Paradigm

Miners will accept EIP-1559, here is why

Feb 22, 2021 | hasu, Georgios Konstantopoulos

EIP-1559 is one of the most highly-anticipated Ethereum upgrades of all time, radically changing how users bid for transactions, among other major benefits.

EIP-1559 has overwhelming community support and is technically ready to be included in the hard fork after Berlin, pending the usual core developer evaluation process. Lately, miners have started to rally against the proposal. This is not too surprising, since the mechanism would burn some of the transaction fees miners would have previously received.

Although it may seem counterintuitive, our hypothesis is that the best strategy for miners is to support the deployment of EIP-1559.

We test this hypothesis by examining the two most effective ways for miners to protest the proposal: (1) fork Ethereum to create an altcoin without EIP-1559; and (2) block EIP-1559 on Ethereum by driving the basefee to zero.

After considering their feasibility & cost, we find that any form of aggressive protesting would damage long-term miner revenue more than cooperating with users.

Miners are structurally long ETH and the Ethereum economy

The revenue miners receive, and that EIP-1559 stands to affect, currently consists of three sources:

  • First, a block subsidy of 2 ETH per block as well as an extra reward for uncle blocks.
  • Second, fees from users who bid for inclusion in the blockspace market (irrespective of their final position in the block).
  • And third, the difficult-to-quantify but very high value miners can extract by inserting (or not inserting) transactions at specific points in the block. This is called miner-extractable value, or MEV, and most miners currently “outsource” it to frontrunning and arbitrage bots, who run bidding auctions with each other in the mempool.

After activation of EIP-1559, miners will continue to receive the same revenue from the block subsidy and MEV. The value from inclusion fees would be burned as long as the system isn’t congested (demand below the maximum gas limit). When demand exceeds the maximum gas limit, there would be an additional first-price auction between transactors, with the proceeds going to miners.

To earn these rewards, miners have to invest in mining hardware, power purchasing agreements, and other capital expenditures. This investment makes them structurally long ETH and the Ethereum economy because they have to mine for the investment to pay off.

While we don’t deny that EIP-1559 has the potential to reduce one of those three sources of revenue, miners will still have enough future revenue “at stake” to protect Ethereum and its users. Even with the entire basefee being burned, MEV and the block subsidy will still be a significant source of revenue for miners. Finally, the deployment of this upgrade could also mark a turning point in user demand for Ethereum, ultimately growing the Ethereum economy as a whole.

Users are the Ethereum economy

To understand the power dynamic in Ethereum, it is important to understand that all three sources of revenue stem from users and the applications and businesses that serve them.

Users create the demand for ETH, which miners then sell to them in exchange for fiat, as well as for other tokens in the Ethereum ecosystem. Their demand to transact, exchange, borrow, lend etc. these tokens creates congestion fees. And finally, their usage of Defi applications, e.g. decentralized exchanges, creates MEV in the form of constant price arbitrage and other opportunities for miners.

Users are the Ethereum economy. Miners provide a service to them in the form of network security. It is a transactional relationship – miners don’t provide this service out of the good of their heart, but in response to the financial incentive that users create for them.

Users have no moral (or other) obligation to pay miners more than is needed for Ethereum to be secure anymore than miners have a moral (or other) obligation to keep mining when it’s not profitable for them.

Ultimately, the power dynamic between users and miners can be explained with replaceability. It is near impossible for miners to replace the current Ethereum users as their main source of revenue. But it is very possible for users to replace some or even most of the current Ethereum miners.

Having established this basic relationship between miners and users, we will apply the framework to various scenarios of how the EIP-1559 activation could play out.

Scenario 1: Miners maintain the old chain without EIP-1559

We mention this only for the sake of completeness because in many other blockchains, upgrades face an inherent uphill battle. That is because it’s generally cheaper for users to do nothing and stay on the existing chain, and hence it becomes easier to block new proposals from going through.

This cannot happen in Ethereum due to the difficulty bomb. In short, if there is no hard fork to reset the difficulty bomb, mining difficulty will increase until Ethereum itself grinds to a halt. That makes it impossible to stay on the old chain ‒ any EIP-1559 opponents would need to incur the same cost of going through a hard fork that at least defuses the difficulty bomb.

