A recursive ETH staking loop deposits wstETH on a lending market, borrows ETH against it, stakes that ETH again, and repeats until the position reaches a target multiple. In late October 2026, the pay is thin: Lido reported a stETH APR of 2.23% on October 7, and borrowing WETH on Aave cost roughly 2% in September and October. Each turn of the loop adds only about 0.4 percentage points of carry, while your liquidation buffer shrinks fast. This article runs the net return and liquidation numbers at each multiple, shows how loops failed in 2022 and 2026, compares Aave, Morpho and Fluid, and helps you decide whether to loop and at what size.
Panaprium est indépendant et pris en charge par les lecteurs. Si vous achetez quelque chose via notre lien, nous pouvons gagner une commission. Si vous le pouvez, veuillez nous soutenir sur une base mensuelle. La mise en place prend moins d'une minute et vous aurez un impact important chaque mois. Merci!
How a Loop Pays: The Net APR Math
If you are new to wstETH itself, read this liquid staking ETH beginner guide first. A loop only adds borrowing on top. You earn staking yield on all the wstETH you hold and pay the borrow rate on all the ETH you owe.
Lido stETH holders earned 2.38% over the first nine months of 2026, down from 5.26% in 2021, and Lido keeps 10% of rewards. WETH borrowing cost 2.08% on Aave Ethereum Core and 1.98% on Arbitrum in September. On Morpho, the 96.5% wstETH/WETH market charged 1.94% and the 94.5% market 1.76% in the snapshot I read.
I use a 2.4% staking yield and a 2.0% borrow cost below. The formula is:
Net APR = stake yield + (multiple − 1) × (stake yield − borrow rate)
At 10x that is 2.4% + 9 × 0.4% = 6.0%. The spread can also flip because Base WETH borrowing sat near 2.50% in late summer, which is above the stETH yield.
Liquidation Math: How Much Room You Have
LTV is debt divided by collateral value. The health factor (HF) is liquidation threshold × collateral ÷ debt, and 1.0 or lower means liquidation. Aave’s ETH-correlated E-Mode allows 93% max LTV, a 95% liquidation threshold, and a 1% penalty.
At multiple x, HF = 0.95 × x ÷ (x − 1). The price drop that triggers liquidation is 1 − 1 ÷ HF.
|
Multiple |
LTV used |
Net APR |
Health factor |
wstETH/ETH drop to liquidation |
|
2x |
50.0% |
2.8% |
1.90 |
47.4% |
|
3x |
66.7% |
3.2% |
1.43 |
29.8% |
|
5x |
80.0% |
4.0% |
1.19 |
15.8% |
|
8x |
87.5% |
5.2% |
1.09 |
7.9% |
|
10x |
90.0% |
6.0% |
1.06 |
5.3% |
|
14x |
92.9% |
7.6% |
1.02 |
2.3% |
Going from 5x to 10x adds 2 points of yield and removes two-thirds of your buffer. On $10,000 of equity at 10x, the extra yield over plain staking is about $360 a year. A 1% penalty on $100,000 of seized collateral is $1,000, nearly three years of that extra yield.
Exit slippage scales with the multiple too. A 0.1% swap cost costs 1% of equity at 10x.
Three Real Failures
June 2022. The stETH/ETH ratio fell to 0.94 on Curve. Celsius and Three Arrows Capital each pulled 50,000 stETH from Aave on June 8 and 9, and Celsius paused withdrawals on June 11. A 6% discount liquidates anything at 10x or above in the table, and spares a 5x loop.
March 10, 2026. A parameter update to Aave’s CAPO risk oracle priced wstETH about 2.85% too low and triggered roughly $26.6 million of liquidations. About 10,938 wstETH across 34 accounts was liquidated, and Aave reported no bad debt. The market never moved, so a 14x loop died to a configuration error.

April 18, 2026. Attackers minted 116,500 unbacked rsETH through a compromised bridge, borrowed about $193 million from Aave V3, and pushed WETH to 100% utilization. Aave’s report modeled WETH bad debt of about $60 million to $161 million, and outside estimates reached $200 million. Aave cut the WETH rate at full utilization to 3% to protect loops from a funding spike. That shows how another asset in a shared pool can threaten your loop.
Aave vs Morpho vs Fluid
|
Platform |
wstETH/WETH setting |
WETH borrow cost |
Risk structure |
Best for |
|
Aave V3 |
E-Mode: 93% LTV, 95% LT, 1% penalty |
About 2.0% to 2.1% |
Shared pool per chain |
3x to 5x with deep exit liquidity |
|
Morpho Blue |
Markets at 94.5% and 96.5% LLTV |
1.76% and 1.94% |
Isolated market per pair |
Careful users, small size |
|
Fluid |
Vaults up to 95% LTV |
Check the live vault |
One collateral and one debt per vault |
Users who understand its liquidation engine |
Aave has the deepest liquidity, but both 2026 incidents hit its wstETH or WETH reserves. Morpho isolates risk per market. The cost is thin liquidity: about 5,682 WETH was available in the 96.5% market and about 1,145 WETH in the 94.5% market. Several large borrowers there sat at health factors between 1.005 and 1.05.
