APY — annualized percentage yield — is the number DeFi protocols lead with. It's prominently displayed, refreshed in real time, and often enormous. Protocols advertising 200%, 500%, even 1,000% APY aren't rare during bull markets. The number seems to speak for itself.
It doesn't.
High APY is not the same as high return. The two can diverge significantly, and understanding why requires separating the number from the mechanism that produces it.
Three Mechanisms, Same Label
Not all APY is the same thing. The same label gets applied to at least three structurally different yield sources, each with distinct risk and sustainability profiles.
Emissions-based APY is paid in newly minted tokens — typically a protocol's own governance token. The protocol prints tokens and distributes them to liquidity providers or stakers. This is how most high-APY numbers are generated. You provide liquidity, receive token rewards, and the protocol's display tells you what that yield would amount to annualized.
The problem: the APY is denominated in tokens whose value isn't fixed. If the reward token falls 90% over the year you're earning a 200% APY, your real return is roughly -70%. The yield and the value are separate. Emissions-based APY attracts early liquidity, but the mechanics push toward token dilution and eventual APY collapse as supply grows and demand fails to keep pace.
Fee-based (or "real yield") APY is paid from actual protocol revenue — trading fees on a DEX, borrowing interest on a lending protocol, liquidation penalties redistributed to liquidity providers. This yield exists because users are paying to use the protocol. Real yield APY at scale tends to be modest (roughly 2–15%) because it's bounded by what the protocol earns. You can only distribute what you've collected.
Staking yield on proof-of-stake networks sits somewhere between the two. ETH staking currently yields around 3–4%, paid in newly issued ETH. This technically dilutes supply, but the dilution applies proportionally across all ETH holders — not just non-stakers. It's not cleanly analogous to an emissions program flooding the market with a low-demand governance token. Ethereum's staking yield is bounded by the protocol specification and predictable.
The Impermanent Loss Problem
Liquidity provision on AMMs (Uniswap, Curve, Balancer) adds another layer. When you provide liquidity to a trading pair and the prices of the two assets diverge, you end up with less value than if you'd simply held the assets separately. The mechanism is called impermanent loss — it's structural, not accidental.
APY figures for liquidity pools are almost always gross of impermanent loss. A pool advertising 40% APY might return 5% net once impermanent loss is factored in — or less, depending on how much the pair's prices move. For volatile pairs, this can easily flip a positive gross APY into a net negative. For stablecoin pairs, where prices don't diverge much, the effect is smaller.
The APY number doesn't tell you which situation you're in. That requires looking at the composition of the pool, the volatility of the underlying assets, and the fee structure.
What High APY Actually Signals
High APY is nearly always a signal of one or more of the following:
- A protocol paying for liquidity acquisition through token emissions. The APY buys market share; it isn't evidence of underlying demand.
- A small pool where even modest fee income produces high yield — which collapses as TVL grows, since your entry dilutes the rate for everyone already in the pool.
- A perception of high risk that the market prices in by requiring higher expected return. Smart contract risk, protocol risk, team risk, counterparty risk — these aren't reflected in the APY number.
The signal is never simply "this is an attractive opportunity." It's "there's something explaining this number, and it's worth finding out what."
The canonical example is Anchor Protocol's 19.5% APY on UST. Through 2021 and into early 2022, Anchor — built on Terra — advertised fixed stablecoin yield of 19.5%. The yield was manufactured: the protocol subsidized it from a reserve that required continuous replenishment from new capital. When the Terra/Luna death spiral began in May 2022, the yield became worthless alongside the principal. What looked like stable, high APY was structurally dependent on inflows it couldn't guarantee.
What's Changing
"Real yield" entered common use in DeFi circles roughly post-2022, specifically as a response to the emissions-era collapse. Protocols began explicitly marketing fee-backed yield to distinguish themselves from emissions-heavy competitors. The framing is imperfect — "real yield" isn't a formal category — but it reflects a meaningful structural distinction that didn't have widely understood vocabulary before.
Liquid staking derivatives (stETH, rETH, cbETH) have given the market a cleaner reference point. ETH staking yield is public, verifiable, and set by the protocol specification. It now functions as a baseline: anything materially higher than roughly 3–4% on ETH-denominated positions implies either elevated risk, emissions, or a small pool effect. None of those things are necessarily disqualifying, but they need to be understood.
Confirmation and Invalidation
A protocol's high APY is more likely to be durable if: yield is paid primarily in tokens with demonstrated demand, fee revenue approaches or exceeds total yield distributed, and APY remains relatively stable as TVL scales. These conditions are visible on-chain.
It's more likely to collapse if: the reward token has a declining price trend, fee revenue doesn't approach the yield being distributed, and the APY spikes around token launch phases and decays afterward. That pattern describes an incentive program, not a yield mechanism.
Timing
Now: The distinction between emissions-based and fee-based yield matters for any active DeFi position. Both produce the same number on a protocol's interface. The difference in outcomes can be total.
Next: Liquid staking yields will increasingly serve as the reference rate for ETH-denominated DeFi activity — transparent, protocol-governed, and publicly auditable.
Later: As DeFi protocols mature and compete more on fee structure than liquidity incentives, high-APY emissions programs may concentrate in earlier-stage launches where bootstrapping liquidity is most urgent. The pattern won't disappear — it'll move earlier in the protocol lifecycle.
Boundary Statement
This post explains what APY measures and what it doesn't. It doesn't recommend any specific protocol, yield target, or risk tolerance. Whether a given yield is worth pursuing depends on factors this post can't assess on your behalf.
High APY is information. What you do with it is a separate question.
