Analyze the trading package market under EIP-1559 from the perspective of MEV
Looking at all the activities surrounding the Ethereum London hard fork this week, I noticed that people have various ideas about EIP-1559 (the main protocol change for this hard fork). Therefore, I want to try to outline my views on this EIP here, hoping to help integrate these different ideas.
- The difference between packaged market and sorted market
- First price bidding based on fixed supply iteration
- Miner income type
- Fee market model
- Impact on MEV
- Basic expenses as agreement income
If you want to see the summary of the article, you can skip directly to the summary section
In order to understand the motivation and impact of EIP-1559, we must first look at the situation of Ethereum today. In the entire blockchain mechanism, the parts that are relevant to us are those transaction pricing and packaging mechanisms, because these are the main areas that this EIP will bring about changes. In the first part, we will formalize it as a transaction packaging market and distinguish it from the recently emerging transaction sequencing market (on Flashbots).
Then we will go deep into this market, because it is a first-price auction mechanism based on a fixed supply iteration, and we will build a simple model for it.
Finally, we will dive into the various types of miner income distinguished by this model, and study under what conditions they can be regarded as MEV (miner extractable value).
The difference between packaged market and sorted market
Like most blockchains, Ethereum’s fee payment system is directly built into the protocol. First, each transaction specifies the price it is willing to pay per unit of gas. Then, when the transaction is packaged into the block, the account that sent the transaction will be charged the total cost of executing the transaction, that is, total_fee = gas_price * gas_consumed. Transaction fees are issued to miners along with user transactions to encourage them to package those transactions into blocks, thus forming a transaction packaging market.
Since a few years ago, the emergence of specialized use case patterns has made it necessary not only to package general transactions, but also to control the exact location of transactions in a block. One of the most prominent examples is the “snap-off transaction”, that is, a user who targets another user’s sending transaction and wants his own transaction to be packaged just before the user’s transaction. By setting the chain state in an accurate and specific use case (that is, pushing the trading market in the desired direction), users can use their information on the impact of subsequent transactions to profit.
Initially, these specific sorting use cases completely relied on the same built-in transaction fee mechanism, because the sorting algorithm of miners is usually predictable and therefore will be affected by some successful use cases. However, as the competition in the block space becomes more fierce, the situation soon leads to a considerable amount of transaction fees being wasted due to the transaction being placed in the wrong block location. Flashbots is built as an intermediate service between users and miners to solve this problem and create a fair market for transaction sequencing. This market is basically independent of the packaged trading market, and is basically not affected by the changes introduced by EIP-1559. Therefore, in the following we only focus on the trading package market. For the sake of brevity, we will also ignore those remaining sort-related activities that are still based on the local transaction fee mechanism.
It should also be noted that the term “miner extractable value” (MEV) refers to the value generated based on the miner’s control of the block’s construction of secondary levels (such as transaction ordering). Next, we will explore how more traditional miner income in packaged transactions can be viewed as a form of MEV. However, this is not a common usage of this technology, so it is best seen as a thought experiment limited to the context of this article.
First price bidding based on fixed supply iteration
In order to better understand the nature of the trading package market, it will be helpful to look at this relevant supply and demand curve:
The above picture shows the short-term supply and demand curve of the trading package market before 1559 when the block is full. Please note that this picture has been simplified in many ways to balance accuracy and legibility. It should also be noted that if you are not familiar with the concept of supply and demand curves, and do not understand how to interpret these charts, you may as well stop and try to really understand what you are reading (including background knowledge in the following chapters), because the rest Most of the content is based on this knowledge.
The somewhat unusual shape of this chart is the particularity of the blockchain background that I want to illustrate. The best way to think about the market is the first price auction mechanism based on fixed supply iterations. The following will explain these three attributes one by one, and combine them with the above figure:
Like most blockchains, Ethereum has a hard limit on the size of new blocks, which is set in the form of a gas cap. Therefore, the “block space supply” (that is, the number of transactions that miners are willing to pack into a block) is fixed at this upper limit. The vertical part of the supply curve in the chart represents this limit.
