Ethereum's exit queue: 72.56% of the ETH is consolidation

In Ethereum epoch 482,252, read on Oct-14-2026, the exit queue held 18,124 validators and 985,952.44 ETH, and 715,384.29 of that ETH —72.56%— is not going to leave staking. These are sources of an EIP-7251 consolidation, the operation introduced by Pectra to merge multiple 32 ETH validators into a single one of up to 2,048: the source balance transfers to the destination validator when its turn arrives, without passing through a withdrawal account. Real exits amount to 3,583 validators and 270,568.16 ETH, or 27.44%. The separation changes the count because the Electra specification gives each of the two its own churn limit, and measured for that epoch, they are 256 ETH per epoch for exits and 416 for consolidations: two limits that do not subtract from one another. With each queue under its own limit, consolidations take 7.64 days to clear and exits 4.70. Below is the full arithmetic, including the division that resulted in the "14 days and 18 hours" figure circulated on Oct-2-2026.

Of those 715,384.29 ETH, 455,979.08 —14,246 sources— belong to six Lido operators migrating their validators from the Curated v1 module to their entities in Curated v2, in blocks of up to 63 sources of 32 ETH toward a destination that ends at 2,048. Counting the 2,190 Lido sources that have already left the queue —2,177 waiting for their withdrawable epoch and thirteen with the withdrawal completed— the program totals 16,436 sources and 525,656.09 ETH. Among those 16,436, there are none from Consensys, the operator under which Lido registers MetaMask Staking: of its 11,213 keys in the v1 module, the registry shows 11,105 have already exited, with 41 remaining in the queue, totaling 1,312.16 ETH.

Why might an Ethereum validator in "active_exiting" not be leaving staking?

The active_exiting status means the validator has a set exit epoch, and this occurs for two reasons with different outcomes. The first is the obvious one: someone requested to withdraw that ETH and, upon reaching its withdrawable epoch, the balance becomes available in the corresponding credential. The second was created by EIP-7251 with Pectra: the validator is the source of a consolidation, and at its withdrawable epoch, its effective balance is added to that of another validator with a 0x02 credential —the one that allows up to 2,048 ETH— while the excess is swept to the operator's withdrawal vault.

From the outside, they appear as the same row from the same endpoint, with the same status and the same epoch. What separates them is the cross-reference with the list of pending consolidations on the chain, which in epoch 482,252 had 16,750 entries toward 417 destinations.

Real exits consist of 3,354 validators with the old 0x01 credential and 229 with 0x02; among the sources, 14,376 carry 0x01 and 165 already have 0x02. Another 2,209 sources left the queue before the anchor: 2,196 are waiting for their withdrawable epoch and thirteen already had the withdrawal completed at the time of reading, outside the 985,952.44 ETH in the table. The network's staked ETH does not decrease due to the 715,384.29 in the second row —they change validators within the same operator— but the number of validators does, which is the goal of consolidation.

Which six Lido operators are moving the 455,979 ETH in consolidation in epoch 482,252?

Of the 14,541 sources in the queue plus the 2,209 already exited, 16,436 have the Lido vault as their withdrawal credential, totaling 525,656.09 ETH and pointing to 262 destinations. Cross-referencing each public key with the Lido Keys API, which publishes the module and operator, there are six source-destination pairs, and in all cases, the destination bears the name of the same operator in Curated v2.

PGO, EEO, and PTO are the labels used by the Keys API to name these entities and are left unexpanded. The median and maximum number of sources per destination both match at 63, with 2,016 ETH to move; the destination receiving the least has 32 sources and 1,024 ETH, and the highest final balance measured is 2,048.163 ETH. Sixty-three sources of 32 ETH plus the 32 from the destination equal 2,048, the effective balance ceiling per validator set by Pectra. None of the 262 destinations end up more than one ETH above that ceiling; all remain active and were activated between Sep-13 and Sep-26-2026. The residual amount swept to the vault instead of moving to the destination is 120.09 ETH out of 525,656.09.

Lido announced the program on Jul-27-2026: "more than 265,000 existing Curated Module validators" would move from 0x01 to 0x02 via consolidation with the new module. Epoch 482,252 shows this program in execution, with Lido's weight in institutional staking concentrating 63.7% of the ETH in consolidation still in the queue (455,979.08 out of 715,384.29). The other 314 sources total 294,586.01 ETH toward 155 destinations and do not carry Lido credentials, so they are not attributed.

Where does the "14 days and 18 hours" wait time circulated on Oct-2-2026 come from?

