Written by: imToken

If you have recently participated directly in Ethereum's native staking, you may have noticed that while ETH was deposited and on-chain transactions succeeded, the validators did not start working for a long time and remained "pending."

Moreover, this waiting period is highly likely to be over a month.

As of September 16, Ethereum still has over 1.8 million ETH waiting to enter the validator set, with an estimated waiting time of about 32 days; in July this year, the waiting time once exceeded 43 days, and during the peak queue in March, about 3.4 million ETH were stuck at the entrance, with waiting times nearing 60 days.

In other words, "queuing for a month" is not an exaggerated statement.

But the problem arises accordingly.

Ethereum generates a Slot approximately every 12 seconds, a regular transaction can quickly get confirmed, so why does it take dozens of days for staked ETH worth billions of dollars to remain idle? More importantly, is there any yield during these days? Why does entering staking now require over a month of waiting while exiting may only take a few minutes?

The answer lies in a very important yet often overlooked mechanism in Ethereum: the Validator Queue.

1. How is the "one-month queue" calculated?

First, it is easy to misunderstand; Ethereum does not specify that newly staked ETH must be locked for 40 days before they can start working.

One month is simply a dynamic result.

In simple terms, it can be understood as the ETH waiting to enter divided by the amount of ETH Ethereum allows to enter daily, which gives a rough estimate of the queue time.

Ethereum PoS does not allow all new validators to enter the network at once, but instead has set a speed limit mechanism called Churn Limit, controlling how much stake can join the validator set in each Epoch.

An Epoch consists of 32 Slots, each taking about 12 seconds, making an Epoch approximately 6.4 minutes long, and there are about 225 Epochs in a day.

After Pectra/Electra, Ethereum started to calculate this quota in terms of "ETH balance" rather than just the number of validators; currently, the upper limit for new stakes per Epoch is 256 ETH. If all were converted to traditional 32 ETH validators, it would mean a maximum of 8 standard validators can be released every 6.4 minutes.

This makes it easy to calculate how much ETH can enter each day: 256 ETH × 225 ≈ 57600 ETH/day.

Thus, if approximately 1.83 million ETH are waiting to enter, that’s 1.83 million ÷ 57,600 ≈ 32 days, which is the core source of the "one-month queue."

Therefore, this number changes daily.

If the amount of new ETH entering the queue daily is less than 57,600, the backlog will gradually be digested, and 40 days could become 30 days or 20 days;Conversely, if more than that amount continuously enters, the queue will not shorten, but rather lengthen further.

This year's March had about 3.4 million ETH queued, waiting nearly 60 days, which is an extreme example.

So, it is a completely different matter from our usual understanding of "Ethereum being congested."

Low Gas fees and plenty of block space do not mean validators can immediately join, because what limits them is not the execution layer's transaction capacity, but a speed limit set by Ethereum for the validator entrance.

2. Why not simply open the entrance and let all ETH in at once?

Since waiting for several days means funds cannot generate staking yield, the most straightforward solution seems simple: just raise the upper limit of 256 ETH per Epoch, right?

The problem is that validators are not ordinary accounts.

After a transfer of ETH is completed, it generally does not add a continuous burden to the entire network; however, once a validator is activated, it will enter Ethereum's consensus system long-term, participating in block proof, committee allocation, and network consensus.

Currently, Ethereum has over 900,000 active validators with more than 42 million ETH staked. Each new validator becomes part of this vast consensus system.

This is why Ethereum has always actively controlled the speed at which the validator set changes.

The EIP-7514 proposed in 2023 specifically set an upper limit on the speed of validator activation. The concern behind it is not that "too many stakers is bad"—quite the opposite; rather, if the number of validators grows excessively in a short time, it increases the consensus layer state size, validator management, and network communication and processing pressure.

After Pectra, Ethereum allows the effective balance of single compounding validators to be raised to a maximum of 2048 ETH, partly to prevent a large amount of staked funds from needing to be split into thousands of 32 ETH validators, thus reducing additional burdens from the continuous inflation of the number of validators.

Moreover, the Churn Limit has an even deeper purpose: to prevent the group that maintains Ethereum's security from undergoing drastic changes in a short period.

After all, the security of a PoS network relies on the currently active validator set.

If a large amount of staked funds can instantaneously join or leave, then the foundation upon which the network judges who is qualified to participate in consensus and how much economic security remains in the system will also change rapidly. Therefore, Ethereum needs to impose a speed limit on such changes artificially.

This is why the "queue" is not a bug waiting to be fixed.

