What is Yield?

Introduction

Yield in DeFi is the return earned on crypto assets after you put them to work inside a protocol. That sounds straightforward, but it hides the question that matters most: *who is paying that return, and why does the protocol have value to distribute at all? * If you do not answer that question, an advertised APY is just a number on a screen.

The useful way to think about yield is not as a feature, but as a cash-flow pattern. Assets go into a system. Some activity inside that system produces fees, interest, or incentives. Those flows are then allocated back to depositors, liquidity providers, or vault shareholders. The hard part is that different protocols produce those flows in very different ways, and the risks are tied to the mechanism, not the label.

That is why lending yield feels different from liquidity-provider yield, even when both are shown as an annual percentage. In a lending market, the mechanism is usually borrowers paying interest. In an automated market maker, the mechanism is usually traders paying swap fees. In a rewards program, part of the return may come from newly distributed tokens rather than from underlying economic activity. Those are not cosmetic differences. They determine whether the yield is likely to persist, how it changes with market conditions, and what can go wrong.

What does 'yield' actually mean in DeFi?

The first thing to separate is yield from the products that generate it. Yield is not itself a protocol. It is the increase in value or balance that a user earns by supplying capital to some DeFi mechanism.

That sounds obvious, but a lot of confusion comes from treating “yield” as though it were a standalone box you can deposit into. In reality, yield is always downstream of some other activity. A lending protocol can generate yield because someone wants to borrow. A DEX pool can generate yield because someone wants to trade through it. A vault can generate yield because it routes capital through one or more strategies that themselves depend on lending rates, trading fees, or reward emissions.

So the right first-principles question is: **what real on-chain action causes value to move from one set of users to another? ** If no one is paying borrow interest, no lending yield appears from nowhere. If no one is swapping, LP fee yield does not appear from nowhere. If the only source of return is token emissions, then the yield depends on the value and persistence of those emissions rather than on organic usage.

This is the simplest organizing principle for DeFi yield: there are two broad economic sources of return, and many products are mixtures of them. The first is activity-based revenue such as interest and trading fees. The second is incentive-based rewards such as governance-token emissions. Research surveys of DeFi describe yield in essentially these terms: interest and accrued fees on one side, token rewards on the other. That distinction is the difference between revenue generated by users of the protocol and subsidies paid to attract users to the protocol.

How do DeFi protocols generate and distribute yield?

A protocol cannot distribute what it does not somehow obtain. That statement is almost embarrassingly simple, but it is the key to making yield click.

Suppose users deposit a stablecoin into a lending pool. Another group of users borrows that stablecoin and pays interest over time. The protocol keeps some spread or fee, and the rest accrues to suppliers. That is yield as a transfer from borrowers to lenders, mediated by smart contracts.

Now suppose instead that users provide two assets to an AMM pool. Traders use the pool for swaps and pay fees. Those fees accumulate to liquidity providers in proportion to their share of the pool. That is yield as a transfer from traders to liquidity providers.

Now add token rewards. A protocol may distribute governance tokens to depositors or LPs on top of interest or fees. That creates an additional source of return, but it is economically different. The value is not coming from borrowers or traders directly. It is coming from the protocol issuing or allocating tokens in order to incentivize participation. Sometimes that is a temporary subsidy to bootstrap liquidity. Sometimes it is tied to governance rights or claims on protocol earnings. Sometimes it is simply inflationary compensation that looks attractive until the token price falls.

So when someone asks, “What is the yield here?”, the complete answer usually has the form: *this much comes from activity, this much comes from incentives, and here is the mechanism by which each part reaches the user. *

How does lending produce yield for suppliers?

Lending markets are the cleanest place to see DeFi yield as a mechanism rather than a slogan. Protocols such as Compound and Aave let users supply assets into pools and let other users borrow against collateral. Suppliers earn interest because borrowers pay for access to liquidity.

Compound’s whitepaper describes each asset market as a pool with algorithmically derived interest rates based on supply and demand. That phrase matters. There is no bilateral negotiation between a specific lender and a specific borrower. Instead, the protocol aggregates all supply and all borrowing into a market, then adjusts rates according to market conditions.

The central variable is usually utilization: the fraction of available liquidity that has been borrowed. In Compound’s formulation for asset a, utilization is U_a = Borrows_a / (Cash_a + Borrows_a). Here Cash_a means unborrowed liquidity in the pool, and Borrows_a means the amount currently borrowed. When utilization rises, liquidity is becoming scarce relative to borrowing demand. A rational protocol then tends to raise borrowing rates to discourage additional borrowing and attract more supply. When utilization falls, borrowing is easier and rates usually decline.

