Subscribe to Overcollateralized Thoughts
Share Dialog
Despite most websites being concentrated among a few major server providers, the internet remains fundamentally decentralized and fragmented in its architecture. The internet is essentially a vast network of interconnected computers that use open standards to communicate (TCP/IP, HTTP, DNS, etc…). Technically, anyone can self-host their own server and participate in the World Wide Web without any middle-men.
While that is the case, not every business or application has its own website. Many businesses and apps share the same website, while others spin up their own. The first example that comes to mind is e-commerce storefronts. While many larger stores have their own front-end, many merchants share their storefront with others on platforms like Amazon. Businesses have the liberty of launching their own storefront when it makes sense, or distribute their product on a shared platform to reach a wider audience. Many do both. This is not only true for different merchants, but for different applications.
There has been a rise in “super apps” over the years. To use Amazon as an example again, they offer several services such as faster delivery, music, video streaming and more to their Prime customers. Another popular example of this is the Chinese messaging app WeChat that has incorporated payments, shopping, loyalty programs and much more all in one app. These companies have been able to form an ecosystem of products and services on one platform.
How do all of these examples relate to value networks and blockchain? Well, the same is true with crypto networks. There’s been lots of development from infrastructure providers to offer the ability for any organization to spin up its own network (i.e. Gelato Network, Optimism Superchain, Arbitrum Orbit, and many more). This allows for the freedom of experimenting with what’s possible for their specific use case. Breakthroughs achieved by these experiments will proliferate to other networks that can benefit from these discoveries. This can be seen with the incoming Cambrian explosion of rollup development that has been exhibited in 2024 and will continue in 2025 and beyond. In the last year we’ve witnessed the launch of the first Ethereum rollup running the Solana Virtual Machine (SVM). We’re also seeing experimentation with cross-chain liquidity with the announcement of Uniswap’s Unichain, a rollup in the Optimism Superchain ecosystem looking to solve liquidity fragmentation (explained later in this article).
Right now, the majority of rollups (i.e. Layer 2) that have gained mainstream adoption have been developed for the intention of scaling the Ethereum network. This has been done by deploying permissionless rollups running a copy of the Ethereum Virtual Machine (EVM), that mimics most of the properties of Ethereum Mainnet but scaled to orders of magnitude faster than the Layer 1. General purpose permissionless rollups and based rollups are typically focused on offloading execution costs away from the Layer 1 and further cements the Layer 1 as the main network for the recognition of value and state of accounts as the Layer 2s depend on the main chain for validation and asset bridging. Because of this, people often confuse rollups as strictly scaling solutions, but they can be practical for different purposes than scaling, such as creating bespoke environments geared for specific applications (further elaborated in the next section).
The business case for running a rollup versus launching a new Layer 1 blockchain is becoming increasingly clear. While many founders have historically been drawn to launching their own Layer 1 networks, primarily because they could attract higher valuations from venture capitalists by selling an unlimited growth story, the reality is that running a rollup offers more sustainable advantages. This phenomenon will continue until VCs realize that the L1 play will saturate and the return on capital will diminish over time as the market realizes that most L1s will fail to ever gain any meaningful adoption.
The economics of running a rollup are particularly compelling. Take Coinbase's Base rollup for example, it generates steady revenue from transaction fees while leveraging Ethereum's security, avoiding the massive costs of maintaining an independent blockchain network. If you were to annualize their past 30 days of revenues, Coinbase would be generating $137.55M per year from the transaction fees generated by their rollup alone. That paired with an annualized cost of $38.64M for running the network, and you have a potentially 9-figure generating product with a 72% profit margin. (Token Terminal, Jan 5, 2025) Keep in mind that transaction fees are not the only way a company can monetize a rollup. Coinbase benefits from having its own L2 network by having a wide range of builders deploying apps on their network, attracting users to interact with other Coinbase products such as their staked Ether token (cbETH) and wrapped Bitcoin (cbBTC), along with countless other benefits to their brand and visibility.
