This repository hosts supplementary code for the paper "Finding the Right Curve: Optimal Design of Constant Function Market Makers" by Mohak Goyal, Geoffrey Ramseyer, Ashish Goel, and David Mazières.
-
The script
cfmm_lp2.pycomputes for any given belief function (on the future valuations of assets) optimal CFMM designs and liquidity allocations.- Invoke as
python3 cfmm_lp2.py - Custom belief functions can be added in
psi_func_custom(px, py). - The framework is general; alternative objectives (as in e.g. Section 7 of the paper, when accounting for liquidity provider fee revenue and divergence loss) follow from modifications to the objective functions.
- Invoke as
-
The script
lp_functions.pyprovides auxiliary functions and implements various belief functions.
The optimization toolkit relies on CVXPY, which can be installed via
python3 -m pip install cvxpy