There is a subtlety that transaction packaging is not completely controlled by miners. Especially when increasing transactions will make the block larger as a whole, this will slow down the speed of transactions broadcast on the network, and will eventually slightly increase the risk of miners producing uncle blocks. Therefore, miners usually set a minimum gas price (the default value of the geth client is 1 Gwei) and reject transactions below this value, although this will lead to block saturation. Therefore, part of the supply curve at the bottom, slightly above 0, is a horizontal straight line.
Market demand is composed of users’ willingness to pay (reflected by their transaction gas prices) and follows a typical demand curve (marked in the chart as “reveal demand”-“true demand” see below one period). Its intersection with the supply curve represents the market clearing gas price of the transaction package (as shown by the horizontal dashed line).
Under the current system, transactions must pay the full amount of gas they specify, even if it is higher than the market clearing value. This feature is similar to the first price auction mechanism (or more generally any situation with perfect price discrimination). As a result, most users pay exorbitant fees for packaging their transactions.
(Translator’s Note: Differential pricing is a microeconomic pricing strategy, that is, the same or basically similar goods or services are sold by the same supplier at different prices in different markets.)
The last relevant aspect of the market is its iterative nature. Because a block is generated every 15 seconds or so, transactions have multiple opportunities to be packaged. If the initial price is not enough for the transaction to be packaged, they still have a chance to increase the gas price. In addition, the clearing price of the previous block can roughly show the gas price required for packaging.
For these reasons, many users will not set the gas price as the highest price they are willing to pay at the beginning. This leads to a gap between the observed (shown) demand (translator’s note: the original text here is supply, which means supply, but according to the context, the translator believes that this should be demand) and the theoretically true demand. To show this, the chart also adds a “real demand” curve, above the display demand curve.
Miner income type
An ingenious aspect of the supply/demand diagram is that the area between the curves naturally represents consumer surplus and producer surplus, and the division depends on the degree of difference in price. (To illustrate this point: the high under the demand curve indicates the gas price (eth/gas) that the packaged exchange pays, and the width of the supply curve indicates the block gas upper limit (in gas). Therefore, high The product of and wide represents the transaction fee denominated in eth and paid to the miner.)
(Translator’s Note: Consumer surplus refers to the purchaser’s willingness to pay minus the purchaser’s actual payment; producer surplus is the difference between the lowest price in the mind of the producer and the actual selling price.)
Our situation here is a completely different price, so the two areas under the curve (cyan and purple) represent the surplus of miners. Because this is pure miner income and conforms to the general definition of “the value obtained by miners based on the control of block production”, it is reasonable to call this the residual MEV. However, the nature of these two areas is completely different, and the reasons will be explained below.
Please note that the area at the bottom (orange) also represents the income of the miners. However, since it reflects the “production cost” of filling blocks with transactions, it is not a profit.
It should also be noted that the top (green) area does not represent any actual money paid. This is a theoretical user surplus based on the user’s lower gas price preference. The area of this part is for illustrative purposes only and will not be discussed further below.
“Basic MEV” (purple part)
This part of MEV represents the income of miners under the uniform price auction setting. If each transaction only pays the gas price cleared by the market, miners will still get this part of the profit.
When people point out that EIP-1559 does not make Ethereum transactions cheaper, they are referring to this: the purple area in the figure cannot be reduced because it is balanced in any market with fixed supply and high demand. The natural result of inside.
“First Price MEV” (cyan)
This part of the MEV represents the additional miner’s income due to the ability of miners to achieve complete differential pricing.
When people point out that the current charging market overcharges many users, they are referring to this.
After understanding the current trading package market in detail, we are now ready to move to the changes introduced by EIP-1559. We will first look at the history of this EIP (including its cause and its design), and then build a model similar to the one above. We will compare the two models to see how EIP-1559 affects the different MEV forms distinguished above. Finally, we take the payment of basic fees as the first form of agreement income, one of the side effects of this EIP.
