Banks do not just lend to households and firms — they lend heavily to each other, through overnight interbank loans, repo agreements and derivatives exposures. This creates a dense web of mutual claims. When one bank fails, it cannot repay what it owes to its interbank lenders. Those lenders take a loss on that exposure; if the loss is large enough relative to their own capital buffer, they too become insolvent and stop paying their own creditors — and the failure cascades outward across the network. This is financial contagion, one of the central concerns of systemic-risk regulation since the 2008 crisis.
After Lehman Brothers' collapse in 2008 froze interbank lending worldwide, regulators introduced capital surcharges for "systemically important" banks and stress tests that explicitly model contagion through interbank exposure networks — the same mechanism this simulation reproduces in miniature.
A 3D network of banks linked by simulated lending exposures, where triggering one bank's default lets you watch losses propagate — or get contained — across the interbank web.
When a bank defaults, its interbank creditors absorb a loss on their exposure to it. If that loss exceeds a creditor's own capital buffer, it defaults too — cascading through the network in rounds until losses are absorbed or the whole system fails.
Adjust network size, interconnectedness, capital buffers and loss-given-default, then click any bank (or use the trigger button) to fail it and watch the contagion round-by-round.
After Lehman Brothers collapsed in 2008, regulators began explicitly modelling these interbank exposure networks in stress tests to estimate how far a single failure could spread.