How will the application layer reshape crypto value capture from fat protocols to chain abstraction?
Cryptographic applications are the main beneficiaries of the transformation in the way infrastructure is built.Original source: Adrian X account
Author: Adrian
Original compilation: Shenchao TechFlow
Alt L1s - Accelerating Again?
In every crypto cycle, the most successful investments are often early bets on new foundational layer infrastructure primitives (such as PoW, smart contracts, PoS, high throughput, modularity, etc.). If we look at the top 25 assets on Coingecko, only two tokens are not native tokens of L1 blockchains (excluding stable assets), which are Uniswap and Shiba Inu. Joel Monegro first explained this phenomenon in 2016, proposing the "fat protocol theory". The theory states that the biggest difference in value accumulation between Web3 and Web2 is that the value accumulated at the crypto foundational layer exceeds the total value obtained by the applications built on top of it. This value comes from:
Blockchains provide a shared data layer for settling transactions, which fosters positive-sum competition and supports permissionless composability.
The positive feedback loop driven by token appreciation is: token appreciation > attracts speculators to participate > converts speculators into actual users > user growth and token appreciation attract more developers and users, forming an ever-expanding ecosystem.
The original fat protocol theory
By 2024, the original fat protocol theory has undergone multiple industry debates and has been challenged by structural changes in the industry dynamics:
1. Commoditization of block space ------ With the realization of infrastructure premiums, the emergence of successful alternative L1s (such as Solana in high throughput and Celestia in data availability) has made them "category definers," attracting builders and investors to participate in alternative L1 investments in each cycle. Each cycle sees new blockchains that excite investors and users due to their differentiated characteristics, but they may ultimately become "ghost chains" (e.g., Cardano), which lack actual users and application support. Overall, this has led to an oversupply of block space in the market, but a lack of sufficient users and applications to utilize these resources.
2. Modularization of foundational layers ------ As more dedicated modular components emerge, the definition of "foundational layer" becomes increasingly complex, not to mention how to decompose the value accumulated at each layer of the stack. However, what can be clearly stated in this transition is:
In modular blockchains, value is dispersed throughout the stack. For a single component (like Celestia) to be valued higher than the integrated foundational layer, that component (like data availability DA) must become the most valuable part of the stack, and the "applications" (modular blockchains) built on it need to generate more usage and fees than the integrated system;
Competition between modular solutions drives more economical execution and data availability solutions, further lowering user fees.
3. Moving towards a "chain abstraction" future ------ Modularization has led to fragmentation of the ecosystem, making the user experience cumbersome. For developers, this means facing too many choices in deciding where to deploy applications; for users, this means overcoming many obstacles when jumping from application A on chain X to application B on chain Y. Fortunately, many smart individuals have recognized this issue and are working to build a future where users do not need to understand the underlying blockchain when interacting with crypto applications. This vision is known as "chain abstraction," and it is a theory that excites me. The question now is, how will value accumulate in the future of chain abstraction?
I believe that crypto applications are the main beneficiaries of the shift in how infrastructure is built. In particular, intent-driven transaction supply chains, as well as intangible assets like exclusivity of order flow and user experience, branding, will increasingly become competitive barriers for these applications, allowing them to achieve profitability more effectively than existing models.
Exclusivity of Order Flow
Since Ethereum completed its merge and introduced Flashbots and MEV-Boost, its MEV (Maximum Extractable Value) landscape has changed significantly. The once seeker-dominated "dark forest" has evolved into a commoditized order flow market. In this market, the current MEV supply chain is primarily dominated by validators, who capture about 90% of the generated MEV through bidding forms from each participant in the supply chain.
Ethereum's MEV supply chain
Validators capture most of the extractable value, leaving many participants in the transaction supply chain dissatisfied. Users want to be compensated for generating order flow, decentralized applications (dapps) want to retain the value brought by user order flow, while seekers and builders want to increase profits. As a result, participants eager to capture value have begun experimenting with various strategies to extract excess returns, one of which is the integration of seekers and builders—this strategy centers on achieving higher profit margins by increasing the determinism of the seeker’s packages being included. A wealth of data and literature shows that exclusivity is key to value capture in competitive markets, and applications with the most valuable traffic will have pricing power.
This phenomenon is also reflected in retail stock trading conducted through brokers like Robinhood. Robinhood maintains "zero-fee" trading by selling order flow to market makers and profits by capturing rebates. Market makers like Citadel are willing to pay for this order flow because they can profit through arbitrage and information asymmetry.
