An Authoritative Record of Ethereum Market Research

Ethereum Market Research Center

Original research and sharp opinions on Ethereum's protocol, economics, and systems. Read by the people building and using it.

stakefish — The State of Ethereum in 2026

stakefish's mid-year reading of Ethereum captures a network in institutional transition: 37M ETH staked, 900,000+ validators, a 71-day entry queue at peak, and yields compressing toward 2.8-3.3% APR. The report covers the staking landscape, client diversity progress, the Glamsterdam and Hegota upgrades, and what March 2026's commodity classification means for institutional staking.

stakefish — The State of Ethereum in <em>2026</em>

stakefish’s mid-year reading of Ethereum’s network state captures a system in transition: more capital, more validators, more institutional infrastructure — and compressing yield as the natural consequence.

The headline metrics are staggering. Roughly 37 million ETH is now staked, comfortably above 30% of supply, distributed across more than 900,000 active validators. The validator entry queue peaked at 71 days in February as institutional capital surged in. Staking APR has compressed to 2.8–3.3%, down from above 5% in early 2023 when the staked supply was less than half today’s level.

The institutional staking phase is now visible in the calendar. The Ethereum Foundation began staking 70,000 ETH from its own treasury on February 24, channeling rewards into protocol research and grants. BlackRock launched ETHB, its staked Ethereum ETF, on Nasdaq on March 12 — staking 70–95% of holdings and distributing 82% of rewards as monthly payments. Bitmine launched MAVAN on March 25 with 3.14 million ETH (roughly $6.8 billion) staked at inception. Lido remains the largest liquid staking provider but has seen its share decline from above 32% to roughly 24%.

Client diversity tells a more nuanced story. The consensus layer is in healthy shape — Lighthouse leads at ~43%, Prysm at ~31%, Teku at ~14%, with Nimbus, Grandine, and Lodestar splitting the remainder. No supermajority risk. The execution layer remains the persistent concern: Geth is down to ~50% from its former 85%, but that is still above the threshold where a single-client bug could halt finality. Nethermind sits at ~25%, with Besu, Reth, and Erigon making up the rest. stakefish itself runs more than half its validators on Nethermind as a deliberate decentralization stance.

On the developer side, Ethereum retains the largest community of any chain — 31,869 active developers per Electric Capital, nearly double Solana’s count — but more than half now work primarily on L2s. Base alone accounted for 42% of new Ethereum code in 2024, raising live questions about where value accrues as activity migrates upward.

Two protocol upgrades define the year ahead. Glamsterdam, shipping in H1, introduces enshrined Proposer-Builder Separation (ePBS) and Block-Level Access Lists (BALs) — laying groundwork for parallel transaction processing, a 200-million gas limit, and reduced MEV extraction. Hegota in H2 brings Verkle Trees, cutting node storage requirements by approximately 90% and enabling stateless clients that validate without downloading full state history.

The regulatory backdrop turned decisively favorable on March 17, when U.S. regulators classified ETH as a digital commodity — providing the clearest framework yet for institutional staking adoption.

The takeaway for stakers is uncomfortable but clear: as ETH supply migrates into staking and yields compress toward a new baseline, validator quality starts to matter materially. The gap between operators measured in basis points of effectiveness compounds meaningfully across multi-year positions, particularly for institutions managing nine-figure stakes.

Read the full report at stake.fish →