The State of Ethereum L2s
Everything happening in the race to scale Ethereum
Welcome to the alpha please newsletter. We curate alpha for you. That’s it; that’s the newsletter.
This newsletter is going to be about Ethereum Layer 2s. There is so much happening across L2s that I can barely keep up.
I’ve done my best to sum up the activity and how the landscape is looking currently. It's a high level snapshot. I am inevitably going to miss a ton of stuff.
Every major rollup will have their own token, and I’ll provide some resources at the end for how to receive any potential airdrops.
Today’s newsletter is brought to you by Artemis.
Artemis is building a suite of products that makes crypto data more accessible for investors and builders. Many of the charts featured in this newsletter have come from Artemis dashboards. They are one of my go to destinations for analysing L1s and L2s.
What’s to come?
• A year in retrospect
• The Optimistic Ones
• The ZK Ones
• The Polygon Ones
• L2 value capture
• What’s the play?
A year in retrospect
It’s been more than a year on from the launch of the major L2s. The chains have matured and ecosystems and communities have clearly formed.
The rollup centric future remains in tact. I posted this thread almost a year ago, which was coincidentally just before the market started its descent to the depths of goblin town.
I’m going to pull out a few references from that thread to show some of the progress made by L2s.
Fees were relatively high across the major ORUs.
The fees situation is very different today and reflects the progress all L2 teams have made.
In terms of TVL marketshare, L2s have moved up the rankings in the last year.
This was November 2021.
This is October 2022.
We now have two L2s in position six and seven. L2s have taken marketshare away from various alt-L1s.
And the two major ORUs are holding up better than most during this bear market.
Over the course of the last year we have a respectable numbers of users that have bridged to L2s. These are lifetime stats for Arbitrum, Optimism, zkSync and StarkNet.
Metis DAO also emerged as a community owned rollup. They forked Optimism and use METIS is the underlying gas token. An ecosystem is starting to shape up on this L2, although it remains to be seen if it can progress beyond being home for degens speculating on low caps. I’m not going to talk about Metis beyond this, as there is just too much going on in the L2 space, but it is worth checking out.
However, despite the technological progress, if we are being critical and look at the two most dominant rollups by activity compared to Ethereum, both L2s still have a way to go. Despite the lower fees and speed offered by Arbitrum and Optimism haven’t overtaken Ethereum in daily actives or daily transactions.
That was a high level data snapshot of where L2s are right now in terms of adoption. We can definitely see progress has been made, but I’m going to look in more granular detail now.
The Optimistic Ones
Outside of application specific rollups, optimistic rollups were the first general purpose rollups to see large numbers of users bridging funds from L1. They were the first real opportunity for people to have the security guarantees of Ethereum without the high gas fees.
Arbitrum has shipped their flagship upgrade: Nitro. They have gained the most adoption of any general purpose rollup so far. And despite the bear market they have seen their daily transactions 4x and daily uniques double.
Here are the rankings for the projects with the most transactions on Arbitrum.
Two other significant things to note in respect to Arbitrum:
• Arbitrum Nova is an addition to the suite of products offered by Arbitrum. A new chain which is an ideal fit for gaming and social applications. Reddit’s community points will live on Arbitrum Nova.
• They have acquired Prsym Labs. Prysmatic Labs developed and still operates the computer software Prysm, which over 43% of Ethereum nodes use to run the Ethereum network.
Stephen Goldfeder and Preston Van Loon spoke about the acquisition here:
I would suspect Arbitrum acquired Prysm because the future sequencer can then be built similarly to Prysm, making it easy for the majority of Ethereum validators to run it.
If validators can run the Prysm client on a single device then they are quite likely to run Ethereum and Arbitrum sequencers using the same hardware (and earn the Arbitrum token).
And Optimism?
It’s clear to see that Optimism’s native token OP has had the intended effect, with daily transactions also up and to the right.
There are now over 150,000 OP holders. OP is being distributed in many different ways, but it’s mainly being used to support the ecosystem growth right now.
