Ethereum
Executive summary
Ethereum in 2026 is best understood not as a single execution venue but as a settlement and data-availability backbone for a fan of L2 rollups that now host the bulk of user activity. The Pectra upgrade landed in May 2025, raising validator effective balance to 2,048 ETH and adding EIP-7702 account abstraction, while Fusaka is on track for late-2026 with PeerDAS, the most ambitious data-availability scaling step since proto-danksharding. Spot ETH ETFs, approved in July 2024, have absorbed roughly $14B of net inflows on the way to a circulating ETH supply that briefly turned net inflationary in late 2025 as L2 fee draining cut burn. Solana has narrowed Ethereum's lead in retail trading volume, but Ethereum remains the dominant venue for stablecoin float, RWA tokenization and institutional DeFi, with $58-62B of L1 TVL plus another $48B+ across rollups. The investable thesis hinges on whether the L2-centric roadmap recaptures L1 fee value via blob demand, EIGEN restaking and EigenLayer-style middleware before competing monolithic chains erode developer mindshare.
Origin and architecture
Ethereum launched on July 30, 2015 from a 2014 ICO that raised roughly 31,000 BTC, designed by Vitalik Buterin and a co-founder cohort that included Gavin Wood, Joseph Lubin and Charles Hoskinson. Wood's Yellow Paper formalised the EVM as a quasi-Turing-complete state machine: a 256-bit-word stack VM with persistent storage trie per account, deterministic gas metering, and the global state transition function that has been the template for nearly every smart-contract chain since. Architecturally, post-Merge Ethereum is split into a consensus layer (originally Beacon Chain, now simply the CL) and an execution layer that runs the EVM, communicating via the Engine API. Five execution clients (Geth, Nethermind, Besu, Erigon, Reth) and five consensus clients (Prysm, Lighthouse, Teku, Nimbus, Lodestar) provide redundancy that no other major chain matches. After Dencun (March 2024) introduced EIP-4844 blob transactions and Pectra (May 2025) raised validator max effective balance to 2,048 ETH plus shipped EIP-7702 for ephemeral EOA-to-smart-account upgrades, the L1 has explicitly repositioned as a data-availability and settlement layer rather than the primary execution venue. Roughly 80-85% of user transactions now occur on L2s that post compressed batches and proofs back to L1, making blob throughput, not L1 gas, the binding scaling constraint.
Consensus and validator economics
Ethereum's PoS uses Gasper, combining Casper FFG finality every two epochs (~12.8 minutes) with LMD GHOST fork-choice. Validators stake 32 ETH (or up to 2,048 post-Pectra) and earn issuance roughly inversely proportional to total stake, with the curve currently producing ~3.0% gross issuance yield against 35.4M ETH staked (~29% of supply, split across 1.1M+ active validators). Real validator yield in Q1 2026 sits near 2.7-3.1% APR, after factoring MEV-Boost tips that average 0.05-0.08 ETH per proposal and execution-layer priority fees. Liquid staking via Lido (~26% of stake), Rocket Pool, Coinbase, Binance and a long tail of restaking-aware operators dominates, while solo home stakers represent a stubborn but small share that the Foundation is trying to grow via Distributed Validator Technology (Obol, SSV) and lower bandwidth requirements. EigenLayer restaking, with around 4.6M ETH restaked at peak in 2024 and roughly 3.1M in early 2026, layers slashable middleware-security commitments on top of the base validator set. The Pectra-era debate over MEV burn (EIP-7732 enshrined PBS variants), issuance curve modifications to penalise excessive LST concentration, and the 'rainbow staking' proposal for a two-tier validator economy will define monetary policy through Fusaka and the subsequent Glamsterdam fork.
Ecosystem, TVL and economic activity
Ethereum L1 hosts roughly $58-62B of TVL across DeFi as of April 2026, split roughly $30B in lending (Aave v3, Morpho Blue, Spark, Compound), $14B in DEX liquidity (Uniswap v2/v3/v4, Curve, Balancer), $7B in liquid staking depositories beyond Lido's stETH, and $9-11B across LRTs, perp DEXes and structured products. Counting Arbitrum, Base, Optimism, zkSync Era, Linea, Scroll, Blast, Polygon zkEVM and Mantle as the L2 perimeter, the broader Ethereum economic zone holds $108-115B of TVL, the largest single-chain lineage by an order of magnitude. Stablecoin float on Ethereum L1 is about $98B (USDT $48B, USDC $34B, DAI/sky-USDS $9B, PYUSD/FDUSD/other $6-7B); including L2s pushes it past $130B, which is roughly 53-55% of all stablecoin supply across chains. Daily L1 transactions oscillate between 1.05M and 1.45M, while combined L2 daily transactions exceed 18-22M, dominated by Base (~7-9M) and Arbitrum (~2.5-3M). L1 gas fees averaged $1.50-3.50 per simple ERC-20 transfer through Q1 2026, with priority fees spiking only in NFT mints, airdrop claims and large stablecoin migrations.
