Daniel PoschThere are three different Ethereums, and their fates are diverging.
Just last week, we saw ETH hit a new all-time high. Meanwhile we learned that Stripe and Circle are each launching a new chain built on Ethereum’s technology stack. However, both are “alt-L1s”, independent of Ethereum L1 and ETH.
What does this mean? Why are these big institutions embracing the tech while shunning the chain? And given that they’re doing that, why is ETH up? What is the larger trend here?
We argue that Ethereum the chain, Ethereum the asset, and Ethereum the technology are untangling. The technology has critical mass and looks like the winning stack. We predict tons of real-world economic activity running on EVM chains. Meanwhile the asset and the chain may succeed, but their fates look less certain and less coupled to each other than they once were.
Our goal in this article is not to judge, but simply to observe and understand where each of these three pieces are going.
We love the core idea behind Ethereum as a freedom technology: giving users ownership and letting builders ship great products fast without gatekeepers. We’ll examine how this core idea is affected by the decoupling.
1. Ethereum Mainnet
TLDR; stablecoins and other issuer-backed assets have unique properties that make settling to Ethereum optional rather than essential. New chains are taking advantage of this. The tech winning no longer implies that the original chain will thrive.
On Aug 12, Circle announced Arc, an EVM L1 centered on USDC. This week, Stripe and Paradigm announced that they are incubating another payments-focused, EVM-compatible L1 called Tempo.
Meanwhile, tokenized treasuries and real world assets keep expanding. RWA.xyz shows almost $300B in stablecoins and $30B in non-stablecoin RWAs on-chain.
These new assets are what’s decoupling new EVM chains from mainnet. Why?
First, issuer-backed assets don’t need mainnet for security
Mainnet remains the preferred network for high-value issuance, but its moat is thinner when distribution giants can launch EVM chains and bring large flows with them. In particular, inheriting mainnet security is less important for chains designed for issuer-backed assets, including RWAs and all major stablecoins, because those assets are fundamentally governed by their issuers. The chain remains important for machine consensus, but “layer zero” is no longer social consensus. Instead, the issuer decides when and on which chain or fork to honor redemptions.
The harbinger of this was USDT on Tron, five years ago. Payment surveys show Tron USDT to be the dominant stablecoin rail in the developing world, because trusting the issuer matters more than the validator set (CoinDesk on Artemis survey). Tron is a centralized chain, but it doesn’t matter much: when using USDT, your counterparty is Tether, not Justin Sun, regardless of chain. We are now seeing the same dynamic on a potentially much larger scale with Circle, Stripe, Robinhood and more.
Second, issuer-backed assets don’t need mainnet for bridging, either
So EVM chains are proliferating. How do you move assets between them? Crucially, assets can now teleport across chains via mint-and-burn protocols:
- USDC via CCTP: burn on chain A, mint on chain B a few seconds to minutes later, 1:1, any amount, minimal to no fees.
- LayerZero/USDT0: provides similar functionality for USDT and other assets.
The best UX for moving assets across chains is mint-and-burn. This dramatically decouples “Ethereum the technology” from mainnet. For issuer assets, you don’t need a hub-and-spoke system with native bridges from mainnet to each rollup. There is a possible future where the world standardizes on EVM-based asset settlement while barely touching L1.
At Daimo, we’ve gained a lot of experience building apps, infrastructure & contracts for cross-chain stablecoin transactions. A future where the world runs on L1-bridged assets looks unlikely to us, at least until withdrawals are much faster. Mint-and-burn is cheaper, faster, and more reliable today. For example, we recently dealt with a partial outage on Linea, a chain that stopped settling to L1 for several days. CCTP still worked, while L1-bridge-reliant protocols like Across used their governance power to halt transfers. Even when bridge-based protocols are running, they have variable fees and limited liquidity. Future tech improvements will help somewhat; in particular, using validity proofs to eliminate the 7-day withdrawal period from optimistic rollups. Even then, mint-and-burn remains the clean engineering solution: if the asset has an issuer anyway, like for USDC/CCTP, then the fewest-moving-parts bridge is via that issuer.
What would it mean for Ethereum the chain to win?
So issuer-based assets are the big growth category, and they don’t strictly need mainnet. Mainnet remains unbeatable in other ways—it’s decentralization, robustness, 10-year zero downtime track record, and more. Can it still win? What does victory look like?
We can list possible criteria:
- Blockspace and blobspace are in high demand
- Mainnet secures a growing, productive TVL of assets beyond ETH
- Mainnet settles an important part of the world’s real economic activity
Contrast this to Bitcoin. If you’d asked a Bitcoiner in 2014 “what does winning look like”, they might’ve also said that last line about settling economic activity. For Bitcoin, this did not happen at all. The big-blockers lost and even today’s small blocks are often half-empty. The Bitcoin chain succeeded in one way and one way only: as the native settlement chain of BTC, which has grown as a store of value.
