A token shows a 12% staking yield on every dashboard. The chart is flat. Holders are told they are earning while they wait, and for a while that feels true.

Then a quarter passes. The balance in the wallet is larger, but the price is lower, and the position is worth about the same or less. Nothing dramatic happened. The supply simply grew underneath the number everyone was watching.

Key Takeaways

  • Most staking rewards are newly minted tokens, so yield is paid by expanding supply
  • Stakers are compensated for inflation while non-stakers absorb it as dilution
  • Nominal APY overstates the real return once supply growth is subtracted
  • Part of every reward is sold to cover costs, creating steady, mechanical sell pressure

The Common Misunderstanding

The common view is that staking rewards are income. A network pays you for securing it, the same way a bank pays interest on a deposit. The yield is treated as a return that appears from somewhere outside the system.

It is an intuitive picture, and it is incomplete. In most proof of stake networks, the reward is not taken from a pool of existing money. It is created at the moment it is paid.

A bank pays interest out of income earned elsewhere. A proof of stake protocol, in the simplest design, pays rewards by minting new tokens. The payout does not come from a revenue stream. It comes from the issuance schedule.

That difference matters because it changes who is paying. If the reward is new supply, then the cost is spread across every holder of the existing supply, whether they staked or not.

A second misunderstanding follows from the first. Many traders read a high APY as a sign of strength, as if a larger number means a healthier token. Often a high APY simply means a high issuance rate. The network is paying more because it is printing more.

One observation a week on liquidity, flow, and structure. 4 minutes. No price calls.

Subscribe →

What Actually Happens

Rewards are a transfer, not a gift

Take a simple network with 100 tokens in circulation. Half of them, 50 tokens, are staked. The protocol pays a 10% reward to stakers, which is 5 new tokens.

After the payout, total supply is 105. The stakers now hold 55 tokens, or about 52.4% of supply. The non-stakers still hold 50 tokens, but that is now about 47.6% of supply, down from 50%.

The network did not create value for stakers out of nothing. It moved ownership. Five tokens of claim on the network went from the existing base to the stakers, and the mechanism was dilution.

The real return for a staker is roughly the nominal yield minus supply growth. Here that is about 10% against 5% inflation, so around 4.8% in terms of ownership share. For a holder who does not stake, the same math gives a loss of about 4.8% in share. The reward and the dilution are two sides of one entry.

Participation rate decides who bears the cost

The split depends on how much of the supply is staked. If almost everyone stakes, almost everyone receives the reward, and dilution roughly cancels out across the whole base. In that case the yield is close to cosmetic. Holders gain tokens and lose value per token in about equal measure.

If staking participation is low, the effect is sharper. A small group captures the new issuance, and a large group of idle holders funds it. The lower the participation, the more the reward looks like a transfer from passive holders to active ones.

This is why staking participation and inflation are tied together in many designs. Some networks raise issuance when too few tokens are staked and lower it when too many are, trying to hold participation near a target. The supply schedule is a policy, not a constant.

The sell-side channel

Dilution is the accounting effect. The market effect comes from what recipients do with the new tokens.

Validators and node operators have real costs: hardware, hosting, staff, taxes. Those costs are paid in fiat currency, so a portion of the rewards has to be sold. That selling is not driven by a view on price. It is driven by invoices, and it recurs on a schedule.

Larger operators and staking services also take a fee. That fee is often paid in the reward token, and the operator may sell part of it to fund the business. Individual stakers add to this when they sell rewards to realize income or to cover tax obligations tied to the payout.

None of this requires bearish sentiment. It is a constant, low-level flow of supply reaching the market whether or not anyone wants to sell. Buyers have to absorb it before price can move up, which acts like a headwind on every rally.

Locked supply changes the picture, but not the sign

Staking also removes tokens from liquid circulation, which can offset some of the pressure. Tokens that are locked cannot be sold, and a high staking ratio shrinks the float. We covered how that shapes behavior in How Staking Locks Supply and Shapes Market Cycles.

But the two effects work on different layers. Locking reduces the supply available to trade today. Inflation increases the total supply that will eventually be claimed. A smaller float with a growing total can look stable while the overhang quietly builds.

