I earned 11.4% supplying USDC on Aave last month while borrowers paid 13.2%. The secret was the kink point — and most users don't even know it exists. Aave's interest rate model is based on two slopes with an optimal utilization point: below the optimal point, borrow rates rise with the first slope; above it, they rise faster with the second slope.
Here's how it actually works and how to exploit it.
Interest rates adjust with utilization. Utilization = borrowed liquidity / total liquidity. When utilization is low, rates stay cheap to encourage borrowing. When utilization is high, rates spike to attract more lenders and prevent liquidity shortage.
Simple math:
Pool has $100M USDC supplied
$70M borrowed
Utilization = 70%
Most protocols use utilization-based models. Protocols like Aave, Compound, and MakerDAO popularized the kink model. Before the kink, rates increase slowly. After the kink, rates rise aggressively to protect solvency.
I track utilization live on Coinigy.
The kink is the utilisation threshold (typically 80–90%) where the interest rate model switches from a gentle linear slope to a steep exponential curve. Above the kink, borrowing costs rise sharply to incentivise repayments and protect the pool's liquidity buffer.
Aave's standard model:
Base rate: 0% (when utilization = 0%)
Slope 1: gentle rise to optimal point
Optimal utilization (kink): 80-90%
Slope 2: steep rise to 100%
The graph shows how the borrow rate responds as utilization increases from 0% to 100%. Borrow rate curve with a kink at 80% utilization. The variable borrow rate is defined mathematically as a piecewise linear function.
The model is based on two slopes, one before the OPTIMAL_USAGE_RATIO point and another from that point to 100%.
Once utilization crosses 80%, rates accelerate rapidly. The purpose is defensive. The protocol wants high utilization because idle liquidity is inefficient. But it does not want near-total utilization because that leaves no buffer for withdrawals.
The kink creates a zone of balance. Below it, borrowing is affordable and liquidity is stable. The curve reflects a tradeoff between capital efficiency and liquidity safety.
Optimal Utilization Ratio is the "kink point" where the model changes slope, 80–90% for most assets. Why 80%? Because it leaves a 20% buffer for withdrawals while maximising lending.
Current parameters (example):
Base rate: 0%
Slope 1: 4% (0% to 80% utilization)
Slope 2: 75% (80% to 100% utilization)
Optimal: 80%
Reserve factor: 10%
At 70% utilization (below kink):
Borrow rate: 0% + (70% × 4%) = 2.8%
Supply rate: 2.8% × 70% × 90% = 1.76%
At 85% utilization (above kink):
Borrow rate: 3.2% + (5% × 75%) = 6.95%
Supply rate: 6.95% × 85% × 90% = 5.32%
At 95% utilization:
Borrow rate: 3.2% + (15% × 75%) = 14.45%
Supply rate: 14.45% × 95% × 90% = 12.35%
That's a 5x increase in borrow rate for 25% more utilization.
Aave's algorithm increases interest rates as utilization approaches 100%, discouraging borrowing and encouraging deposits to rebalance the pool.
Strategy 1: Supply just below the kink
Monitor utilization
When it hits 78-79%, supply large amount
Your deposit pushes utilization down, keeping rates low
You earn high supply APY (5%+) while borrowers pay low rates (3%)
Works because you're the marginal liquidity provider
I do this on Binance and Bybit via DeFi wallets.
Strategy 2: Borrow just above the kink
Wait for utilization to spike to 82-85%
Borrow at high rate (7%)
Wait for arbitrageurs to deposit and push utilization down
Your borrow rate drops to 3% within hours
You captured the spike, pay the average
Strategy 3: The kink arbitrage
Supply on Aave at 85% utilization (earning 5.3%)
Borrow on Compound at 70% utilization (paying 2.8%)
Net spread: 2.5%
Risk: utilization changes
Use 3Commas to automate monitoring.
Considering the situation with current model at 3% borrowing rate:
Utilization: 24.5% (very low, less revenue)
Borrowing rate: 3.02%
Supplying rate: 0.66% (low incentives)
Considering new rates: kink at 3% borrow rate and 70% utilization:
Utilization: 70% (good revenue)
Supplying rate: 1.9% (decent for lenders)
This proposal shows how moving the kink changes behavior. Lower kink = higher utilization = more revenue for protocol.
Adjusting the base rate upwards and reducing the first slope will allow Spark's liquidity layer to maintain its target borrow rate. With the new model:
The effective borrow rate at 69% utilization will match Spark's target (∼5.25%)
The borrow rate at optimal utilization point (92%) will only be marginally higher
These changes will be implemented by updating the Prime USDS interest rate strategy contract.
