APY is an annualized yield metric that combines the nominal interest rate and the compounding frequency into a single number. The figure states what a given amount of capital grows to within twelve months when the holder immediately reinvests every payout. More frequent payouts therefore produce a higher APY at the same nominal rate.
Legally, the term remains defined for banks only. The Truth in Savings Act became law in December 1991 as part of the Federal Deposit Insurance Corporation Improvement Act. The accompanying Regulation DD took effect eighteen months later, in June 1993, and applies explicitly to depository institutions, not to crypto platforms. Crypto providers publish the number anyway, for example alongside offers for staking or lending. That calculation rule, however, does not apply to them.
How is APY calculated
Two calculation routes lead to APY. The first starts from the nominal annual rate and divides it by the number of payout periods n. Each payout subsequently flows back into the capital and earns interest in the following period. Therefore the formula reads APY = (1 + APR/n)^n - 1. The higher n, the wider the gap between APR and APY. However, this gap converges toward continuous compounding rather than growing without limit. An example with USD 10,000 and a nominal rate of 8% shows how much payout frequency alone contributes.
| Payout | Periods n | APY | Value after 12 months |
|---|---|---|---|
| annual | 1 | 8.00% | USD 10,800.00 |
| monthly | 12 | 8.30% | USD 10,830.00 |
| weekly | 52 | 8.32% | USD 10,832.20 |
| daily | 365 | 8.33% | USD 10,832.78 |
| per second | 31,536,000 | 8.33% | USD 10,832.87 |
Moving from annual to monthly payouts adds USD 30.00. From monthly to per second, a comparatively small USD 2.87 comes on top. So anyone weighing daily against per-second interest accrual is comparing rounding noise.
The second route appears in Appendix A to Regulation DD and reverses the logic. Rather than extrapolating from a rate, the rule sets the interest actually paid in relation to the capital. It then annualizes that amount over a year of 365 days. The rule reads APY = 100 [(1 + I/P)^(365/days) - 1]. Any offer becomes verifiable as soon as the payout figure stands. Banks must also state the value to within 0.01 percentage points. No such requirement exists for crypto platforms.
What is the difference between APY and APR
APR is the nominal rate without compounding, APY the effective rate with it. Given identical earnings, APY consequently sits at or above APR. Still, neither metric ranks above the other in reliability. Both measure the same earnings, only differently. Comparing two platforms thus holds only when both disclose the same metric.
Whether the quoted rate actually holds matters just as much. Regulation DD forces banks with variable-rate accounts to assume the initial interest rate alone and carry it forward across the full year. Crypto platforms, however, face no such duty. In its case against BlockFi, the SEC recorded how investors lent their crypto assets. They received the promise of a variable monthly interest payment in return. The provider subsequently paid USD 100 million to the SEC and to US states.
A third difference concerns compounding itself. Yet not every mechanism requires action from the user. With automatic reinvestment or a rebase mechanism, the position keeps growing on its own. Where the platform demands manual reinvestment, however, only those who return every payout immediately reach the stated APY.
Why is staking APY lower than advertised
With a savings account, a balance-sheet counterparty pays the interest from its own earnings. Staking works differently. The reward comes from protocol issuance and from priority fees, not from a debtor's earnings. Ethereum derives a validator's base reward from its effective balance and from the square root of the total active deposit. Total issuance therefore rises with the square root of the number of validators. Per validator, however, it falls with the inverse of that root.
The effect is substantial. Suppose the amount locked in stake grows from 35 to 45 million ETH. The gross yield of the same validator then falls by roughly 12%. Yet the validator has changed nothing and delivers the same service. As a result, an advertised staking APY is not a commitment but a snapshot of the amount currently staked.
Notably, the protocol denominates the reward in its own currency, not in francs or dollars. An ETH holding may grow by a few percent while the price falls by a double-digit figure. At the end, a loss remains. Anyone comparing a staking APY with a savings rate therefore places two different quantities side by side.
What makes up APY in lending and liquidity mining
In crypto lending, the interest comes from borrowers. Where their repayments fail to arrive, the yield takes an immediate hit. What matters, then, is whom the platform lends the capital on to, and against which collateral. On the Celsius Earn program, the US Federal Trade Commission (FTC) recorded unsecured loans of USD 1.2 billion. The platform had issued them contrary to its own account. At Gemini Earn, the provider additionally deducted an agency fee from the yield before it reached the investor. A lending APY is thus a gross figure before counterparty risk and before fees.
In liquidity mining and yield farming, APY consists of three components. Pro-rata trading fees form the core: Uniswap v2 passes 0.30% of trading volume on to liquidity providers. Moreover, token issuance from the protocol comes on top. Working against this, however, is the valuation effect of a diverging price pair, and that appears in no advertised number. Yet once the price pair diverges far enough, this deduction exceeds the entire fee income.
Some protocols pay the APY rather than earning it. Anchor paid a stable rate of 19.5% on UST deposits, which the protocol's lending rates never covered. The Terra community therefore decided to lower the rate in steps. Terra collapsed shortly afterward. So an APY from subsidy or issuance dilutes the token that funds it.
What MiCA and Swiss tax practice mean for advertised APY figures
In the EU, Regulation (EU) 2023/1114 applies, better known as MiCA. It prohibits issuers of asset-referenced tokens and e-money tokens from granting interest on them. Article 50 also brings crypto-asset service providers into scope. Notably, interest here means any benefit that depends on the holding period, including discounts and compensation from third parties. A stablecoin APY advertised in the EU can therefore not come from the token itself. It can only come from a separate service with its own counterparty risk.
For the advertising itself, MiCA requires marketing communications to be identifiable as such and to contain fair, clear and non-misleading information. They must also match the whitepaper and carry a note that no authority has reviewed them. These duties have applied since 30 December 2024. Since then, anyone advertising an APY in the EU must be able to explain where the number comes from.
For Swiss retail investors, tax also comes on top. The working paper of the Swiss Federal Tax Administration (ESTV) sets out the classification. Staking compensation generally qualifies as income from movable assets under Art. 20 para. 1 DBG. It is therefore subject to income tax at the moment of accrual. An advertised APY is ultimately a pre-tax figure, and fees reduce it further.
