What APY Really Means in Staking (and Why It Moves)
APY is an estimate, not a promise. Learn what staking yield is built from, why the number moves, and how to read a rate without fooling yourself.

Almost every staking page leads with a percentage. It is the first thing you see and the number people compare, which makes it the number most worth understanding properly. A lot of disappointment in staking comes from treating a projected rate as a promise, and from assuming that a percentage in tokens is the same thing as a return in money.
This article explains what APY is measuring, where the rewards come from, why the figure moves on its own, and how to read a quoted rate without misleading yourself. None of it is financial advice.
What APY is actually measuring
APY stands for annual percentage yield. It answers a narrow question: if the current rate of reward continued unchanged for a full year, and rewards were reinvested along the way, what proportion would the holding grow by?
Two assumptions are buried in that sentence, and both matter.
The first is if nothing changed. APY is a snapshot extended into a projection. It is not a record of what happened and it is not a commitment about the future. In staking, the inputs change continuously, so the projection is being recalculated all the time.
The second is compounding. APY assumes rewards are put back to work and start earning too. Some designs compound automatically. Others pay rewards into a claimable balance that sits idle until you do something with it. When rewards are not reinvested, the realistic figure is closer to a simple annual rate than to the compounded one, which is one reason APY and APR can differ for the same underlying reward stream.
You will also see rates described as estimated, projected, variable or historical. Those words are doing real work. They are the difference between a measurement and a forecast.
Where staking rewards come from
In proof-of-stake systems, the network needs participants to put tokens at stake and use them to help produce and confirm blocks. Doing that job costs time, hardware and attention, and doing it dishonestly needs to be expensive. The reward is what makes honest participation worth the trouble.
Those rewards are typically funded from two directions. Some come from newly issued tokens, created by the protocol as part of its issuance schedule. Some come from fees paid by people using the network. The balance between the two differs from network to network and shifts with how busy the chain is.
This has an implication beginners often miss. Where a share of the reward comes from new issuance, everyone holding that token is being diluted while stakers are being compensated. Staking is partly a way of keeping pace, not purely a way of getting ahead. That is a reason to be sceptical of the idea that heavy issuance is straightforwardly generous.
Why the number moves
A staking rate is an output of a live system, not a setting someone chose for your benefit. Several things push it around, usually at the same time:
- How much total stake is participating. Reward budgets are broadly shared among stakers. When more tokens are staked network-wide, each participant’s slice tends to shrink. When stake leaves, slices tend to grow.
- Network activity. Where fees make up part of the reward, busy periods and quiet periods produce different outcomes.
- Protocol issuance rules. Many networks vary issuance according to their own formulas, and those rules can themselves be changed through governance.
- Validator performance. Rewards depend on the validator actually doing its job. Missed duties mean missed rewards.
- Commission. Operators take a cut for running the infrastructure, and that cut is not fixed forever.
- Penalties. On many networks, serious misbehaviour by a validator can reduce a stake rather than grow it.
Because of all this, a rate quoted anywhere is a reading taken at a moment. Treat a large gap between one quoted rate and another as a question to investigate rather than an opportunity to grab, since the difference often comes from different assumptions rather than a better deal.
Rewards in tokens are not returns in money
This is the biggest gap between how staking is advertised and how it feels to hold.
Staking rewards are usually paid in the same token you staked. If you stake a token and earn that token, your percentage growth is measured in that token. Whether you are better off in the currency you pay rent in depends entirely on what that token is worth when you eventually value it.
A holding can grow in token terms and shrink in purchasing power at the same time. It can also do the reverse. The percentage tells you nothing about that, because it is not measuring it. Anyone presenting a staking rate as though it were interest on a savings account is being careless at best.
This is also why unusually high advertised rates deserve suspicion rather than enthusiasm. A very large number often signals heavy token issuance, an unproven design, a short promotional window, or a scheme where early participants are effectively paid with later participants’ money. Reward-based scams lean on exactly this instinct, which is why we track their patterns in scam alerts.
What sits between the headline and your wallet
Even taking the rate at face value, several things reduce it before anything lands.
Commission. Operators deduct a percentage of rewards. Some quoted figures are shown after commission and some are not, and the label does not always make this clear.
Transaction costs. Staking, claiming and unstaking can each involve a fee. On smaller balances these matter a great deal, and frequent claiming can quietly consume much of the benefit.
Waiting time. Many networks have a warm-up before a new stake starts earning, and an unbonding period before withdrawn tokens become usable. During unbonding, tokens commonly earn nothing while remaining fully exposed to price movement.
Tax. Treatment of staking rewards varies by jurisdiction, and is worth checking locally rather than guessing.
How to read a quoted rate sensibly
A few questions turn a number back into information:
- Is this rate estimated, historical or promotional, and over what period?
- Is it shown before or after commission?
- Does it assume compounding, and does compounding happen automatically?
- What is the lock-up or unbonding period, and do tokens earn during it?
- Who is quoting it, and do they benefit from you acting on it?
- What would this look like if the token’s value fell significantly?
That last question is the useful one, because it forces the rate back into proportion. Rewards are compensation for taking on risk and giving up flexibility. They are not income, they are not guaranteed, and they can be outweighed by movement in the underlying asset.
If you want to see how we assess the things we cover before writing about them, that reasoning is set out in how we vet, and the entries we consider worth watching sit on the radar. TokenSpin is informational only. We take custody of nothing, we never ask you to connect a wallet or sign a transaction, and nothing on this site is financial advice.
Get The Spin
The week's vetted rewards + the scams to avoid — free, every week. Informational. Not financial advice. We never ask for your keys.