With the current model of the expense market, it is now worth revisiting the proposal originally proposed by Vitalik, which later became EIP-1559. As Vitalik said, the motivation is to transform the first price bidding system into a unified price system (or at least to achieve this goal as much as possible). In other words, the goal is to remove or minimize the “first price MEV” part (cyan area) in our first chart.
To promote the realization of this goal, we need some kind of oracle built into the agreement for market clearing package prices. The proposed mechanism requires the minimum price (ie the base price) to be written into the agreement. At the same time, the maximum block capacity will be doubled, and a simple proportional controller is used to calculate the basic cost to achieve 50% full block. Simply put, since most of the blocks are no longer fully saturated (there are not enough transactions willing to pay the basic fee), the usage of the block capacity will be used as the current basic fee (and can make the block exactly 50 % Full “ideal level”) is an indicator of whether it is too high or too low. The basic cost will be adjusted accordingly, and its current level can be used as an ideal price prediction machine.
Please note that this mechanism is possible because the main problem encountered by the block size of Ethereum is the long-term overhead caused by the transaction (that is, the main state expansion). Although the block size also has short-term constraints (such as broadcast and verification time), these problems are not so severe, so doubling the maximum block size is acceptable. This is also the reason why chains like BSC have higher throughput. They only optimize short-term throughput and do not care about long-term effects.
With this price oracle, we can now turn the first price auction system into a unified price system by charging only the basic transaction fees. Since this will make the basic fee a possible target for miners to manipulate, it becomes impossible to directly pay the basic fee to the miners because they can add additional transactions to their blocks for free, distorting the basic fee adjustment signal (see details) Vitalik’s article). Since the basic fee is calculated through the agreement, the above situation can be easily avoided by directly charging the basic fee from the user without sending it to the miner. This effectively turns these payments into direct protocol revenue (the first ever in Ethereum!). For further discussion, see the discussion in the next section.
Although the above mechanism is almost feasible, it lacks an important part: the incentive for miners to pack blocks. Because miners do not receive basic fees, they lack the income that offsets their costs in the previous model. Therefore, the final version of EIP-1559 adds a new fee category (priority fee, or tip) to the transaction, and only this part of the fee restores the first price auction mechanism. Since the lowest tip price is quite low and well understood, this small return to the first price system has limited impact.
Fee market model
When all the elements in this EIP are available, we can now construct an updated cost market model:
The above picture shows the short-term supply and demand curve of the trading package market after 1559, when the block is not full (slightly greater than the gas target of 1/2).
Compared with the previous figure, the main changes come from the introduction of the basic fee, the doubling of the block size limit, and a new and effective gas price payment method.
The basic fee is now equivalent to the reserve price of the transaction fee, and miners can only get the part above the reserve price. Therefore, the supply curve rises due to the tip portion.
Maximum block size
As the block size has doubled, the supply curve has been widened twice as much. For most blocks, this results in insufficient levels of demand to fill the entire block, as shown by the new intersection of the horizontal part of the supply curve and the display demand curve.
In the case shown in the figure, miners can fill 2/3 of the block, which results in a slight increase in the basic cost of the next block.
Effective gas price
The last change is the introduction of effective gas price payment methods. Unlike just a gas price, the 1559-style transaction specifies a fee cap and tip (both are priced per unit of gas). Unlike paying the full fee cap, the transaction only pays the basic fee and the full tip (although there is an entire fee cap). This makes most of the effective price (effective price) curve in the graph flat, and only the slightly curved part reflects the tip auction mechanism of the first price.
Impact on MEV
Since the original intention of this EIP is to reduce one of the two types of MEV, it is very meaningful to re-examine these types and see how they are affected by these changes:
As observed in the original model, the purple “basic MEV” rectangle cannot be removed or reduced because it is a basic attribute of the limited supply market. However, it turns out that one thing we can do (in fact, we must do this for this EIP to be feasible), let the protocol capture this type of MEV, and then effectively convert it to “PEV” (protocol extractable value, the protocol can be Extract value). Although this does not reduce the user’s transaction price, it completely gets rid of a form of MEV and redistributes it directly to the agreement itself.