Scenario 2: Miners create an altcoin with Ethereum’s state

A more viable proposal would be for miners to simply fork Ethereum and create their own altcoin, similar to how ETC once forked from ETH or how BCH forked from Bitcoin. Whether it can make sense to fork would depend on the opportunity cost of doing so(1). In the case of EIP-1559, miners would have to decide between mining their new altcoin chain and the existing Ethereum chain.

This opportunity cost is no joke, since—as we previously established—in order to pay miners any revenue, a blockchain first needs to create value for users for there to be a valuable block subsidy, congestion fees and MEV. Both Bitcoin and Ethereum have been forked dozens (or even hundreds) of times, but most of these forks never got any traction with users.

Building that traction is a lot easier when you can also fork a blockchain’s state, and all successful forks of the past have done just that. In Bitcoin, the state is simply a list of coin ownership. BCH forked this list to leverage Bitcoin’s existing supply distribution and airdrop new coins to all BTC holders.

But Ethereum’s state is more complex, containing not just the distribution of Ether, but also thousands of different tokens, smart contracts, applications, and so on. These would also be copied in a fork, but they would be mere skeletons on another chain.

For example, many of the largest tokens on Ethereum such as stablecoins or WBTC are claims on an asset in the real world. Duplicating the claim wouldn’t duplicate the asset. These claim tokens would continue to operate on the EIP-1559 Ethereum chain, but be worthless on the fork chain.

As a consequence, remaining Defi apps on the fork chain that rely on collateral would also unwind, for example the collateral-backed stablecoin DAI or any form of AMM pool. In short, everything other than ETH, including important off-chain infrastructure like oracles, liquidation bots, etc. would blow up and it would create a huge mess on the fork chain.

While ETC was able to fork from Ethereum in 2016, a similar event is no longer possible today. The emergence of tokenized assets and Defi has made Ethereum’s state unforkable(2).

Scenario 3: Miners create an altcoin with a fresh state

If Ethereum’s state cannot be forked, what about an altcoin that copies only the safe elements of Ethereum’s state (e.g. the distribution of ETH) or even starts from a completely fresh state?

This is more viable than scenario 2), as is proven by other “stateless” forks of Ethereum such as Tron and most recently Binance Smart Chain (BSC). Especially the success of the latter proves that there is tremendous value in leveraging Ethereum’s Virtual Machine (EVM), the existing wallet infrastructure (like Metamask), and developer tooling. Further, while dapps wouldn’t be copied automatically, they would be trivially deployed and could later be populated with new assets.

Given the rapid success of BSC, wouldn’t there possibly be market demand for a “permissionless” version of it, with PoW mining instead of a centralized operator? The new chain could even increase the gas limit, to target the same demographic of users who are currently priced out of using Ethereum due to the high gas prices.

But on further reflection, this approach is also loaded with problems, and the issue is around supply distribution.

If the new chain decided to reset the supply distribution of ETH and start from 0, it would lose the existing supply distribution. Bootstrapping a new supply distribution would require years of high inflation, making the asset unattractive to hold. BSC, in comparison, doesn’t have this issue, since Binance is the only block producer and needs no additional mining incentive.

But if the new chain, alternatively, copied the distribution of ETH, then a lot of the new ETH would be in the hands of potentially hostile users who could use it to depress the price for a long time. This would render any block reward for miners on the new chain worthless and shows that even “stateless” forks require some amount of support from existing users.

Scenario 4: Miners join the new chain but block EIP-1559 there

As we have laid out, any attempts to create an altcoin are basically doomed to fail. This leaves one more possibility, which is also the most discussed by miners. In this scenario, miners would join users on the new Ethereum blockchain, but then suppress the EIP-1559 mechanism from burning any ETH by manipulating the basefee to zero.

The method would work as follows: The EIP-1559 controller determines the next block’s basefee by observing the size of the previous block. If the previous block exceeded the target gas limit (50% of the maximum), the basefee would increase to throttle transactor demand. If it was smaller than the target gas limit, the basefee would decrease to encourage demand.

Miners can technically control how many transactions they include, and hence can control the blocksize, and hence can control the basefee. If they only ever mined blocks that are less than half-full, the basefee would never increase above zero and hence none of the fees would be burned. However, competition between different miners makes this strategy impossible in practice.

