What this page traces, and whose money it is
Most explanations of staking stop at the button. This one follows the transactions. On 11 September 2026 a stake of 3,412,194.03873 NIM was created on the Nimiq staking contract and delegated to the validator we run. Below is every transaction that produced it, with block heights, timestamps and hashes, so you can pull the same rows out of a public endpoint and check us.
Two addresses appear on this page and both are ours:
- The staking wallet,
NQ27 NCB1 3CYU 9P4L EM2V D7L2 28QE 36PA EXB1. This is the account that holds the NIM and sends the staking transactions. - The validator,
NQ08 ACT8 T0FE PTG8 P5RL H2S3 QGXH V15R NVXY. ImpactZero, which we operate, and which is printed in the footer of every page on this site.
We could have traced a stranger's stake. Staking history is public and nothing would have stopped us. We have not, and we will not: the point of a case study is that somebody accepts responsibility for what it exposes, and the only history we are entitled to publish in that much detail is our own. Everything here was read on 14 September 2026.
Chapter one: getting out of the old position
The story does not start with a deposit. It starts with an exit, because this wallet already had NIM delegated elsewhere. Worth saying at the top: we took the long road here, and you usually should not. Nimiq has an update staker transaction that moves a delegation from one validator to another without the round trip, covered in how to stake NIM. It was no use to us on 10 September, because the validator we were moving to did not exist yet. It was registered the following morning, in chapter two.
| When (UTC) | Block | Operation | Amount |
|---|---|---|---|
| 10 Sep 2026, 15:14:42 | 61,243,155 | Set active stake to zero | 0 NIM, a signalling transaction |
| 11 Sep 2026, 07:04:50 | 61,300,946 | Retire stake | 0 NIM, a signalling transaction |
| 11 Sep 2026, 07:04:57 | 61,300,953 | Remove stake, out of the contract | 3,357,339.87779 NIM |
That is the three step staker exit described in how to unstake NIM, performed rather than explained. The first two transactions move no value at all, which is what a signalling transaction is: it changes the state of your staker record and nothing else.
The wait, measured
The gap between the deactivation and the retire was 15 hours, 50 minutes and 8 seconds. That is not an arbitrary delay, and the epoch arithmetic explains it exactly. The deactivation landed in batch 963,120, which is batch 480 of epoch 1338. The lock is the rest of the current epoch plus one whole epoch, so the earliest valid retire was the first batch of epoch 1340, batch 964,081. Our retire went in at batch 964,083, two batches later, which is as close to the floor as a human clicking a button gets.
Then the withdrawal took seven seconds. Once a stake is retired there is no second wait: removing it is an ordinary transaction. Everything that feels like waiting happens before the retire, not after it.
This wallet has been through that cycle five times, which makes it a small but real dataset on a number the rest of the internet gets wrong:
| Deactivated | Wait before the retire | Payout after the retire |
|---|---|---|
| 23 July 2025 | 39 h 25 m | 1 second |
| 25 July 2025 | 33 h 13 m | 1 second |
| 5 August 2025 | 20 h 37 m | 2 seconds |
| 7 August 2025 | 25 h 46 m | 1 second |
| 10 September 2026 | 15 h 50 m | 7 seconds |
The long waits are ours, not the protocol's: we retired when somebody got round to it. The short one is close to the floor. Nothing in five cycles comes anywhere near the "7 to 21 day unbonding period" that generic proof of stake articles import from other chains, and nothing exceeded two days either.
Chapter two: the validator had to exist first
Our own validator had been deleted the previous year, so before anything could be delegated to it, it had to be registered again. Registering costs a locked deposit of 100,000 NIM, and that money came out of the same wallet:
| When (UTC) | Block | What happened |
|---|---|---|
| 11 Sep 2026, 07:20:19 | 61,301,883 | 100,000 NIM, wallet to the validator address, a plain transfer |
| 11 Sep 2026, 07:21:08 | 61,301,932 | 100,000 NIM, validator address to the staking contract, Create validator |
Forty nine seconds between the two. That deposit is not a stake and it is not delegated to anybody: it is the operator's collateral, it belongs to the validator record, and it comes back only when the validator is deactivated, retired and deleted. If you are a staker rather than an operator, this chapter is the part of the story that does not apply to you.
Chapter three: the delegation itself
Four minutes later, one transaction did the actual staking:
| Field | Value |
|---|---|
| Time | 11 September 2026, 07:24:27 UTC |
| Block | 61,302,133 (batch 964,103, batch 23 of epoch 1340) |
| Operation | Create staker |
| Amount | 3,412,194.03873 NIM |
| Recipient | NQ77 0000 0000 0000 0000 0000 0000 0000 0001, the staking contract |
| Fee | 0 |
| Hash | 84c79ff2fa793f250e24b2159f0f730a3f2586edd02261c8aff166df063276b2 |
Where that amount came from is worth a line, because it is the only arithmetic on this page that is not a single transaction. The retired stake returned 3,357,339.87779 NIM. Two top-ups had landed in the hours before: 8,439.94308 NIM at 01:36 and 146,393.68593 NIM at 06:12 on the same morning. Take 100,000 NIM off that for the validator deposit and what was left is what went in.
One thing the chain does not hand you here, and we are not going to pretend otherwise: which validator a stake names is inside the data payload, and the history endpoint exposes only its first byte, which is the operation type. So a transaction list can tell you that a stake was created and how much it was for, but not where it points. For that you read the staker record, which is the next section. It also means this page cannot tell you where the delegation we exited in chapter one had been pointing, so it does not say.
