3
@dankrad and @CamiRusso debate Ethereum's fee model and L2 ecosystem challenges
3d • timdaub.eth •

The statement that Ethereum is currently making zero fees from both L1 and L2 transactions is true, and it is true that it does so because it chooses to do so.

Ethereum has consistently made several billion dollars annually in fees: in 2021 $9 billion, 2022 $4.2 billion, 2023 $2.4 billion, 2024 $2.4 billion. In 2025, extrapolating from January, February, and March's data, it only manages to generate $200 million in revenue. So it has become cheaper by a factor of 10.

Many people have said that Solana was at fault for this, but if we look at Token Terminal financial statements, we see that while Solana has successfully increased its fee accrual, for example, in 2024 they had $750 million, in 2023 only $25 million, and in 2022 only $26 million, in 2025 they are going to probably have $350 million in fees accruing to the network owners, which is still lower than Ethereum.

But if you were to argue that Solana is competing with Ethereum for fees, this would be wrong, because Ethereum has consistently made between $2B and $9 billion in fees between 2024 and 2021. So just because Solana made $750 billion in fees in 2024 alone, it doesn't compete away Ethereum's fees.

While it makes sense that the ETH price and the Solana price was probably fluctuating a lot in those periods, and this also explains the fee accrual differences in those years, the fact of the matter is, and Denkrad also shows this, that the low amount of fees accruing to Ethereum stakers is a problem that Ethereum has created for itself. And it has done so because it wants to offer cheap transactions for its users.

So Solana is really not at fault for Ethereum's low fee accrual, it's really not a zero or negative sum game in this regard, because otherwise we probably would see the fee accrual from Ethereum go down and the fee accrual of Solana go up. We are seeing this to a degree, but not in a capacity where it would be significant.

But I think Denkrad's plans are all wrong. Shorter block times, single-slot L2 interoperability and maximizing the value of Ethereum DA will not increase fee accrual in the short term.

The inputs to fee accrual is basically an increased Ethereum price, higher costs of transacting (which we don't want) or more user activity. And so scaling the L1 to make sure that integration with Ethereum remains attractive does not necessarily increase the number of users that are transacting on Ethereum.

Scaling DA to make sure that we lower the incentives for alt DA is obviously an input into making more people use Ethereum, but at the end of the day, getting more people to use Ethereum is really a task of the ecosystem.

The issue is that when we solely focus on Ethereum being a casino and a badly usable one too, then it's no wonder that when the casino isn't running because prices are depressed in Memecoins like how it happened in February 2025 or in several other periods where Memecoin trading just led to many losses, then this is far less sustainable than if we had a diverse set of use cases which made money for Ethereum that were not subject to seasonality.

I guess there are many other things that could generate more stable revenues other than Memecoin trading. Also: Solana is basically competing for Memecoin trading fees. But in general: lowering the user experience friction would also potentially onboard more people to Ethereum which would mean that more people are transacting on it and so more fees would accrue to the protocol.

It's really the task of the applications which are using Ethereum that end up generating fees for it that then make revenues go up and make the Ethereum evaluation priced more attractively.

Sources:

🥝
🔥
👀reactor
💯
🤭