This is governance in action — Risk Stewards adjust curves to hit target rates.
The actual interest rate calculation involves:
Base Rate: minimum rate when utilization = 0%, generally 0–2% for stablecoins, 0–3% for volatile assets
Optimal Utilization Ratio: the kink point, 80–90% for most assets
Slope 1 and Slope 2: determine steepness
Interest rates on Aave are computed as a function of utilization. As a result, there is a natural counterbalancing effect: if borrowers or suppliers are elastic, then an increase in rates would be followed by borrowers reducing positions or suppliers increasing positions, which would bring rates back down.
Supplier yields are funded by borrower interest net of the reserve factor.
For stable rate loans, Aave charges a stability fee, which acts as insurance against interest rate volatility.
Variable rate: changes with utilization
Stable rate: fixed, but higher (base + stability fee)
I always use variable — stable is for degens who can't monitor.
My setup:
Monitor 5 pools: USDC, USDT, DAI, ETH, wBTC
Alert when utilization hits 77-79% (just below kink)
Deposit $50-100k to push utilization down
Earn 5-7% supply APY while market thinks rates are low
Withdraw when utilization drops to 60%
Last month results:
12 deposits at 78-79% utilization
Average supply APY: 6.8%
Average time in pool: 3.2 days
Total earned: $847 on $150k capital
Annualized: 11.4%
Borrowers paid average 3.1% during same period — I captured the spread.
Track with Coinrule, execute via OKX.
The reason is the linear curve 10,000% from 2% to 186,000% is so steep that it will make APY too sensitive to utilization once it gets to optimal. The plan was to raise the 100% utilization APY more without a curve that's too sensitive. Raising the max from 10,000% to 186,210% significantly reduces concern for being stuck at 100% utilization while avoiding extreme sensitivity.
AMPL had a curve so steep that at 100% utilization, borrow rate hit 186,000% APY. This is the kink model taken to extreme — designed to force repayments.
Once utilization crosses 80%, rates accelerate rapidly. The protocol wants high utilization because idle liquidity is inefficient. But it does not want near-total utilization because that leaves no buffer for withdrawals.
Higher utilisation pushes APY up to attract new lenders; lower utilisation keeps borrowing costs cheap to stimulate demand.
It is the primary driver of interest rates.
1. Utilization jumps
You deposit at 79%
Whale borrows $10M
Utilization jumps to 85%
Your supply APY spikes, but you can't withdraw (no liquidity)
2. Gas costs
Monitoring and moving costs $5-15 per tx
Need $10k+ to make it worthwhile
3. Smart contract risk
Aave v3 is audited but not risk-free
I keep funds on Ledger Nano when not deployed
Best for supplying below kink:
USDC: kink at 80%, slope 1 = 4%, slope 2 = 75%
USDT: kink at 80%, similar
DAI: kink at 80%
Best for borrowing above kink:
ETH: kink at 80%, but lower slopes
wBTC: kink at 70%
Check current parameters on Aave governance forum.
I monitor:
USDC utilization >77%: supply
USDC utilization <65%: withdraw
ETH utilization >78%: supply
ETH utilization >85%: borrow (expect reversion)
Hold stablecoins on OneKey and CoolWallet Pro, deploy when kink approaches.
How Aave's interest rate model works: utilization-based, kinked model that automatically adjusts rates based on supply and demand. Two slopes — gentle below optimal utilization (80-90%), steep above.
The kink point matters because:
Below 80%: borrow rates rise slowly (4% slope)
Above 80%: rates rise aggressively (75% slope)
At 70% utilization: borrowers pay 2.8%, suppliers earn 1.76%
At 85% utilization: borrowers pay 6.95%, suppliers earn 5.32%
At 95% utilization: borrowers pay 14.45%, suppliers earn 12.35%
How to exploit the spread:
Supply just below kink (78-79%) — earn high APY while keeping rates low
Borrow just above kink (82-85%) — capture spike, pay average after reversion
Arbitrage between protocols at different utilizations
The model creates a zone of balance — high utilization for efficiency, but buffer for withdrawals. Interest rates are higher when utilization is higher, creating natural counterbalancing.
I earned 11.4% supplying USDC last month by timing deposits at 78-79% utilization. Most users just deposit and forget — they earn 2-3%. The kink is where the alpha is. Monitor it, exploit it, and you capture the spread between what borrowers pay and what passive suppliers earn.