TL;DR
This paper tackles the challenge of designing fair and efficient transaction fee mechanisms (TFMs) in blockchains where block producers (BPs) can extract "Maximal Extractable Value" (MEV). The authors prove fundamental impossibility results: no non-trivial TFM can be both incentive-compatible for users and active BPs simultaneously. They then introduce a new framework incorporating "searchers" to overcome these limitations, proposing a mechanism that achieves desirable properties and significant welfare guarantees.
What Problem Does This Paper Address?
Blockchains rely on transaction fee mechanisms (TFMs) to decide which transactions get included in blocks and how much users pay. Traditionally, TFMs assumed "passive" block producers who only cared about the fees they collected. However, in modern blockchains, especially those with decentralized finance (DeFi) ecosystems, block producers are "active" – they can extract additional value (MEV) from the order and content of transactions beyond just the fees. This MEV can arise from activities like arbitrage or liquidations.
The core problem this paper addresses is how to design TFMs that remain incentive-compatible for both users (ensuring they bid truthfully) and active block producers (ensuring they follow the mechanism's rules) in the presence of MEV. The authors highlight that existing TFMs, like EIP-1559, struggle to maintain these properties when BPs are active and can leverage MEV. This leads to a fundamental question: can we design TFMs that are both user-friendly (DSIC) and robust to active block producers (BPIC) in a non-trivial way?
Methods
The paper first establishes foundational impossibility results by modeling active block producers with private valuations for blocks, in addition to transaction fees. They prove that no non-trivial TFM can be simultaneously Dominant-Strategy Incentive-Compatible (DSIC) for users and Block Producer Incentive-Compatible (BPIC) for active BPs. This means either user experience suffers (users can't easily determine their optimal bid) or BPs can exploit the mechanism.
To overcome these limitations, the authors introduce a more sophisticated model that incorporates "searchers." Searchers are entities that identify MEV opportunities and compete to include them in blocks. This creates an "MEV oracle" that informs the TFM. They then propose and analyze the "Searcher-Augmented Knapsack Auction" (SAKA) mechanism. SAKA uses a first-price auction among searchers for bundles of transactions and their associated MEV extraction logic. It then uses a threshold-based allocation rule to include transactions and bundles, aiming to balance user payments, searcher bids, and block size constraints.
Datasets
Not discussed in the paper.
Key Insights and Ideas
- The MEV Dilemma: Active block producers fundamentally break the assumptions of traditional TFM design. Their ability to extract MEV creates a tension that makes it impossible to satisfy both user incentive-compatibility (DSIC) and block producer incentive-compatibility (BPIC) simultaneously in a non-trivial way.
- Searchers as a Solution: Introducing searchers, who specialize in finding and bundling MEV opportunities, significantly expands the design space. They act as a valuable signal to the TFM about potential value extraction.
- SAKA: A Practical Approach: The SAKA mechanism demonstrates that by leveraging searcher bids, it's possible to design TFMs that are DSIC for users, DSIC for searchers, BPIC for the proposer, and sybil-proof. This is a significant step beyond the impossibility results.
- Welfare Guarantees: While SAKA doesn't achieve perfect social welfare maximization, it provides a strong approximation guarantee (around 50% of optimal welfare), which is a notable achievement given the inherent complexities.
Key Results
- Impossibility Theorems: The paper proves that no non-trivial TFM can be simultaneously DSIC for users and BPIC for active block producers (Theorem 3.1). Furthermore, such mechanisms cannot guarantee any significant fraction of maximum social welfare (Theorem 3.3).
- SAKA Mechanism Properties: The proposed SAKA mechanism is shown to be DSIC for users and searchers, BPIC, and sybil-proof (Theorem 5.3).
- SAKA Welfare Guarantee: SAKA achieves at least a (1-γ)/2 welfare guarantee, where γ is the fraction of block capacity consumed by a single transaction or bundle (Theorem 5.5). This guarantee is shown to be near-optimal among mechanisms satisfying the desired properties (Theorem 5.7).
Limitations and Future Work
The authors acknowledge that their impossibility results hold even under simplified assumptions (e.g., additive valuations, single feasible block). The SAKA mechanism's welfare guarantee is an approximation, not an exact optimum. Future work could explore:
- Bayesian Analysis: Investigating reserve prices and their trade-offs between revenue and welfare in a Bayesian setting.
- More Complex Collusion: Analyzing the robustness of mechanisms against more sophisticated forms of collusion between users and searchers.
- Practical Implementations: Exploring how these theoretical mechanisms can be practically implemented and deployed on live blockchains, considering factors like gas costs and network latency.
- Dynamic Reserve Prices: Developing dynamic strategies for setting reserve prices in mechanisms like SAKA to improve welfare guarantees.
Summary updated