On-chain lending rates are usually generated by protocol rules that respond to how much liquidity is supplied and borrowed. In pooled markets such as Aave, utilization is a major input: as more of a reserve is borrowed, the rate curve typically raises borrowing costs and supplier returns to attract liquidity and discourage additional demand. But the curve itself—including its target utilization, slopes and other parameters—is set by protocol design and governance. A higher displayed rate therefore reflects both market conditions and protocol rules, not a guaranteed better investment.
How do on-chain lending rates actually work?
Most pooled DeFi lending markets connect two sides of the same reserve. Suppliers deposit an asset into a shared pool. Borrowers draw from that pool, usually after posting collateral under the protocol’s rules. The protocol then uses an interest-rate model to determine what borrowers pay and what suppliers may earn.
The key variable is utilization: the share of supplied liquidity that is currently borrowed. In plain English, utilization asks how much of the pool is already in use. If a reserve has 100 units supplied and 70 units borrowed, its utilization is about 70%.
That number matters because a lending market needs to balance two competing goals. Borrowers want available liquidity at a reasonable cost. Suppliers need enough compensation for making assets available. If utilization climbs too high, the protocol may raise borrowing rates sharply to discourage new borrowing and encourage additional supply.
Supplier APY and borrow APR are related, but they are not the same number. Borrowers pay according to the market’s borrow-rate model, while supplier yield depends on borrowing activity, utilization and the protocol’s own economic parameters. Some markets may also display incentive rewards on top of base supplier yield.
Why utilization changes the rate
Utilization is not just a descriptive statistic. In major pooled protocols, it is an input to the rate curve.
At low utilization, a protocol may keep borrowing relatively inexpensive because there is plenty of idle liquidity. As utilization approaches a target or “kink,” the model can increase rates more aggressively. Above that point, the slope may become much steeper because the pool needs to protect withdrawal liquidity and reduce additional borrowing pressure.
Aave’s current documentation explains that supplier and borrower rates move dynamically with utilization and governance-controlled parameters. Aave governance discussions in 2026 also show that optimal utilization and rate-curve slopes can be actively adjusted. That is why the sentence “supply and demand set the rate” is directionally useful but incomplete.
The more accurate version is: market utilization changes the input, while protocol design determines how strongly the rate responds.
Who sets the formula?
The logic is implemented through smart contracts and protocol parameters. On public chains, that logic and the current parameter state can generally be inspected, but interpreting it often requires technical context.
Protocols do not all use the same formula. Compound’s foundational model, for example, formalized utilization-based supply and borrow rates, while newer markets may use different curves, caps and governance processes.
Governance also matters. Risk stewards or governance processes may alter target utilization, slopes, caps or other parameters as market conditions change. A rate can therefore move because users changed their behavior, because the protocol changed its rules, or because both happened at the same time.
Why the highest rate is not automatically the best opportunity
A headline APY is a snapshot, not a complete decision. A high rate can be attractive, but it can also be a signal that the market is under stress, highly utilized, incentive-heavy or exposed to risks that are not visible in the percentage alone.
Investor.gov has also warned that crypto interest-bearing products do not necessarily carry the same protections as insured bank deposits. That comparison matters because an on-chain rate should not be treated as a bank-equivalent savings rate simply because both are expressed as a percentage.
Rate comparison: what to calculate before moving funds
Separate base supplier yield from incentives. A temporary token reward can make two otherwise similar markets look very different.
Check utilization and available liquidity. A pool close to its target or kink may react sharply to new borrowing or withdrawals.
Estimate switching costs. Network fees, protocol fees and other transaction costs can erase a small APY advantage, especially over short holding periods.
Review the asset and collateral mix. Yield on a stablecoin, volatile token or thinly traded asset should not be treated as equivalent.
Look at protocol and governance history. A current rate exists inside a system whose parameters and risk controls can change.
Match the comparison to the holding period. A 0.5 percentage-point APY difference may be irrelevant if entering and exiting costs more than the expected benefit.
This is where the fee framework from Bluwhale’s investment-cost guide becomes relevant: the economically useful question is the expected net result after observable costs, not just the largest percentage displayed on a dashboard.
Where an AI investment assistant may help
An AI investment assistant can be useful as a monitoring and explanation layer when the comparison problem becomes too large to handle manually.
A safe role is to collect rates and utilization across supported markets, normalize comparable opportunities, estimate switching costs, flag major liquidity or parameter changes, and summarize why a market deserves review.
That can reduce the operational burden of watching multiple protocols without turning the assistant into an oracle. The assistant can surface evidence and changing conditions; it should not imply that the highest current APY is the best choice for a specific person.
Bluwhale brings connected financial context into one view. Alongside a lending platform's rate and position details, that broader perspective can help you understand how a position fits into your financial picture.
What AI cannot know from a headline rate
Whether the current APY will persist after more capital enters the pool.
Whether a protocol will suffer a future exploit, oracle failure or governance shock.
Whether a stablecoin or collateral asset will maintain its value.
Whether fees, slippage and switching costs will make a move worthwhile over the user’s actual holding period.
Whether a higher on-chain yield is suitable for a particular person’s financial situation.
The durable rule: a rate is evidence to interpret, not a promise to obey.
A better question than “Which rate is highest?”
Ask why the rate is high, how durable the conditions appear to be, what liquidity and protocol risks sit behind it, and what the net economics look like after switching costs.
See how an AI financial assistant can monitor changing opportunities
Get started — https://profile.bluwhale.com/