Ethereum's Layer 2 scaling strategy has produced a measurable economic paradox: the network processes more transactions than at any point in its history while collecting less fee revenue than it did in 2022. Combined L2 daily transaction counts exceed mainnet by a factor of 5:1 to 10:1, according...
"Just as Ethereum had seemingly found a working fundamental mechanism for tying increasing protocol utility and usage to token value accrual, the underlying economics driving this mechanism are way too fluid to form a reasonable assumption set on which to build valuation models." — CoinShares Research, ETH's Value Crisis Amid Scaling and Institutional Interest
Ethereum's Layer 2 scaling strategy has produced a measurable economic paradox: the network processes more transactions than at any point in its history while collecting less fee revenue than it did in 2022. Combined L2 daily transaction counts exceed mainnet by a factor of 5:1 to 10:1, according to L2Beat data, yet daily ETH burn has collapsed from thousands of ETH per day to frequently below 100 ETH. The result is mild supply inflation of approximately 0.23% annually — a structural reversal of the "ultrasound money" thesis that underpinned ETH's monetary premium during the 2021-2023 cycle.
ETH is down 32% year-to-date in 2026, falling from $3,001 on January 1 to below $2,000 on May 28. The ETH/BTC ratio hit 0.027 on May 21, its lowest level since mid-2025. JPMorgan issued a research note on May 19 stating that ETH is "unlikely to reverse its multi-year underperformance against BTC without meaningful improvements in network activity, DeFi adoption, and real-world use cases." The Fusaka upgrade (December 2025) introduced EIP-7918's blob fee floor — a direct intervention to restore pricing power — but the market has yet to price in recovery.
The numbers are unambiguous. Ethereum mainnet's daily gas revenue has fallen from approximately $30 million at peak to roughly $500,000 post-Dencun, according to Token Terminal data. The median transaction fee hit a record low of $0.00554 in May 2026, while monthly transaction volume simultaneously reached a record high of 70 million. More transactions, less revenue per transaction — a pattern that benefits users and L2 operators but structurally impairs ETH's burn mechanism.
The mechanism that drove this: EIP-4844, activated in March 2024, introduced "blobs" — dedicated data space for L2 rollups to post transaction batches. Blob space replaced expensive calldata, cutting L2 settlement costs by 90-99%. Two subsequent Blob Parameter Only (BPO) forks expanded capacity further: BPO1 (December 9, 2025) raised the per-block blob target from 6 to 10, and BPO2 (January 7, 2026) pushed it to 14 target and 21 maximum.
The result: OP Mainnet now charges approximately $0.0007 for a simple ETH transfer. Base's median transaction fee sits at $0.02. These are fractions of a cent — economic negligibility for the mainnet.
The economic asymmetry between L2 operators and Ethereum mainnet has become the central tension in the ecosystem. The data illustrates the scale of value extraction:
Base (Coinbase): $94 million in sequencer profit, contributing $4.9 million to Ethereum mainnet in blob fees — a 19:1 profit-to-contribution ratio. Average daily revenue of $185,291, processing 12.89 million daily transactions.
Arbitrum: Average daily sequencer revenue of approximately $55,025. Combined with Base and Optimism, these three chains process multiples of Ethereum mainnet's daily transaction count while returning minimal fees to the base layer.
Aggregate L2 Ecosystem: 73 active rollups collectively secure more than $48 billion in TVL. Combined throughput exceeds 5,600 TPS, with projections to 24,000+ TPS as the BPO schedule continues through 2026.
The structural issue: users pay transaction fees to L2 sequencers, not to Ethereum validators. MEV extracted on L2 blocks accrues to L2 block builders or sequencer operators. Gas fees paid by DeFi users on Base are denominated in ETH but collected by Coinbase's sequencer infrastructure first. Ethereum mainnet receives only periodic batch posting costs — blob fees that, at times, have measured in single-digit ETH per day.
This is not a bug. The Ethereum Foundation explicitly chose a rollup-centric roadmap. The tradeoff was well-understood in theory. In practice, the magnitude of value leakage has exceeded projections.
Ethereum's supply now sits at approximately 120.7-121.5 million ETH as of April 2026 — higher than the 120.06 million at the time of the Merge on September 15, 2022. Net supply has grown by roughly 950,000 ETH since the Merge, according to Glassnode data.