Given Ethereum’s explicitly broader goals, it would not count as a win if L1 meets a similar fate, with its primary relevance being issuance and settlement of native ETH.
Will L1 succeed beyond that, fulfilling its original goals? Will it be the Schelling point venue to issue new, important assets? Will it remain the largest chain for defi? Will it become the root settlement layer for an appreciable chunk of the economy? All that could still happen, but it’s looking uncertain today.
Big issuers like Circle and Robinhood are launching their own chains. The fastest-growing defi protocol is Hyperliquid, which has its own chain. The major rollups today are externally governed, loosely tied to L1, and could switch to alt-DA if blobs become expensive. All of these and more have material advantages over L1 in terms of cost, speed, UX, and (from the creator’s perspective) sovereignty.
If L1 does not grow much beyond being the native chain of ETH, then the ultra-sound money meme is dead and future protocol revenues of L1 will remain low. Would that mean that ETH is toast? Not necessarily. That brings us to:
2. ETH the asset
TLDR; ETH the asset appears to be increasingly valued as an institutional store-of-value. This means that ETH is increasingly a bet on that dynamic, rather than a bet on on-chain activity, or on Ethereum’s tech stack.
ETH printed a new all-time high over $4,900 last Sunday, Aug 24, as spot ETH ETFs saw large inflows and custody totals climbed past ~6.4M ETH (Cointelegraph). The current bid looks institutional: allocators are increasing “blue-chip crypto” weights.
Store of Value
There are two stories you could tell. The optimistic one is that the live narrative is now monetary premium (“digital silver”) more than activity or discounted future protocol cash flows.
Notice that this narrative is almost entirely disconnected from Ethereum the chain and from Ethereum the technology. The chain could “fail” (= still reliable, but reduced to a Bitcoin-like state where its dominant use is transferring ETH) and the tech could “fail” (= superseded in real economic activity by SVM or other competitor tech stacks), and despite that ETH could still remain valuable. How?
Let’s say you’re an asset allocator. You see a world of increasingly politicized monetary policy and inflation. Following Modern Portfolio Theory, you’re always looking for uncorrelated, unprintable +EV assets. Crypto looks less scary as regulatory clarity increases and the assets gain vintage, size, and track record. As it crypto store-of-value climbs the adoption curve—from family offices, to hedge funds, to endowments and traditional large allocators—you eventually buy. You’ve been hearing about BTC and ETH for years. Other cryptoassets are either much smaller, newer, or more closely-held and therefore more exposed to risk. To paraphrase Michael Saylor, “there is no third best”.
This is a departure from earlier value narratives for Ethereum. Past cycles valued ETH based on hopes of large-scale L1 usage: “digital oil” powering transactions. ETH peaked in 2017 on a narrative around powering new kinds of capital formation, and later in 2021 on a frothy “metaverse” narrative about a new mass consumer behavior. Just as “people will pay for coffee using Lightning” did not happen for Bitcoin, the ICO story and the metaverse story did not pan out for Ethereum. Today, stablecoins and RWAs really are happening, but it’s looking like they require mostly Ethereum the technology, not ETH or the L1. Going forward, ETH looks more like a bet on monetary premium and institutional adoption and less like a bet on on-chain activity.
Value from future protocol revenue
That’s the optimistic story. The less optimistic story is that ETH really is a bet on activity—that it will eventually be marked to the discounted value of its future flows, which is to say L1 protocol revenue. That is, after all, the fate of the vast majority of assets. Durable monetary premium is extremely rare. In that world, either L1 revenue dramatically increases or ETH drops in value, or both. Because rollups can move to alt-DA if blobs get expensive, L1 has limited pricing power; so a dramatic revenue increase is unlikely.
ETH also does not look like bet on “metaverse” or other web3 narratives. Even if the metaverse idea were somehow reanimated, it would necessarily run on low-cost chains. The only way it helps ETH is if it comes with mass ownership of ETH as a kind of in-game currency. Certain teams are trying this with “content coins”, crypto social, etc, but the real-user traction numbers look disappointing.
ETH is also not a bet on stablecoins, since EVM blockspace, for those, is a commodity for the reasons discussed above. The market price for transaction fees is ~$0: even if mainnet succeeds in getting payments and RWAs to happen on full rollups, it will do so by scaling blobs fast enough to keep them near-free, thus avoiding revenue.
Better hope for monetary premium.
