Censorship Resistance in On-Chain Auctions

Smart contracts offer a way to credibly commit to a mechanism, as long as it can be expressed as an easily computable mapping from inputs, in the form of transactions on-chain, to outputs: allocations and payments. But proposers decide which transactions to include, allowing them to manipulate these mechanisms and extract temporary monopoly rents known as MEV. Motivated by both general interest in running auctions on-chain, and current proposals to conduct MEV auctions on-chain, we study how these manipulations effect the equilibria of auctions. Formally, we consider an independent private value auction where bidders simultaneously submit private bids, and public tips, that are paid to the proposer upon inclusion. A single additional bidder may bribe the proposer to omit competing bids. We show that even if bids are completely sealed, tips reveal bids in equilibrium, which suggests that encrypting bids may not prevent manipulation. Further, we show that collusion at the transaction inclusion step is extremely profitable for the colluding bidder: as the number of bidders increases, the probability that the winner is not colluding and the economic efficiency of the auction both decrease faster than $1/n$. Running the auction over multiple blocks, each with a different proposer, alleviates the problem only if the number of blocks is larger than the number of bidders. We argue that blockchains with more than one concurrent proposer can credibly execute auctions on chain, as long as tips can be conditioned on the number of proposers that include the transaction.