That is the mechanism behind lending yield. Supplier yield is not chosen independently. It is derived from the borrow side. Compound’s whitepaper states that the supply rate is a function of the borrowing rate and utilization, minus a spread that the protocol keeps as economic profit. Euler’s vault audit presents the same structure in a compact form: supplyRate = borrowRate * totalBorrows / totalAssets, where totalAssets = cash + totalBorrows. Different protocols implement different rate curves, but the invariant is similar: suppliers are paid out of what borrowers pay, scaled by how much of the pool is actually in use.

A worked example makes this concrete. Imagine a USDC lending pool with 10 million USDC supplied. If 8 million is borrowed and 2 million remains idle, utilization is 80%. Because liquidity is relatively tight, the protocol’s interest-rate model may set a higher borrow rate. Borrowers pay that rate on the 8 million they have borrowed. Suppliers do not all earn that same headline borrow rate, because not all supplied USDC is deployed and because the protocol may take a spread. Instead, the supplier yield reflects the borrower payments distributed across the full 10 million supply base. If borrowing demand later drops and only 2 million USDC remains borrowed, utilization falls to 20%, rates usually fall, and supplier yield drops with it.

This is why lending APY is fundamentally a market-clearing price for liquidity. The protocol is trying to balance two competing needs: enough idle cash that users can withdraw, and enough borrowing demand that suppliers are paid. If utilization stays too low, suppliers earn little. If utilization stays too high, withdrawals become harder and the system becomes more fragile.

Compound III makes this structure more specific. It allows users to supply crypto assets as collateral in order to borrow a single base asset, initially USDC in the first deployment, and it also lets users earn interest by supplying that base asset. In other words, the protocol narrows the market design but keeps the same economic idea: suppliers of the borrowable base asset earn yield because that asset is demanded by borrowers.

How do lending protocols account for accrued interest without per‑wallet updates?

One subtle question is how protocols can make balances “grow” continuously without rewriting every account on every block. The answer is that they do not try to do that.

Compound’s whitepaper explains a gas-efficient accounting pattern built around an Interest Rate Index. The protocol records how the market index changes over time. Each account stores a checkpoint from the last time it interacted. When the user next deposits, withdraws, borrows, or repays, the protocol compares the current index to the stored checkpoint and computes the accrued interest then.

This matters for understanding yield because it explains what is fundamental and what is merely display. The protocol is not literally sending tiny interest payments to your wallet every second. It is updating a global accounting state and using that state to determine what your claim is worth when needed. Research on lending pools describes the same idea more generally through claim tokens whose exchange rate versus the underlying asset increases as interest accrues.

That exchange-rate view is often clearer than “my balance went up.” If you deposit into a lending market and receive an interest-bearing receipt token, what grows over time may be the value of each receipt token in terms of the underlying asset, not necessarily the raw token count in your wallet. Either way, the economic content is the same: your claim on the pool increases because borrower payments are being attributed to suppliers.

How do liquidity providers earn fees from traders?

Yield in AMMs comes from a different mechanism. Instead of borrowers paying for liquidity over time, traders pay for immediate execution.

Protocols such as Uniswap, Curve, Raydium, and Osmosis let users provide assets to trading pools. In return, LPs receive a claim on a share of the pool and on the fees generated when others trade against it. Uniswap’s documentation puts the user-facing point plainly: provide liquidity to a pool using tokens to earn fees. Raydium describes LP yield even more directly as trading fees plus token rewards.

The mechanism here is simpler than in lending. A trader wants to swap asset X for asset Y. The AMM quotes a price based on pool reserves and the pool’s rule, such as a constant-product invariant. The trader pays a fee as part of the swap. That fee stays in the system and accrues to LPs, typically in proportion to the liquidity they supplied.

The economic reason yield exists is that LPs are warehousing inventory and allowing others to trade against it. Traders are paying for that service. But unlike lending, the risks are not mainly about default by borrowers. They are about inventory exposure. If prices move a lot, an LP may end up holding a different mix of assets than they started with. The trading fees can compensate for that, but they may not. This is why LP yield and realized LP profit are not identical.

Concentrated-liquidity designs make this clearer. Raydium’s CLMM pools, like similar designs elsewhere, let LPs choose a price range in which their capital is active. If trading happens in that range, fee earnings per unit of active capital can be higher because the capital is used more efficiently. But the position now depends more heavily on active management. If the market price leaves the chosen range, the position stops earning fees until it is repositioned. So a higher advertised fee APR in a concentrated pool is not “free yield”; it is payment for a more conditional and management-sensitive form of liquidity provision.

Stable-asset AMMs such as Curve push in another direction. They are designed so that similar-priced assets can trade with lower slippage, which tends to make such pools attractive venues for high-volume stablecoin trading. That can produce substantial fee-based yield for LPs. But again, the yield depends on actual flow and on the specific fee and incentive architecture of the protocol.

Are token rewards sustainable as yield or just temporary subsidies?