When unlock conditions change or exits open up, the locked supply and the accumulated rewards can reach the market together. The dynamics resemble what happens in What Happens When Token Unlocks Hit Exchanges, except here the unlock is continuous rather than a single dated event.

The reflexive loop

There is also a feedback loop. When price falls, the fiat value of rewards falls with it. Operators with fixed costs need to sell more tokens to cover the same bills. More tokens sold into a weak market adds to the pressure that lowered the price in the first place.

This does not guarantee a spiral, and plenty of networks absorb it without trouble. But it explains why inflated rewards matter most in weak conditions. The mechanism that is invisible in a rising market becomes visible when buyers step back.

Example from Crypto Markets

Issuance policy varies widely across large networks, which makes the comparison instructive.

Ethereum moved to proof of stake with the Merge in 2022. Its issuance to validators is relatively low as a share of total supply, at roughly under 1% a year in gross terms. Part of that has been offset at times by fee burning, so net supply growth has swung between slightly negative and slightly positive depending on network activity. The yield to stakers is modest, and so is the dilution.

Solana launched with a higher initial inflation rate that was designed to decline gradually toward a lower long-term level. Early on, the network paid stakers a meaningful yield funded by new supply. The designed disinflation means the dilution pressure was front-loaded and then eased over time.

Cosmos Hub, home of ATOM, sits at the other end. Its inflation has been dynamic and has run at double-digit annual rates, adjusting with the share of tokens staked. Stakers see headline yields well above the large-cap norm, but a large part of that number is the network paying itself with new supply. A holder who does not stake sees their share shrink every block.

Three networks, three policies, and three very different relationships between yield and supply. The dashboard number looks like a comparison of returns. The more useful comparison is how much of each yield is real after supply growth.

There is a related risk worth noting for stakers who go beyond plain staking. When rewards are boosted by reusing staked assets or borrowing against them, the same dilution applies to a larger, more fragile position. This is explored in What Is Leverage Staking? The Hidden Risk.

What Traders Can Learn

The first lesson is about reading the number. A staking yield is a gross figure. The figure that describes an outcome is yield minus supply growth, and for a holder who sells rewards regularly, it is yield minus supply growth minus the price impact of that selling.

The second is about who is on the other side. When a token advertises a high yield, ask what pays for it. If the answer is network revenue, the yield is backed by activity. If the answer is new issuance, the yield is a redistribution funded by everyone who did not participate.

The third is about concentration. Staking tends to reward larger holders in absolute terms, and compounding rewards increases their share over time. Dilution therefore tends to pull ownership, and often voting weight, toward those already holding the most. That dynamic overlaps with the points made in Governance Tokens as Leverage: Buying Voting Power.

The fourth is about timing. The pressure from rewards is steady and small, which makes it easy to ignore. It tends to matter most when demand is thin, because that is when the recurring supply has no one to absorb it.

None of this makes staking good or bad. It makes it a transaction with a cost. Understanding the cost is what turns a yield number into something that can be interpreted.

FAQ

Does staking cause inflation?

In most proof of stake networks, yes. Rewards are typically paid in newly minted tokens, so every payout increases total supply. The size of the effect depends on the issuance rate set by the protocol, and some networks offset it with fee burning.

Is staking yield real return?

Only partly. The nominal yield counts the tokens you receive, but the real return also depends on how much total supply grew and what happened to price. If supply grows faster than demand, a high yield can still leave the holder with a lower position value.

Do you lose money if you don't stake?

You are diluted, not debited. Your token count stays the same, but your share of total supply falls as new tokens go to stakers. Whether that shows up as a lower price depends on demand, but your claim on the network shrinks in relative terms.

Why do staking rewards create sell pressure?

Validators and operators have fiat costs, and many stakers sell rewards for income or taxes. That means a portion of new issuance reaches the market on a recurring basis. The selling is driven by obligations rather than opinion, so it tends to continue regardless of sentiment.

Related Concepts

Conclusion

Staking rewards feel like income because they arrive as tokens in a wallet. Mechanically, they are new supply, paid for by dilution across the whole holder base and partly sold into the market by the people who receive them.

The yield is real for the participant, but it is not free for the system. What looks like a return from the network is often a transfer between holders, and the net result depends on how fast supply grows relative to demand.

Yield is only real after supply is counted.