One major development has recently gone largely unnoticed by the market: Lido is “moving” more than 8 million ETH, worth approximately $16 billion.
Of course, the funds are not being transferred from Lido to another protocol. Instead, hundreds of thousands of legacy validators backing stETH are gradually being migrated to the new validator architecture introduced by the Pectra upgrade.
Under Lido’s plan, more than 265,000 validators using legacy 0x01 withdrawal credentials will be consolidated into a smaller number of higher-balance 0x02 validators. Once the migration is complete, the total number of validators across Ethereum is expected to fall by nearly one-third, from approximately 880,000 to around 628,000. The number of attestation messages propagated during each epoch could also decline by roughly 29%.
This will not directly reduce gas fees for everyday users or suddenly accelerate transaction confirmations. The migration itself also carries a cost: Lido estimates that the temporary loss of rewards will amount to approximately 0.28% of the protocol’s annual staking rewards.
If the improvement in returns is limited and the migration comes at a cost, why is Lido still pushing ahead with a restructuring of more than $10 billion in staked assets?
The answer lies in one of the most important changes introduced by the Pectra upgrade in May 2025: compounding validators.
1. What Is Lido Actually Moving?
On May 7, 2025, the Pectra upgrade was activated on Ethereum mainnet.
Among its changes, EIP-7251 raised the maximum effective balance of a single validator from 32 ETH to 2,048 ETH and introduced withdrawal credentials beginning with 0x02. Validators using these new credentials can retain consensus-layer rewards in their Beacon Chain balances, allowing their effective balances to grow and generate additional rewards. They are therefore known as “compounding validators.”
At first glance, the change simply enables rewards to compound at the protocol level. The deeper change, however, is that it breaks Ethereum’s long-standing fixed structure of 32 ETH per validator.
Historically, an Ethereum validator’s maximum effective balance was fixed at 32 ETH. Even if its actual balance later increased to 33 ETH or more, only up to 32 ETH counted toward consensus rewards. Any excess balance did not increase the validator’s weight and was periodically swept to its execution-layer withdrawal address.
For an individual running only one or a few validators, putting withdrawn rewards back into native staking meant accumulating another 32 ETH before activating a new validator. This created a high reinvestment threshold, while smaller balances transferred to withdrawal addresses remained outside the staking system.
Lido, large exchanges, and professional staking providers could aggregate rewards from many users and reach the next 32 ETH threshold more quickly. But they faced a different cost: every additional 32 ETH generally required the creation and maintenance of another validator.
As the amount of ETH staked continued to rise, the validator count expanded with it, increasing the amount of indexing data, keys, signatures, and attestation messages that operators needed to manage.
EIP-7251 was designed to change this structure.
Under the new 0x02 model, the minimum balance required to activate a validator remains 32 ETH, but the maximum effective balance rises to 2,048 ETH. Rewards no longer need to be swept out automatically. Instead, they can remain in the validator, increase its effective balance, and continue generating returns.
Multiple existing validators can also be consolidated. For example, 2,048 ETH previously distributed across 64 validators can be combined into a single higher-balance validator. The total staking weight remains unchanged, but the number of validators, keys, and network messages that must be managed falls substantially. Further reading: “One Year Later, Lean Ethereum Returns: What Is Ethereum Trying to Prove?”
Ultimately, the funds do not leave Ethereum, nor does the economic security they provide decline. What changes is the operational burden, which becomes significantly lighter.
This is precisely the type of consolidation Lido is now pursuing.
Strictly speaking, the “compounding” in compounding validators explains only half their value. The other half lies in allowing validators to be reorganized from vast numbers of standardized 32 ETH units into leaner infrastructure better suited to large-scale operations.
2. How Much Can Returns Actually Improve?
Interestingly, the gains from compounding are not distributed evenly.
In theory, both individual stakers and large institutions can use 0x02 validators to reduce idle balances by keeping consensus-layer rewards staked. But because different participants already have very different capital-management capabilities, the marginal benefit of the upgrade varies considerably.
The June 2026 paper “When Staking Rewards Compound: Measuring the Impact of Ethereum’s Pectra Upgrade” compared the returns generated by 0x01 and 0x02 validators.
Its simulations found that, within the balance range of 32 to 2,048 ETH, 0x01 validators generated an average consensus-layer APR of approximately 2.17%, compared with around 2.26% for 0x02 validators—a relative improvement of roughly 4.7%. At staking balances between 8,192 and 10,240 ETH, however, the relative difference narrowed to approximately 0.3%.
One point is particularly important: the paper’s “roughly 4.7%” figure does not mean that APR increases by 4.7 percentage points. It refers to a relative improvement of around 4.7% on top of an existing consensus-layer APR of approximately 2% to 3%.
The reason smaller stakers see a more noticeable improvement is not that they receive any exclusive reward. It is that they previously found compounding much harder.
Consider a user operating a single 32 ETH validator. Once the rewards are automatically swept to the withdrawal address, the user must either continue accumulating ETH for a long time or combine those rewards with other funds before reaching another 32 ETH and activating a new validator.
Any amount below 32 ETH also remains scattered across separate addresses, leaving little incentive to consolidate it.
