Metrics, Events, Order Flow, and Innovations: Highlights from MEV from 2022 to 2023 Twitter Space
Reflecting on the Past Year and Looking to the Future. And how to prevent citadel PFOF scenario.
2022 is a memorable year for DeFi. Terra Luna, 3AC, Celsius, FTX... Despite all these downfalls in the bear market, MEV activities still contributed $328B. It's about half the total DEX volume of $666B on Ethereum.
Coming to 2023, now Q1 has already passed. Signature Banks, SVB, Credit Suisse, and the potential contagion in TradiFi have been making people pull their hair. The volatility is inevitable, which is good news for MEV.
The volume is an indicator we talked about a lot. But it's not good enough for MEV activities, which would lead to our first topic.
On March 23, EigenPhi hosted a Twitter Space reviewing and discussing the MEV's performance in 2022, where some experts in MEV were invited to give their insights, including:
Christine Kim. Vice President, Research at @galaxyhq
Danning, Lead Data Scientist at @0xProject
Fiddy, Core Contributor of @CurveFinance
James Prestwich, Founder/CTO of @nomadxyz_
Uri Klarman, CEO of @bloXrouteLabs
Dr. Yixin Cao, Data Scientist at @EigenPhi
Here are some of the highlights of their thoughts.
Topic 1: MEV's Key Metrics
Q1: What key metrics and methodologies should we use to study and analyze MEV, and how can we better understand this complex and multifaceted phenomenon?
Christine Kim @galaxyhq: Regarding the key metrics, it depends on the aspect of MEV that one is interested in. After the MEV boost and post-merge, one of the metrics people were interested in was the censorship activity happening on-chain, adoption of relays, and transaction activity moving through the Flashbots relay versus other relays. However, if the focus is on earning the most MEV, the metrics may vary and rely more on analyzing smart contract activity and understanding how DeFi markets are moving. It's crucial to note that MEV is multifaceted, and one's interest may heavily depend on their focus.
Etherscan is an excellent tool for understanding complex MEV strategies. I became interested in MEV after discovering how adept some people are at analyzing Etherscan step by step. Bert C. Miller, part of the Flashbots team, deserves a shoutout for writing informative threads detailing how MEV is extracted on-chain and how to observe it in action.
Danning @0xProject: As Christine mentioned, there are different aspects to consider. For instance, at 0x, we look at slippage, which is the difference between the expected price and the executed price of a trade. Our research found that MEV often causes users' trades to slip to the worst price at the end of the slippage threshold. Another comment on MEV volume is that it's primarily underestimated because most of the activity happens across DEX and centralized exchanges outside our data vision. However, the volume is significant, as shown in the EigenPhi report, that it is 50% of all the DEX volume.
After the MEV boost, it was interesting to see the shift in block production distribution from POW miners to validators and the rising MEV-boost adoption. So another in-depth metric I noticed is the split of MEV revenue across validator and block builders, while the validator percentage has been going up over months. This means block builders have had more and more competition. So they have to give up more.
Fiddy @CurveFinance: The slippage threshold set by users in the UI significantly impacts the ceiling of how much can be extracted. Different protocols have their logic for the default threshold value. For example, at Curve, we monitor trades and look at sandwiches, and iteratively try to reduce the default slippage by a little bit to see if it makes a difference. When calibrating the default slippage in our front end, we consider volume, payout, block producer rewards, and gas costs. We also use EigenPhi to study these things.
James Prestwich @nomadxyz_: Unlike the other panelists who are very concrete, I'm more of an abstract first principles person. My key metric is the balance between MEV extraction and non-extraction transactions. So as MEV transactions consume more gas than non-MEV transactions, they have a higher expected value per unit of gas. As a result, at market equilibrium, we should expect to see a split where MEV extraction consumes the majority of the gas in the block and non-extraction transactions take up a minority of the book. I don't have a strong opinion on where that shakes out, but I am interested in seeing this dynamic development in the wild that we predict from the first principal's perspective.
Uri Klarman @bloXrouteLabs: I would argue that the metrics we extract from the chain are not insignificant but tell half of the story at best.
Denning noted that over half of the MEV trading volume comes from CeFi and DeFi arbitrage. The process is such that CeFi responds to real-time price changes on platforms like finance and Coinbase. At the same time, DeFi arbitrageurs with substantial liquidity compete to capture the arbitrage close to the next block's mining. This is where the significant money is, but we can only see one leg happening on DeFi. Whoever successfully lands their DeFi leg first will also perform the CeFi leg of the arbitrage, and we don't know how much money is left with the arbitrageurs or how much goes to the validators. It's theoretically a competition, and the value should eventually go to the validators and their stakers. However, from my perspective, the CeFi DeFi arbitrageurs are keeping much of the value, and it's not going elsewhere.
