Complete Agent-Based Macroeconomic Laboratory with Spatial Dynamics | 5000 HH, 200 Firms, 5 Banks | ODD Protocol
GDP (Real, Quarterly)
Unemployment (%)
Inflation (Annual %)
Policy Rate (Taylor Rule)
Phillips Curve (π vs u - Stylized Fact)
Okun's Law (ΔY vs Δu)
Wealth Distribution (Households)
Firm Size (Log-Log) (Zipf Check)
📖 Model Documentation (ODD Protocol)
Purpose: Study how monetary/fiscal policy and credit frictions shape GDP, inflation, unemployment,
defaults, bank leverage, and inequality through emergent, decentralized agent behavior.
Agents: 5000 heterogeneous households (wealth, propensity to consume, employment status),
200 firms (pricing, wages, inventory, debt, adaptive expectations), 5 banks (loans, deposits, equity, CAR).
Key Mechanisms:
- Taylor Rule: i = i* + φ_π(π - π*) + φ_y·gap
- Inventory Pricing: Firms adjust prices based on inventory gaps and markup over costs
- Credit Rationing: Banks lend if PD×LGD ≤ risk_limit and post-loan CAR ≥ minCAR
- Adaptive Expectations: D̂_t+1 = (1-α)D̂_t + αD_t + ε
- Labor Matching: Random matching with separation shocks
Stylized Facts Targeted:
Phillips CurveNegative π-u slope
Okun's LawNegative ΔY-Δu relation
Firm SizesZipf tail (α ≈ 1)
Price ChangesLeptokurtic (fat tails)
Bank LeveragePro-cyclical
Wealth InequalityGini ≈ 0.35-0.55
Reproducibility: All runs are seeded. Export CSV time series and microdata for external analysis.
Batch experiments can sweep parameter grids to map policy trade-offs.