A 0x02 validator allows those rewards to continue increasing the effective balance of the same validator. This reduces the idle capital created by the 32 ETH threshold.
Ultimately, what smaller stakers previously lacked was not merely the willingness to reinvest, but the ability to return small amounts of ETH to native staking efficiently.
Large staking providers can also benefit from native compounding, but they already possess much stronger aggregation capabilities. They can quickly aggregate rewards into another 32 ETH and activate a new validator, effectively approximating compounding at the pool level.
The larger the staking pool, the smaller the proportion of fragmented balances relative to its total capital. As a result, the marginal improvement delivered by 0x02 naturally declines with scale.
That does not mean 0x02 validators are unimportant to large institutions.
On the contrary, the central challenge for these institutions is shifting from “How can rewards continue generating returns?” to “How can more ETH be managed with fewer validators?”
For them, 0x02 provides value in two main ways.
First, rewards can remain within validators and continue compounding, reducing the need for frequent aggregation, redeposits, and validator creation. Second, large fleets of existing 32 ETH validators can be consolidated, significantly reducing the cost of managing nodes, keys, and consensus-layer messages.
This change also introduces new trade-offs.
Legacy 0x01 validators automatically sweep rewards above 32 ETH to their withdrawal addresses without requiring an active on-chain operation. With 0x02 validators, rewards remain in the validator by default. Large providers that need to process user redemptions or manage liquidity must initiate partial withdrawals and redesign their accounting, reward-distribution, and liquidity-buffer mechanisms accordingly.
For smaller stakers, the most direct benefit of 0x02 is therefore a lower reinvestment threshold and less idle capital. For large institutions, the improvement in returns may be smaller, but validator consolidation and infrastructure efficiency are considerably more important.
Both groups benefit from the same mechanism, but the sources and priorities of those benefits differ.
3. What Changes—and What Does Not—in the Ethereum Staking Ecosystem?
Viewed solely through the lens of APR, Lido’s migration does not appear particularly compelling.
Large providers may gain less than 1% in relative returns from compounding. They will also experience temporary reward losses during the migration, while their existing accounting, withdrawal, and liquidity-management systems must be adjusted.
Yet Lido has still chosen to proceed with its largest core architectural upgrade since V2 in 2023. Once a protocol manages more than 8 million ETH, the number of validators becomes a cost in itself.
After Pectra, a single validator holding 2,048 ETH can carry the same staking weight as 64 legacy validators. This allows more capital to be managed more efficiently with far fewer validators.
Lido’s upgrade is also about more than validator consolidation.
After migrating to Curated Module v2, or CMv2, its professional Node Operators will be required to post ETH as collateral for the first time. If an operator experiences downtime, is slashed, misallocates rewards, or causes another attributable loss, the bond can be used to cover the resulting losses.
Historically, Lido’s curated Node Operators were trusted primarily based on their reputations and track records. Those factors will remain relevant, but they will now be reinforced by a layer of real capital at risk.
All 34 existing curated Node Operators are expected to migrate to CMv2. None has chosen to leave because of the bonding requirement.
This change may prove even more significant than compounding itself.
It suggests that the basis of competition in staking is changing structurally after Pectra. Differences among staking services may increasingly depend on how effectively they use capital, manage withdrawals and liquidity, allocate validator risks, and balance asset control, operational complexity, and returns.
For wallets and other user-facing gateways, the value proposition can no longer stop at displaying a single yield figure. They must also help users understand the flow of funds and the risk structure behind each staking method.
Take imToken Stake as an example. Users can access staking directly from the ETH wallet page and choose an option based on the amount of ETH they want to stake and their individual needs. Those who want to participate with smaller amounts can access integrated staking services through imToken.
Users with at least 32 ETH can choose a non-custodial validator solution, allowing them to participate in native Ethereum staking while retaining control of their assets—without having to operate the validator infrastructure themselves.
As compounding validators become more widely adopted, these interfaces will need to present more information: whether rewards are automatically compounded, when they can be withdrawn, which withdrawal credentials the validator uses, who controls the funds, and what technical and liquidity risks each option introduces.
This means wallets are no longer connecting users to a single staking yield page. They are connecting them to an increasingly diverse set of validator services.
Final Thoughts
From The Merge to the Shapella upgrade and then Pectra, Ethereum has gradually been completing the full lifecycle of staking.
The Merge made validators central to network security. Shapella addressed how staked funds could exit. Pectra has begun optimizing how capital enters staking, accumulates, and is reorganized.
Compounding validators will not deliver the same improvement in returns to every participant.
For smaller stakers, they reduce idle balances and allow long-term ETH holdings to participate more fully in consensus.
For large institutions, their more important benefit may not be a higher APR, but fewer validators and a lower operational burden.
The migration to 0x02 validators will therefore be gradual. Depending on their capital, liquidity requirements, and operating structures, different participants may choose to retain their existing validators or progressively adopt the compounding model.
But the way Ethereum organizes staked capital is already undergoing a structural transformation.
As validators move beyond fixed 32 ETH units, the staking ecosystem can evolve from a standardized validator model into a more diverse landscape shaped by stake utilization, liquidity needs, risk allocation, and infrastructure design.
This is the long-term paradigm shift in Ethereum staking worth watching.