MEV 2023 Outlook: Crossing the Dark Forest

Original Title: MEV Outlook 2023: Walking Through the Dark Forest

Author: EigenPhi

Compiled by: Peng SUN, Foresight News

* Note: The statistical scope of liquidation transactions calculated in this report is from May to December 2022, while the scope for other types is for the entire year of 2022.

1. A New Perspective on MEV

Currently, most reports and articles on the topic of MEV revolve around the introduction of concepts and how to bring orders into the MEV market. The issues of concern include but are not limited to the fairness of the MEV market, the negative externalities brought by MEV, and the centralization risks or censorship issues related to mechanism design, all of which are directions that many blockchain protocols and application layer builders are striving to address.

In addition, Amber Group and Forbes have also published articles introducing a large number of long-tail MEV cases, the insider strategies of the famous MEV seeker Nathan Worsley, and the two quantitative institutions running MEV strategies, Wintermute and Alameda Research. Their articles mention the concept of "Payment for Order Flow," which seems to be a significant trend that developers are working towards, including Cowswap, 1inch Fusion model, and Flashbots' MEV-Share. Mainstream financial institutions have also begun to pay attention to the MEV field; for example, Nasdaq's news and insights section shared a CoinDesk article on how JIT bots and MEV promote DeFi. Of course, a small number of people are also concerned about the regulation of the MEV market.

In fact, beyond the general focus on the operation of the macro market mechanisms of MEV, we believe there is a micro issue worth attention: as a representative, MEV trading means that liquidity data has unprecedented opportunities and challenges.

As early as May 2022, during the series of events triggered by the UST-LUNA meltdown, financial institutions engaged in crypto business were already under scrutiny by users, publicly disclosing their balance situations. This situation is uncommon in traditional finance, as the transparency and openness of blockchain ledgers allow anyone to access transaction data for free. Through on-chain transactions, we can obtain real-time generated and updated historical price information and liquidity data. Of course, the transparency of liquidity data can also help crypto financial institutions establish real-time risk control mechanisms based on liquidity, allowing for earlier awareness of crises.

This report aims to prompt users to pay attention to the value and risks of existing liquidity data in order to better protect assets or design better protocols to safeguard users.

2. Overview of the MEV Market in 2022

MEV is the most authentic reflection of liquidity in the DeFi market. In 2022, MEV bots generated at least $307 million in revenue on Ethereum. Therefore, this section will outline the MEV market in 2022.

(1) Increase in MEV Trading Opportunities Throughout the Year

In 2022, the number of MEV transactions increased over time. Arbitrage bots were the most active, accounting for 68.3% of the market, while sandwich attacks accounted for about 30.6%. The opportunities for liquidation transactions were far less than the other two types of MEV, although their statistical scope is from May to December, this difference does not significantly affect the overall picture. Therefore, compared to traditional arbitrage trading, liquidation opportunities are more likely to depend on severe market volatility.

At the same time, the total number of MEV transactions per month showed an upward trend, reflecting a steady increase in opportunities for MEV seekers in the market.

(2) Market Share and Trends of Different MEV Types

Seekers can generate MEV income from three different types: arbitrage profits from narrowing price differences between markets, sandwich attack profits from front-running user transactions, and liquidation profits from exploiting the differences between debt and collateral values during market volatility.

Below is an overview of the income, costs, and profits of MEV transactions. (Note: Liquidation data starts from May)

Mainstream MEV bots generated at least $307 million in revenue in 2022, of which MEV seekers earned a total of $145 million from arbitrage trading, accounting for over 47.5% of total revenue, and $128 million from sandwich attacks, accounting for about 41.7%. Although the revenues of the two are similar, the costs of sandwich MEV are significantly higher, and profits are lower than those from arbitrage trading. Arbitrage seekers generated $75 million in profit, while sandwich attacks only generated $26 million in profit. From May 2022, the total revenue from liquidation transactions was $33 million, with costs of $16 million, resulting in liquidation bots generating over $16 million in profit from the market.

In fact, sandwich attackers paid nearly 79% of their income to block builders and validators. As shown in the chart below, compared to arbitrage trading, the "Cost Revenue Ratio" for sandwich trading is higher, averaging 0.78 for sandwich trading and 0.49 for arbitrage trading.