The math is straightforward:
For Ethereum to return to deflation, average gas fees must exceed approximately 16 gwei. Current averages hover near 2 gwei. The gap between current fee levels and the deflation threshold represents the scale of the value accrual problem.
Approximately 4.6 million ETH have been burned since EIP-1559 launched in August 2021, but the rate has slowed dramatically. The "ultrasound money" thesis — premised on net supply reduction — is now conditional rather than structural. It requires mainnet activity growth that has not materialized.
The market has rendered its verdict on Ethereum's relative positioning:
JPMorgan's May 19 research note identified five structural headwinds:
The one counterweight: approximately 30% of all ETH is locked in staking, reducing liquid supply. The validator exit queue has collapsed to near-zero (32 ETH from a peak of 2.67 million ETH), while the entry queue holds 1.3 million ETH — the highest since mid-November. Institutions are staking, even as price declines.
The Ethereum development community has acknowledged the value accrual problem and responded with protocol-level interventions. The Fusaka upgrade, activated December 3, 2025, was explicitly designed to repair the economic link between L1 and L2.
EIP-7918 (Blob Fee Floor): Establishes a minimum blob fee equal to 1/15.258 of the L1 execution base fee, replacing the previous 1 wei floor. Upon activation, the blob base fee surged 15 million-fold — from negligible to economically meaningful.
Fidelity Digital Assets published an analysis concluding that had EIP-7918 been active since the Dencun upgrade, Ethereum would have generated an additional $78.6 million (24,641 ETH) in cumulative blob-fee revenue. They found 93% of days since 2024 would have recorded higher blob fees under the new model.
PeerDAS (Peer Data Availability Sampling): Enables blob verification without requiring full blob downloads, allowing blob capacity to scale further without proportional infrastructure costs. Combined with EIP-7918's pricing floor, the design intent is to create a scalable revenue stream from data availability services.
Glamsterdam (next upgrade, expected 2026): Will introduce parallel transaction processing, targeting 200 million gas per block and 10,000+ TPS on L1. If successful, this would reduce L1's cost disadvantage relative to L2s, potentially recapturing some direct-to-mainnet activity.
The question, as CoinShares notes, is timing. These mechanisms are "real but unproven at scale and measured in years, not months."
Despite the value crisis, institutional infrastructure continues to build on Ethereum:
BlackRock ETHB (launched March 12, 2026): First iShares staked Ethereum ETF. Grew from $107 million seed capital to $254 million AUM within one week. Net yield approximately 2% annualized after fees — below the 10-year Treasury rate of ~4.2%, but offering ETH price upside that bonds do not.
Grayscale Ethereum Staking ETF: First U.S. Ethereum ETP to distribute staking rewards to shareholders, with the first distribution of $0.083178 per share announced January 5, 2026.
BitMine: Added 82,560 ETH to staking on January 3, 2026, bringing total staked position to 659,219 ETH (~$2.1 billion). DAT (Digital Asset Treasury) firms now control over 5.5% of ETH's total circulating supply.
The institutional calculus appears to be: ETH's base-layer security role in a $48 billion TVL ecosystem justifies holding, even if current fee capture is insufficient. The bet is on future value accrual repair, not current yield.
Ethereum faces what CoinShares calls a "value accrual timing risk" — a question of whether the protocol delivers fee market recovery before competitive pressure from Solana, Sui, and application-specific chains erodes its positioning further. The Fusaka upgrade's EIP-7918 is the first explicit acknowledgment that the rollup-centric roadmap created an economic problem, not just a scaling solution.
The data shows an ecosystem in structural transition. L2s process more transactions than ever but Ethereum mainnet captures a diminishing share of the economic activity they generate. Institutional capital continues to flow into ETH infrastructure — staking ETFs, DAT treasuries, validator operations — but price performance reflects the market's skepticism that fee revenue can recover to levels that justify a monetary premium.
The resolution likely depends on two variables: whether Glamsterdam's L1 scaling can recapture direct activity, and whether growing blob demand from 73+ rollups eventually saturates blob space sufficiently to generate meaningful revenue even with EIP-7918's floor. Both outcomes are plausible. Neither is guaranteed. The market is pricing uncertainty accordingly.