Fluid offers up to 95% LTV and penalties as low as 0.1%. The tradeoff is a more complex liquidation design, so verify the exact vault first.
Rate and Exit Risk
Borrow rates jump near full utilization. The Base WETH curve implied about 10.5% at 100% utilization in September. At 10x that would be roughly negative 70% annualized for as long as it lasts.
Do not plan to exit through Lido’s queue. Withdrawals normally take one to five days, but an October 10 snapshot showed about 870,000 ETH waiting to exit, an estimated wait of about 15 days. The same queue is why crypto staking pools and solo staking behave differently under stress. Your real exit is a DEX swap or flash-loan unwind, and its cost grows with your multiple.
Before You Deposit
- Oracle design. Check how wstETH is priced and who can change the parameters.
- Pool collateral list. On shared pools, see what else can be posted at high LTV.
- Exit liquidity. Compare your size to the WETH you could swap into.
- Rate curve. Look at the borrow rate at 95% to 100% utilization.
- Alerts and gas. Set health factor alerts at 1.25 and 1.15, and keep ETH for gas.
- Test exit. Open a small loop and fully unwind it before scaling.
Common Mistakes
- Using the maximum multiple. The LTV limit is a risk-model ceiling, not advice.
- Judging by base APY. At a 0.4% spread, rates and slippage decide the result.
- Treating the peg as the only risk. March 2026 had no market move.
- Relying on Lido’s queue to exit. Waits can reach weeks.
My Take
For most readers, the best loop is no loop. At 4x you earn about 3.6% versus 2.4% for plain wstETH, which is $120 a year on $10,000 for extra contract, oracle, and rate risk.
If you do loop, use Aave E-Mode at 3x to 4x with a health factor above 1.25. Choose Morpho’s 94.5% market instead if isolation matters more than liquidity. Avoid anything above 5x, and read the oracle setup and test an unwind before adding size. Choose plain staking if your goal is simple yield.
Conclusion
Loops pay about 0.4 percentage points per turn at current rates. At 5x you earn about 4.0% with a 15.8% buffer, and at 10x about 6.0% with a 5.3% buffer.
The main risks are oracle errors, short depegs, rate spikes, and contagion in shared pools, and 2022 and 2026 produced all four. Set your maximum multiple from the buffer you need, then check the oracle, liquidity, and exit path before depositing.
FAQs
1. What health factor is safe for a wstETH/ETH loop?
I would target 1.15 or higher, which on Aave’s 95% threshold is roughly 5.7x or less. Go higher if you cannot check the position daily, because the March 2026 oracle error moved collateral values by 2.85%.
2. Can I be liquidated if stETH stays at its peg?
Yes, because oracle errors and parameter changes can lower your collateral value without a market move. The March 2026 CAPO error liquidated 34 accounts while wstETH itself traded normally.
3. Is looping stETH worth it at a 2% staking yield?
Usually not, since each turn adds only about 0.4 percentage points and the borrow rate can rise above the staking yield. At 4x, the extra return is about 1.2 points over plain staking.
4. Is Aave or Morpho safer for looping wstETH?
Morpho’s isolated markets limit contagion from other collateral, while Aave offers deeper liquidity but shares risk across its pool. Pick Morpho for isolation at small size and Aave for exit depth at larger size.
5. How do I exit a loop quickly in a crisis?
Sell wstETH for WETH on a DEX or use a flash-loan unwind, not Lido’s withdrawal queue. An October 2026 snapshot estimated about 15 days in the validator exit queue, so test your exit route with a small position first.
References
Aave, WETH reserve overview (Ethereum Core): https://app.aave.com/reserve-overview/?underlyingAsset=0xc02aaa39b223fe8d0a0e5c4f27ead9083c756cc2&marketName=proto_mainnet_v3
Aave Governance, Risk Stewards September 2026 WETH interest rate adjustment: https://governance.aave.com/t/risk-stewards-september-2026-weth-interest-rate-adjustment-on-base/25667
Aave Governance, CAPO post-mortem on wstETH: https://governance.aave.com/t/post-mortem-exchange-rate-misallignment-on-wsteth-core-and-prime-instances/24269
Aave Governance, rsETH Incident Report (April 20, 2026): https://governance.aave.com/t/rseth-incident-report-april-20-2026/24580
Cet article vous a-t-il été utile ? S'il vous plaît dites-nous ce que vous avez aimé ou n'avez pas aimé dans les commentaires ci-dessous.
About the Author: Chanuka Geekiyanage
Contre Quoi Nous Luttons
Les groupes multinationaux surproduisent des produits bon marché dans les pays les plus pauvres.
Des usines de production où les conditions s’apparentent à celles d’ateliers clandestins et qui sous-payent les travailleurs.
Des conglomérats médiatiques faisant la promotion de produits non éthiques et non durables.
De mauvais acteurs encourageant la surconsommation par un comportement inconscient.
- - - -
Heureusement, nous avons nos supporters, dont vous.
Panaprium est financé par des lecteurs comme vous qui souhaitent nous rejoindre dans notre mission visant à rendre le monde entièrement respectueux de l'environnement.
Si vous le pouvez, veuillez nous soutenir sur une base mensuelle. Cela prend moins d'une minute et vous aurez un impact important chaque mois. Merci.
0 commentaire