Overall, the income generated by arbitrage seekers was relatively high in January, May, June, and November, while the other months showed stable performance. Except for November, the income of sandwich bots showed a downward trend in the third and fourth quarters. Monthly income from liquidation transactions was significantly higher in May, June, and November, as the sharp decline in ETH in May and June, along with the CRV short squeeze event, created liquidation opportunities.

(3) MEV Bots

The chart below shows that a large number of bots generated negative profits, with 68% of arbitrage bots generating positive profits, 67% for sandwich bots, and 51% for liquidation bots.

Regarding the lifecycle of MEV bots, a significant portion are one-time bots, with long-term bots making up a smaller proportion. Seekers can use one-time bots for testing, and once a viable strategy is determined, the bot is transferred to a new address for long-term execution of that strategy. On average, the lifecycle of arbitrage bots tends to be shorter than that of sandwich and liquidation bots.

Additionally, this article ranks the top 10 most profitable bots based on their profits and their percentage of total profits among all bots. The data shows that the profits of the top 10 arbitrage bots account for about 51.3%, the top 10 sandwich bots account for 74.8%, and the top 10 liquidation bots account for as much as 90.2%.

Competition among liquidation bots is fierce, with the top two liquidation bots accounting for over 53% of profits, dominating the market. Other liquidation bots must find ways to surpass these top bots to gain market share. Interestingly, the most profitable arbitrage bot, 0xbad, executed fewer trades because it created the highest profit from a single arbitrage trade, generating over $3 million in profit through a single atomic transaction during the August Nomad cross-chain bridge exploit, indicating that strategy is crucial for MEV.

(4) MEV Before and After The Merge

In 2022, a total of $133.8 million in MEV value was allocated to miners or block builders responsible for collecting transactions and constructing blocks, with MEV values paid for arbitrage trading, sandwich attacks, and liquidation transactions amounting to $45.8 million, $73.9 million, and $14.1 million, respectively. Among these, sandwich attacks contributed more than half of the MEV to miners or block builders, serving as their primary incentive source. After the Ethereum Merge, the MEV contribution percentage from sandwich trading has also been increasing.

In the 100 days before the Ethereum Merge, the MEV income shares of Ethermine and F2Pool were 31.46% and 13.91%, respectively. After the Merge, the top two block builder addresses extracted over half of the MEV, accounting for 30.4% and 29.41%, respectively. Although changes in the off-chain auction market for block space affect the distribution of MEV, the oligopolistic structure remains unchanged.

(5) Calibration of Trading Volume Involving MEV

In 2022, the total trading volume involving sandwich bots was $287 billion, accounting for 87.5%, significantly higher than the arbitrage trading volume. Similarly, in the total monthly MEV trading volume, sandwich trading consistently dominated. Perhaps due to the collapse of FTX leading to a large influx of users into DEX to exchange assets, the growth of sandwich trading in November was particularly notable.

Uniswap V3 is the most popular protocol for both arbitrage bots and sandwich bots, with a total trading volume of $72.8 billion. Compared to the arbitrage trading volume, the sandwich trading volume on Uniswap V3 is an order of magnitude higher. For protocols like DODO and Uniswap V3, sandwich trading volume accounts for nearly half of the total trading volume, which is worth considering for users.

(6) Comparison of MEV Markets on Ethereum and BNB Chain

Regarding different blockchain platforms, MEV opportunities on BNB Chain are much more cost-effective than on Ethereum, as the average cost revenue ratio (CRR) for arbitrage bots on BNB Chain is significantly lower than that on Ethereum. In 2022, arbitrage seekers generated a total income of $9.5 million from BNB Chain, lower than Ethereum's $14.5 million in arbitrage income, but the total profit generated by BNB Chain seekers was $9 million, higher than Ethereum's $7.5 million.

Compared to Ethereum, competition among arbitrage bots on BNB Chain is not fierce. The profits generated by the top 10 arbitrage bots on BNB Chain account for 25%, lower than Ethereum's 51.3%. Therefore, arbitrage bots on BNB Chain have fairer opportunities than those on Ethereum. Newcomers are more likely to find a friendly environment to explore BNB Chain.

3. Current Status and Outlook of the MEV Market

Based on the above data analysis, this section will reveal the current major trends of MEV and new opportunities for the future.

