The current Ethereum Gas auction model is difficult to achieve effective resource allocation. EIP-1559 is the solution proposed to solve the inefficiency of Ethereum Gas transactions.
Recommended reading: ” Understand the Ethereum EIP-1559 proposal, why is it controversial? 》
Authors: Li Lianxuan and Chen Yuxuan, working at Huobi Research Institute
The main goal of the Ethereum EIP-1559 protocol is to improve the efficiency of gas fee auctions. Previously, Ethereum adopted the one-price sealed auction model, but due to the information asymmetry in the market, it caused the problems of low resource allocation and complicated quotation strategies under Bayesian games. For this reason, EIP-1559 introduces the fixed fee Base Fee that must be paid for each transaction, and formulates it in-line, thereby improving the problem of information asymmetry and increasing the efficiency of auctions; at the same time, in order to prevent miner fraud and abstract economic problems, EIP-1559 destroys the Base Fee part and introduces TIP fees quoted by users independently to incentivize miners to package transactions. Of course, the Base Fee destruction mechanism inevitably affects the interests of the miner group, but it is good for the ETH holders. In addition, EIP-1559 cannot fundamentally solve the problem of high transaction costs.
On the evening of August 5th, Beijing time, Ethereum will officially usher in the London hard fork. Among the contents of this upgrade, the EIP-1559 proposal has received extensive attention from the industry. However, when it comes to EIP-1559, the Chinese community’s awareness of it is obviously insufficient. For example, many readers mistakenly believe that the EIP-1559 proposal is to reduce transaction fees on the chain or to achieve deflation of ETH. In the endless controversy, the true motive of EIP-1559 has gradually been buried by the noise. For this reason, on the eve of the upcoming upgrade of EIP-1559, it is necessary to lead readers to re-understand EIP-1559 and understand its motivation, principles and influence.
Background and motivation of EIP-1559
I believe that most readers understand that EIP-1559 mainly modifies the method of collecting transaction fees on Ethereum. The transaction mechanism of Ethereum is essentially an auction. There are currently four popular auction methods, namely British auction (increased price auction), Dutch auction (reduced price auction), sealed one-price auction and sealed two-price auction, as shown in the table below. Among them, Ethereum adopts a sealed one-price auction model: users put forward fee quotations under the premise of mutual confidentiality of bids, and miners preferentially pack higher bid transactions into blocks to obtain transaction fees.
The biggest feature of auctions is that the price is determined by the way of competition, not the seller has the final say, nor is it done by the buyer and the seller in bargaining. The superiority of price determination by competition stems from asymmetric information (it’s important to understand this, and we can see later that EIP-1559 is just to optimize this problem). The seller does not fully know the true price that a potential buyer is willing to give. A potential buyer does not know the bids of other buyers. Of course, for the special economic mechanism design of auction, there are two core goals that need to be completed in any case:
Pareto Optimal: The auction can realize the effective allocation of resources, which means that the auction item should be sold to the users who have the highest evaluation of it, because the person with the highest evaluation is often the person who needs the auction item the most;
Profit maximization: The auction allows the seller to obtain the maximum expected profit
Under the assumptions of traditional economics, the sealed one-price auction model can achieve Pareto optimal (that is, high transaction efficiency). However, the reality of Ethereum has exceeded the assumptions of traditional economics, making the current Gas fee auction mechanism deeply criticized by the market, which is mainly reflected in the following two points:
(1) Information asymmetry in the Ethereum Gas auction, resulting in inefficient allocation of resources and low transaction efficiency
Under the sealed one-price auction rule, every buyer will have the following considerations: if his true evaluation of the item is 100 US dollars, then he will never bid 100 US dollars, because in doing so his profit is 0; he will not bid Above $100, it means he will lose money. Therefore, he had better bid less than 100 yuan. However, at this time he faces the following trade-offs: the higher the bid, the greater the winning side, but the less profit after winning; the lower the bid, the smaller the winning side, but the greater the profit after winning. Therefore, he must make a trade-off between the size of the winning side and the size of the profit after winning to find an optimal bidding strategy.