My perspective differs slightly in that the metrics we use to measure MEV are only part of the story. I am more concerned with how much of the extracted value ends up with validators and stakers, which is uncertain. So it is difficult to know, and it's unclear whether DeFi is becoming too predatory and all the value is being extracted, leading to unintended results.
Yixin @EigenPhi: I'd like to propose two metrics. The first is we can construct some metric to see whether the MEV has replaced the original incentive mechanism to bring a critical impact on the security of Ethereum. As mentioned by Danning and Uri, we can calculate the MEV as a percentage of the builder and validator's revenue to see if it has already surpassed the original gas fees they get.
And another indicator I am interested in is the scale of cross-chain maybe, which should also indirectly reflect the development of multi-chain ecology. And I think expanding this metric or the cross-chain MEV scale will also promote further competition and reshuffling among block builders. So I'm glad to see these two metrics and how they evolve in the long run.
Topic 2: Most impactful Events.
Q2: Besides The Merge, in 2022, which MEV-related events would you consider the most impactful?
Uri Klarman @bloXrouteLabs: The most significant change in the MEV landscape is undoubtedly the Merge and the transition to MEV-boost. While cross-chain MEV exists, it is dwarfed by CeFi and DeFi arbitrage. The Merge and MEV-boost have created a world in which validators constantly strive to extract the most value from users and liquidity pools, resulting in users receiving the worst execution possible. With MEV-boost and the Merge, the value extracted from users by CeFi and DeFi arbitrage now goes to validators and their stakers. While some may view this as positive, it does not make DeFi more user-friendly or competitive with CeFi. From my perspective, this is a concerning direction for DeFi, and we must consider the impact on users. This has been the most significant event in the MEV landscape, and the outcome is still uncertain.
James Prestwich @nomadxyz_: I both agree and disagree with Uri. Five years ago, I wrote an article about MEV. We didn't use that term at the time. The article was called "Miners are Not Your Friends." The article aimed to explain to people why MEV extraction would eventually happen to every user all the time.
Five years later, as Uri points out correctly, that is the situation we are in. All validators essentially extract MEV to the best of their abilities. This is precisely what we predicted five years ago. Where I disagree with Uri is that we are not trying to build the best DeFi experience. Still, we are trying to build the best DeFi experience while being mindful that people will always try to extract money from us. This is something that we cannot change, but what we can do is try to minimize the amount we overpay for that value. This is why proposers extracting value was inevitable; we predicted this confidently five years ago. Our task is not to prevent them from extracting money but to try to extract equivalent value from them and minimize the amount we overpay for that value.
Uri Klarman @bloXrouteLabs: I agree that we need to incentivize participants to make money in ways that benefit users and minimize their ability to harm users. While we cannot wholly eliminate MEV extraction, we should aim to minimize it. However, I observe that the focus is now on maximizing MEV extraction and distributing it among everyone rather than minimizing it. We should prioritize MEV minimization and create incentives that align with this goal. It is essential to remember that we cannot ignore the reality that people will try to extract money from us, but we can work towards reducing the amount we overpay for that value. As we continue to navigate this landscape, we should focus on incentivizing positive behavior that benefits users rather than solely maximizing profits.
Topic 3: Order Flow & PFOF
Q3: Starting from last year and during the ETH Denver, especially Flashbots' mev [re]search-athon, Order Flow is a hot topic.
1. What is order flow, and why does it matter for MEV?
Danning @0xProject: My current understanding of MEV order flow mainly involves the DEX order flow or the orders that can potentially be extracted with MEV value, primarily related to DEX trading. Some other events such as long liquidations may also be included. The order flow concepts proposed by Flashbots are focused on accepting more orders from application teams or RPC providers, potentially exposing them to users, and then distributing them to a group of searchers who can bid for the highest potential extracted value bundle on top of each order. Then, we can decide how to take the bid and send it to a group of block builders to execute the transaction. Many researchers are currently figuring out how to provide a pro-privacy layer while exposing this order flow information. This is because the value of this information is significant, and once it is shared, it can potentially be leaked. The topic of discussion is how to track or prevent this from happening.