From the user's perspective, it does sacrifice some capital efficiency; but from the protocol's perspective, it resembles an intentionally designed speed bump, indicating that Ethereum would rather let millions of ETH wait at the entrance for several weeks than allow the validator set to suddenly swell or shrink in just a few days.

Even from the future roadmap perspective, this notion has not changed.

The current plan to incorporate EIP-8061 into Glamsterdam intends to significantly increase the processing capacity for exits and validator merges to alleviate exit congestion, yet still plans to retain the activation cap of 256 ETH per Epoch.

In other words, Ethereum is considering making "exits" more flexible but is not yet ready to fully open the floodgates for "entries."

3. Is there yield during the waiting period? Why is exiting so quick now?

So, when ETH is deposited, does the waiting month count as staking?

From the perspective of protocol rewards, it does not.

Only when a validator truly enters the Active, that is, "active" state, does it begin to execute proofs, participate in block proposals, and receive corresponding Ethereum protocol rewards.

Until then, even if the staking transaction has been completed and ETH has entered the staking process, as long as the validator remains Pending, there will be no normal validator staking rewards produced. Ethereum.org considers "Active" as the point at which validators start working and generating yield.

This is also where long queues truly impact ordinary users.

According to a rough estimate based on the current overall staking APR of about 2.6%, a 32 ETH validator waiting for 40 days means missing out on approximately 0.09 ETH of potential protocol earnings.

Individually, this does not seem particularly exaggerated, but if it's an organization that needs to deploy tens of thousands or hundreds of thousands of ETH, the cost of capital idleness over 40 days can quickly amplify.

This also explains why Ethereum's staking queue has recently been attracting increasing attention from institutions and ETH treasury companies. Previously, when comparing different staking solutions, people would most easily notice a few percentage points difference in APR or fee amounts; however, once queuing itself could potentially take over a month, "when can we really start generating yields" also becomes part of capital efficiency.

However, there is an interesting contrast here.

If you check the exit queue, you will often find the waiting time is only a few minutes to a few hours. Does Ethereum allow people to exit quickly but not to enter quickly?

Actually, it’s not. Both entry and exit have Churn Limits; both are subject to protocol speed limits, but the queue length ultimately depends on how many are in line.

Currently, over 2 million ETH want to enter, but the ETH actually waiting to exit is very little, thus it’s the same limited-speed road where one side has a queue stretching for dozens of kilometers, while the other is almost clear.

Moreover, "a short exit queue" does not mean ETH can immediately return to wallets in a few minutes; after a validator completes its exit, it still needs to wait for a fixed 256 Epochs, about 27 hours, to enter a withdrawable state, after which it must wait for Ethereum's automatic Withdrawal Sweep to send the funds to the withdrawal address, and the entire process may still take several additional days.

This indicates that for ordinary users, there is an easily overlooked difference emerging among different staking methods.

If choosing to run a validator themselves or to create an independent validator through a non-custodial ETH staking service like imToken, then it must genuinely enter the Ethereum validator set, and thus must undergo this protocol queue. For example, in imToken's non-custodial staking, users hold the withdrawal keys while the node service providers are responsible for running the validators; when the page shows "pending," it essentially means the validator is waiting for Ethereum to release it, and will only start generating protocol earnings once its status turns to "active."

On the other hand, some liquid staking pools, due to the presence of numerous active validators and liquid tokens, can to a certain extent abstract this waiting period at the product level, so users may not directly feel the "zero-yield waiting period" of several days.

However, the cost is the introduction of another set of trust boundaries, including smart contracts, LST liquidity, protocol governance, and asset exchange risks. This is a problem often masked by APR, that the true costs of staking are never just about fees and yields.

When funds enter, when they truly start working, when they can exit, and during this process, who holds the asset control rights are all integral parts of the staking product.

In Conclusion

On the surface, it seems quite inefficient for Ethereum to let several million ETH queue outside for more than a month.

However, if you shift the perspective to the entire network, it becomes understandable, after all, regular transactions seek to be executed as quickly as possible, while validators are about who maintains Ethereum's consensus at the next moment.

The former can pursue getting faster, while the latter needs to exercise restraint.

So, this is actually a queue visibly manifesting itself, as the demand for staking far exceeds the entry speed set by the protocol, which was originally an invisible safety barrier.

In a sense, the longer this queue, the more it indicates that at this stage, the demand to enter the Ethereum validator set continuously surpasses the rate of entry permitted by the protocol.

But regardless of how many people are standing outside, Ethereum is still willing to open its doors according to its own pace, one Epoch at a time.