How Lending Protocols Work
Having covered swapping and liquidity provision, this module turns to the third major DeFi activity: lending and borrowing. Protocols like Aave and Compound let two distinct kinds of users interact through a shared, pooled smart contract: lenders, who deposit assets to earn interest, and borrowers, who deposit collateral and borrow against it. Unlike a traditional bank, there's no loan officer evaluating your creditworthiness or a committee approving your application — the entire relationship is governed by code and collateral.
Here's how it works mechanically: lenders deposit an asset — commonly a stablecoin, though many protocols support a range of assets — into a shared pool. In return, they receive a token representing their deposit plus accruing interest, and they can withdraw at any time, subject to the pool having enough available liquidity (if nearly all of the pool is currently borrowed out, a lender may need to wait for borrowers to repay before withdrawing, which is one reason utilization is tracked closely). Borrowers, separately, deposit a different asset as collateral — for example, ETH — and can borrow up to some percentage of that collateral's value (covered in depth in the next lesson on collateral ratios) in a different asset, like a stablecoin, paying interest on what they've borrowed.
The interest rates on both sides are not set by a human — they adjust algorithmically based on utilization, meaning the percentage of the pool's total deposits that are currently borrowed out. When utilization is high (most of the pool is borrowed), interest rates rise automatically on both sides, incentivizing more lenders to deposit (attracted by the higher rate) and discouraging additional borrowing. When utilization is low, rates fall, encouraging borrowing and discouraging new deposits. This self-balancing mechanism is what allows the protocol to function entirely without human rate-setting.
Worked Example
Imagine a stablecoin lending pool currently showing a supply APY of 4% and a borrow APY of 6%, with a utilization rate of 70% displayed on the protocol's dashboard — meaning 70% of all deposited stablecoins are currently borrowed out by other users. A lender who deposits 10,000 USDC earns that 4% annualized, paid continuously as the pool accrues interest, and can withdraw at essentially any time as long as the remaining 30% of the pool (plus any new deposits) covers it. A borrower, separately, deposits ETH as collateral and borrows some amount of that same USDC pool, paying the 6% borrow rate — the 2-percentage-point gap between the 6% borrowers pay and the 4% lenders earn is the protocol's "spread," which typically funds a reserve buffer and, in many protocols, a share that goes to the protocol's treasury or token holders.
If a wave of new borrowing suddenly pushes utilization from 70% to 95%, both rates would rise automatically and immediately — the supply APY might jump to 7% and the borrow APY to 11%, reflecting the increased scarcity of available liquidity to borrow and the increased reward needed to attract new lenders to replenish the pool, all without any human adjusting a setting.
It's also worth understanding why borrowers bother with this arrangement at all, rather than simply selling their ETH outright if they need stablecoins. Borrowing against collateral lets a holder access liquidity without giving up their underlying asset's potential price appreciation (and, in some jurisdictions, without triggering a taxable sale event the way an outright sale would) — a trader who believes ETH will rise further can borrow stablecoins against it to fund another opportunity while still holding their ETH position, accepting interest cost and liquidation risk (covered in depth in the next two lessons) as the price of that flexibility. This is also the core mechanic behind "leveraged" DeFi strategies: borrowing against one asset to acquire more of another, which multiplies both potential gains and potential losses, and significantly raises liquidation risk relative to simple, single-asset lending.
Common Pitfalls
- Assuming supply APY and borrow APY are fixed numbers rather than live, utilization-driven rates. Both can change meaningfully within hours during periods of high market activity — check current rates immediately before acting, not from memory.
- Not distinguishing between simple lending (deposit only, no borrowing) and borrowing against collateral. The first carries essentially no liquidation risk; the second does, and the next two lessons focus specifically on managing that risk.
- Overlooking protocol-specific risk differences. Not all lending protocols are equally audited, equally liquid, or equally battle-tested — the mechanics described here are general, but real risk varies significantly by specific protocol.
Exercise
Look up the current supply APY and borrow APY for a major stablecoin on a real lending protocol's dashboard (or use a cached example if live access isn't available), and note the utilization rate shown for that same asset.
Explain, in one sentence, why the borrow APY is structurally always higher than the supply APY for the same asset on the same protocol — think about where the interest borrowers pay actually goes, and who it has to cover. Then write a second sentence predicting, directionally, what would happen to both rates if utilization suddenly spiked toward 100% — you'll use this intuition directly in the next lesson on collateral ratios and liquidation.
Recap
- Lenders deposit to earn interest; borrowers deposit collateral to borrow a different asset, with rates set algorithmically by utilization rather than a human underwriter.
- Borrow rates are structurally higher than supply rates on the same asset — the spread covers the pool's buffer and, often, protocol revenue.
- Rising utilization pushes both rates up automatically, balancing the incentive to lend against the incentive to borrow without manual intervention.