Raoul Pal holds Bitcoin, Solana, Ethereum, Sui, and Zcash. He calls Ethereum wildly undervalued. The argument he uses to get there is one I have heard in various forms since 2017, and it has always bothered me, though it took me a while to articulate why.
Here is the version he offers: if Ethereum went offline tomorrow, seventy to seventy-six percent of stablecoins would stop moving. Most of DeFi would freeze. The RWA market, the digital art market, every Layer-2 built on top would seize. He puts the damage at $1 trillion to $1.5 trillion. Therefore, the market has mispriced ETH, and it is undervalued relative to what is coming.
The logic feels airtight. It is not. It is a stress test that proves systemic importance and then quietly converts that into a claim about token price. Those are two different questions, and I want to spend this piece pulling them apart, because the confusion between them is where most Ethereum bull cases quietly break.
Systemic importance is not the same as value capture. This is the sentence I would underline in every tokenomics memo I write for the fintech desks I work with in Tel Aviv. TCP/IP is arguably the most economically important protocol in human history. There is no TCP token. The value accrued to the applications built on top, to the hardware vendors, to the service providers, not to the protocol itself. A system can be load-bearing for a trillion dollars of activity while its native asset captures a thin slice of that value. Ethereum is currently living through exactly this tension, and the L2 architecture is making it worse, not better.
When I audited tokenomics for those four hundred ICO whitepapers back in 2017, the recurring failure mode was never a lack of technology. It was a missing mechanism that connected usage to price. I would read pages of elegant consensus design and then find a single sentence at the bottom explaining that the token "captures value from network growth" without specifying how. I have a mental checklist for this now, and Ethereum's current position fails one of its items: the entity capturing the most activity is not always the entity capturing the most fee revenue.
Consider the actual mechanics. EIP-1559 tied ETH issuance to burn, and the burn is tied to L1 gas consumption. Then the rollup thesis emerged: move execution to L2s, settle on L1, and let the settlement layer capture value through data availability fees. The problem is that those data availability fees are a fraction of what L1 execution used to cost. As rollups matured through the Dencun upgrade and blob fees collapsed, L1 fee revenue compressed in ways that nobody modeling the burn three years ago anticipated.
Pal mentions L2s as part of the $1.5 trillion blast radius. He does not mention that L2s are, from the perspective of ETH holders, cannibalizing the very revenue stream that makes the burn thesis work. That is the blind spot, and it is not a small one. It is the central structural question in the entire Ethereum valuation debate.
I want to be careful here, because I am not arguing ETH is worthless. I hold Ethereum. I think the settlement layer position is genuinely defensible and the network effects are real. What I am arguing is that a stress test about activity disruption is not a valuation framework, and treating it as one is a category error.
The second layer of the argument bothers me even more. Pal notes that the market has not priced the tail risk of an Ethereum outage. He reads this correctly as an observation and then incorrectly as a buy signal. If a risk is unpriced, it can be an opportunity to buy because the market is complacent. It can also be a warning that the asset is overvalued and the risk is latent. The same observation supports both conclusions. When I see a piece of evidence that can be pointed in either direction but is only ever pointed in one, I look for the incentive structure underneath.
The incentive here is not hidden. Pal is the CEO of a media and research company whose revenue depends on crypto narrative growth. He holds the assets he recommends. He was speaking at a Sui event when he made the most aggressive claims. None of this makes him dishonest, and I respect that his words match his positions. He bought Ethereum and Solana at the 2022 low and has held for two years, which is more discipline than most commentators show. But honesty and correctness are different things, and a bull case delivered by the CEO of a business that benefits from bullish sentiment deserves a forensic read, not an affective read.
There is also the deeper methodological problem with stress-testing for tail events and reading the result as valuation. The Ethereum mainnet has run since 2015 without a prolonged outage. The scenario Pal constructs is engineering-implausible at the probabilities that would justify the asset price he implies. You can use an extreme scenario to demonstrate systemic importance. You cannot use it to derive a discount rate. That is a rhetorical move, and it is the same rhetorical move I remember reading in the 2017 whitepapers that predicted trillions in tokenized asset markets by 2020.
What would actually make the Ethereum undervaluation case? A model that ties L1 fee revenue, validator yield, and the relationship between L2 activity growth and settlement revenue into a discountable cash flow. I have not seen one, and I have looked. What I see instead are macro narratives: stablecoin supply going to $1 trillion, RWA tokenization, AI agents driving on-chain activity, liquidity rotating back from AI infrastructure into crypto. Each one is plausible. Stack them together and the probability of all of them landing on schedule collapses fast. The implied assumption is that capital stops flowing into AI, or that both can grow simultaneously, and the timeline for that shift keeps sliding right.
There is one part of the argument I think deserves more credit than the commentary usually gives it, and it is the contrarian angle nobody wants to discuss. Zcash. Pal holds it for privacy demand, and the industry generally treats privacy coins as regulatory toxic waste. That is exactly why the position is interesting. If AML regimes tighten to the point where conventional stablecoin flows become highly surveilled, the demand for private settlement does not disappear. It moves to whatever implements it with the least friction. This is a bet against the regulatory trajectory of the next three years, and it is the most genuinely counter-consensus position in the entire interview, far more than the Ethereum call. It is also the most dangerous one, because privacy coins face active delisting pressure across exchanges in Japan, Korea, and parts of Europe.
I understand the appeal of the Ethereum thesis. It is a beautiful narrative. But a system that everything depends on does not automatically have a token that everything accrues to, and the industry has spent a decade pretending otherwise. The value capture question is the question. The stress test was never an answer to it.
What happens to ETH if the L2s eventually capture the settlement layer too, and the mainnet becomes a data availability service with a burn rate approaching zero? Nobody has answered that, and until somebody prices it properly, the word "undervalued" is doing a lot of work it has not earned.