Uri Klarman @bloXrouteLabs: When considering order flow, the idea is that searchers and validators could benefit from taking advantage of transactions. For instance, a user making a transaction could choose to send it exclusively to a particular searcher in exchange for a fee. However, this could result in centralized mechanisms favoring whoever has the most order flow, with Metamask potentially becoming a kingmaker. The problem here is twofold. First, we want users to receive value back. However, if the value goes to the searcher, it doesn't go to the validator, creating a tug-of-war between the two ends. Second, this can lead to substantial centralization effects, where one builder dominates all the blocks. Thus, we must find a way to enable users to extract value without turning Ethereum or DeFi into a super centralized setup with one builder. An auction or bidding process based on the intent of a transaction could be a potential solution where not all transaction information is shared. This should be the focus of our thinking around order flow.
Fiddy @CurveFinance: My understanding of order flow in the context of decentralized exchanges (DEXs) is as follows: when a user wants to make a trade, they offer their intent to trade, approve their tokens, etc. Then, someone else ensures that the user gets the best possible price. The problem of how to execute the trade is not the user's concern. Instead, it's the responsibility of an entity or group of entities that compete to execute the trade at the best price. These entities include searchers, block builders, and solvers. Solvers are a new addition to the ecosystem who determine the best possible execution venues, including DEXs, RFQ, and quotes from 0x. They may even leverage other sources of liquidity that one cannot imagine.
But by the way, this approach is becoming increasingly centralized as more entities with specialized infrastructure and skills can capture multiple verticals of the MEV landscape. So this is quite an existential centralization threat to any blockchain.
2. How to prevent citadel-like PFOF?
Uri Klarman @bloXrouteLabs: Bloxroute is a major block builder and relay, accounting for between 5 and 15% of Ethereum's blocks and propagating between 15 and 25% of Ethereum's blocks. However, as a major block builder and relay, I find the position undesirable. I don't want to be responsible for building and relaying blocks, and I certainly don't want to be in a place where I could lie about it. Is this centralizing? Absolutely.
Flashbots present a valid argument that some validators will be better at extracting MEV than others, creating a self-feeding cycle that reinforces centralization. Validators with higher API attract more stakers, making them more prominent. Searchers send bundles to the most prominent validators to guarantee to land on the chain. As a result, some individuals become better at MEV extraction than others. This makes MEV centralized.
From my perspective, this is the strongest argument for the centralizing effect of MEV. While Flashbots' idea is good, I think we can find ways to improve the current MEV infrastructure.
Topic 4: MEV innovation
Q4: Order flow, or Flashbots' SUAVE. These are all MEV-related innovations. How has MEV driven innovation in the blockchain space, and what examples of MEV-based innovations might we see in 2023?
Danning @0xProject: I don't have much to share, but James, the earlier speaker, shared an exciting design for an MEV wallet where users can propose how much MEV they want to request while making a trade through the wallet. I found that idea fascinating, although I am not sure if it is currently in use or still in development. The Order Flow Auction (OFA) is also an innovative idea that could introduce some value redistribution.
Fiddy @CurveFinance: I find the long-tail MEV quite interesting; as James nicely put it, the MEV extracted must provide some value in return. However, this difference should be minimized as much as possible so that we receive enough value in return for delivering MEV opportunities. For example, at Curve, we plan to conduct liquidations involving automatic market makers, which is only possible because of the MEV incentive to do so. This is an example of how we take an incentive and try to innovate on top of it.
Uri Klarman @bloXrouteLabs: My approach differs slightly from James's and Fiddy's. Rather than focusing on extracting value, I want to offer a different paradigm for DeFi. The problem with DeFi is that it's slow compared to CeFi, and arbitrage opportunities can be missed. So we want DeFi to be an alternative that can compete with CeFi in real-time trading.
Imagine validators providing free confirmation during the 12-second block time instead of waiting for the next block. They would confirm transactions in real time, turning the entire DeFi domain into something similar to CeFi. This would allow for more volume and a more valuable trading experience. The idea is to make DeFi more useful to users in a way that pays better than extracting value.
Currently, DeFi is small compared to CeFi. If we make it bigger and more valuable, validators could capture more value for their service than through front-running and arbitrage. So my first-principle approach is to incentivize and make DeFi more valuable than just extracting MEV, which is a different paradigm than trying to get value back while it's being extracted.
The above summary does not fully reflect the guests' insightful opinions. Click here to listen to the full recording. And please stay tuned for more!
Follow us via these to dig more hidden wisdom of DeFi: