Orderflow: The Status Quo, Observation, and Competition Equilibrium

Opening Remark

Source: Flashbots

I’ve referred to this figure for quite some time when I shared my investment thesis with people — it’s fascinating to examine various projects through the lens of the MEV supply chain, or more simply, the transaction lifecycle.

Thanks to many researchers’ persistent work, we’ve moved from the MEV Dystopia of a few years ago to a relatively clear supply chain, though some challenges still need to be addressed.

In the context of this supply chain, orderflow plays a key role. The orderflow can be likened to the running water that flows over the supply chain riverbed. In this article, we aim to summarize some key aspects of orderflow.

Orderflow…wat do?

Source: IOSG

Quintus defined “order” as anything that allows one to change the state of the blockchain in the context of Ethereum. In the above figure we try to illustrate the orderflow’s typical journey through the MEV supply chain. However, the actual case may look like this…

Source: Frontier Research

This article will not attempt to dissect all the details. The entire supply network has become so huge and complex, making it difficult to figure out everything within a short time. Instead, we will focus on two main topics: private orderflow and orderflow auction (OFA).

Private Orderflow Dominance

Public mempools will dry up… — Hasu

We usually come across two terms — private orderflow and exclusive orderflow, which are semantically different. For instance, MEVBlocker sends the orderflow to top builders, bypassing the public mempool. In this case, the orderflow is private but not exclusive to a single entity. Importantly, OFAs are not ‘private’ transactions as well since people can subscribe to the stream (though could be gated) and monitor the transactions, which maximizes competition among each party.

There has been much discussion in the community about the centralizing force that private orderflow may pose.

Source: mempool.pics

Within the entire MEV supply chain, the builder segment possesses a substantial competitive landscape. Orderflow, being the fundamental means of block building, holds a key position within this. Under this competitive market, builders offer a variety of services to lure orderflow from the upstream:

• Transaction pre-confirmation. For instance, consider what Vitalik mentioned in his SBC 2022 speech: Builder publicly agrees that if a user sends tx with priority fee >= 5, builder will immediately send an (enforceable) message agreeing to include it. If priority fee >= 8, the user gets a post-state root. This allows for some use cases that involve contention for the block space.
• Front-running protection, which is a key selling point for many OFA platforms.
• Revert protection. If a certain bundle fails or reverts, the builder will not include it. For users, MEVBlocker’s RPC endpoint offers revert protection by sending the regular transaction as a bundle. Blocknative’s data shows that only approximately 4.2% of private transactions reverted on-chain, compared to 13.8% of public transactions, which emphasizes the importance of revert protection.

Source: Danning Sui (@sui414)

In the figure above we can easily find the correlation between landed blocks and private orderflow in the builder landscape.

Source: Blocknative

Recently, many studies have shown that private orderflow is intensifying. The data from Blocknative reveals that the percentage of private transactions grew from 5% pre-Merge to nearly 15% today. This indicates that approximately 15% of Ethereum transactions flow through private channels.

Basically, private orderflow can be divided into searcher flow and user flow.

Searcher Flow

• Atomic Flow

Source: Titan

Atomic searcher flow refers to DEX-DEX arbitrage, sandwiching and liquidation.

A while ago, Titan complained that even though with ~15% market share (7d) and over 50k blocks built, they’re still not receiving bundles from ~50% of all searchers.

In the figure above, we found the market share of builders is positively correlated with the number of searchers connected to them. This correlation is often exponential, as revealed in Titan’s recent research.

• Non-atomic Flow

Non-atomic searcher flow refers to CEX-DEX arbitrage.

Source: searcherbuilder.pics

In the atomic MEV case, we can find all searchers evenly submit the bundles to the top three builders. However, when it comes to the orderflow of CEX-DEX arbitrage, Titan, as a neutral builder and one of the top three builders, only accounts for 8.8% orderflow share for all searchers. There is a noticeable difference in the two figures above. This highlights the exclusive access to orderflow by vertically integrated searcher-builders.

Overall, the inclusion guarantee is the most crucial factor for searchers when deciding which builders to submit their bundle to. It determines whether the opportunities captured by the searchers can be realized on chain.

In addition, the searcher faces some trade-offs. If the searcher sends the bundles to all builders, it maximizes the chances of including their bundles in the block. However, according to Titan, this also exposes the searcher to more risks, including unbundling and information leakage of long-tail strategies.

For searcher flow, currently most searchers submit the bundle to top builders. When submitting the bundle to searcher-builder entities, potential conflicts of interest need to be considered. The mid-field builders now also hold over 16% of the market share — surely it is necessary to send the bundles to them as well.

• To mitigate EOF…

Exclusive orderflow has a solution. — Stephane

Source: Danning Sui (@sui414)

Quintus proposes several methods here.

• Large infrastructure neutrality. The community needs to closely monitor the actions of large infrastructure, acting like a watchtower for optimistic rollup. In the face of potential public pressure, we believe Metamask and Infura will proceed with caution in their actions. But ultimately “it is not in the spirit of blockchain to rely on agents to do the right thing”.
• Decentralizing the builder role. Vitalik shared his perspective on this at the SBC MEV Workshop. Jon later elaborated this in his article, discussing algorithms, resources, builder services, and orderflow. Indeed, distributed builders can provide censorship resistance and stronger trust guarantees than a single entity. However, to achieve specific goals, tech often involves trade-offs. While it may bring certain benefits, there is inevitably a sacrifice in other aspects, such as inefficiency or increased latency. While a distributed builder may be a potential solution, we reckon it would be challenging to establish a leading position in a competitive market, given that people in the supply chain are time-sensitive, especially for builders. It’s the cost of decentralization after all.
• Defaults settings. It is crucial for upstream players to have the flexibility to send their orderflow, as it can be beneficial for new builders who are entering the market.

In terms of compliance that people may overlook in the space, Alex pointed out here that integrated searcher-builders who get private orderflow are risking ending up (unknowingly) insider trading. It is illegal to use Material Nonpublic Information (MNPI) for one’s advantage in trading stocks or other securities in TradFi.

Evan in this article, provided a detailed analysis of the OFAC risks in the MEV supply chain. For builders, their OFAC risk outlook is similar to searcher. However, they would have to more actively review the orders they receive from earlier block-building participants and reject those that touch the SDN (Specially Designated Nationals And Blocked Persons) List.

In TradFi, Pay For Order Flow (PFOF) has always been controversial. In the crypto space, we believe that there is still a need for further exploration in terms of compliance.

User Flow

Users today don’t really have control over the MEV value they leak, as the submission is through wallets, dApps, and RPCs, who then expose users’ order to public mempool, or privately for searchers to capture MEV value.

Source: mempool.pics

Frontier’s research shows that within a six-month period, approximately 71% of all validator payments are made from transactions that include at least one swap. In the above figure, we can also see that, when breaking down all private orderflow count by type, swaps from the user flow still represent the majority compared to atomic searcher flow.

Source: Dune@angelfish

Based on the dashboard by angelfish, it’s evident that the main sources of orderflow continue to be dominated by DEXs such as 1inch and Uniswap. Plus, it is worth noting that vertical integration within wallets, e.g., Metamask Swap and Tokenlon (by imToken), also plays a role in driving orderflow.

Moreover, there has been a notable surge in the popularity of Telegram Bots, including Unibot, Maestro and Banana Gun. TG Bot’s main selling points are ease of use, abstraction of wallet, and simplification of dApps interaction. It also provides some common functionalities such as revert protection and front-running protection.

These emerging platforms have witnessed rapid growth and are gradually gaining traction in the orderflow space. Months ago, Maestro signed an exclusivity agreement with bloXroute. This agreement allows Maestro to route their orderflow through bloXroute’s private RPC. Stephane (Frontier) and Vadym (Kolibrio) have recently launched TG Bot Alfred as well, receiving a lot of attention.

Telegram has a wide user base. Although it is still premature to discuss achieving mass adoption through TG Bots, with the continuous iteration of different bots’ functionalities, we expect to see fascinating developments in this field and how they will impact the supply chain.

Current State Summary of Orderflow Types

Source: Danning Sui (@sui414)

To summarize, the table above illustrates the distinctions between different types of orderflow.

As mentioned earlier, integrated searcher-builders have an advantage in their non-atomic searcher flow compared to neutral builders. This is because they have an exclusive internal flow and can fully utilize the margin generated by their own searcher to support the builder, e.g. bidding higher in the block building process.

By consolidating their market share, they can obtain more atomic searcher flow. On the other hand, neutral builders typically do not receive non-atomic searcher flow, and smaller atomic searchers also lack a builder and must send orders to everyone. In a word, various factors contribute to the dominant position of integrated searcher-builders in the block building market.

Orderflow Auction (OFA)

• Overview

Source: Frontier Research

OFA is a method that enables price discovery of orderflow. The interaction between buyers and sellers determines the appropriate pricing for the orderflow.

Basically, both Frontier and Monoceros already provided a comprehensive landscape of OFA today, including its design space, tradeoff, and other aspects. The design decisions shared by Frontier regarding OFA enabled us to consider the design possibilities of OFA from scratch. Several companies, including Kolibrio and DFlow, have been actively exploring this field. In this chapter, we will not reiterate these details.

Source: Dune@cowprotocol

On the data side, according to the MEV Blocker Dashboard, transactions processed through MEV Blocker currently represent approximately 3%-4% of all Ethereum transactions. There are currently around 473,337 users, and a total of 738 ETH has been returned as kickback. Among them, builder0x69 and beaverbuild have returned the highest amount of ETH, both exceeding 200 ETH.

So far, there have been several OFAs actively operating, proactively talking to upstream platforms like wallets and doing “community-wide marketing”.

While the comprehensive protection promised by OFA may seem like the ultimate solution for users, a recent article by Blair indicates that private transactions with front-running protection don’t necessarily guarantee improved settlement or the comparable execution speed as public transactions.

As mentioned in the article, OFA ensures that the user’s trades are protected from front-running, but this does not mean “all-around” optimal execution, and the user’s trades can still be accidentally affected by slippage.

This article also highlights the distinction between sending private transactions using “eth_sendBundle” and “eth_sendRawTransaction”. A bundle that does not include MEV is not competitive compared to other bundles. However, this necessitates the builder to conduct unnecessary simulations, potentially leading to slower execution speed.

Afterwards, Blair highlighted the existing observability gap in OFA in a thread. As a user (or wallet that adopts OFA), how do I know/ can verify that both the auctioneer and the builders are acting honestly? In the observability case study mentioned in the thread, they found 19 different transactions from their user that “could” have received a refund that did not and instead the builder’s backrun bot claimed the opportunity. This implies the observability gap does exist in OFA today.

• Takeaways

Source: IOSG

In this part, we’ll share our insights based on our market observation in the region.

• Users

According to imToken’s 2023 Crypto Wallet Report, 63% of surveyed users find trading on exchanges to be either more convenient or cheaper than trading on wallets. Additionally, 38% of users perceive wallets to be less secure than exchanges due to fears of being hacked (29%) or losing crypto due to their own mistakes (18%).

This suggests that many users may still lack sufficient knowledge about things on chain. Therefore, users may experience friction when changing from the wallet’s default RPC settings to the OFA. There is still much to do in terms of user education.

• Wallets

To achieve OFA adoption, the wallet may need to make decisions on behalf of the user. Based on our conversations with wallets, we found:

• Most wallets see OFA more as a vitamin than a painkiller.
• Some are not fully aware of MEV and its potential effects.
• Some wallets think OFA is still in an early stage, with a lack of transparency. This lack of transparency is one of the key factors that influences their decision on whether to integrate OFA, as they are unable to evaluate the associated risks under the hood.
• Although some do recognize it as a monetization opportunity, integrating with OFA is still not their priority. In fact, since its launch in April this year, the total kickback of MEV Blocker is only 783 ETH, which is not a significant amount to wallet overall. For these wallets, transaction inclusion speed and success rates are more important than the potential rebates.
• Overall, wallets lean towards cautious optimism — there is currently not a strong motivation for them to proactively adopt OFA.

Based on the The 2023 Geography of Cryptocurrency Report by Chainalysis, countries in Central and Southern Asia are leading the world in grassroots cryptocurrency adoption. According to our observations in the APAC market, wallet adoption often exhibits geographical trends as well. For instance, mobile wallets in APAC usually provide timely support for multiple chains and consumer apps. This makes them highly suitable for the Southeast Asian market.

As a result, in the case of key infrastructure like Metamask remaining neutral, long-tail wallets focusing on specific regions are more likely to adopt OFA first compared to the others. This should be the primary focus for OFA platform’s business expansion. In a market with multiple existing OFAs, the competition in the wallet sector will become extremely intense.

OFA has various trade-offs in its design, and the main challenge is to persuade the upstream of the supply chain to adopt this solution. Public education, of course, is a slow and long-term process. Moving forward, we believe that OFA will be gradually adopted, potentially accounting for over 30% of all transactions on Ethereum. Wallets will play a crucial role in facilitating this transition. It is important to be mindful of the centralizing driving force in this process.

Closing Thoughts

Source: Max Resnick (@MaxResnick1)

In the community discussions, there has been a lot of attention given to the risks of centralization posed by large mining pools/validator pools. However, in the upstream of the supply chain, the risks of centralization related to orderflow are even more concealed. Large infrastructure entities, like Metamask and Infura, need to be mindful of potential bias in their decisions. Small-scale infrastructure can conduct more cutting-edge exploration, facilitating ecosystem discovery.

In the future, the competition among various stakeholders for orderflow will become even more fierce. Those in the upstream of the order flow will firmly grasp and gradually strengthen their pricing power. The emergence of OFA and TG Bot indicates that infrastructure is exploring the upstream in the supply chain. In this process, we found developers are increasingly focused on user-oriented and MEV-aware design.

Numerous challenges still exist within orderflow. At the MEV researchathon, hosted by Flashbots during SBC ’23 week, navigators proposed the co-creation of a transaction supply map: orderflow.pics. Recently, Toni published mempool.pics, a site that presents statistics about private orderflows and the participating actors. And winnsterx launched transparency dashboard searcherbuilder.pics. An increasing number of researchers and developers are focusing on the advancement and future of orderflow. This focus is driving the competitiveness, fairness, and censorship resistance in the supply chain. We will also actively keep an eye on it.

Stay tuned!

Disclaimer：

1. This article is reprinted from [medium]. All copyrights belong to the original author [Jiawei]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.