3. Ethereum the technology
Ethereum, the technology, is the piece that’s most clearly winning. All of those new chains with big distribution potential run or will run on EVM. The most successful apps likewise run on EVM. Hyperliquid launched HyperEVM to invite builders onto its high-throughput L1. Polymarket, the most visible consumer crypto app, runs on Polygon, another EVM chain.
Worse is better
Solidity is the JavaScript of smart contracts. Is it a platonically ideal language? No. But it gets the important parts right, it’s Good Enough, and its ubiquity gives it escape velocity. Overall, the Ethereum tech stack has the best dev tools, the most tested custody options, and by far the largest developer and auditor ecosystem. Worse is better.
The technology has stronger network effects than the chain
In addition to the dev ecosystem, Ethereum has the most smart contracts that have already been thoroughly tested in production. These can be reused and composed.
All of this adds up to a compelling package for new serious chains. You get devs, wallets, support for onchain infrastructure (Aave, Pendle, Uniswap and so on) and a whole lot of associated offchain infrastructure. Modern Ethereum wallets, from Safe to Rabby to new ones like Splits, are natively multichain.
The biggest non-EVM smart contract system is Solana, but that’s a single chain, not a deployable stack. You lose the benefits of sovereignty. Beyond that lies a cautionary land of bespoke-VM ghost chains poorly connected to the main ecosystem and to tradfi. For institutions launching chains, EVM is the clear winner.
(New chains can of course extend EVM. Hyperliquid has HyperEVM, but also a fast special-purpose order book called Hypercore. Tempo is EVM-based, but with a dedicated “payments lane” for cheap transfers. Regardless, EVM compatibility is key.)
Not done yet
There are still areas where “Ethereum, the technology” needs improvement. For example, we need chain-specific addresses. A very common way for users to lose money is to send to the correct address on the wrong chain. We are also overdue for EIP-7708 to make native transfers emit logs. Until then, most wallets will fail to display ETH transfers correctly, which is an odd problem for Ethereum to have. In short, there are a few remaining user and developer-experience paper cuts; those are being fixed.
Flood warning
But fundamentally, the tech is ready. Live players with big distribution and a high bar for product quality recognize this. They will likely ship products that increase EVM-based DAU by orders of magnitude in the next few years. This is especially true if you count DAU conservatively—not daily addresses, but daily non-token-incentivized real organic users. By this standard, EVM has (very roughly) a million DAU today. A single successful fintech integration can 10x it.
We are shifting from “build it and they will come” to “ready or not, here they come”.
Conclusion
All three Ethereums can thrive, but they’re pulling apart. The fastest-growing segment is stablecoins and real-world assets coursing through Ethereum technology, without relying strongly on mainnet or ETH the asset.
Meanwhile, ETH the asset is also increasingly abstracted from mainnet, both in form (an increasing percentage is in ETFs or other big custodial pools, transacting offchain) and in narrative (monetary premium, not on-chain activity).
Ethereum, the idea
The core idea of Ethereum is about freedom. We see both good and bad news. The good news is that Ethereum, the technology will materially increase freedom for lots more people soon. Stablecoins on fast chains enable “WhatsApp for money”: UX-optimized apps that regular people enjoy, giving them free and instant borderless rails. Cross-border businesses benefit from instant settlement. And EVM chains, even the corporate ones, allow permissionless interop: a small team in Argentina or anywhere can ship an app that seamlessly runs on any of the chains we’ve discussed, without doing any BD deals or hiring licensing lawyers first. This translates to more freedom for entrepreneurs and other builders, more competition for financial incumbents, and better apps for users.
The bad news is increased custodial & governance points of failure. Imagine a transfer of USDC (issued by Circle, a US company) on Arc (governed by Circle) from an email-based embedded wallet (keys stored on Turnkey servers) to a Lemon account (fully custodial Argentine Venmo, built on stablecoin rails). The purist weeps! The silver lining is transparency and competition. We keep these counterparties honest if user-hostile behavior is publicly visible and users have sufficient ability to exit in practice. Another silver lining is that empirically, this scenario results in far less lost user funds than traditional seed-phrase wallets. We can further improve with better engineering. Instead of a custodial account, you could use Porto to provision a passkey account, or otherwise tap into the OS-native cloud backup system.
Finally, the great news is that, ten years in, the tech is finally positioned to power great products for a large audience. Whether these products are incremental upgrades—”PayPal with extra steps”—or truly great depends on whether they embody the conviction and freedom ethos of Ethereum, the idea.
Thanks to Liam Horne, Andrew Liu, Gianluca Minoprio, Yi Sun, Will Minshew, Georgios Konstantopoulos and Dankrad Feist for feedback and edits.