Skip to content

Simulations paper

Empirical validation of the protocol's properties across capacity dynamics, cascade dynamics, TWAP manipulation, and bad-debt analysis.

Topics

Capacity dynamics

Simulation of the cap function over many operations confirms:

  • O(√n) Sybil scaling matches the theory.
  • A patient attacker with capital retains ~48.6% share after 60 weeks of repeated operations.
  • Honest small users are not artificially blocked — the cap floor and √(n+2) divisor combine to give meaningful first-deposit access.

Cascade dynamics

Simulation of 1,000 borrowers under price shocks across different lock fractions:

Shock0% locked25%50%75%100%
10%8.6%8.2%7.8%7.8%7.4%
25%85.7%55.8%37.1%33.3%29.4%
30%99.1%84.9%61.2%47.9%38.6%

Up to 80% reduction in liquidation count at full lock adoption under stress shocks.

TWAP manipulation

60 Foundry-based scenarios test single-block, sandwich, and sustained manipulation. Two-tick immunity is verified: no single-block manipulation can affect the reported price.

Bad-debt Monte Carlo

10,000 paths under Merton jump-diffusion confirm:

  • At default LTV (66.67%) with 50% buffer: bad debt at 99% CVaR = 0.02% of pool.
  • At conservative floor (33% LTV, 200% buffer): zero bad debt for crashes up to 50%.

Where to find the paper

Where to go next