Moreover, an increasing number of trades are conducted through private memory pools, which have reached a historical high of 30% on Ethereum. Decentralized applications (dapps) realize that the value of user order flow is being extracted and leaked into the MEV supply chain, while private transactions provide more opportunities for customization and monetization around high-stickiness user flows.
(Chart source X)
As we move towards a future of chain abstraction, I expect this trend to continue. In an intent-driven execution model, the transaction supply chain may become even more fragmented, with applications limiting their order flow to solver networks that can provide the most competitive execution, which will intensify competition among solvers, thereby lowering profit margins. However, I believe that most value capture will shift from the foundational layer (i.e., validators) to the user-facing layer, where middleware components, while valuable, have lower profit margins—meaning that front-end applications capable of generating valuable order flow will have pricing power over seekers and solvers.
On how value will accumulate in the future
We have already seen this trend manifest in certain types of order flows that leverage specific applications' sorting mechanisms, such as Oracle Extractable Value (OEV) auctions (like Pyth, API 3, UMA Oval), which provide lending protocols a way to reclaim liquidation bids that should belong to validators.
User Experience and Branding as Sustainable Moats
If we further analyze the previously mentioned 30% private transaction sources, we find that most come from front-ends, such as Telegram bots, decentralized exchanges (Dexes), and wallets:
Breakdown of transaction sources through private memory pools
Although cryptocurrency users are often considered to have short attention spans, we are finally seeing a degree of user retention. Applications demonstrate that branding and user experience can become powerful moats------
User Experience: Alternative front-end forms introduce entirely new experiences, starting with connecting wallets on web applications, which naturally attracts users with specific experience needs. A great example is Telegram bots like Bananagun and bonkbot, which have generated over 150 million dollars in fees, allowing users to trade memes easily in Telegram chats.
Branding: In the cryptocurrency space, a well-reputed brand can increase fees by earning users' trust. Built-in exchange functions in wallets, although known for high fees, represent a successful business model because users are willing to pay for convenience. For example, Metamask's exchange function generates over 200 million dollars in fees annually. Lastly, Uniswap Lab's front-end fee switch has netted 50 million dollars since its launch. Transactions interacting with Uniswap Labs contracts through unofficial front-ends do not incur this fee, yet their revenue continues to grow.
This indicates that the Lindy effect exists in applications as well, and may even be more pronounced than in infrastructure. Typically, the adoption of new technologies (including cryptocurrencies) follows some S-curve, as we transition from early adopters to a broader mass audience—the next wave of users may be less tech-savvy and thus less price-sensitive, allowing brands that reach critical scale to profit in creative (or simple) ways.
The S-curve of cryptocurrencies
Conclusion
As someone primarily focused on infrastructure research and investment, this article is not meant to undermine the status of infrastructure as an investable asset in cryptocurrency, but rather to encourage a shift in mindset when thinking about the new types of infrastructure that can support the next generation of applications. These applications will serve users at the higher end of the S-curve. New infrastructure needs to showcase entirely new use cases at the application level to attract attention. At the same time, there is already sufficient evidence at the application level that user ownership directly promotes sustainable business models for value accumulation. Unfortunately, we may have passed a phase where every new hot L1 project brings exponential returns; however, those with significant differentiation may still be worth attention and value.
Instead, I spend more time thinking about and understanding the "infrastructure" that includes the following aspects:
AI: This includes the agent economy (AI Agent Economy), which can automate and improve end-user experiences; markets for computation and reasoning that continuously optimize resource allocation; and verification technologies that expand the computational capabilities of blockchain virtual machines.
CAKE** stack:** Many of my previous points suggest that we should move towards a future of chain abstraction, while there is still considerable room for design choices in most components of the stack. As infrastructure supports chain abstraction, the design space for applications will naturally expand, potentially blurring the lines between applications and infrastructure.
DePIN: I have always believed that DePIN is the key application of cryptocurrency in the real world, second only to stablecoins, and this view has not changed. DePIN leverages the advantages of cryptocurrency: achieving permissionless resource coordination, launching markets, and decentralized ownership through incentive mechanisms. Although each type of network still faces specific challenges, significant progress has been made in addressing cold start issues. I am very much looking forward to seeing founders with industry expertise leverage crypto technology to launch their products.
If you are developing projects related to the above content, feel free to reach out to me; I would be happy to discuss. I also welcome all kinds of feedback or dissenting opinions because, frankly, if I am completely wrong, investing will become simpler.
"The most exciting applications on the Ethereum blockchain may be those we haven't even thought of yet." ------ Vitalik Buterin, 2014
