The risks that published staking yields leave out

Staking is presented as the low-risk corner of crypto: lock an asset, help secure a network, receive a yield. The mechanics are genuine, and the yield is real. What makes the headline number misleading is everything it does not mention.

What staking actually is

Proof-of-stake networks require participants to post collateral in order to validate transactions. Honest validation earns newly issued tokens and a share of fees; dishonest or unreliable validation is penalised. The yield is compensation for taking on that obligation and that risk.

This matters because it reframes the number. A staking yield is not interest on a deposit. It is payment for accepting a specific set of risks, and understanding the payment requires understanding the risks.

Slashing

Slashing destroys part of the staked collateral when a validator double-signs or is offline for a prolonged period. It affects principal, not merely rewards.

Networks differ enormously. On some, delegating to a validator carries no slashing exposure at all — the worst outcome is missed rewards. On others, a delegator’s stake is slashed alongside the validator’s. If you are staking through a provider, how losses are allocated is a matter of that provider’s terms rather than the protocol’s, and some indemnify while others do not.

Any advertised rate that does not sit beside a statement of slashing exposure is an incomplete figure.

You cannot always leave

Many networks impose an unbonding period between requesting your stake back and receiving it, during which the assets earn nothing and cannot be moved or sold. Periods range from none to several weeks.

The risk this creates is specific: an unbonding delay is longest in effect exactly when you most want out, because that is when everyone else wants out too. A yield earned on an asset you cannot sell during a sharp decline may not compensate for the price move you were forced to sit through.

Who is actually holding it

Staking through an exchange or a liquid-staking provider adds a counterparty between you and the protocol. That party’s solvency, operational competence and terms now sit inside your risk. Convenience is real — no minimum, no infrastructure, often no lock-up — and so is the substitution of protocol risk for company risk.

Liquid staking, where you receive a tradeable token representing the staked position, adds a further layer: the token can trade below the value of the underlying, particularly under stress, precisely when you would want to exit.

The denomination problem

This is the omission that costs people the most. A yield quoted as a percentage is a percentage of the token. If the token falls further than the yield pays, the position is a loss in any currency you actually spend.

A high advertised rate frequently accompanies a token with heavy issuance, and heavy issuance is itself a source of downward price pressure. The rate and the risk are not independent; a very high number is often a description of the risk rather than compensation for it.

APR, APY, and what the number assumes

APR excludes compounding and APY includes it, so APY is always the larger and more marketable figure. Comparing one platform’s APY against another’s APR is not a comparison.

Both are extrapolations from current conditions. Staking yields move with the proportion of supply staked — more stakers means the same issuance divided among more participants — so a rate quoted today is not a rate promised for a year. We publish no staking rates ourselves for exactly this reason: a per-asset APR is a number we cannot source and date reliably, so our tools ask you to supply the rate you have actually been offered.

It is also worth checking whether rewards compound automatically or must be claimed manually, because a manual claim usually costs a transaction fee and, on a small position, that fee can consume a meaningful share of the reward it is collecting. Some networks restake automatically; others leave it to you, and the published rate rarely distinguishes between the two.

Choosing a validator is a real decision

When you delegate, you are choosing an operator, and the choice affects both your return and your risk. Commission is the obvious variable — a validator takes a percentage of rewards before passing the rest on — but the lowest commission is frequently not the best outcome, because a validator with poor uptime earns fewer rewards to share and, on some networks, exposes delegators to penalties.

Two less obvious factors matter. A validator’s share of total stake affects network decentralisation, and delegating to whoever is already largest concentrates it further. And on networks where governance votes are cast by validators, your delegation carries an implicit political proxy. Track record over several months tells you more than any single published figure.

The accounting nobody mentions

Staking rewards create a record-keeping obligation that surprises people. In many jurisdictions rewards are treated as income at the moment they are received, valued at that moment, and then separately subject to capital gains treatment on eventual disposal. That means a position paying rewards frequently can generate a large number of small taxable events, each needing a value at the time it occurred.

The practical consequence is that a tax liability can arise on rewards you never sold, denominated in a currency you must find elsewhere — and if the token has fallen since, the liability can exceed the current value of the rewards that created it. Rules vary substantially by country and change, so this is a matter for someone qualified in your jurisdiction. What is universal is that the record-keeping is far easier to do as you go than to reconstruct afterwards.

What to establish before committing

Whether delegation carries slashing exposure and who bears it. The unbonding period. Whether you are staking with the protocol or with a company, and what that company’s terms say about loss. What the rate is denominated in and what the issuance schedule does to that denomination. And whether the quoted figure is APR or APY.

None of that makes staking a bad idea. It makes the headline number one input among several rather than the whole proposition.