What it is
Peer-to-pool lending is the dominant model for on-chain lending (money markets like Aave and Compound): lenders deposit an asset into a shared pool and borrowers draw from that same pool by posting collateral, with no need to find a matching counterparty. Interest rates are set algorithmically by a formula tied to pool utilization, not negotiated between individuals.
How it works
- Lenders (suppliers) deposit an asset — say USDC — into the pool and receive an interest-bearing receipt token representing their claim on the pool plus accrued interest.
- Borrowers deposit a different asset as collateral (often over-collateralized, e.g. depositing $150 of ETH to borrow $100 of USDC) and draw liquidity directly from the shared pool.
- An interest-rate model reads the pool's utilization ratio (borrowed / total supplied) and sets both the borrow rate and supply rate — as utilization rises toward 100%, rates spike sharply to discourage further borrowing and attract more supply.
- Interest accrues continuously; borrowers' debt balance grows over time, and suppliers' receipt-token balance (or exchange rate) grows correspondingly, funded by that borrower interest.
- Each collateral asset has a loan-to-value (LTV) limit and a liquidation threshold. If a borrower's collateral value falls (or debt value rises) past that threshold, anyone can trigger a liquidation: repaying part of the debt in exchange for a discounted chunk of the borrower's collateral.
- Because lenders share one pool, withdrawal requires enough unborrowed liquidity sitting idle — if utilization is near 100%, some lenders may be temporarily unable to withdraw until borrowers repay or get liquidated.
Why designers use it
- No need to match specific lenders with specific borrowers or negotiate terms — anyone can supply or borrow instantly at the prevailing pool rate.
- Deep, fungible liquidity: a single pool aggregates capital from many lenders, giving borrowers reliable access and lenders passive yield.
- Rates respond automatically to supply and demand, removing the need for a human rate-setting committee.
- Composable: pool receipt tokens and borrowing positions can be built on top of by other protocols (collateral for other loans, yield strategies, etc).
Failure modes
- Utilization spikes to near 100% during stress, freezing withdrawals for lenders even though their funds are technically "there" — just lent out.
- A sharp price drop in a collateral asset can trigger cascading liquidations faster than liquidators (and the chain) can process, leaving the pool undercollateralized and lenders holding bad debt.
- Oracle price feeds that lag or can be manipulated let attackers borrow against inflated collateral or trigger unjust liquidations.
- Long-tail or illiquid collateral assets can be listed with LTVs too generous for their real market depth, meaning liquidation sales themselves crash the price and leave the protocol with unrecovered debt.
- Interest-rate model mis-tuning (too flat near full utilization) fails to discourage borrowing in time, letting utilization pin at 100% and locking lender funds for extended periods.
What to check before using it
- Verify the interest-rate curve steepens sharply well before 100% utilization, not just at the edge.
- Confirm collateral assets use liquid, manipulation-resistant price oracles, ideally time-weighted or from multiple independent sources.
- Check liquidation incentives (discount, bot competition) are large enough to actually get positions liquidated during fast market moves, not just on paper.
- Model a worst-case simultaneous price crash across correlated collateral assets and confirm the protocol survives without insolvency.
- Confirm withdrawal behavior under high utilization is disclosed to lenders — "instant liquidity" claims should be tested against real utilization history.