Perpetual Protocol is seeing the most transactions on Optimism right now, but Uniswap and Velodrome are seeing the most unique addresses interact with them.
The other interesting metric for both Optimism and Arbitrum has been their developer adoption over the last year. These stats are taken from Alchemy’s recent web3 report:
Optimism:
• >1000% growth in active developer teams YTD
• >1000% growth in active developer teams Y/Y
• 460% growth in API consumption YTD
Arbitrum:
• 516% growth in active developer teams YTD
• 795% growth in active developer teams Y/Y
• 121% growth in API consumption YTD
The ecosystem of dapps on both Arbitrum and Optimism has increased a fair bit since their launch. There are over 280 projects listed on Arbitrum’s ecosystem website and over 200 on optimism’s.
GMX has emerged as one of the top fee earning dapps in all of crypto, demonstrating L2 native dapps can capture value.
The harder to quantify metric is the strength of both communities. I think Arbitrum has managed to grow the largest cult of all the L2s so far. Looking at the size of the community and engagement on Twitter, and the fact that certain native dapps (GMX and Magic) have their own die hard fans, it seems Arbitrum have a powerful base to grow from.
I think Optimism are also doing a good job with community initiatives. I think their recent launch of quests to earn NFTs has been well executed. And the design decision of the non-token Citizen House was a great governance move.
The OP Stack and the Superchain
Optimism have just announced another major upgrade is dropping soon: Bedrock. And they introduced the “OP Stack” in a blog post a few days ago.
The OP Stack is a modular, open-source blueprint for highly scalable, highly interoperable blockchains of all kinds. Not just rollup. Not just optimistic.
The OP Stack will allow developers to build modular blockchains. It’s a series of modules, with each component implementing a specific layer of the stack: Consensus, execution and settlement. Theoretically you could swap out Ethereum’s DA layer for Celestia and run Bitcoin as the execution layer.
One team has already built something kind of cool using the OP Stack. OPCraft runs on an OP Stack custom rollup.
“OP-Chains” will be highly compatible and will be able to leverage atomic cross chain composability if an OP chain decides to share a Sequencer set. To the end user, this should all just feel like one logical chain: The Superchain. This will be fun to see play out.
This is a phenomenal talk on the OP Stack.
Last word on the ORUs. Both Arbitrum and Optimism have a path to use validity proofs and can evolve to remain competitive if they need to. This is worth remembering.
The ZK Ones
Zero-Knowledge Rollups have come on tremendously in the last year. I’m not going to explain zero knowledge proofs in this piece, but click here if you want a great list of resources to help with your understanding of this technology.
During ETHSeoul in August this year, Ethereum co-founder Vitalik Buterin projected that ZK-Rollups would win the Layer 2 scaling warfare against Optimistic Rollups.
That image is a simplified ecosystem. There are over 100 dapps building on ZkSync alone.
zkSync
zkSync has managed to garner a healthy number of unique addresses.
My guess is that this is largely a result of Argent.
I am talking about Argent’s mobile app which allows anyone to create a zkSync wallet from their phone, use DeFi quickly, cheaply, with the security guarantees of Ethereum whilst navigating around with great UI/UX. All the complexity that crypto natives are used to is abstracted away into the background.
Account abstraction is going to change the way people onboard into crypto and transact. Both StarkNet and zkSync support it. The below thread explains the power of AA. TLDR: Better UX. Greater security. More flexibility, and much more. Seed phrases will be a thing of the past.
zkEVM
The zkEVM was once thought to be years away, but zkSync are now on the cusp of launching zkSync 2.0. They say it will be the first fully working end-to-end zkEVM.
The zkEVM is the first zero-knowledge scaling solution that’s fully compatible with Ethereum. It makes it super easy for developers to migrate their dApps from chains compatible with the EVM. Developers can simply deploy their existing smart contracts without making any additional changes.
Yup, big deal. This thread outlines zkSync’s vision.