On Oct-2-2026, with the queue at 850,736 ETH and described as the highest of the year, a wait time of "14 days and 18 hours" to exit Ethereum staking was published; the report signed by Tron Weekly and published in the news section of edgeX, a perpetuals exchange, is one of those that picked it up, using queue data from beaconcha.in as the source. The calculation is reproduced exactly, and reproducing it reveals its assumptions: a day has 225 epochs, and at 256 ETH per epoch, the daily drain is 57,600 ETH; 850,736 divided by 57,600 gives 14.7697 days, which is 14 days and 18 hours.

Both factors are correct separately: the numerator is the entire queue from that day and the denominator is the network's exit limit. The division assumes the entire numerator drains through that denominator, yet in epoch 482,252, 72.56% of the queue drains through the other limit of the same specification. The report mixes units by writing "256 validator exits per epoch" and "roughly 57,600 validator exits per day": 256 is ETH per epoch —eight 32 ETH validators— and 57,600 is ETH per day, not validators.

These are two different snapshots: the 850,736 ETH is the queue from Oct-2-2026, and the measured breakdown is from epoch 482,252, twelve days later. The 270,568.16 ETH of real exits is equivalent to 31.80% of that headline figure, but they compare two readings separated in time and do not describe the composition of the Oct-2 queue, which cannot be reconstructed: the historical state of the beacon chain is not available in the public nodes used. It is confirmed that as of that date, the 262 Lido destinations were already active since Sep-13 to Sep-26.

What two churn limits does the Electra specification set and what wait times do they yield in epoch 482,252?

The Electra specification calculates a churn budget proportional to the staked ETH —total active balance divided by 65,536, with a floor of 128 ETH— and splits it into two portions: activations and exits receive the minimum of 256 ETH and that budget, while consolidations receive the remainder. A consolidation does not consume budget from the exit portion, nor vice versa.

The two limits are measured in the data. Grouping the effective balance of the queue by exit epoch, real exits yield 256 ETH per epoch in 376 of the 596 occupied epochs, with a median of 256; consolidation sources yield 416 in 937 of the 1,254 occupied epochs, with a median of 416. The consolidation portion is 62.5% larger than the exit portion.

The first two rows are the wait times the network applies to each queue; the last two are the result of pushing the entire queue through the exit limit. A validator requesting to exit in epoch 482,252 entered the first one: 4.70 days, and the furthest exit epoch in that queue falls on Oct-20-2026, five days after the anchor.

The reconciliation aligns with the formula itself. With an exit limit of 256 and a consolidation limit of 416, the churn budget is 672 ETH per epoch, which, given the 65,536 divisor in the specification, implies a total active balance between 44,040,192 and 44,105,728 ETH: the actual order of magnitude of staked ETH.

On the other side, the entry queue had 16,988 pending deposits and 1,267,180 ETH in the epoch following the anchor, 482,253: 22.0 days at the same limit of 256 ETH per epoch. More ETH is entering than exiting, and both sides share the proportional budget that EIP-8363 tapered issuance discusses through another route.

Where are the MetaMask Staking validators registered in Lido on Oct-14-2026?

Lido identifies MetaMask Staking as operator 21 of the Curated v1 module and operator 6 of v2. In the on-chain registry as of Oct-14-2026, operator 21 has 11,213 keys deposited and 11,105 exited, and operator 6 has 133 and 133. Of those 11,213, the beacon chain shows 11,075 with the withdrawal completed, 34 waiting for their withdrawable epoch, 41 in the queue, 4 in activation, and 59 not yet registered. The two exit figures measure different things: 11,105 is what the operator has reported in the registry, and 11,109 —the 11,075 with withdrawal completed plus the 34 waiting for it— is the key-by-key status on the beacon chain, four more.

The 41 in the queue total 1,312.16 ETH, 0.13% of the ETH in the queue, and their furthest exit epoch is 482,471: Oct-15-2026. Of the 7,141 exit epochs, 7,066 fall before Oct-7, the deadline Lido published on Sep-30 when announcing precautionary out-of-order exits for those keys; the remaining 75 range from Oct-13 to Oct-15.

Among the 16,436 Lido consolidation sources, there are 0 keys from that operator. What does exist are 136 Lido validators among the real exits in the queue, with 4,352.83 ETH: 93 from operator 436 of the community module, the 41 from operator 21, and two from operators 318 and 574 of the same module.

Taking the validators of operator 21 with an exit epoch starting from 00:00 UTC on Oct-2-2026 —5,900— and the maximum effective balance of 32 ETH from their 0x01 credential, the maximum attributable to those keys within the 850,736 ETH headline is 188,800 ETH, or 22.19%. This is an upper limit and covers only the keys registered in Lido: whatever that service operates outside is not identifiable from the primary sources used. Third-party figures published for the first week of October 2026 describe a different universe and are not used as data here.