ETH token economics and supply trajectory
EIP-1559, live since August 2021, splits transaction fees into a base fee that is burned and a priority fee paid to proposers. Cumulative burn has erased roughly 4.7M ETH (~$15-18B at prevailing prices), but post-Dencun the L2 migration cut blob and L1 fee revenue, reducing daily burn from 4,000-7,000 ETH in 2022-23 to 600-1,400 ETH today. Net issuance therefore turned slightly positive in mid-2025 for the first time since the Merge, and circulating supply has crept from 120.2M post-Merge to roughly 120.95M by April 2026. The bull case for ETH-as-asset rests on three pivots: (1) Fusaka's PeerDAS and subsequent full Danksharding lifting blob throughput so L2 demand scales without per-blob price collapse; (2) restaking and DA monetisation via EigenLayer / Avail-style markets producing a new usage tier that pays L1 in ETH; (3) ETF-driven institutional ETH balances creating durable price floors and tightening float. The bear case argues that L2 sovereignty, bridgeless app-chains and Solana-style monolithic execution starve L1 of fee revenue, leaving issuance the marginal flow and turning ETH into a programmable bond rather than ultrasound money. Spot ETH ETFs, approved July 2024 with staking inclusion permitted in Q1 2025 after the SEC's revised guidance, hold roughly 4.1M ETH between BlackRock's ETHA, Fidelity's FETH and Grayscale's ETHE/ETH products.
Notable protocols and applications
Uniswap remains the canonical example of EVM composability, with v4 (deployed January 2025) introducing hooks that let pools embed custom logic for limit orders, dynamic fees, MEV-internalising auctions and TWAMM execution. Aave v3.2 dominates lending, with the GHO stablecoin reaching $640M supply and the Horizon RWA-collateralised facility live for tokenised treasuries. Maker / Sky's split into the USDS rebrand and SubDAO architecture (Spark, Andromeda, NewSubDAOs) routes more than $5.5B of T-bill collateral through onchain vaults. Liquid restaking via ether.fi, Renzo, Kelp and Puffer manages a combined $7-9B even after the post-2024 LRT correction, while EigenLayer's first AVS cohort (EigenDA, AltLayer, Hyperlane, Lagrange) shipped slashing in late 2025. NFT activity has migrated mostly to Blur and OpenSea on L2s, but blue-chip ERC-721 trading and onchain provenance still prefer L1. Onchain RWAs on Ethereum total roughly $13B led by BlackRock's BUIDL ($1.9B), Ondo's OUSG/USDY ($1.3B) and Franklin's FOBXX. Privacy and identity primitives (Aztec Network, Polygon ID, Worldcoin's World ID) are increasingly anchored to Ethereum settlement even when execution lives elsewhere.
Competitive position vs Solana and modular rivals
Solana has overtaken Ethereum on raw daily transaction count (60-90M including vote transactions, ~25-40M non-vote) and on retail-meaningful volume metrics like DEX trade count and active address fan-out. In 2025-26 Ethereum's defensive moat is composed of: (1) institutional and stablecoin issuer preference, since Tether, Circle, PayPal and Visa all default to Ethereum/L2s for new pilots; (2) deep developer mindshare with the Solidity, Vyper, Foundry and Hardhat tool stack; (3) bridges and account-system network effects that make ETH/USDC the universal settlement assets across EVM L2s; (4) the regulatory clarity that ETH itself is not an SEC-deemed security (a position the SEC effectively conceded by approving spot ETH ETFs without contesting commodity status). The competitive risk vector is bifurcated: Solana captures retail trading and consumer apps, while specialist app-chains (Hyperliquid for perps, Berachain for DeFi-native PoL, Monad for parallelised EVM) chip away at category leadership. Modular rivals like Celestia, EigenDA and Avail vie to be the cheap DA layer for sovereign rollups that no longer settle to Ethereum, threatening blob revenue capture. Vitalik's response — codified in the 'L1 simplification' essays and the Glamsterdam fork drafts — is to accelerate proposer-builder separation enshrinement, native rollup primitives and zk-EVM precompiles to keep Ethereum the default settlement venue.