This stands in stark contrast to many Layer 1 networks that struggle with unsustainable token emissions and minimal fee generation relative to their security costs. Let’s use Solana as an example, the most active L1 after Ethereum (in terms of DeFi and size of ecosystem). The Solana blockchain has generated $57.71M in fees in the last 30 days, while issuing $455.32M in its SOL token to pay for the network’s security. On an annualized basis, the blockchain is inflating its token by $5.46B (5.3% of its current market cap) while only generating 12.6% of that in revenues for stakers. (Token Terminal, Jan 5, 2025) Therefore, the vast majority of the staking revenue for stakers comes from token emissions rather than fees generated by transaction volume (barring MEV). All these numbers are substantially more accentuated for newer (and less adopted) L1s in the space.
With that being said, Solana is pushing the needle on many innovations in the space, and there’s a strong possibility that the network generates more in fees than issuing its token sometime in the future. Unfortunately, this is not going to be the case for most L1s that currently exist, and there’s a strong likelihood that many L1s will sunset or convert into L2s in the coming years. This has already been the case for some networks already such as Lisk and Canto.
Rollups also offer unprecedented flexibility in customizing the network environment. Banks and financial institutions are taking notice. Deutsche Bank's launching their own rollup highlights how traditional institutions can tailor their blockchain environments to meet regulatory requirements and specific business needs. Want to implement KYC gates? Need transaction reversal capabilities? These features can be built directly into the rollup while still maintaining connection to the broader ecosystem. Users can then choose whether to participate in these more controlled environments based on their preferences.
Perhaps most importantly, launching a rollup allows businesses to tap into an existing ecosystem rather than building everything from scratch. They get immediate access to established liquidity, a large user base, and battle-tested infrastructure like wallets and development tools. This ecosystem advantage helps explain why we're seeing major companies across industries launching their own rollups - from crypto exchanges like Coinbase and Kraken to traditional technology giants like Sony (yes, you read that right). This trend is likely to accelerate as more businesses recognize the advantages of having their own customizable blockchain environment that remains connected to a broader network.
While we've focused primarily on Layer 2 rollups, it's important to understand that networks can take several forms, each with its own tradeoffs. Layer 2 rollups, which directly inherit Ethereum's security, represent just one category in this evolving landscape.
Based rollups are a type of scaling solution that leverages the underlying Layer 1 blockchain (like Ethereum) for transaction sequencing. Instead of relying on their own sequencers, they utilize the L1's infrastructure for ordering transactions, enhancing security and decentralization. Think of it as a Layer 1.5.
We're also seeing the emergence of Layer 3 networks, which are rollups built on top of existing rollups. Gaming provides a perfect example of where this might make sense. Imagine a gaming-specific L3 built on top of a Layer 2, inheriting the L2's security and liquidity while providing specialized features like processing an enormous amount of micro-transactions. The reason why this makes sense is because bridging assets such as USDC and Ether may be useful for these micro-transactions, but the chain does not need the settlement guarantees offered by a massive decentralized network like Ethereum Mainnet (L1).
Finally, there are validiums - a variant of rollups that store their data off-chain while still using Ethereum for bridging assets. This approach can dramatically reduce costs but comes with different trust assumptions about data availability. For certain applications where data availability is less critical or can be guaranteed through other means, validiums offer an attractive scaling solution.
The choice between these different types of rollups ultimately depends on the specific needs of the application - whether you prioritize maximum security, flexibility, cost efficiency, or some balance of these factors.
Application developers are bound by the direction and vision of an ecosystem if they choose to stick to one network. Even though a feature that the application wants/needs may be on the development roadmap of a certain network, it may not be the priority of the greater mission of the L1 network. This can lead to a dislocation in priorities of a network and the applications that live on it. This dilemma is mostly alleviated if application developers have a wide choice of networks that they can build on, as this increases the probability that the application will find an ecosystem that is closer in line with its interests. These interests are not limited to technical ones, but also philosophical ones that curate an ecosystem of users that may be most in-line with the target market that the application is addressing.
Fundamentally, a monolithic model restricts the imagination of builders, as they are confined by what is capable in the environment that they must operate in. This restraint is already being recognized in the Solana ecosystem with the concession of needing “network extensions” to allow builders to develop outside of the confines of the SVM.
If every rollup has its own execution environment, how are users expected to bridge their funds constantly between these networks? If there’s deeper liquidity on a token on rollup A, how will a user on rollup B get the best price if they cannot access that liquidity from the rollup they are on? Do apps need to deploy and maintain their application on every individual network? How is any of this sustainable as more and more rollups get released over time? With the oncoming of dozens of rollups, many have expressed valid concerns over fragmentation of user experience, liquidity, user base and applications.