The premise of the sealed one-price auction to achieve Pareto optimal is that the reserve price is equal to the true evaluation. But Ethereum’s Gas auction mechanism failed to achieve this, mainly due to the asymmetry of information. A typical case: Consider that there are only two users, and the real evaluations of Gas are $50 and $30 respectively. If user A with a higher evaluation (50 US dollars) mistakenly believes that another user B’s true evaluation is 10 US dollars, then user A will bid 20 US dollars to win the goods; however, user B bids 30 US dollars, resulting in User B, who rated the highest gas fee, did not win the product, which obviously did not realize the effective allocation of resources, resulting in low transaction efficiency.
(2) The sealed one-price auction model complicates the bidding strategy and increases the bidding cost of users
In the Chinese community, many interpretation articles mention EIP-1559 in order to improve the user experience. But it didn’t explain why. From the perspective of game theory, the sealed one-price auction mechanism is a typical static Bayesian game, that is, the user’s probability of winning the auction depends not only on his own bid, but also on the bidding strategies of other buyers, resulting in users not having an advantage. Strategies (the so-called dominant strategy, that is, no matter how the competitors react, they have an optimal strategy, similar to “changing with the same”) can be selected, which complicates the bidding strategy.
Here, Ethereum’s Gas fee auction mechanism can be seen as a simplified eagle-dove game model, that is, a single static Bayesian game under the condition of incomplete information: Suppose that user A and another user conduct Gas fee auction at the same time. But I don’t know whether the other party provided a higher or lower price than my own. Suppose that the probability that the opponent is a user with a high reservation price (M-type user) is m, and the probability that the opponent is a user with a low reservation price (N-type user) is 1-m. Then the game strategy and profit matrix of the two parties are as follows. It can be seen from the figure that any user faces four situations when making a quotation. You must first speculate whether others are users with high reserve prices or low reserve prices, and then you must evaluate whether you should quote a low price or a high price. In each case, there are different coping strategies and benefits, and there is no dominant strategy at all.
Of course, we can also use a simple mathematical formula to describe the static Bayesian game: there are n buyers, the private value of buyer i is vi, his strategy is to bid bi, and the strategy range is in the interval [0, + ∞). His profit depends on his own bid bi, and it also depends on other people’s bid bj. We can express the expected return function of buyer i as:
U_i=(v_i-b_i)Pr(b_i≥b_j,j≠i)
Where Pr(.) is the highest probability of b_i among all bids. After a series of complex derivations, we can get the equilibrium price of the Bayesian game:
Where v^((1)) and v^((2)) are the highest and second highest values of all buyers’ private values (v1, v2, …, vn) respectively.
From this, it can be seen that the user’s final equilibrium bid under the Bayesian game is not only related to the highest and second highest values of all buyers’ private values, but also related to the number of people participating in the auction and the value distribution function F(). Therefore, Ethereum’s Gas auction mechanism increases the complexity of the bidding strategy and reduces the user’s auction experience.
In general, the current Ethereum Gas auction model can neither realize the effective allocation of resources of the “higher real price”, which brings the difficulty of complicating the bidding strategy. EIP-1559 is the solution proposed to solve the current low efficiency of Ethereum Gas transaction.
The content and principle of EIP-1559
The EIP-1559 proposal actually includes two major blocks: one is the reform of the Gas auction mechanism, and the other is the block relaxation mechanism. Here we mainly analyze the Gas auction mechanism.
Through the analysis in the first section, we can see that the main reason for the inefficiency of Ethereum Gas auction is the information asymmetry:
As a result of information asymmetry, the reserved price is not equal to the true evaluation, and the one-price sealed auction cannot realize the effective allocation of resources;
The Bayesian game formed by information asymmetry, complicates the bidding strategy.
Let us see how EIP-1559 solves these two problems. The proposal introduces a fixed fee that must be paid for each transaction, namely the Base Fee. Specifically, the Base Fee fluctuates based on the situation of the previous block, with the goal of an average block utilization rate of 50%. When the current usage rate of a block exceeds 50%, the fixed fee will increase proportionally (the upper limit of each block is +12.5%), and when the usage rate is lower than 50%, the fee will decrease. The formula for calculating the Base Fee is specifically written in the agreement. After each block, the Base Fee will be automatically updated:
Here, b stands for Base Fee, t stands for block height, T stands for block size, the target block load is T/2, and g stands for the number of transactions contained in the block. Since g depends on the basic fee b, we will gt | bt represents the number of transactions included in Bt when the basic fee is equal to bt. D is the adjustment factor, and the current setting is d=0.125.
