MEV: A Journey Into the Dark Forest with Hasu
Join me as I learn the state of MEV (miner extractable value) through the Bell Curve show hosted by @MikeIppolito_ and @hasufl.
✍️ My notes on Ep0: A Journey Into The Dark Forest.
1. What is MEV?
MEV refers to the value extracted by privileged actors in any system.
Example in Web2: central bank printing money (@Balaji BTC = 1 million bet) 👀
The perfect MEV solution should:
Avoid centralization.
Minimize bad MEV.
Return MEV to the protocol.
The main principles to reduce MEV
Competition.
Privacy.
Redistribution MEV to users.
2. Order Flow Auctions
MEV searchers will compete to execute users' txs through auction & return part of the captured fees to users.
Previously, blockspace auctions happened mostly at the validator level. Now, order flow auctions will shift them closer to the user & wallet layer.
3. Current MEV State in L1 & Modular World
MEV in L1 is extracted by:
Validators
MEV searchers
Other participants.
Ethereum can pay its security budget for validators through inflation and should not rely on MEV.
MEV in the modular world:
MEV will be accrued to the execution layer (including L2/L3) since it controls the sequencers and the ordering of txs.
When L2s switch to decentralized sequencers, they will struggle with inefficient MEV extraction problems like the early days of Ethereum.
4. MEV Searchers Landscape
Arbitrage is king (80% of the MEV)
Sandwich attacks: ↓ by aggregators, flashbots protect, & @Uniswap v3.
Liquidation.
5. The Role Of Latency
Trading firms need microsecond latency → Might become block builders.
If we minimize the latency's impact → the competition will shift toward price, where parties compete on who pays the most to fill a certain MEV opportunity (aka auctions).
6. Fair Ordering Protocols
Txs are ordered in the order they are received → Fairer.
Easier to make them private than to make the auctions private.
7. MEV on Solana and Cosmos
MEV on Solana
Too centralizing → Low latency → Very big advantages for validators & bad for users.
Bad mempool design → 58% failed txs.
MEV on Cosmos: fun & very innovative
Osmosis & Skip Protocol: very fun & innovative approaches to MEV capturing through block building & searching, & protorev module.
Anoma & Penumbra: privacy + intent-based txs frameworks → similar to SUAVE (The future of MEV is SUAVE).