I do think we should temper some expectations. It’s actually expected to launch in alpha fully at the end of the year. No users will be able to externally bridge until then. Steven Goldfeder, co-founder of Arbitrum, shared his skepticism of the timeline that is being portrayed.
We have two other teams in the zkEVM race. Scroll and Polygon.
Scroll is the project I am the least familiar with. I don’t think they have spent much time marketing themselves.
They have just released an upgraded version of the Scroll Pre-Alpha Testnet. Their previous version onboarded over 10k users to test their bridge and demo forks of dapps like Uniswap. Their new testnet upgrade enables smart contract deployment on Scroll, allowing developers to write and deploy their own contracts on Scroll using the same Ethereum developer tools they are familiar with. Their alpha testnet will come next.
StarkNet
If you want a StarkNet primer, I wrote a bit about it a while ago.
I’m going to quote Vitalik again:
"Starkware is made up of very smart cryptographers who are actually sane".
Every time I hear Eli and Uri (StarkWare founders) talk I become more convinced that their vision has a great chance of becoming reality.
StarkNet has been in alpha for while. Alpha 0.10.1 of StarkNet has just been successfully released on the Goerli testnet. They are still restricting the amount that can be bridged to StarkNet while they focus on improving performance.
I don’t think the StarkNet metrics matter too much right now, as it’s still in Alpha.
But, the metrics that showcase what is possible with StarkWare’s technology is the scale achieved by the StarkEx prover and verifier. Over 70 million NFTs have been minted across Sorare and ImmutableX. That is orders of magnitude more than the combined total of NFTs minted across other L1s and L2s. This scale hasn’t been matched by any other system. The combined StarkWare (StarkEX and StarkNet) transactions last week were 1.6x Ethereum.
Cairo 1.0 is a big release, which will take place at the end of 2022. Cairo 1.0 has been “rebuilt for a decentralised network.” It is designed to support StarkNet’s requirements as a permissionless network, allowing the protocol to become simpler and safer.
StarkNet Alpha is now progressing towards Regenesis. The main motivation for Regenesis is capitalising on the new possibilities brought about by Cairo 1.0 — namely sequencers DOS protection, censorship resistance and gas metering, which are essential for StarkNet as a decentralised network. After Regenesis, StarkNet will work only with Cairo 1.0-based contracts, and will start from a new genesis block but with the existing state.
Some have been skeptical about whether developers would commit to learning Cairo in order to build on StarkNet.
Interestingly, AAVE and Maker are currently building on StarkNet and neither are transpiling their solidity code into Cairo. They are starting from a clean slate, which shows they believe Cairo offers their protocol enhanced performance.
The StarkWare founders thesis is essentially the majority of L2 projects will be written in native L2 code in two years time.
They have also announced the STARK token:
StarkNet’s decentralization involves a native token and a new foundation.
The StarkNet token is used for governance and as the network’s payment and staking asset.
Ten billion tokens have been minted and their allocation has begun.
I am genuinely excited to see StarkNet take off the shackles in 2023.
Last bit of alpha on StarkNet:
A quick word on ImmutableX.
ImmutableX is an application-specific StarkEx zk-rollup designed to go after gaming. Immutable already offers incredible scale, which is needed to support games (9k tps). One of the blockchain games I’ve been following for a long time is Illuvium and I think it was a big vote of confidence for ImmutableX that they decided to make that the home of their games.
Aztec
OFAC has done a lot of damage to privacy projects recently. A lot of crypto natives are quite concerned about this. Aztec offers a glimmer of hope.
Aztec is a zkRollup that allows users to interact with protocols on Layer 1 Ethereum privately.
Projects do not need to re-factor codebase to re-deploy on a rollup. They don't even need to use a transpiler to deploy onto the zkRollup. Aztec basically acts like a VPN and allows users to use protocols natively on L1. If you want to dig deeper, Jason has put together this great thread.
The Polygon Ones
I feel like I had to just put Polygon in its own section. Polygon has evolved to become a suite of its own Ethereum scaling tools. You can see that they have a number of L2 solutions, that have mostly been independent teams/projects that were acquired.
