銀bank · team project

ENCS

500 of the most connected US and EU banks, wired by their interbank exposures. Shock one, and the engine plays out the next ten rounds: creditors run on banks that look weak, fire sales push asset prices down, and Eisenberg-Noe clearing settles who can pay whom.

Shock this bank (or click one) Shock size share of its outside assets wiped out Fire-sale impact α how hard forced selling moves prices Run threshold creditors pull funding below this solvency signal

Circuit breaker halt trading once prices fall this far

A finding worth trying: in this model, which bank you shock matters far less than α. Most banks run about 10× leverage, so a few percent off asset prices tips hundreds of them; between α ≈ 0.012 and 0.03 the outcome flips from a handful of defaults to nearly all. The engine is the team's Rust core, unchanged, compiled to WebAssembly; the network is the top 500 of the dataset's 4,556 banks, with their exposures to the rest counted as ordinary assets.

round: 0 / 10
defaulted: 0
distressed: 0
asset price: 100%
equity lost: $0

defaulted
distressed
price

US bank
EU bank
distressed
defaulted
size: total assets

loading 500 banks and the engine…