DLT Interoperability and More ⛓️#3 ⛓️ — Unity is Strength: A Formalization of Cross-Domain Maximal Extractable Value

In this series, we analyze papers on blockchain and interoperability (and both).

This edition covers a paper from December 2021 on cross-domain MEV

Photo by Shubham Dhage on Unsplash

➡️ Title: Unity is Strength: A Formalization of Cross-Domain Maximal Extractable Value
➡️ Authors: Alexandre Obadia , Alejo Salles , Lakshman Sankar, Tarun Chitra , Vaibhav Chellani , and Philip Daian

➡️ Contributions:

• the authors define the maximal extractable value (MEV) in the context of several blockchains
• the authors informally define the domain concept.

💪 Strong points:

• Important problem: MEV is starting to get exploited in a cross-blockchain environment, where extraction implies moving assests across blockchains.
• Figure 2: depicts a cross-chain arbitrage opportunity across two heterogeneous blockchains, by conducting two transactions: exchanging MATIC for WETH on the Ethereum blockchain; (bridging WETH on Ethereum to WETH on Polygon), and exchanging WETH for WMATIC, on Polygon. This what we call as a cross-chain transaction that is composed by two local transactions.
• Figure 3: depicts a cross-chain arbitrage opportunity across three heterogeneous blockchains, by conducting four transactions: exchanging MATIC for WETH on the Ethereum blockchain; exchanging WETH for USDC on the Ethereum blockchain; (bridging USDC to the Binance smart chain); exchanging USDC on the Binance Smart Chain for ETH, (bridging ETH to WETH), and exchanging WETH for WMATIC, on Polygon. This what we call as a cross-chain transaction that is composed by three local transactions.

😞 Limitations:

• As an introductory paper, there are no limitations to point

🔥 Points of interest:

• MEV in a cross-domain setting, but adds some layers of complexity. First, transaction ordering happens in multiple protocols, implying that MEV operators need to own nodes in different protocols. Second, the assets our trader starts and ends are not necessarily the same. Third, as bridging operations are not atomic, our cross-chain transaction needs additional mechanisms to ensure the atomicity of the operation. The trade-off typically is between safety and liveness (we can have a finalized transfer in a larger ammount of time, or assume the transaction will go through and lowering latency).
• The definition for extractable of a certain participant is the difference of utility (in the form of balance) from two different points in time, taking into account a set of actions (or transactions). Now, the MEV that can be obtained from a set of blockchains corresponds to the difference of balances, that were obtained with a sequence of transactions, in several blockchains. Well, more specifically, it is the profit that can be obtained in a blockchain while only manipulating another (e.g., reordering transactions). We can see that it is possible to compose these operations such that the sequence of actions in one blockchain will influence the “initial state” on another, and create a sort of consequence chain that propagates across multiple domains. This implies that the profit that you can do by operating in multiple chains may be superior to the profit on transacting in each chain.
• The authors say “We intentionally allow for fuzzier distinctions in actions that affect multiple domains simultaneously, to allow for modeling cross-chain communication protocols, bridges, and other interactions between domains as their own actions that act simultaneously on multiple domains” — this is a good idea, since formally representing cross-chain transactions can be quite cumbersome (see for example Hermes: Fault-Tolerant Middleware for Blockchain Interoperability).
• Cross-Domain Maximal Extractable Value is then the problem of choosing a set of actions such that it outputs the highest final balance. Choosing those across several systems has some challenges. First, the modelling and representation of cross-chain state (or, at least, a consistent state representation across chains). We solve this problem in our BUNGEE paper, to the best of our knowledge, the first solution to do so. After that, choosing these actions seem to be an NP-complete problem. More research needs to be done in this domain, and the fact is, the more ubiquitous, secure, and cheap bridges become, the higher the potential for MEV.
• The authors say “We expect that, given the deployment of AMMs and other MEV-laden technologies across multiple domains, the benefit of extracting MEV across multiple domains will often outweigh the cost of collusion”
• One does not necessarily need to use bridges to maximize the MEV: having funds on each chain might suffice.
• We can therefore say that the cross-chain MEV problem is a generalization of MEV (but considerably harder due to the composition of different blockchains, new security assumptions, and different system model).
• It is interesting to see that authors consider that the notion of state is not tied to any specific blockchain, but there is no formal desciprtion. In our work in progress paper, BUNGEE, we define a notion of arbitrary state that can be used to represent information in any blockchain. We also define the concept of domain, as the subset of the global state that a certain participant can see. With arbitrary sate, one can construct what we call blockchain views. A blockchain view is a subset of the global state, from the perspective of a blockchain participant, that has an initial and final timestamp. It can be a good construct to reason about cross-chain MEV.
• Cross-chain MEV where sequencers/miners collude might have a nefast consequence: the perception of fairness on that chain diminuishing and, as a result, the token value as well.
• Cross-chain MEV is a nascent area of research. Be sure to check some good references on MEV: https://arxiv.org/abs/2101.05511; https://ieeexplore.ieee.org/abstract/document/9519469; https://arxiv.org/abs/2106.07371; (for more generic blockchain interoperability see also https://dl.acm.org/doi/abs/10.1145/3471140 and https://www.techrxiv.org/articles/preprint/Do_You_Need_a_Distributed_Ledger_Technology_Interoperability_Solution_/18786527/1)

🚀 How does it relate to our work at Técnico Lisboa, INESC-ID, and Blockdaemon? (views are my own and do not necessarily reflect the opinions of my employer)

• When the need for interconnecting blockchain emerged, we saw a new research area: blockchain interoperability. We are working to understand how to create robust and secure solutions, as well as new technologies. Studying MEV in the context of cross-chain will allow exciting new directions.

🚀 What are the implications for our work?

• This work inspires us and makes us think about the implication of blockchain interoperability on the application layer.