Staking is a supply question before it is an income one

Staking bonds units to the operation of a network. In exchange, the protocol pays rewards, and while bonded the units cannot be transferred until an exit process completes. The familiar framing treats this as an income product with a rate. The more accurate first framing is that staking moves supply between two states, and the reward is a redistribution of newly issued units among holders rather than a payment from anyone outside the system.

Bonding, queues and exit

Under proof of stake, a validator puts capital at risk to gain the right to propose and attest blocks. Misbehavior can be punished by slashing, which destroys part of the bond, and that possibility is what makes the bond meaningful rather than decorative.

Most holders do not run validators. They use delegation, assigning stake to an operator who runs the infrastructure and keeps a commission. The units usually remain under the holder's control in the sense that the operator cannot spend them, though the arrangement varies by chain and the distinction between custodial and non-custodial staking is not always made clearly in marketing.

A third route is staking through a custodian, usually an exchange, which bonds the units on the customer's behalf and passes on part of the reward. Here the customer's position is a claim on the custodian rather than a bonded position they control, and the difference is invisible in the account balance and decisive if the custodian fails. The 2022 failures of several centralized lenders and of FTX made the distinction concrete for a large number of customers who had understood their balances as holdings rather than as claims.

Exit is where the supply mechanics become concrete. Chains impose an unbonding period, and often a queue, so unstaking is not instant. The size of the queue determines how quickly bonded supply can become transferable, which matters most precisely when many participants want to exit at once. A chain where a large share of supply is bonded behind a long queue has a different liquidity profile from one where exit is quick, even when the two report a similar staking ratio. Queues are also symmetrical: entry queues delay the point at which newly bonded capital starts earning, which is why a rate quoted today is not the rate a new participant meets.

Why the reward is not income in the ordinary sense

A dividend is paid by a company out of earnings to a shareholder, in a currency the shareholder did not previously own. A bond coupon is paid by a borrower. In both cases an external party transfers value in.

A staking reward has two components with different characters. One part comes from newly issued units, which no one pays: it expands the supply, and it increases a staker's unit count while every non-staker's proportional share falls. The other part comes from transaction fees, tips and, on some chains, MEV, all of which are paid by users of the network. Only the second part represents value entering from outside the holder base.

The practical consequence is that a headline staking rate cannot be compared with a bond yield, and the comparison is made constantly. Following the next lesson's arithmetic, a nominal rate roughly equal to the issuance rate leaves a staker's proportional share of the network approximately unchanged, which is a rather different result from receiving the same rate in cash. It is also worth naming that the reward is paid in the same asset, so its value in any other currency varies with that asset's price. A rate quoted in units says nothing about a return measured in a currency.

Staking as a change in float

Bonded units are not available to sell without first unbonding, so a rising staking ratio reduces the readily tradable float even though circulating supply is unchanged. Most data conventions count staked units as circulating, since the holder can withdraw them by choice, and that convention is defensible but leaves the immobilized portion invisible in the headline figure. Staked supply is the series that makes it visible.

Liquid staking complicates this in a way worth understanding rather than skipping. A staker deposits units with a protocol, receives a transferable token representing the staked position, and can sell that token while the underlying stake stays bonded. The stake is locked; the economic exposure is not. So the reduction in effective float from staking is partly undone by liquid staking, in proportion to how much of the stake is wrapped this way and how deep the market for the receipt token is. The receipt can also trade away from the value of the underlying, particularly when exit queues are long, which is a discount that reflects the cost of waiting.

Restaking extends the same idea further, reusing bonded capital to secure additional systems. It adds obligations and additional slashing conditions to the same units, which is a risk statement rather than a supply statement, but it is another reason a single staked figure understates what is happening to those units.

Concentration, and what the ratio does not say

A staking ratio describes participation, not decentralization. The same ratio is consistent with stake spread across thousands of independent operators and with stake concentrated in a handful, and only the distribution answers questions about censorship or coordination. Validator count is a starting point and an imperfect one, since one operator can run many validators. The Nakamoto coefficient attempts a sharper reading by counting the smallest number of entities that could jointly disrupt the network.

Nor does a high staking ratio mean the network is more secure in a linear way. Above the level needed to make an attack impractical, additional bonded capital adds cost to the attacker without proportionate benefit, and it does so by paying issuance to stakers. Where that trade sits is a design question chains answer differently.

The staking pages carry the ratio, the bonded quantity, exit-queue conditions where they are published, and the nominal and real rates side by side. The next lesson does the arithmetic that separates those two rates.