How Interest Rates Work in DeFi
How Interest Rates Work in DeFi
Interest rates are the mechanism through which lending protocols balance supply and demand. In traditional finance, a bank's economist might set a savings rate, and the market accepts it or moves their money elsewhere. In DeFi, rates are determined algorithmically and update continuously. Understanding how these algorithms work and why they differ from traditional finance is essential for anyone seeking to optimize yield or minimize borrowing costs.
The Fundamental Problem: Matching Supply and Demand
Before diving into mechanisms, consider the fundamental challenge: how much should the protocol pay lenders and charge borrowers? If rates are too low, no one will deposit capital, and the protocol lacks funds to lend. If rates are too high, borrowers can't afford to borrow, and capital sits idle earning nothing.
The traditional finance solution is for an institution's management to adjust rates based on their understanding of market conditions and competitive dynamics. But this approach is subjective and slow. A bank might change its prime rate quarterly; in the interim, rates might be too low (capital leaves) or too high (loans can't be made).
DeFi's solution is transparency and market mechanisms. The utilization rate—the percentage of capital currently lent—is a transparent signal of supply-demand balance. High utilization means capital is scarce; low utilization means capital is abundant. By making interest rates a function of utilization, protocols create automatic, transparent incentives.
Utilization Rate: The Core Signal
Utilization rate is calculated as total borrowed divided by total supplied:
Utilization = Total Borrowed / (Total Supplied)
If a lending pool has 1,000 units of capital supplied and 800 units borrowed, utilization is 80%. This single number captures the capital market's state—is capital abundant or scarce?
When utilization is low (say, 20%), the protocol is oversupplied with capital. Many lenders are competing for limited borrowing demand. Competition for borrowers drives down rates. Lenders accept lower yields because the alternative is earning nothing.
When utilization is high (say, 95%), capital is scarce. Borrowers are competing for limited available capital. The protocol raises rates to ration capital, discouraging new borrowing and incentivizing lenders to supply more.
Utilization as a signal is powerful because it's objective, transparent, and responsive. No committee vote is required; the algorithm acts automatically based on on-chain data.
The Interest Rate Curve
Interest rates in DeFi protocols are modeled as a function of utilization. Different protocols use different functional forms, but the underlying principle is consistent: interest increases with utilization.
The simplest model is linear: interest rate increases proportionally to utilization. If utilization is 50%, the rate might be 5%; at 100%, it might be 10%. Linear models are simple to understand but poorly reflect capital scarcity dynamics. At low utilization, they underdifferentiate—the rate barely increases despite abundance. At high utilization, they may not increase sharply enough.
Most modern protocols use kinked or piecewise linear curves. Below a "kink point" (often 80%), rates increase gradually. Above the kink, rates increase sharply. This creates two regimes:
Normal regime (low to moderate utilization): Rates increase slowly, creating stable yield expectations and reasonable borrowing costs. This regime encourages efficient capital deployment and attracts both lenders and borrowers.
Emergency regime (high utilization): Rates spike, rationing scarce capital to the most critical uses. This protects the protocol from over-leverage and incentivizes repayment or additional collateral.
The kink point is a governance parameter. Set it too low (say, 60%), and the protocol frequently enters expensive emergency rates, discouraging borrowing. Set it too high (say, 95%), and the protocol allows dangerously high utilization before rates spike, risking insolvency if withdrawals are requested en masse.
Comparing Protocols' Rate Models
Different protocols implement different rate curves. Aave uses kinked curves with steep increases above the kink. Compound uses similar models but with different parameters for each asset. Newer protocols experiment with alternative models—some use exponential rather than piecewise linear curves; others implement complex formulas that adjust not just to utilization but to historical volatility, asset correlations, and other factors.
These differences matter. A borrower might find Aave's rates reasonable on one asset while Compound offers better terms on another. An algorithmic yield optimizer might constantly move capital between protocols to capture the best rates.
Dynamic Rate Adjustment and Market Evolution
Interest rate curves are not immutable. Protocols adjust them—through governance votes—in response to market conditions. When a protocol is launching and seeking to attract capital, governance might lower the kink point or slope, making rates attractive even at low utilization. Conversely, after the protocol has grown and accumulated capital, governance might increase rates to reward lenders for their liquidity.
These adjustments reflect a deeper reality: markets evolve. In 2020, DeFi was nascent and lending rates were sky-high—20%, 30%, sometimes 50%+ annually—because capital was extremely scarce and demand was explosive. By 2024, as capital has accumulated, competition has intensified, and excitement has cooled, sustainable DeFi yields have fallen to 5–10% for stablecoins, matching or slightly exceeding traditional finance alternatives.
This normalization is healthy. Unsustainable yields were subsidized by governance token distributions and reflected speculative demand rather than fundamental scarcity. As these incentives ended, yields fell toward levels supported by real economic activity—borrowers earning returns on deployed capital, spreads between lending and borrowing rates covering protocol risk.
