âĄDynamic Fees
Last updated
Last updated
This is an ongoing case study on TWAMMs where fees can be adjusted based on future orders that are queued to be executed. Below is a problem and related parameters we're researching.
What is the optimal swap fee discount for dedicated arbitrageurs (toxic, atomic swaps, short term) to maximize price updates for traders (non-toxic, non-atomic swaps, long term) and MEV kickback to LPs?
LPs: maximize swap fees & MEV remittance to compensate for IL and toxic flow from arbitrageurs
Traders: maximize price updates of assets to keep parity with external venues and get a smoother order fill
Arbitrageurs: maximize MEV opportunity and minimize swap fee for a profitable arbitrage opportunity
Pool parameters:
ETH/USDC
Partner Fee: Dedicated Arbitrageur (0.025%)
Short Term Fee: Mempool Users (0.05%)
Long Term Fee: Traders (0.15%)
Gas (transaction) fees can vary on number of virtual orders that need to be executed
For the purpose of this problem, assume a fixed amount
Trades last multiple blocks and can be cancelled at any time
Mempool arbitrageurs will always exploit opportunity if dedicated arbitrage donât act in time
Dedicated arbitrageurs get a fee discount to get a first right arbitrage opportunity
TWAMMs have two swap interfaces: non-atomic swap (traders), atomic swap (arbitrageurs)
TWAMM pools are 2 asset, 50/50 Uni V2 style pools that follow X*Y=K invariant
Dedicated arbitrageur fees can be changed in-flight and are customizable per pool
Long term trades create back running opportunities as prices move on the bonding curve
Traders are primarily uninformed flow (non-toxic), benefit from arbitrageurs keeping prices fresh
Dedicated arbitrageurs remit share of MEV extracted to LPs for the swap fee discount
Arbitrageurs are informed flow (toxic), profit from correcting prices and pay gas to write traders virtual orders on-chain
LPs get swap fees from long term traders, and MEV + discounted short term swap fees from arbitrageurs
MEV remitted to pools are automatically invested and distributed pro-rata to LPs similar to swap fees, but on a weekly basis
Private relays: ineffective because first right advantage is negated after the first block
Rook auctions: group of arbitrageurs bid on the arbitrage opportunity, 90% of bid is remitted back to LPs. Note bid != potential MEV opportunity
SLAs: require off-chain agreements and trust with 3rd parties to behave ethically
