Original title: The Fat App Thesis: L1s as Commodities, Not Monopolies Original author: @0x_Arcana, Crypto Research Institute Original compilation: Rhythm Little deep
Editor’s note: The 'Fat Application Thesis' argues that as the cost of block space approaches zero, L1 blockchains transition from monopolies to commoditization, with value shifting from foundational protocol layers (like Ethereum, Solana) to application layers. Successful applications capture more revenue through vertical integration, controlling order flow and MEV, becoming sovereign application chains. The market is repricing L1/L2, and the future winners will be applications that are close to demand and focus on utility, rather than chains pursuing high TPS.
Below is the original content (to facilitate readability, the original content has been reorganized):
The crypto infrastructure phase is entering a world of zero marginal costs. Just like bandwidth and computing power, the price of block space will rapidly trend towards zero. The only chains that can survive are those that can:
· Achieving growth today through subsidies
· Tomorrow's capture of non-inflationary income
· Providing infrastructure that is difficult for applications to easily replicate or abandon
But in this new environment, L1 is no longer defined by early advantages or native ecosystems as monopolists. Instead, they have become commodities — interchangeable tools competing economically based on performance, interoperability, and cost efficiency.
Their value now depends on how well they can embed into application processes and provide indispensable or non-outsourcable services. The 'protocol premium' that once drove high valuations is fading, replaced by demand for genuine utility and performance. The current market repricing of many L1/L2 reflects this trend.
The chart includes: BERA, MOVE, SCR, STRK
Did the Fat Protocol Thesis get it wrong?
In 2016, Joel Monegro proposed the Fat Protocol Thesis, pointing out that most of the value in crypto networks is concentrated at the foundational protocol layer (like Ethereum, Solana, etc.), rather than at the application layer. This is in stark contrast to the Web2 model, where applications like Facebook, Google, and Amazon capture most of the value, while protocols like HTTP and TCP/IP are commoditized.
The Fat Protocol Thesis has indeed been correct over the past eight years. This can be seen from the huge discrepancy in valuation and revenue multiples between infrastructure and applications. On average, the trading valuations of applications relative to revenue are still far lower than those of infrastructure.
In this model, crypto infrastructure has received massive funding and venture capital. In fact, this situation is so common that founders and developers are almost incentivized to launch another alternative L1 or general Rollup, knowing that venture capital will support them at any time.
In a recent report, I mentioned that data availability (DA) is being commoditized and is inevitably trending towards zero. Based on the same logic, we can assume that all parts of the infrastructure stack will eventually be commoditized and value extracted. Why is that?
1. Fat Application Thesis: Applications realize that by becoming sovereign 'application chains' and vertically integrating the entire stack, they can capture more value.
2. Application-specific sorting: Applications can control their own transaction sorting and inclusion processes. This is an alternative path for those applications that do not want to build application chains from scratch.
Fat Application Thesis
The Fat Application Thesis posits that successful crypto applications will capture more value than the underlying blockchain protocols. The simple reason is: applications are business entities, and business entities prioritize maximizing revenue.
The most successful applications in the space are those that continuously generate revenue, such as: pumpfun, Hyperliquid, Jupiter, and Uniswap. What do they have in common? Fee income. These business entities want to control their own order flow and MEV capture, or in many cases, become sovereign application chains, which is entirely reasonable.
Vertical integration seems to be the most cost-effective way for applications to plug value leaks. As applications scale, the opportunity cost of not doing so only increases. This is good for applications, but not necessarily for underlying infrastructures like Ethereum. We have already seen clear signs of this trend in Unichain and JupNet.
What remains of the protocol layer?
There are two viewpoints regarding the future value accumulation of the foundational protocol layer:
1. Base fees and transaction fees will trend towards zero over time. MEV, as the only remaining source of income, will be abstracted by applications seeking to internalize all value. The protocol layer (e.g., Ethereum, Solana) will provide value as a settlement layer but will not capture any value — similar to HTTP and TCP/IP.
2. Cheap block space will lead to increased demand and a surge in applications. Transaction volume will increase as a result to offset low base fees and will re-accumulate value to the protocol layer.
Let's break down the first scenario:
A possible series of occurrences: · SOL surpasses ETH · We all realize that no one is special, just technology · SOL is surpassed · The value of L1 provides increasingly more to the world, but the value captured by its tokens relatively decreases · BTC reigns supreme
This viewpoint is based on the assumption of complete commoditization of infrastructure. Regardless of data availability, fees, or computation costs, all parts of the stack will trend towards zero over time. Cheap and abundant block space for Rollup and DA layers is eroding Ethereum's transactional monopoly.
Blob-based data inclusion (EIP-4844) will decouple execution from settlement, with L2 choosing alternative DA solutions, further reducing the residual value of sorting and data storage over the past year.
But the main evidence in this direction is the decline in the share of MEV captured by L1 block proposers. In 2024, most MEV will be captured by searchers and relayers through systems like Flashbots, rather than Ethereum validators. Currently, 90% of Ethereum blocks are proposed through MEV-Boost, a substantial portion of which is processed by relays associated with Flashbots.
This does not yet consider applications like CoW Swap, which use solver networks to handle matching and execution off-chain, completely bypassing the public mempool and its associated MEV.
The second scenario heavily relies on the demand and transaction volume surge brought about by near-zero fees. It assumes that the abundance of cheap block space will lead to increased consumption rather than a deflationary effect.
Just as the decline in computation costs spurred the internet boom, the reduction in transaction fees will unlock new categories and use cases for applications. The main analogy here is that general-purpose computing and coordination layers are more akin to AWS or Linux, rather than HTTP. Ethereum and Solana are not just 'settling' transactions but supporting large-scale programmable state coordination.
As usage increases and cost barriers decrease, this support for trustless computational power becomes more valuable, not less. Low fees will not push value to zero; rather, they expand the addressable market for block space.
· Low fees > Increased network demand
· Increased network demand > Total fee revenue
Token valuation — what does this mean for my investments?
If there is one point to summarize, it is this: capital allocation will shift in ways that are unfamiliar to many since 2016/17.
Unfortunately, the Fat Protocol Thesis has implanted a false sense of L1 premium subsidized by hundreds of millions in venture capital. However, we are currently at a turning point in the value distribution curve, where the income of applications relative to the protocol layer is evident.
Fat Application Thesis > Fat Protocol Thesis
In terms of L1 valuation, we have abused the narrative to the point where these tokens can no longer maintain their prices after TGE. Hundreds of millions in financing and billion-dollar valuations before mainnet launch have become the norm for L1/L2. A common trend among most new protocols is: prices only fall, not rise.
This does not mean that infrastructure will become irrelevant; however, the signs of market maturation are evident. Yet, the transactions of L1/L2 have saturated. The sentiment of low throughput and high FDV reflects this. The FDV of newly launched L1s is several orders of magnitude higher than in the previous cycle. Monad, Bera, and Story Protocol all raised nine-figure funding before launch, while Solana raised only $45 million (including public token sales).
The next cycle will not be led by chains competing to reach 100,000 TPS. It will be driven by focused and composable applications that prioritize usage over architecture and sustainability over speculative hype. The winners will be those applications closest to the source of demand.
「Original link」