Variable vs. Stable Rates
Most modern DeFi lending protocols support two types of interest rates for borrowers:
Variable rates fluctuate automatically with the protocol's rate curve. If you borrow at a 5% variable rate and utilization spikes, your rate might jump to 15% within minutes. This exposes borrowers to rate risk but aligns their costs with actual capital scarcity.
Stable rates are fixed for the duration of the loan (or until the protocol governance changes them). You might borrow at 5% fixed for a year, knowing your cost regardless of what happens to utilization. Stable rates protect borrowers from shocks but expose lenders and the protocol to duration risk—if rates should rise but they're locked in at low levels, capital becomes misallocated.
Providing stable rates requires the protocol to hedge duration risk or incentivize borrowers to repay and reborrow at new rates. Aave manages this by allowing governance to reset stable rates periodically, preventing them from becoming too detached from equilibrium.
Spreads and Protocol Revenue
The difference between the interest rate paid to lenders and charged to borrowers is the spread. In traditional banking, a large spread (paying 0.5% on deposits, charging 4% on loans) is normal and funds the bank's operations and profits. In DeFi, where overhead is minimal, spreads are narrower.
A typical DeFi spread is 10–20% of the borrowing rate. If borrowers are charged 10% and lenders receive 9%, the 1 percentage point difference (10% of the 10% borrowing rate) goes to the protocol. This revenue is captured by the protocol's treasury or distributed to governance token holders.
These spreads vary by asset. Stablecoins—the most stable and most traded assets—have lower spreads (sometimes just 2–5%) because competition is fierce and risk is low. Altcoins with lower volume and higher volatility have higher spreads (sometimes 30–50%), reflecting higher risk and lower competition.
Oracle Impact on Interest Rates
Interest rates affect liquidation mechanics, which depend on oracle prices. If an oracle reports an inflated price for a collateral asset, borrowers appear more solvent than they are. If the true price is lower, they're undercollateralized and should be liquidated. This creates a feedback loop: oracle accuracy affects risk, which affects sustainable rate curves, which affects yields.
Protocols respond to oracle risk by using conservative risk parameters. They accept lower utilization thresholds, higher collateral requirements, and lower liquidation discounts than would be optimal if oracles were perfectly accurate. This safety margin is a drag on efficiency but essential for protocol stability.
Yield Aggregation and Rate Chasing
Sophisticated users employ yield optimization strategies. Rather than depositing in a single protocol, they monitor rates across platforms and move capital to wherever yields are highest. Automated protocols (robo-advisors on the blockchain) facilitate this, automatically reallocating funds when spreads shift.
This rate-chasing behavior has important effects. First, it smooths rate discrepancies across protocols. If Aave offers 6% and Compound 5.5% on the same asset, capital flows to Aave until yields equalize. Second, it creates competitive pressure on protocols to manage rates effectively. Poor rate policies drive capital away.
However, this behavior also creates fragmentation. Capital scatters across protocols rather than concentrating in a single deep market. This can reduce stability and liquidity in each individual protocol while increasing systemic interconnectedness—a crash on one protocol triggers cascades on others that hold its tokens as collateral.
Sustainability and Long-Term Rate Trends
The trajectory of DeFi yields over the past several years points toward stabilization. Yields have fallen from 30%–50% in 2020 to 5–10% in 2024–2025 as capital has accumulated and governance incentive programs have ended or been scaled back. This normalization reflects the ecosystem's transition from speculative growth phase to mature operation.
What yields should DeFi sustain long-term? Theory suggests they should approximate real-world lending yields. If a business can borrow at 8% in traditional finance to fund productive activities, they should be willing to borrow at similar rates on-chain. DeFi yields above this are unsustainable absent additional incentives or risks.
However, DeFi will likely maintain a risk premium over traditional lending due to smart contract risk, oracle risk, and regulatory uncertainty. Yields 2–3 percentage points above comparable traditional rates might be sustainable long-term.
Future Directions
Interest rate mechanisms will continue to evolve. Some protocols are experimenting with AI-driven forecasting, adjusting curves not just based on current utilization but predictive models of future supply and demand. Others are implementing negative interest rates (the protocol pays borrowers to take loans) during periods of extreme oversupply, a mechanism that pushes capital allocation in novel directions.
Cross-protocol yield optimization is becoming more sophisticated. Protocols that coordinate on rate curves might reduce competition and increase stability, but they also risk collusion and reduced efficiency.
As DeFi matures, interest rate mechanisms will become increasingly sophisticated—incorporating more data, adjusting more dynamically, and managing risk more effectively. Understanding how they work is essential for participants who want to optimize yields and for protocol developers building the systems of the future.
Key Takeaways
- Interest rates in DeFi are determined algorithmically based on utilization—the percentage of capital currently borrowed.
- Kinked curves create two regimes: normal rates for healthy utilization, emergency rates when capital is scarce.
- Rates adjust continuously without human intervention, creating market efficiency but also volatility.
- DeFi yields have fallen from speculative highs toward sustainable levels as capital accumulates and governance incentives end.
- Different protocols use different rate models; comparing them helps optimize lending and borrowing decisions.