Staking gets sold as free money. Park your coins, earn 5%, sit back. I used to believe that too. Then I actually ran the numbers on my own holdings and the picture got a lot less shiny.
The short version: that advertised APY is a nominal number, and nominal is not what ends up in your pocket. Inflation, lock‑ups, slashing, and taxes all take a bite before you see any real gain. Let's walk through each one, and then do the math nobody puts in the marketing.
What staking actually is
On a proof‑of‑stake blockchain, validators lock up coins as collateral to help confirm transactions and produce blocks. In return, the network pays them newly minted tokens plus a share of fees. When you stake, you're either running a validator yourself or delegating your coins to someone who does. You get a cut of the rewards for putting your capital at risk.
It replaced the energy‑hungry mining model that proof‑of‑work chains like Bitcoin still use. Ethereum switched over in 2022. Since then, staking has become the default way people try to earn passive income on their crypto. If you want a wider view of that, here are other passive income options worth comparing against.
Nominal yield vs real yield
This is the part that trips up almost everyone. When a network pays you 5% in new tokens, it's also printing those tokens for everyone. That printing is inflation, and it dilutes the value of every coin already in circulation, including yours.
So your real yield is roughly the nominal reward rate minus the network's inflation rate. If a chain pays 5% APY but is inflating its total supply by 4% a year, your real yield is closer to 1%. You've grown your token count, sure. But your slice of the total pie barely moved.
Here's a concrete example. Say you stake $10,000 of a token at an advertised 5% APY. After a year you hold tokens worth 5% more of that token. But if 4% more tokens got created network‑wide, your purchasing power inside that ecosystem only climbed about 1%. That's $100 of real gain, not $500.
Some chains get this right. A network with 6% nominal staking rewards and only 1% net inflation gives you a genuine 5% real yield. Others are basically running fast to stand still. The trick is to find the actual inflation schedule, which is public data, and subtract it. Marketing pages almost never show you that subtraction.
Lock‑ups and unbonding periods
Staked coins usually aren't liquid. Most networks make you wait before you can move or sell them, and that waiting has two flavors.
- Bonding or lock‑up: some chains require your coins stay staked for a minimum period.
- Unbonding: when you decide to unstake, there's a cool‑down before the coins are free again. Cosmos chains often run around 21 days. Ethereum's exit queue can vary. Others are shorter.
Why does this matter? Because crypto moves fast. If the price crashes 30% during your 21‑day unbonding window, you're stuck watching it happen. You can't sell. That illiquidity is a real cost, and it's the reason liquid staking tokens like stETH exist. They give you a tradeable receipt for your staked position, though they add their own smart contract risk.
Slashing: the risk people forget
Slashing is the network's way of punishing bad validators. If a validator double‑signs blocks or stays offline too long, the protocol destroys part of its staked coins. If you delegated to that validator, your share can get burned right alongside theirs.
On Ethereum, slashing events are rare and the penalties are usually small for honest mistakes. On some other proof‑of‑stake chains, the punishment is harsher. The lesson is simple: the validator you pick matters. A cheap validator with a spotty uptime record can quietly cost you more than the fee you saved by choosing them.
Custodial vs self‑staking
You've basically got two paths, and they trade off control against convenience.
Custodial staking means an exchange or platform stakes on your behalf. It's dead simple. You click a button. But you're trusting them with your coins, paying a commission that often runs 10% to 25% of rewards, and you carry counterparty risk. If the platform blows up, and we've all watched a few of those, your coins can vanish. Not your keys, not your crypto still applies.
Self‑staking, or delegating from your own wallet, keeps you in control of your keys. You pick your validator, you pay lower fees, and you're not exposed to a platform going insolvent. The cost is that you have to actually understand what you're doing. For most people the honest answer is somewhere in between: delegate from a self‑custody wallet to a reputable validator.
Taxes eat the rest
Here's the quiet killer. In the US and many other countries, staking rewards count as ordinary income at their fair market value the moment you gain control of them. You owe tax on that value even if you never sell and even if the price later drops.
Then, when you eventually sell those reward tokens, you can owe capital gains tax on top. So the same coins can get taxed twice: once as income when earned, once as gains when sold. If you're in a 24% income bracket, that 5% nominal reward is really more like 3.8% before you even touch the inflation problem.
Let's do the full math
Stack it all up on that same $10,000 at a headline 5% APY:
- Start with $500 in nominal rewards for the year.
- Subtract 4% token inflation. Real value of the reward drops to roughly $100 of actual purchasing power.
- Now apply income tax on the $500 you received. At 24%, that's $120 gone.
- You're left with a real, after‑tax gain that can be near zero, or even negative, on a high‑inflation token.
Run the same math on a low‑inflation chain with a reasonable validator and no custodial fee, and it looks much better. A 6% nominal reward with 1% inflation and 24% tax still leaves you with a solid real return. The point isn't that staking is bad. It's that the headline number tells you almost nothing.
So is it worth it?
For me, yes, on the right networks. If I already plan to hold a token long term, staking beats letting it sit idle and get diluted while I earn nothing. Passive is passive. But I stopped treating APY as the deciding number a long time ago.
Do the subtraction first. Check the inflation rate, the unbonding period, the slashing history of your validator, and how your country taxes rewards. If real yield after all of that is comfortably positive, stake away. If it isn't, you might do better with dollar‑cost averaging into assets you actually want to own. Either way, know what you're really earning, not what the banner says.