Many DeFi protocols add another layer on top of interest or fees: they distribute tokens to users who supply capital. This is often called liquidity mining, but the underlying economic point is broader than that specific term.

Aave documents Incentive Programs as a first-class part of its markets. Compound III has a dedicated rewards contract that can hold tokens such as COMP or WETH and lets users claim rewards according to tracking indices. Raydium supports permissionless farms that add token incentives to pools. Curve’s ecosystem is famous for governance-linked reward allocation. Across these systems, the common idea is that the protocol wants to attract and retain liquidity, so it pays users extra in the form of token distributions.

This is still yield in the narrow sense of “additional return on supplied capital,” but it is different from activity-based yield in an important way: it is often a subsidy rather than organic revenue. If a pool is offering 3% from fees and 20% from token emissions, the sustainability of the total headline APY depends heavily on whether those emissions continue and on what the reward token remains worth when recipients sell or hold it.

That does not make incentive yield fake. Protocol-native token rewards can matter economically, especially when the tokens grant governance rights or claims on protocol earnings. Research on DeFi often compares these rewards to equity for that reason. But it does mean the user should ask a second question: *is this yield being earned from users of the protocol, or is it being paid by the protocol to shape behavior? *

That distinction explains why very high APYs are often temporary. Incentives can bootstrap a market, deepen early liquidity, or direct capital to strategic pools. But unless the protocol eventually generates enough underlying activity to replace subsidies with durable fee or interest revenue, the yield is unlikely to remain at the same level.

How do vaults and aggregators generate and compound yield, and what risks do they add?

Some protocols do not generate yield directly. Instead, they package, route, and compound yield from other protocols.

Yearn Vaults v2 describe themselves as a way to generate risk-adjusted returns by composing lending protocols, liquidity pools, and yield-farming strategies on Ethereum. Beefy describes its core product similarly: vaults tie into strategies that automatically compound reward tokens from LPs, AMMs, and other farming opportunities back into the originally deposited asset.

The key idea here is that a vault is an automation layer. Rather than asking a user to deposit into a farm, harvest rewards, sell them, add liquidity again, and re-stake, the vault handles that loop in smart contracts. That can improve convenience and may improve returns through more consistent compounding. But the vault does not create value ex nihilo. It depends on the underlying protocols’ interest, fees, and incentives.

This is where composability becomes both powerful and dangerous. A vault may hold LP tokens from one protocol, borrow against them somewhere else, farm rewards in a third place, and periodically swap through a DEX aggregator to compound. Each step can be sensible on its own. Together, they create more moving parts, more smart contracts, more price dependencies, and more assumptions about execution. Research on DeFi emphasizes that composability increases systemic risk precisely because failures can cascade across connected protocols.

A practical example is automatic compounding of LP rewards. Suppose a vault deposits assets into an LP farm that pays trading fees plus reward tokens. The vault harvests those reward tokens, swaps them back into the underlying pair, adds more liquidity, and restakes the resulting LP position. The user’s balance grows because rewards are repeatedly converted into additional productive capital. That is a genuine mechanism. But it also introduces swap risk, smart-contract risk in the vault and farm, and dependency on reward-token liquidity.

How does blockchain choice affect yield mechanisms and risk?

The idea of yield is not specific to Ethereum. The same economic patterns show up across blockchain architectures.

On Solana, Raydium offers yield through liquidity provision, where LPs earn trading fees plus token rewards. Jupiter, while not a yield protocol itself, affects where trading flow is routed, which indirectly affects which pools generate fee revenue for LPs. On Cosmos, Osmosis frames itself as a cross-chain DEX and DeFi hub where users can swap and provide liquidity, and its tooling exposes AMM math and pool logic that underlie LP returns. The cash-flow source remains the same: traders pay fees, and those fees accrue to liquidity providers, sometimes augmented by incentives.

On the lending side, the same utilization-based intuition travels well across implementations. Whether a protocol is on Ethereum or another chain, if lenders supply capital into a pool and borrowers draw from that pool, supplier yield generally depends on borrower demand, collateral policy, and the protocol’s interest-rate model. The interface and contract layout may differ, but the economic structure does not.

This is why comparisons across chains should focus less on the raw APY number and more on the mechanism beneath it. A 12% yield funded mainly by active borrowing demand is not the same thing as a 12% yield funded mainly by short-lived token emissions, even if both are technically “yield.”

What specific risks can make a quoted APY worthless?

Because yield is downstream of protocol mechanisms, the failure modes are also mechanism-specific.

In lending protocols, the central assumptions are that collateral remains sufficient, prices are measured correctly, and liquidation works when needed. Compound’s whitepaper describes liquidation as automatic and permissionless: if an account’s collateral falls below the required ratio, anyone can buy that collateral at a discount by repaying the debt. This is the system’s way of enforcing solvency. If liquidations fail, or if the oracle price is wrong, suppliers may discover that the promised yield was paired with hidden loss risk.

