What is Looping in DeFi?

What is DeFi looping?

DeFi looping is a leveraged lending strategy that increases exposure to an asset by repeatedly borrowing against collateral and putting the borrowed capital back to work. A user deposits collateral, borrows against it, converts the borrowed asset into more collateral, and deposits again. Each cycle adds both assets and debt to the position. If the collateral earns more than the debt costs, leverage can increase the return on the user's starting capital.

A lending protocol lets a borrower deposit collateral and borrow another asset against it. The loan is overcollateralized, so the value of the debt must remain below the borrowing limit set for the collateral. Looping takes that ordinary loan one step further.

Example: Assume a user starts with $10,000 of a yield-bearing stablecoin. The user deposits the $10,000 as collateral, borrows $8,000 of another stablecoin, converts that $8,000 into more of the original yield-bearing asset, and deposits it. The account now has $18,000 of collateral and $8,000 of debt.

The user can borrow again against the larger collateral base. If another $2,000 is borrowed and redeposited, the position grows to $20,000 of collateral and $10,000 of debt. Repeating the process creates a larger position, although each new round adds less collateral than the one before it because lending markets require overcollateralization.

Why DeFi looping exists

Looping exists because DeFi collateral can often do more than secure a loan. A yield-bearing asset may keep accruing staking rewards, lending income, vault returns, or a fixed-yield return while it sits inside a lending position. Borrowing against that asset creates a financing leg, where reinvesting the borrowed capital creates the loop. The strategy becomes economically attractive when the return on the financed asset exceeds the cost of debt by enough to compensate for the added risks.

Assume a yield-bearing asset earns 7% while a correlated debt asset costs 4%. An unleveraged $10,000 position earns an estimated $700 per year. A 2× loop creates $20,000 of gross collateral exposure and $10,000 of debt.

The simplified annualized return becomes:
Collateral yield: $20,000 × 7% = $1,400
Borrow cost: $10,000 × 4% = $400
Estimated net return: $1,000

That equals an estimated 10% return on the original $10,000, compared with 7% without leverage. The extra 3 percentage points did not come from a new yield source, as the user financed a larger asset position with debt.

This is why productive collateral has become closely connected to looping. Curvance's Productive Collateral in DeFi research explains how assets such as liquid staking tokens, yield-bearing stablecoins, vault shares, and tokenized real-world assets can keep generating value while supporting borrowing. The same logic applies when the goal is directional exposure rather than yield. A borrower can post ETH, borrow a stablecoin, buy more ETH, and redeposit it. In that structure, the loop functions as an onchain leveraged long.

How a DeFi looping strategy works

A standard long loop has four actions:

1. Deposit collateral.
2. Borrow another asset against it.
3. Convert the borrowed asset into more collateral.
4. Deposit the new collateral and repeat.

The main types of DeFi looping strategies

Yield-bearing stablecoin loops

A yield-bearing stablecoin can be posted as collateral while a borrower takes debt in another stablecoin.
Example:

• Deposit a yield-bearing dollar asset earning 8%.
• Borrow USDC at 5%.
• Convert USDC into more of the yield-bearing asset.
• Redeposit and repeat.

The intended return comes from the 3 percentage-point spread applied across the leveraged balance sheet.
The main danger is assuming two dollar-denominated assets are economically identical. They may differ in redemption rights, collateral composition, liquidity, oracle design, and behavior during stress.
A move from $1.00 to $0.96 is only 4%, but a highly levered stablecoin position may have little room to absorb it.
Curvance's Avant savUSD Yield: Looping With USDC strategy article provides a protocol example of this structure and explains how collateral yield, the USDC borrowing rate, LTV, and liquidation buffer interact.
The hyAUSD Collateral: Borrow AUSD or USDC research covers another stablecoin-loop design, including same-asset AUSD borrowing and cross-stablecoin USDC borrowing. The underlying vault mechanics are covered separately in The Curvance High Yield Vault: Optimized AUSD Yield with hyAUSD.

Liquid staking token loops

A liquid staking token loop uses an asset such as a staked representation of ETH as collateral, then borrows an ETH-linked asset and reinvests it into more staking exposure.

If the staking asset earns 3.5% and the financing asset costs 2.5%, a 3× simplified loop produces:
(3.5% × 3) − (2.5% × 2) = 5.5%
The trade depends on more than that 1 percentage-point starting spread. The staking token can deviate from its underlying asset, liquidity can thin out, withdrawal mechanics can slow an unwind, and borrowing costs can rise with utilization. Aave governance materials have documented leveraged staking structures in which a staking token is posted as collateral, the underlying asset is borrowed, restaked, and repeated to a target LTV.

