Ethereum: Flashbot’s Impact on the Ethereum Ecosystem
Introduction
User-focused updates to Flashbot’s MEV-Boost protocol via transaction-level MEV bundles
Background: Flashbot’s MEV-Boost is an out-of-protocol implementation of Proposer-Builder Separation (PBS) in Ethereum 2.0. The key change here is that validators (proposers) source blocks from relays, which are populated with complete blocks by builders. This increases Maximum Extractable Value (MEV), the amount of fringe value that can be economically drawn out of each block/transaction via adding/censoring/rearranging transactions,1 by allowing MEV searchers to specialize in constructing the most profitable blocks and proposers to efficiently auction off their block space.
Research Overview: What’s not clear is whether it’s possible for users to (at least in naive cases) capture their own MEV. For example, say a user performs a token swap on an AMM, causing a pool imbalance. Rather than a searcher exploiting this opportunity for MEV, can a user (via personal liquidity or flash loans) exploit this for any profit at all? Given a pool containing USDC <> ETH, what happens when a trade is executed? If the user is buying ETH, the amount of USDC in the pool is increased while the amount of ETH drops, causing the price of ETH to rise relative to USDC in this particular pool. In situations where another user has recently swapped in a pool containing ETH (say, ETH <> DAI) and dropped the ETH price, an arbitrage opportunity exists for someone to purchase ETH in the ETH <> DAI pool and sell it at a profit in the ETH <> USDC pool. Currently, these opportunities are exploited by bots, but perhaps some simple rules could allow a user (either with local liquidity or a flash loan) to exploit them themselves as part of a bundled transaction. Alternatively, for less time-sensitive transactions, an order book in which users match trades to immediately rebalance each pool might help reduce opportunities for searchers to exploit everyday users. Significant research into the economics underpinning these swaps will need to be undertaken to determine their feasibility.
(1) As a very naive example, consider a trade on some AMM (e.g., Uni) for 100 USDC <> Eth. Default slippage is usually around 2%, meaning that a trade with 0% slippage is leaving value that someone was willing to pay for on the table. MEV searchers capitalize on this by scanning for trades with a slippage/price delta and frontrunning those transactions to capture the slippage delta. Similar logic applies to, e.g., the inter-AMM arbitrage opportunities arising from the unbalancing of a given liquidity pool after a large trade, etc.