🏦 2D Bank Run — Depositor Fear Contagion Network

What It Demonstrates

This simulator models a bank run as an epidemic spreading across a network rather than as a single institution's crisis. Every depositor is an individual agent with its own fear level. Banks are nodes in an interbank lending graph; when depositors at one bank start pulling out cash, fear diffuses along the graph's edges to depositors at connected banks — exactly like a contagious disease spreading through a contact network. A bank genuinely fails only when its depositors' real withdrawal requests exceed its real available reserves — there is no scripted collapse, only the aggregate outcome of thousands of independent, probabilistic agent decisions.

How to Use

• Click Spark a Run to panic a random bank's depositors and seed the contagion.
• Watch fear (dot color: green → amber → red) spread outward along the grey network edges to neighboring banks.
• Raise Network Connectivity to see how a denser interbank web accelerates system-wide cascades.
• Raise Reserve Ratio to make each bank more resilient to a given contagion strength.
• Increase Contagion Strength to see fear spread faster between connected banks, even before any bank fails.

Did You Know?

Network epidemiology and financial contagion share the same math: both are studied with SIR-style compartment models where "infection" is fear or insolvency risk instead of a virus, and network topology — how densely nodes are interconnected — determines whether a local shock stays local or cascades system-wide. Central banks use exactly this kind of network-stress-test model to evaluate how interconnected a banking system can safely be before a single failure threatens the whole system.