(1) Current Status: MEV Infrastructure

The most critical infrastructure of the MEV market is the auction market, which ensures that MEV bot transactions are packaged on the blockchain. To achieve a fairer MEV market (at least this is the vision of Flashbots), this auction market has begun to evolve from the initial dark forest (Mempool) to an off-chain professional auction market represented by Flashbots' MEV-geth.

After the Merge, the Ethereum Foundation began implementing the pre-PBS scheme. In the PBS scheme, block builders like Flashbots and builder0x69 bid for the right to build the next block, while validators propose the highest-bid block. This method effectively segments the process of packaging MEV transactions into two auction markets: one opened by block builders for MEV bots to bid for block space, and another opened by validators for builders to bid for the right to construct the next block. Block builders and validators can also route through third-party relays (such as MEV-boost) to achieve data privacy and maintain fairness. The level of competition between the two auction markets will affect the income distribution of MEV bots.

The confrontation between centralization and decentralization. As shown in the chart from mevwath.info, blocks that comply with OFAC (Office of Foreign Assets Control) regulations are in fierce competition with those that do not.

The financial engineering infrastructure is often overlooked by the market. MEV bots have the potential to provide higher-end and specialized financial engineering services for the DeFi ecosystem in the future, relying on data services and financial derivatives to enhance their performance.

Flash loans are a typical financial tool favored by MEV bots.

(2) Discussion: Outlook for New Opportunities in 2023

In 2022, MEV bots increasingly extracted value from DeFi protocols, which is a strong trend. Arbitrage trading remains the dominant type in the MEV market, accounting for a major share in terms of income and profit. Moreover, arbitrage bots were the most active in trading during most months, with a smaller proportion of income paid to rent-seekers compared to sandwich bots. A comparison of the total number of independent bot addresses also indicates that the types of arbitrage bots are more diverse, suggesting that sandwich bots and liquidation bots may have other barriers.

A new type of MEV bot called JIT has shown an increasing trend in trading opportunities over the past few months. As JIT bots provide a large amount of concentrated liquidity, there seems to be a growing trend of independent exchange users benefiting from JIT activities. Data on slippage from actual exchange transactions and simulations also validates this. Compared to other types of active liquidity management strategies, JIT bots are attempting to provide liquidity in a more innovative and capital-efficient manner. It is worth considering that stakeholders such as AMM protocol designers directly provide similar functionalities, which can connect exchange users and liquidity providers in a new way, improving user experience and increasing the income of liquidity providers.

In addition to these mainstream MEV bots, another off-chain MEV market has opened in some DEXs and aggregators. Users' trading orders are directly transmitted to the routers created by these protocols in the form of dark pool orders or limit orders. Seekers can tap into these invisible orders in the memory pool under a set of auction rules and extract value by providing optimized settlement services. Payment for Order Flow seems to be a significant trend that developers are working towards, including Cowswap, 1inch Fusion model, and Flashbots' MEV-Share. Although this model generated controversy when it was adopted by Robinhood early on, there is currently no clear regulatory policy.

If users are rewarded for providing information, Payment for Order Flow is essentially a democratization of MEV. MEV democratization is a topic frequently discussed in the community. Another suggestion is to distribute a portion of the MEV income of validators to the Ethereum community through liquid staking, although this has not yet been realized given the current rates of liquid staking platforms like Lido Finance. The thriving MEV market has also prompted developers of DeFi protocols to think about how to leverage MEV to enhance customer experience and protect customer interests.

4. Notable MEV Transactions in 2022

(1) The Beauty of Arbitrage

On August 1, 2022, the arbitrage bot 0xbaDc profited $3.197 million from a single space arbitrage between two Uniswap WETH/WBTC pools, incurring a cost of $2,057, by running a large swap transaction in reverse during the Nomad cross-chain bridge exploit.

(2) Diversified Sandwich Trading

The sandwich attack model has become increasingly complex, with a sandwich attack involving 56 victims occurring in 2022, as well as sandwich attacks during liquidity addition/removal.

Sandwich Attack with 56 Transactions: On Uniswap V2, a block contained 56 transactions swapping WETH for MEME, and the sandwich bot attacked them together. This indicates that sandwich bots do not necessarily only attack a single transaction; splitting a large order into multiple smaller orders does not guarantee that the trader can escape the fate of being attacked.