Luckily, most points of friction that come with fragmented account states will be a concern of the past with the number of cross-chain messaging protocols that are being developed and introduced to the market. The most notable native cross-chain solution would be Optimism’s Superchain stack. By connecting different rollups that are using their OP Stack, the Superchain enables an interconnected network that allows for seamless cross-chain transactions and bridging between networks in the Superchain. With this, if chain A has the deepest liquidity for a certain token, any user on any rollup in the Superchain ecosystem that wants to execute a swap can route their order to rollup A to get the best price, without needing to manually bridge their assets between the chains. Uniswap’s Unichain is trying to be the liquidity hub for the Superchain and have users benefit from standardized cross-chain messaging.
Optimism is not the only rollup ecosystem developing a solution to fragmentation. ZK Sync is building their Elastic Chain, and Polygon is building their AggLayer.
Beyond cross-chain messaging embedded into the rollup’s stack, there are plenty of third-party bridging protocols looking to work their way into applications. These third-party protocols will provide functions that require more precise solutions to applications than the cross-chain messaging standards offered by these network stacks. Examples of these third-party bridges are Chainlink’s CCIP, LayerZero and Hyperlane. Velodrome Finance, Optimism’s largest decentralized exchange, is currently using Hyperlane to deploy its voting mechanism across the Superchain.
We are still in the early stages of solving this issue, but the technology already exists for the most part, and we’re at the stage of needing to agree on industry standards and implementation in applications. We’re currently in a period of growing pain, as the network level is adapting to the needs of the application level and users. Many criticisms are warranted on how things work today, but there are also many disingenuous takes from actors in the industry on this subject. Some highlight that this fragmentation is a critical issue for Ethereum and will be the crux of its failure. Luckily, there are countless of highly competent teams building solutions for this exact problem that are being implemented at every level of the stack, and this issue in a few years from now will be a memory from an era of the quirks of a rapidly evolving industry
I believe that the future of value settlement will look like a wide range of satellite networks that interoperate through a base network. The base network (aka the Layer 1) will be seen as connective tissue in terms of users, applications and representing value for all these networks.
This does not close off any possibility of having competing networks and standards, for that will always exist and is important for a healthy ecosystem. There will be room for monolithic chains to capture specific market shares that relate to what they do best.
I predict that there will be competing rollup ecosystems and that Ethereum will not be the only network that will have this model. Another prediction I have is that once cross-chain messaging becomes more standardized and implemented in popular apps, we’ll see another wave of innovation as founders build projects that take advantage of this functionality. We’ll be seeing more app-chains that are specially designed to maximize one purpose, and we’ll see a future of users routing execution to those specialized chains without thinking about it.
Anyone looking to launch their own network in the future shouldn’t ask themselves why they should consider spinning up a rollup, but instead ask why shouldn’t they launch it as one?
L2BEAT is a platform that tracks and ranks Layer 2 blockchain networks:
L2BEAT is an analytics and research website about Ethereum layer 2 scaling. Here you will find in depth comparison of major protocols live on Ethereum today.
Uniswap Unichain Explainer:
We're excited to announce Unichain - a fast, decentralized Superchain L2 that's built to be the home for DeFi and liquidity across chains.
Sota Watanabe’s thoughts on Soneium, co-building with Sony Block Solutions Labs:
Sony Block Solutions Labs is launching the Soneium Ethereum rollup in a joint venture with Watanabe's company.
Breakdown of what are Based Rollups by mteam:
Optimism Superchain Explainer:
Learn about Optimism Superchain components, features, and roadmap.
ZK Sync Elastic Chains, a network of ZK rollups designed to address high fees and fragmented liquidity in current multi-chain systems:
TLDR: Today, ZKsync unveils the Elastic Chain - an ever-expanding network of ZK rollups, secured by math and natively interoperable under a uniform, intuitive UX. Ethereum's rollup-centric roadmap successfully reduced transaction fees, but fragmented liquidity and user experience.
Polygon Aggregation Layer Explainer:
https://polygon.technology/blog/aggregated-blockchains-a-new-thesis
<100 subscribers
Overcollateralized Jonny
No comments yet