First, imagine that a single mining pool with 5% of hashpower tried to implement this strategy. They would only mine blocks that are half-full or smaller, even if demand far exceeds that level. Meanwhile, the other 95% of hashpower would mine larger blocks, make more revenue from fees, and the basefee would increase anyway. The 5% mining pool quickly would realize it is bleeding money and give up or lose all of its hashpower. This shows that self-interested miners want to include as many transactions as possible, as long as there is competition between them.

So what if there was less competition? Imagine that instead of 5%, 60% of miners would agree to implement this strategy. The outcome would be the same, since for every half-full block the 60% cartel mines, the other 40% would get to mine full blocks, and get all that extra revenue from congestion fees and MEV, and basefee would still increase over time. We call this an unstable coalition.

The strategy only works if the hostile miners could find a way to eliminate that competition, so that no one else can mine large blocks either. With 60% of hashpower they could do that by implementing a so-called miner-activated soft fork (or MASF). This MASF would dictate that blocks over half-full are now invalid, and hence the 60% of miners should simply ignore them. Now the 40% could still technically mine larger blocks, but the 60% would refuse to build on them, and so all the transactions and block rewards distributed by the minority cartel would evaporate.

Now you have to understand that MASFs are nothing new. Miners can already form such a cartel today to e.g. drive up fees by restricting the gas limit, charge higher fees from larger transactions, or set a price floor. All of these strategies seem more profitable at first, but there is good reason miners don’t attempt to implement them.

First, they require cooperation from many mutually distrusting parties, which is very difficult to achieve. But more importantly, a MASF would be an unprecedented attack on the Ethereum network and its users. It would both destabilize the network at the consensus level and disrupt the trust of users into Ethereum. This already threatens future miner revenue, but users can also oppose the censor in more active ways. For example, we would expect users would start broadcasting their transactions directly to a friendly mining pool to withhold fees and MEV from the censoring pool.

In summary, basefee manipulation is not a stable equilibrium for miners without a MASF. But if miners did implement a MASF, it would be an unprecedented and self-destructive attack on Ethereum and hence their own investment(3).

Scenario 5: Miners join the new chain and smoothly implement EIP-1559

In light of the unsatisfying outcomes for miners in scenarios 1 to 4, we are convinced their dominant option is to simply cooperate with users.

Even if miners made less money on the new chain—which is not a given—that would still be far more than they could by trying to create an altcoin. Any such altcoin would have near-zero value in relation to ETH, no transaction fees from congestion, and no MEV from Defi arbitrage opportunities.

Further, implementing a MASF to suppress the basefee would be an unprecedented and transparent attack on Ethereum and its users. We have never seen such an attack in the wild, and for good reason. It would likely disrupt user confidence and the value of ETH as well as the economic activity happening in the system, and hence goes directly against miners’ interests.

Possible concessions

On top of the five scenarios discussed above, there has also been discussion of different concessions users could make to appease miners. The main ones mentioned were:

  1. Raising the block subsidy on the new chain to compensate miners for the burning of basefee
  2. EIP-969: Changing Ethereum’s PoW algorithm to remove ASIC miners from the network
  3. Instead of burning the basefee, distribute it to the miners of the next N blocks

We however re-emphasize that it is already in the best interests of miners to cooperate with users on the upgrade. Hence users don’t need to meet the miners’ demands and make any further concessions to them.

Conclusion

This is how we expect the upcoming EIP1559 transition to play out, and we are confident in our analysis. We look forward to discussing these arguments with the community at the upcoming EIP-1559 roundtable (Friday, February 26, 2021, at 14:00 UTC).

(1) There are also other large costs involved with bringing a new blockchain to market.
(2) We strongly recommend you read Ethereum is now unforkable, thanks to DeFi, by Haseeb Qureshi. His article lays out in great detail what would happen if someone tried to fork Ethereum’s state, concluding that it is utterly impossible today.
(3) For a formalized version of this same argument, see Tim Roughgarden’s analysis Chapters 6.2 and 7

Resources & Further reading

Acknowledgments: Thanks for feedback to Tim Beiko, Pintail, mewny, and Jing Wang.

Disclaimer: This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. This post reflects the current opinions of the authors and is not made on behalf of Paradigm or its affiliates and does not necessarily reflect the opinions of Paradigm, its affiliates or individuals associated with Paradigm. The opinions reflected herein are subject to change without being updated.