Or put another way.
Before the proliferation of L2s, Polygon PoS was called an L2, but now defined as an EVM sidechain.
The Polygon zkEVM is certainly in the running to become the first zkEVM live in production.
As you would expect from a founder, Mihailo is fully backing the Polygon zkEVM as the most comprehensive solution.
Polygon recently announced that their zkEVM is now the first zkEVM to reach a major milestone: We are entering our testnet with a complete, open-source ZK proving system.
It’s worth noting the testnet version of Polygon zkEVM has limited throughput capacity, which means it’s far from its final form as an optimized scaling machine. It still has a long way to go.
If you want to learn more about zkEVM, Polygon have a great free introductory course.
One of the things that is so impressive about Polygon is their business development team. They really do manage to bring the big brands onboard. This isn’t specific to L2s, but they’ve recently partnered with Starbucks and Robinhood.
They have an impressive track record of gaining brand adoption, so I’m sure that pipeline is going to help with the adoption of their L2 solutions.
MATIC is the token that I’m presuming will be used as the fee token across their entire suite of scaling solutions. That could bring quite a bit of intrinsic value to the token.
L2 value capture
There are many predicting L2s will be a more profitable play than the alt L1 rotation trade in the coming years.
But how do the L2 tokens capture value?
OP’s revenue model is based on its sequencer processing transactions it sends to the base layer. The sequencer fee consists of two parts, the L2 execution fee, and the L1 data fee.
Currently, they are investing that revenue in public goods, in order to build the ecosystem, which in turn should bring in more users and revenue.
Optimism or any L2, could decide to distribute that revenue to stakers, but that simply won’t make sense until an L2 has matured sufficiently.
Chains that build with the OP stack could decide to use the OP token to power their chain, which creates more demand for the token.
STARK’s token will be used as a fee token instead of ETH. Although Starkware say that users will also be able to pay in ETH “for good UX”. This means ETH will likely get converted automatically into STARK via a DEX.
ImmutableX already require 20% of the protocol fee on every transaction to be paid in IMX. If you do not currently own IMX, they convert the fee by purchasing IMX on the open market (preserving a seamless user experience). These fees are then distributed to stakers.
MEV is going to be another way L2 tokens can capture value. L2s could see more sophisticated MEV and coordination by validators and sequencers. MEV Auctions sell the power of sorting transactions, winners gain the sorting power of blocks. Proceeds raised by the auction can then be sent to the DAO’s public goods treasury or redistributed to tokenholders through a buyback or burn.
This is still the best piece I have seen written on the topic of L2 tokenomics.
If rollups onboard millions of users then it’s clear there are many ways for them to capture significant value.
What’s the Play?
All L2s are going to reward their community for being loyal and active participants. L2s want to get their tokens into the hands of the community.
Right now you can earn OP from using the dapps. There will be ongoing airdrops for ecosystem participation, for example OP have an ongoing quest right now as well.
In terms of other L2s, you should be exploring all of the ecosystems, bridging, playing with dapps, using test nets, being an active member of the community in discord etc. This thread is a handy starter guide:
The rollup centric roadmap will pay dividends for ETH and I expect the strongest L2s to eventually capture value. However, the distribution of the tokens will take years, just like it took years for effective distribution of Ethereum. We will see low circulating supply for a good while and you will have to consider the fully diluted valuation when assessing value.
L2 tokens will also have the shiny new status going into the next cycle. We know retail likes shiny new tokens…
And that’s your alpha.
Thanks again to today’s sponsor Artemis! ➡️ Sign up for their beta.
If you are looking to trade on-chain, we’re big fans of GMX.
Use our referral link to get a 10% rebate on fees: gmx.io/?ref=alphaplease
Not financial or tax advice. This newsletter is strictly educational and is not investment advice or a solicitation to buy or sell any assets or to make any financial decisions. Do your own research.
My days this is good stuff. Thanks for your time and effort 🙏
Amazing content, thanks!