Sandwich Attacks During Liquidity Addition/Removal: Sandwich bots do not necessarily target swap transactions on DEXs; adding and removing liquidity can also be attacked. In previous studies, sandwich attacks targeting liquidity addition can be seen as risk-free arbitrage based on the slippage changes before and after adding liquidity to the pool. Both the attacked liquidity providers and other liquidity providers may incur losses. Similarly, removing liquidity can also be attacked, and liquidity providers may face similar losses. Therefore, liquidity providers should carefully set checkpoints and be aware that sandwich attacks can profit from their losses.

(3) The Dilemma of Leveraged Sandwiches

In some skilled sandwiches, flash loans provide the necessary funding for leverage, allowing traders to invest only a small portion of funds into front-run transactions, thereby pushing up the victim's transaction price and extending potential gains. However, the downside is that the risks for sandwich bots also increase. Particularly, if the front-run transaction is successfully packaged while the corresponding back-run transaction fails, the sandwich bot will incur losses. Sandwich bots using this strategy may need to pay higher MEV costs to prioritize transaction processing.

(4) Long-Tail MEV Transactions

MEV types may not be limited to arbitrage, liquidation, sandwich, or JIT; multiple attacks in 2022 can also be classified as long-tail MEV.

In the Abracadabra Arbitrage incident, the attacker exploited the protocol's price caching design and utilized the protocol's price oracle update mechanism to complete the arbitrage. In the Mongo Squeeze and CRV short-squeeze incidents, attackers manipulated market liquidity, creating bad debts on the corresponding lending vaults. In the Ankr Attack incident, the attacker used a stolen private key from Ankr deployers to exploit the token issuance mechanism and successfully profited.

At the same time, the large flow of funds accompanying transactions often provides opportunities for traditional MEV bots, while large swaps within AMMs can also present significant arbitrage opportunities. As mentioned earlier, months involving significant risk events often accompany higher overall arbitrage profits. Additionally, the massive price changes related to liquidity manipulation of CRV shorts provide more opportunities for liquidation bots. For example, during the CRV short squeeze event, a liquidation bot implemented a profitable liquidation strategy, earning over $1 million during the sharp price fluctuations of CRV.

5. MEV Bots and the DeFi Ecosystem

Since the concept of MEV became popular, most people's first impression of it has been negative. MEV incentivizes miners to engage in rent-seeking to meet the demand for packaged transactions, and the PGA competition among MEV bots has raised the average transaction fee levels on Ethereum. More importantly, MEV may also threaten the security and decentralization of the blockchain protocols themselves. MEV bots can extract value by monitoring pending transactions and executing front-run or sandwich attacks, which are negative externalities that many protocol developers are trying to mitigate.

However, the fact that must be faced is that due to the nature of blockchain protocols in achieving openness and transparency, MEV is fundamental. From the perspective of DeFi protocols, MEV bots can also provide various values. The most common arbitrage bots detect price differences between DEXs or CEXs and profit by buying low and selling high. Their calculations of price deviations in liquidity pools enhance the efficiency of market price discovery. Liquidation bots monitor the health ratios of loans and initiate liquidations when collateral values plummet, aiding the deleveraging process, especially during market volatility.

In the future, just like the JIT bots and statistical arbitrage market makers that have emerged in this market, more different bots will appear, proficient in calculations, profiting from data analysis and financial engineering, and becoming an essential part of the smooth operation of the DeFi system.

On the other hand, the interaction between MEV bots and other participants in DeFi protocols has also become more complex and important. As revealed by alternative sandwich bots and JIT bots, the victims of sandwich bots are not necessarily just exchange users, but also liquidity providers. Due to the presence of sandwich bots, exchange users may not necessarily incur more slippage losses but may benefit from the slippage discounts brought by JIT bots.

In 2022, EigenPhi conducted an in-depth analysis of the impact of MEV on the two major DEX protocols, Uniswap V3 and Curve. On Uniswap V3, the income scale of mainstream MEV bots accounted for 25% of the income of liquidity providers. On Curve, MEV trading volume accounted for 20% of the total trading volume on most days.

According to EigenPhi's analysis, in 2022, arbitrage and sandwich bots contributed a total trading volume of $328 billion to DEXs, accounting for about 49% of the total trading volume of $666 billion generated on DEXs that year.

In summary, we can see that MEV bots have become an indispensable part of the DeFi community. Understanding the trading relationships between MEV bots and other entities can help stakeholders better understand the long-term impact of MEV on AMM.