a production network to create aggregate economic volatility. Their key insight is the concept of 'overlapping adjustment': in the real economy, new productivity shocks arrive before the system has fully equilibrated from previous ones. This creates continuous 'productivity waves' that travel and interfere with each other across the supply chain, meaning the static equilibrium models used in many AI agents (like those built on GPT-4 or Claude 3) fundamentally misrepresent economic dynamics.
The paper's technical core shows that macroeconomic fluctuations emerge from the interference pattern of these waves, which is governed by the 'dominant transient eigenvalue' of the production network—a measure of how quickly shocks dissipate. Crucially, this dynamic model reveals that the 'tail of the degree distribution' (the influence of highly connected 'hub' firms) is a weaker determinant of volatility than in static models. The implication is stark: once time-averaging of shocks is accounted for, granular firm-level shocks may account for 'only a small fraction' of empirically observed aggregate volatility. For AI developers, this means agents designed for economic prediction or supply chain simulation must move beyond static network analysis and incorporate these wave-interference dynamics to be accurate, impacting everything from autonomous trading algorithms to large-scale policy simulations.
- Introduces 'overlapping adjustment' model where productivity shocks create interfering 'waves' in networks, governed by a 'dominant transient eigenvalue'.
- Finds granular firm shocks may explain only a small fraction of real aggregate volatility, challenging established economic theory.
- Shows network 'hub' influence is weaker in dynamic regimes, forcing AI economic models to move beyond static analysis.
Why It Matters
Forces AI economic agents and forecasters to model dynamic network interference, not just static connections, for accurate predictions.