Greedy Transaction Fee Mechanisms for (Non-)myopic Miners

Decentralized cryptocurrencies are payment systems that rely on aligning the incentives of users and miners to operate correctly and offer a high quality of service to their users. Recent literature studies the mechanism design problem of the auction serving as the transaction fee mechanism (TFM). We show that while the protocol that requires a user to ``pay as bid'' and greedily chooses among available transactions based on their fees is not dominant strategy incentive-compatible (DSIC) for users, it has a Bayesian-Nash equilibrium (BNE) where bids are slightly shaded. Relaxing this incentive compatibility requirement circumvents the impossibility result of [16] and allows for an approximately revenue and welfare optimal, myopic miners incentive-compatibility (MMIC), and off-chain-agreement (OCA)-proof mechanism. We prove its guarantees using different benchmarks, and in particular, show it is the revenue optimal Bayesian incentive-compatible (BIC), MMIC and 1-OCA-proof mechanism among a large class of mechanisms. We move beyond the myopic model to a model where users offer transaction fees for their transaction to be accepted, as well as report their urgency level by specifying the time to live (TTL) of the transaction, after which it expires. We show guarantees provided by the greedy allocation rule, as well as a better-performing non-myopic rule. The above analysis is stated in terms of a cryptocurrency TFM, but applies to other settings, such as cloud computing and decentralized ``gig'' economy, as well.