## Hyperdrive Economic Nerdsnipes

1. move LPandArb to rust
2. make shit go faster
3. forever-sim (add and remove agents, evaluate discretely)
4. multi-pool (what can we do currently? what can't we?)
5. dead-zone analysis
- new results (stylised or high-fidelity)
- compare to previous LVR papers
6. LPandArb logic to choose between LP and Arb
- do pools reach equilibrium?
7. develop expert system strategies
- trend-following, relative value, mean-reversion
8. recreate state without re-running txn's
- attempt is on test_get_max branch in agent0