Ethereum ended the quarter with more capital securing the network than at any point in its history, while the reward for securing it moved lower. Those two observations sound like a contradiction, but they are the same mechanism seen from opposite directions. Reading them together says more about how proof-of-stake economics work than either number on its own.
A record that keeps compounding
Bitwise’s Q4 2026 staking report puts the record in precise terms: 43.80 million ETH, about 35.9% of the total supply, now sits locked in validators. Net inflows of 1.99 million ETH over the quarter extended the run to a sixth consecutive quarter of growth. The curve is the kind that looks unremarkable month to month and significant only when it is drawn out over years.
Why the yield moved the other way
The gross staking yield fell from 2.81% to 2.63% quarter over quarter. The decline is small in absolute terms — 0.18 percentage points, or about a 6% relative cut to validator revenue before node costs and fees — but the direction is the point. It fell for the same reason stake rose.
New issuance is shared across the entire staking pool, so when total stake grows, each validator’s slice of that issuance shrinks even if the network’s nominal rewards are unchanged. More capital chasing a fixed stream of rewards means a lower rate per unit of capital. Bitwise frames the result directly: growth got cheaper, and it was mostly by design.
A smaller validator count is not a weaker network
The most counterintuitive line in the report is that validator counts fell on both major networks even as stake climbed, and for opposite reasons. On Ethereum, the set dropped from roughly 1.1 million to 881,000 as operators consolidated under the Pectra upgrade. Fewer validator records now carry the same ETH behind them, so the change reflects operational efficiency rather than anyone pulling out.
Solana’s decline had a different cause. Its validator count fell from 696 to 659, roughly 30% below the prior year, under the cost pressure of running a validator. Two networks, two downward lines on a chart, and two unrelated explanations — a reminder that a single “validators are leaving” headline would mislead.
Three numbers worth keeping separate
The practical lesson is to avoid conflating three measures that moved in different directions. Total stake measures how much capital trusts the network, and it rose. The reward rate measures what that capital earns, and it fell. The validator count measures how many operators are running the infrastructure, and it fell for structural reasons rather than security ones.
For a staker, the relevant question is not whether the yield fell but what it fell to and what it is being compared against. A lower nominal rate on staked ETH carries a different risk and liquidity profile than the alternatives a holder might consider, so the comparison is never one of rates alone. For an observer, the quarter is a clean example of a design doing what it was built to do: security growing while the cost of attracting it declines.
Adapted from Ethereum Staking Hits Record 43.8 Million ETH as Yields Fall: Bitwise Q4 Report.