Regulatory and institutional treatment
ETH's regulatory status improved markedly during 2024-25. The CFTC has long treated ETH as a commodity. The SEC's July 2024 spot ETF approvals, followed by the December 2024 acceptance of staked-ETH ETPs in Europe and the SEC's Q1 2025 no-action posture on US issuers including staking, effectively foreclosed the 'ETH is a security' theory in mainstream legal discourse. The MiCA regime in the EU (Title III crypto-asset rules effective December 2024) treats ETH as a non-stablecoin crypto-asset subject to whitepaper-equivalent disclosures by listing venues, while the FCA's UK regime maintains ETH on the FSMA Designated Activities Order list. Outside the West, Hong Kong's SFC, Singapore's MAS and Japan's FSA all permit retail ETH access and ETP listings. The lingering regulatory question is restaking and LRTs: the SEC has hinted at tokenised LRT shares as potentially security-resembling, and the OCC's January 2025 guidance for national banks holding crypto explicitly excluded restaked positions until clarity emerges. In terms of institutional rails, BlackRock's BUIDL, Franklin's FOBXX, Hashnote's USYC and Ondo's offerings all default to Ethereum, and JPMorgan's Onyx / Kinexys, Citi's Token Services, HSBC's Orion and Goldman's GS DAP either tokenise on Ethereum directly or settle on private chains that bridge to L1.
Risks and disruption vectors
The most acute technical risk is client diversity collapse: Geth still drives ~52% of execution layer share, and a consensus-affecting bug would risk a finalisation halt under the supermajority threshold. The Foundation's 'no client should exceed 33%' stance has slowed Geth's relative share but not eliminated tail risk. The second risk is restaking-induced systemic exposure: poorly designed AVS slashing conditions or correlated operator failures could cascade into ETH staked across LRT depositors, with second-order effects on lending markets that accept LRTs as collateral. Third, L2 economic fragmentation may turn structural — if applications find that running their own DA on Celestia or app-rollup on Eclipse offers strictly better margins than Ethereum blobs, the L2 toll-collector thesis weakens. Fourth, the political risk of any 'Ethereum tax' framing — issuance changes, MEV burn, or socially-coordinated slashing of misbehaving LSTs — could fracture validator alignment. Fifth, a Solana-style monolithic competitor that demonstrates 100k TPS with strong decentralisation and lower latency would force Ethereum to either ship far faster on rollup-centric throughput or accept retail attrition. Operational risks remain: the Inclusion List proposal in Glamsterdam aims to neutralise builder-level censorship, but if PBS markets degrade, OFAC-style filtering could re-emerge.
Outlook through 2027
Through 2026-27 the base case is that Fusaka ships PeerDAS in late 2026, pushing L1 effective DA throughput from ~3 blob-equivalents per slot today to 16-32, which collapses average L2 fees to fractions of a cent and supports a 5-10x expansion in non-Solana-substitutable activity. ETF flows continue to absorb 4-7% of free-float supply per year, and stablecoin float on Ethereum and its L2s crosses $200B by year-end 2027 as PSD3, MiCA's stablecoin Title III, and a US stablecoin act standardise issuance. Restaking matures into a $30-50B AVS market with EigenDA and a handful of decentralised bridges generating recurring fees. Bear-case scenarios feature client bug induced finality outage, an LRT cascade triggering Aave/Morpho insolvencies, or accelerating L2/Solana migration that turns ETH net inflationary on a sustained basis and raises the political cost of issuance. The asymmetric upside scenario is full Danksharding by 2028 plus enshrined zkEVM precompiles cutting L1 verification costs, which would let Ethereum simultaneously serve as DA layer, settlement layer and verifier-of-rollups, monetising every layer of the modular stack. Investors and builders should treat ETH as the chain whose monetary thesis is contingent on protocol execution — not on adoption, which is structurally locked in for at least the next cycle.
Watch points
- Fusaka mainnet timing and PeerDAS blob count progression
- Net issuance trajectory after burn collapse (turning ultrasound or inflationary)
- Restaking AVS market growth and slashing incidents at EigenLayer scale
- Spot ETH ETF flow direction including staked-ETH ETP adoption in US
TL;DR
Ethereum is the settlement layer for ~$110B of DeFi TVL across L1 and rollups, with Pectra shipped, Fusaka and PeerDAS due late 2026, and ETF-absorbed float underwriting an L2-centric scaling thesis whose success depends on blob demand and restaking monetisation.
Get DeFi Intel research in your inbox
Weekly long-form coverage of papers, incidents, jurisdictions, chains, tokens and the people building them. Free tier covers headlines; Pro adds the analyst-grade breakdowns.