If we analyze carefully, we can see that because the Base Fee is embedded in the agreement and is automatically adjusted according to the current status of the block, this is essentially a price reference standard introduced by the official for the auction mechanism: if the user’s reserved price is low For Base Fee, he will choose not to bid, and wait until the Base Fee is lower than his reserve price in the future before bidding; if the user’s reserve price is higher than Base, then he will choose to offer a higher price than Base Fee. Therefore, EIP-1559 provides traders with a predictable trading quotation by moving the Base Fee into the protocol, reducing the asymmetry of information, and alleviating the inefficiency of resource allocation and the complexity of the quotation strategy to a certain extent. .
However, if EIP-1559 only has Base Fee, it will bring another serious problem: miner fraud. There are two most common ways of this kind of problem: one is that miners can raise the base fee by generating transactions by themselves (after all, the base fee is regulated by a fixed formula), thereby making themselves profit, but this seriously harms the interests of users , And it is easy to cause network congestion; the second is to benefit oneself by conspiring with users to achieve “off-chain transactions”, but this makes ETH’s position on Ethereum weakened.
Based on the above reasons, EIP-1559 designed the most controversial Base Fee destruction mechanism. Base Fee is not paid to miners but burned. This perfectly solves the above-mentioned miner fraud problem: First of all, it is difficult for miners to make a profit by generating transactions themselves, because they can’t just stuff their own transactions into a block. No fees are consumed; secondly, “off-chain transactions” will no longer be feasible in the future. It consolidates ETH as the only way to pay for the use of Ethereum, preventing economic abstraction.
However, if all transaction fees are completely destroyed, then miners will have no incentive to continue packaging transactions; more importantly, when the Ethereum network is blocked, only the Gas fee mechanism of Base Fee will not be able to achieve the optimal allocation of resources. For this reason, it is necessary to introduce a tip on the basis of Base Fee as an incentive for miners, this is Tip. Tip quotation is also a one-price sealed auction mechanism, but compared to Base Fee, which accounts for most of the Gas fee, Tip can only be counted as a small tip, so the transaction efficiency issues caused by it will not be the Gas auction mechanism. main problem.
The size of the Tip is determined by the user. When Ethereum is congested, the higher the Tip, the easier it is to pack and chain. But Tip’s quotation is not unlimited, because EIP-1599 has a price cap (Fee Cap), which represents the highest fee that users are willing to pay for the transaction to be packaged into the block. The Fee Cap is determined, and the tip’s upper limit is also determined.
At this point, the EIP-1559 auction mechanism has formally taken shape: the Gas fee is split into two parts, Base Fee and Tip, of which the Base Fee part is calculated by the protocol formula based on the block situation, and is finally destroyed; the size of the Tip part is determined by the user , And finally paid to the miner.
Of course, it should be noted that in addition to the most important auction inefficiency, EIP-1559 also solves two other inefficiencies: high fluctuations in gas fees and long user delays, both of which are essentially It is caused by the upper gas limit of the block, so it is finally relieved by the block relaxation mechanism. I won’t repeat it here for reasons of space.
Impact of EIP-1559
From the above we can see that EIP-1559 is a new solution proposed to solve the current low efficiency of Ethereum Gas transactions. Of course, due to the controversial Base Fee destruction mechanism, the existing profit distribution pattern of Ethereum has been changed to a certain extent. In reality, EIP-1559 can bring about the following effects:
direct impact:
Improve the efficiency of Ethereum Gas auctions and improve user quotation experience;
Prevent miners from cheating and ensure the smooth progress of transactions
Solve the problem of “economic abstraction” and strengthen ETH’s position in Ethereum
Other effects:
The destruction of the Base Fee leads to a decrease in the income of the miners, which seriously damages the interests of the miners;
It may lead to ETH deflation, but overall it reduces the circulation of ETH, which is good for ETH holders;
Unsolvable problem
EIP-1559 cannot solve the problem of high gas fee. From an economic point of view, the level of gas fee depends on the relationship between market supply and demand, so reducing the handling fee is fundamental to expansion. The main goal of EIP-1559 is to solve the pricing problem of the auction mechanism, but it cannot fundamentally solve the problem of high gas fees.
