Ethereum Value Accrual

The standard bull thesis for ETH holds that Ethereum is digital infrastructure whose native token captures value from the activity running on it, similar to how a toll road operator captures value from traffic. As stablecoins, DeFi, and tokenized real-world assets grow, the thesis goes, ETH fees grow and ETH's value grows with them.

This thesis is empirically broken in its current form. Fees have not followed volume growth, activity is migrating off Ethereum's main chain, and the institutional demand story has no supporting evidence.

The fee-volume disconnect

Since 2020, stablecoin transaction volume and tokenized asset value have grown 100x to 1000x on-chain. ETH daily transaction fees in dollar terms are essentially flat over the same period.¹ This is the single most important fact for the debate: the mechanism the bull thesis assumes, volume growth leading to fee growth leading to ETH value growth, is empirically absent.

Three structural reasons fees did not follow volume:

Network upgrades deliberately reduced per-transaction fees. The EIP-1559 upgrade in 2021 restructured the fee market, and the EIP-4844 upgrade in 2024 dramatically reduced layer-2 transaction costs by creating separate blob storage for layer-2 data.¹ These are genuine user wins, cheaper transactions, but they are structural fee-per-transaction deflation: as volume grows while fee-per-transaction falls, total revenue can stay flat or decline.

Activity is migrating to competing chains. Solana has captured a disproportionate share of stablecoin activity, DeFi trading, and new tokenized asset launches, particularly at the retail and high-frequency end of the market where low latency matters. Layer-2 networks reduce main-chain fees by doing most computation off-chain, so users pay layer-2 fees rather than Ethereum main-chain fees, meaning Ethereum captures only the compressed settlement cost; the success of its own layer-2 ecosystem is structurally bearish for its fee capture by design. Tether, the issuer of the dominant stablecoin, is actively building two separate chains, Plasma and Stable, specifically to move transaction volume away from the Ethereum main chain.

Tokenizing low-velocity assets generates near-zero fees. The real-world asset tokenization narrative is structurally weak for fee capture: a hundred million dollar bond that trades once every two years generates, on one back-of-envelope estimate, something like ten cents of fee revenue regardless of the notional value tokenized. Value tokenized is not proportional to fee revenue: a trillion dollars of infrequently traded tokenized assets might add only around one hundred thousand dollars of value to ETH.¹

Whose roadmap this is

The fee deflation described above is not an accident of adoption. It is the consequence of design decisions taken by the network Vitalik Buterin co-founded and still steers: scaling through layer-2 rollups rather than through the base layer was the stated plan, and the upgrades that cut per-transaction costs executed it. What the plan assumed was that layer-2 activity would settle back to the base layer and pay for the security it borrowed. By 2026 one major layer-2 network was reported to be paying more in licensing fees to the foundation behind its technology stack than it paid the Ethereum base layer, by roughly two to three orders of magnitude, which is this page's fee problem restated as a ratio rather than as a chart.²

Buterin's public turn back toward base-layer development across 2025 and 2026 is read as a direct response to that leak, and the reader is directed to the file on Ethereum layer-2 value extraction for the fuller account, including the centralization half of the same complaint.² The connection matters here because it locates the broken mechanism in a governable choice rather than in market forces. If the leak came from a roadmap decision, another roadmap decision can in principle close it, which is exactly the wager the base-layer turn represents and exactly what remains unproven.

The digital oil fallacy

Bulls compare ETH to oil, a necessary input commodity for blockchain operations whose price reflects demand for the underlying network. The problem: oil is a commodity that has traded in roughly the same real range for over a century, with prices governed by supply and demand curves that return to cost of production. The digital oil framing, taken seriously, is actually bearish: it predicts ETH trades in a multi-decade range with mean reversion, like crude oil, which is broadly consistent with ETH's multi-year trading range, between roughly 1,000 and 4,800 dollars, and repeated failed breakout attempts since 2021.¹

Institutional staking demand

The bull thesis claims banks and financial institutions that tokenize assets on Ethereum will buy and stake ETH as operating capital and to contribute to network security. The evidence for this is zero: no major bank has bought ETH for its balance sheet, none has announced plans to, and no bank treasury committee has argued for ETH as operating capital.¹ Banks pay for network inputs as needed, at market price, in small amounts, the way they pay for electricity; they do not stockpile the inputs. The staking argument does not translate to institutional treasury logic.

The financial illiteracy premium

The sharpest framing belongs to the pseudonymous analyst @Rewkang, whose thread is this page's sole source: "Ethereum's valuation comes primarily from financial illiteracy. Which to be fair, can create a decently large market cap. Look at XRP."¹ XRP is the direct comparison, a token with no fundamental fee-capture mechanism and persistent claims that banks will use it that never materialized, yet one that has maintained a large market cap for years because retail investors believe the narrative. What sustains the premium is retail narrative plus broader macro liquidity during crypto risk-on cycles, both real forces that can maintain prices above fundamental value for years. What limits the premium is that the valuation derivable from financial illiteracy is not infinite, and Ethereum's narrative has not significantly evolved since 2021.¹

What would change the thesis

For ETH value accrual to work as advertised: fee capture would need to return to the main chain, either through layer-2s paying more back to Ethereum or the ecosystem winning back activity from competing chains; high-velocity assets, not slow-moving treasury bonds, would need to dominate the tokenized asset mix; Ethereum's competitive positioning against Solana would need to improve; and a major bank would need to actually announce an ETH treasury position, which has not happened as of the current data.

ETH versus Bitcoin context

ETH has underperformed Bitcoin in every major macro cycle since 2021 to 2022, consistent with the read that Bitcoin has the hardest narrative, digital gold, fixed supply, no real competition, while ETH's narrative is contested and its fundamentals have not caught up to its valuation.¹