Agent-Based Modeling of Economic Systems

Alex Makassiouk
ETH Zurich, D-MTEC, Chair of Systems Design
15.05.2024

---

This repository is dedicated to the exercises of the course Agent-Based Modeling of Economic Systems taught Spring 2024.

The course itself was taught by Prof. Dr. Frank Schweitzer ([email protected]), full professor for Systems Design at ETH Zurich since 2004. The exercises were organized and taught by Dr. Luca Verginer ([email protected]) and Dr. Giona Casiraghi ([email protected]), both senior researchers at the chair of Systems Design. I want to thank them all and the class for a very interesting course and interesting, engaging discussions.

Course content

1. Agent-Based Modeling Basics
2. Random growth models
3. Coupled growth and growth rate distributions
4. Modeling entry and growth, data-driven modeling
5. Models of entry and adoption, Polya Process
6. Models of competition, Inequality
7. Models of failure cascades, Systemic risk
8. Selected PhD projects utilizing ABM
9. Strategic network formation: Efficiency vs. stability
10. Strategic network formation: Reciprocity
11. Systems design: Network interventions
12. Systems design: Induce cooperation

Exercise content

From exercise 4 and onward, Mesa was used as the ABM framework. It's an open-source Python library ideal for simulating complex systems.

1. Python basics
2. Random growth models - CLT, Bimodal mixture distribution, firm growth simulation
3. Firm growth dynamics - Kesten model, Simulations with target size
4. Yule-Simon distribution
5. Polya Process and technology adoption - linear and nonlinear, preference- and technology utilities, lock-in effect
6. Firm growth with interactivity - bilateral competition with firm in size proximity, transaction dynamics and resource exchange, parameter exploration
7. Cascading processes on networks - Capacity, loads and failing nodes, inward and outward models
8. No exercise 8
9. Distance-based utility model - undirected network. sequential edge formation as best response for maximum utility
10. Severance costs and leading eigenvalue utility - Introduction of severance costs to network formations.
11. Network convergence and equilibria