Two more registry entries: 59 keys from operator 21 deposited between Sep-14 and Sep-17-2026 remain in the entry queue, with 1,888 ETH, and four of its validators moved to the activation queue at the anchor epoch.

What dates clear the epoch 482,252 queue between October 15 and 23, 2026?

The control sample consists of ten Lido destinations with 63 sources in the queue each, with their thresholds in the final section, and vote #206 is the ballot that sets the MetaMask Staking deposit limit in Lido to zero. Epochs are deterministic: each lasts 6.4 minutes and expires at a time known in advance, so the depletion chronology is written before it happens.

Real exits end on Oct-20; Oct-22 and 23 belong to the other queue. After 17:07 UTC on Oct-23-2026, the 715,384.29 ETH in consolidation will have moved to the balances of their destinations, except for the residual above the effective balance, which is swept to the withdrawal credential.

Was there a cost for stETH holders between September and October 14, 2026?

The yield perceived by an stETH holder is published every day in the token rebase: the annualized rate is the growth of ETH per share between two rebases, scaled to the year. With 55 rebases between Aug-21 and Oct-14-2026, the breakdown by windows is as follows.

The week of Oct-8 to 14-2026 yields 2.2231% compared to 2.2464% in September: a 2.3 basis point difference, within a daily range that in September was 15 basis points between the minimum and maximum rebase. The Lido API gave a seven-day moving average of 2.224% and a final value of 2.209% on Oct-14. The series shows no cost for stETH holders between Aug-21 and Oct-14-2026, and Lido's Sep-30 notice stated that no action was required on their part, with the exit, withdrawal, and re-staking cycle estimated by Lido at up to approximately 45 days. This is the channel through which the staking paradox in Ethereum ETFs is also measured.

What can a reader verify before October 23, 2026, regarding Lido destinations and vote #206?

The breakdown allows for a falsifiable verification with three dates: Oct-16, 18, and 23-2026. There are ten Lido destinations set as a sample, each with a balance of 32.04 to 32.09 ETH at the time of reading and 63 sources still in the queue moving 2,016 ETH to it, with an expected final balance of 2,048.04 to 2,048.09 ETH. Two epochs after the last withdrawable epoch of its sources —Oct-16 for five destinations, Oct-18 for three, and Oct-23 for the last two— each destination should exceed 2,047 ETH and no expired source should retain more than 1 ETH. The contrary threshold is set similarly: if two or more of the ten remain below 1,900 ETH with their sources expired, or if the Lido withdrawal vault receives full withdrawals of 31 ETH or more with the index of those sources, they were exits and the 72.56% figure is disproven.

An early reading outside the sample points in that direction: a destination with 63 sources, thirteen of them expiring in each of the epochs 482,253 to 482,256, went from 800.039 ETH in epoch 482,253 to 1,216.039 in 482,254: exactly 416 ETH, the thirteen with withdrawal in 482,254. The thirteen from the previous epoch already appeared with between 0.003 and 0.057 ETH in the first reading. That is the signature of a consolidation: the ETH appears in the balance of another active validator in the same transition in which it disappears from the source's balance, without passing through a withdrawal to the vault.

The other open date is for governance and is not yet decided. Lido vote #206, opened on Oct-14-2026 based on the Oct-5 proposal, sets the MetaMask Staking validator limit to zero in both modules: no new deposits would be assigned to it, without touching the keys already operating. In the contract reading on Oct-14, it had 19,764,917 LDO in favor and none against, with a required quorum of 50 million and minimum support of 50%. In a re-reading on Oct-15-2026 at 13:37 UTC, identical across two nodes, the count in favor was 30,658,965 LDO and still without any votes against: 10.9 million more in twenty-four hours and 19.3 million below the quorum, with the contract still returning that the proposal is not executable. The main phase closes on Oct-17 and the full vote on Oct-19: the contract sets 120 hours of voting with the 48-hour objection period inside, not added at the end. If the quorum is not reached, the proposal is not executed; in both cases, the 59 deposits already in the entry queue and the four validators in activation remain outside its scope.

Anyone wishing to verify has the three dates, the two thresholds, and the starting figures. The epoch 482,252 queue clears between Oct-20 and Oct-23-2026, and the balances of those 262 destinations on Oct-24 resolve the 72.56%.

Sources and links: Ethereum beacon chain public API — exiting validators · Electra Specification (consensus-specs) · EIP-7251, consolidation and maximum effective balance · Lido Keys API · Lido, migration to Curated Module v2 (Jul-27-2026) · Lido Research, notice on MetaMask Staking (Sep-30-2026) · Lido Vote #206 · Lido stETH APR API · stETH rebase events on Blockscout · Curated v1 module operator registry on Etherscan · edgeX, Oct-2-2026 report with the 850,736 ETH queue