What the staker record says today
A staker record is the chain's view of one address's stake. Read at block 61,513,979 on 14 September 2026:
| Field | Value | What it means |
|---|---|---|
| balance | 3,412,194.03873 NIM | The active stake. This is what earns, when there is anything to earn. |
| delegation | NQ08 ACT8 ... | The validator this stake backs. One address can back exactly one validator. |
| inactiveBalance | 0 NIM | Nothing deactivated. A pending exit would show here. |
| inactiveFrom | null | No lock running, so no countdown to a retire. |
| retiredBalance | 0 NIM | Nothing sitting withdrawable. This is the balance a remove stake takes. |
Read that balance against chapter three and you get the honest headline of this whole page. It is the same number. 3,412,194.03873 NIM went in on 11 September, and 3,412,194.03873 NIM is what the record held on 14 September, after four election blocks had come and gone. Not a luna of difference.
That is exactly what we expect, and it is why this case study is published at this stage rather than after a flattering month. Our validator is registered and carries no inactivity flag, but it is not being elected: on the same reading it held about 3.57 million NIM of total stake, our deposit plus five stakers, against a network where the largest single validator holds over 750 million. A validator with no slots produces no blocks and earns no rewards, so there is nothing to distribute, so the balance does not move. Where Nimiq staking rewards come from explains that chain in full.
The same story, drawn
Everything above is a table because tables are checkable. The same data as a picture is what NimMap is for, and both addresses open directly:
- The staking wallet, which draws as a hub with a thick cyan arrow into the staking contract and a gold one coming back out of it.
- The validator address, a much smaller map: the deposit in, the Create validator out, and the earlier lifecycle behind it.
Three things on those maps are worth knowing how to read:
- Cyan is money going into the staking contract. The arrow is labelled with the exact operation, so chapter three appears as an arrow named Stake and chapter two as one named Create validator.
- Gold is money leaving it. Retired stake and any rewards that accrued on it are paid out in the same transaction and the chain does not separate them, so neither does the label. Our 3,357,339.88 NIM return is a gold arrow.
- Hairline cyan arrows are signalling. The deactivation and the retire carry no value, so they draw as the thinnest line the map has. They are easy to miss and they are the two transactions that set the whole timetable.
The staking contract itself is drawn but never opened, because it is a counterparty of tens of thousands of wallets. A stake therefore looks like an arrow into a wall rather than a path onward, which is a deliberate limit rather than a gap. How to use NimMap covers the rest of the controls.
What happens at the next election block
A stake does not start working the instant it is confirmed. It starts working when the next validator set is elected, and that is a scheduled event rather than a per-stake one.
- An epoch is 720 batches of 60 blocks, so 43,200 blocks, about 12 hours at roughly a block a second. Our delegation landed 23 batches into epoch 1340, so it waited out the remaining 697 batches of that epoch before the boundary.
- The election block ends an epoch and shares 512 slots among validators in proportion to total stake, which is each validator's deposit plus everything delegated to it. The set is fixed for the epoch that follows.
- Slots are the unit of earning. The batch reward is spread across slots, so a validator with no slots in an epoch earns nothing that epoch, no matter how long its stake has been sitting there.
Which is the answer to the obvious question about this trace. Four election blocks passed between the delegation and the reading, and the balance is unchanged, because our stake is a small fraction of what an average slot costs at current network totals. Being registered, active and elected are three different states, and only the third one pays. Our validators page shows which operators are in which state from live data, including us.
What the reward phase would look like
This is mechanism, not a forecast, and nothing in it is a promise. If and when the validator is elected, here is the sequence the protocol actually runs:
- The protocol pays the validator, every batch. Newly issued NIM plus the transaction fees collected in that batch go to the elected validator's reward address, with the payment for one batch made at the end of the following one.
- Stakers are not paid by the protocol. The chain sees one payment to one address. There is no protocol level split, which is the single most misunderstood fact about Nimiq staking.
- Distribution happens off chain, under the validator's own policy. A validator that restakes sends an add stake transaction that increases your delegated balance. A validator that pays directly sends you NIM. Both are the operator's choice, not the protocol's.
Our published policy in the official validators list is restake at a 0% fee, which means that if this stake ever does earn, the visible signal will be the staker record's balance rising above 3,412,194.03873 NIM rather than a payment arriving in the wallet. That is the specific number to watch, and it is why we printed it to five decimal places. We are stating an intent we have published, not a result, and a 0% fee on rewards that do not exist is worth exactly nothing. Choosing a Nimiq validator makes that argument against our own interest at more length.
Where this stake actually is
Plainly, so that nothing above reads as more than it is:
- The stake is real and it is early. Three days old at the time of reading, through four election blocks, with a balance that has not changed by a single luna.
- It has earned nothing. Not a small amount. Nothing, because the validator it backs has not been elected. We are not going to dress that up.
- What it does prove is the plumbing. Every step of the self custodial path is in the table: the exit, the lock, the withdrawal, the deposit, the delegation, and a staker record that matches the transaction that created it to five decimal places. None of it went near an exchange or a third party deposit.
- This page gets updated, not quietly rewritten. When production starts, the reward section stops being mechanism and becomes measurement, and the date at the top changes with it. If it never starts, this page keeps saying so.
If you want to run the same trace on your own address, start with how to stake NIM for the flow itself, then put your address into NimMap and read the arrows. The rest of our guides are on the Learn hub.
Nothing on this page is financial advice.