“First Price MEV”
The motivation behind EIP-1559 is to eliminate or at least reduce the “first price MEV”, this type of MEV comes from miners that effectively overcharge most users. Under this EIP, this type of MEV has indeed been greatly reduced, and it is now limited to the first price tip withdrawal. Previously, the largest part of MEV became user surplus, that is, the money left in the user’s wallet.
Please note that in all our considerations, we are concerned with the “normal situation” in the unsaturated zone. In some rare cases, there will be a sudden surge in demand peaks, and there will be continuous full blocks in a short but continuous period of time. During this time, the tip will effectively play the role of the previous gas price, and the “first price MEV” will absorb all the extra user surplus, which will be released through this EIP under normal circumstances. However, these extreme peak demand periods are expected to account for only a very small part of the entire blockchain activity.
Basic expenses as agreement income
Regarding the topic of EIP-1559, the last aspect worth discussing is that it created the first source of protocol revenue. For the first time in the history of the Ethereum protocol, it is not only expenditures (payment of block/staking rewards by issuing new ETH), but also actively generates income.
It is worth noting that this is a somewhat unusual behavior, so there is no strong social consensus on how to deal with these incomes. The decision now is to burn these ETHs to offset the additional issuance for security. This is generally in line with the “minimum circulation” principle, which is highly supported by the community. Although I personally very much agree with the choice of using these funds in this way (the reason I will expand in the next article), I think it is wrong to simply classify it under the principle of “minimum circulation”. It remains to be seen whether it will be added as a principle of Ethereum’s social consensus (“burn all protocol income”) and reach a strong level similar to the consensus of the general community.
Another possible use of this revenue is research and development spending, although I agree with Vitalik’s view that for reasons of recognition, the Ethereum base layer should abandon this discretionary spending (discretionary spending).
The third option worth exploring is additional security spending. This will require some kind of fee smoothing mechanism to avoid the problems surrounding miner manipulation pointed out above. Most importantly, a mechanism like this will reduce the instability of incentives created by purely fee-based funding methods that guarantee security. In my opinion, this is not suitable for blockchains like Ethereum, because it has already targeted (and paid for) a specific level of security. But it would be interesting for blockchains like Bitcoin to adopt this option, because Bitcoin aims at zero inflation and can accept a variable security level as a result trade-off. In this case, a mechanism like EIP-1559 can not only solve the problem of incentive instability, but also bring all the benefits of removing the first-price MEV and the price oracle in an agreement. Once Ethereum finally proves the universal feasibility of this EIP, I will not be surprised if this mechanism appears on Bitcoin.
So far, transaction packaging and transaction sequencing are two almost separate and independent markets. Although MEV is a term in the field of transaction sequencing, it can also be applied to the analysis of simple transaction packaging. Before EIP-1559, the transaction package included two forms of MEV-“basic MEV” (withdrawable fees due to the limited supply of block capacity) and “first price MEV” (available due to differential pricing). Extract value).
EIP-1559 achieves its goal of drastically reducing the “first price MEV” by introducing basic fees and using them to charge only the effective gas price. The remaining MEV is the result of the need for a separate tipping mechanism.
As a side effect of this EIP, “basic MEV” is captured by the agreement and converted into agreement income. This is the first time that the Ethereum protocol has generated revenue. Although burning these withdrawal fees is a reasonable option for Ethereum, other blockchains may choose other uses. In particular, Bitcoin can use a mechanism like EIP-1559, but pay the income to the miners through the fee smoothing mechanism.
As a minor side effect, this agreement now also discloses a price oracle (to be precise, including tipping) for the lowest level of package fees. This oracle can be accessed on the chain via the newly added BASEFEE opcode, or via any client off the chain. This will promote the emergence of new use cases and improve the user experience of existing Ethereum-related applications.
Posted by:CoinYuppie，Reprinted with attribution to:https://coinyuppie.com/analyze-the-trading-package-market-under-eip-1559-from-the-perspective-of-mev/
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.