Research on lending pools highlights exactly this issue: LP-style claim tokens may be theoretically redeemable for the underlying, but in stressed states their practical value depends on whether the pool actually has enough liquidity. High utilization can mean the pool is solvent in accounting terms yet short of immediately withdrawable cash. That does not always imply insolvency, but it does mean a headline APY says nothing by itself about withdrawal conditions.

In AMMs, the classic misunderstanding is to equate fee APR with total return. LPs can earn fees and still underperform simply holding the assets if price movements change the inventory mix unfavorably. Concentrated liquidity increases this management sensitivity further.

In incentive-driven systems, the main weakness is reflexivity. High emissions attract capital because the APY looks high. That extra capital dilutes rewards per dollar deposited, and users often sell the reward token, which can push down its price and reduce the real value of the advertised yield. The nominal APY may still look impressive while the actual economic value decays.

Composability adds another layer of breakage. The Alpha Homora V2 exploit is a sharp example of how a yield-oriented system can fail through accounting and integration flaws rather than through a simple missing permission check. The post-mortem describes a flash-loan-assisted attack that exploited debt-share accounting and reserve-handling issues, creating roughly $38 million in bad debt between protocols. The lesson is not that all yield systems are flawed in the same way. The lesson is that a yield strategy often inherits the deepest assumptions of every protocol it touches.

How do protocol control and custody affect yield and risk?

One last layer of understanding is operational rather than economic: how protocol control is implemented.

Many DeFi systems are upgradeable and parameterized. Compound III, for example, uses a fixed market proxy for user interaction, a Configurator for setting and updating parameters and deploying implementations, and a rewards contract for token distributions. That architecture exists partly to save gas and partly to support controlled upgrades, but it also means users are relying on governance and admin processes as well as on immutable code.

This matters for yield because parameters such as interest-rate models, supported collateral, incentive schedules, and liquidation settings directly shape returns. Yield is therefore not just a formula embedded forever on-chain. It is often a formula plus a governance process.

Where protocols or exchanges use threshold cryptography for settlement, the custody model changes in a useful way. **Cube Exchange uses a 2-of-3 Threshold Signature Scheme for decentralized settlement: the user, Cube Exchange, and an independent Guardian Network each hold one key share. No full private key is ever assembled in one place, and any two shares are required to authorize a settlement. ** This is not a yield mechanism by itself, but it is a concrete example of how real systems can reduce single-point-of-failure custody risk around asset movement. That matters because the value of yield is only meaningful if the path to holding and settling assets is itself robust.

How should I evaluate an APY before investing? (4‑question checklist)

If you want a durable intuition, ask four questions every time you see an APY.

First, **what activity generates the return? ** Borrowers paying interest, traders paying fees, token emissions, or a mix? This is the root question.

Second, **what state variable drives it? ** In lending, it is often utilization. In AMMs, it is trading volume relative to deployed liquidity. In reward systems, it is the emission schedule and token price.

Third, **what assumptions must keep holding? ** Correct oracle prices, healthy collateral, liquid reward-token markets, smart-contract safety, and functioning liquidations are common examples.

Fourth, **what part is organic and what part is subsidized? ** A yield made mostly of fees or interest has a different sustainability profile from a yield made mostly of token emissions.

These questions do not eliminate risk, but they turn yield from a marketing number back into an economic mechanism you can reason about.

Conclusion

DeFi yield is value distributed to capital providers because a protocol is doing something others will pay for, or because the protocol is subsidizing participation. Everything else is detail.

In lending markets, borrowers fund suppliers. In AMMs, traders fund LPs. In vaults, automation packages and compounds those underlying flows. And when token rewards are added, the return may rise sharply; but part of that rise may be subsidy rather than durable revenue.

The shortest version to remember tomorrow is this: **yield is not magic income from idle coins; it is a claim on some underlying mechanism of payment, and the quality of the yield is only as good as that mechanism. **

How do you trade through a DEX or DeFi market more effectively?

Trading through a DEX or other DeFi market effectively means matching execution strategy to on-chain liquidity and expected price impact. Use Cube Exchange to fund the trade and place the order, but first inspect the pool depth, recent volume, and slippage risk on the target DEX so you pick the right execution approach.

1. Fund your Cube account with the asset or fiat you will use to trade (on‑chain deposit or fiat on‑ramp).
2. Inspect the target pair’s on‑chain liquidity and recent volume on the DEX explorer or an analytics site; note pool depth and a realistic price‑impact estimate for your trade size.
3. On Cube, choose an execution: use a limit order to avoid slippage on thinner markets or a market/swap for immediate fills; if the fill might move the price, split the order into smaller slices.
4. Set a slippage tolerance and deadline for any on‑chain swap, review estimated fees and total price impact, then submit the order.