Directional crypto loops

A user can deposit a volatile asset, borrow a stablecoin, buy more of the volatile asset, and redeposit it.
For example:

• Start with $10,000 of ETH.
• Borrow $5,000 USDC.
• Buy $5,000 more ETH.
• Hold $15,000 of ETH collateral against $5,000 of debt.

This is economically closer to a leveraged long than a pure yield trade. If ETH falls 20%, the $15,000 collateral falls to $12,000 while the USDC debt remains close to $5,000 before interest. Equity drops from $10,000 to about $7,000, a 30% decline before transaction costs. The same asymmetry works upward, where Leverage magnifies the asset move because debt does not move in lockstep with the collateral.

Fixed-yield and productive-collateral loops

Some DeFi assets already earn yield before they enter a lending market. Examples include principal tokens, vault shares, tokenized Treasury products, and other value-accruing assets.
Using productive collateral changes the financing equation because the collateral can continue generating a return while supporting debt. Curvance's productive-collateral research explains this through the simplified relationship:
Net financing cost = borrowing interest − collateral yield
A loop extends that logic by increasing both the yield-bearing collateral and the debt used to finance it.
For a principal token, the return may come from price convergence toward a known redemption value. Curvance's Using Pendle PT-AUSD as Collateral article shows how a principal token can support borrowing before maturity and how that can feed into a looped position.
For a vault share, return may come from lending or strategy income. The source of the yield matters because each source introduces a different failure mode.

Risks and trade-offs of DeFi looping

Liquidation risk

Liquidation is the most obvious risk. Debt grows through interest while collateral can move in price. If the account crosses the market's liquidation threshold, part of the collateral can be sold or otherwise closed to repay debt.
Leverage reduces the distance between the current position and that threshold. Aave tells borrowers to monitor health factor because collateral adequacy changes with both asset prices and accrued debt. Euler's Multiply documentation states that higher leverage means a smaller adverse price move can cause liquidation.

Borrow-rate risk

Borrowing costs can change faster than collateral yields. A position earning 8% against 4% debt has a 4 percentage-point spread. If the borrow rate rises to 10%, the spread becomes negative even if the collateral performs exactly as expected. This risk matters most in crowded trades. When many users want the same debt asset, utilization can rise and financing becomes more expensive.

Yield-source risk

The collateral APY may not be fixed. A liquid staking token depends on staking economics. A yield-bearing stablecoin may depend on lending, basis trades, Treasury income, or another strategy. A vault can change allocations, and token incentives can expire. If a 10% collateral yield falls to 4%, a loop funded at 6% has changed from positive carry to negative carry without any liquidation event.

Oracle risk

Lending protocols need prices to calculate borrowing capacity and liquidations. An oracle that updates slowly, uses an unsuitable market, or fails during volatility can affect when a position becomes liquidatable. Productive collateral can require more than one valuation input because the token may have both a market price and an internal conversion rate. Curvance's productive-collateral research describes this issue directly: yield-bearing collateral may require the protocol to account for market price, conversion value, liquidity, redemption mechanics, price-feed freshness, and depeg risk.

Liquidity and execution risk

A loop eventually has to be unwound. Repaying debt may require selling collateral, redeeming an asset, or swapping through a DEX. A $5,000 position can be easy to exit while a $5 million position in the same market faces meaningful slippage. Liquidity also matters during liquidation. A theoretically well-collateralized loan can still create losses if collateral cannot be sold quickly enough at a price near the oracle value.

Smart contract and composability risk

Looping often connects several systems. A position might depend on a lending protocol, a collateral issuer, an oracle, a DEX, a bridge, and a vault. Each dependency creates another route through which value or liquidity can change. Euler's documentation lists smart contract risk among the standard risks of leveraged Multiply positions. Curvance's productive-collateral framework treats layered protocol dependencies as part of collateral evaluation.

Incentive risk

Some loops look profitable because rewards are temporarily subsidizing one side of the trade. Suppose the collateral earns a 5% native yield plus 6% of token incentives, while borrowing costs 7%. The displayed collateral APY is 11%, but the persistent economic spread before incentives is negative. If the 6% reward campaign ends, the loop goes from positive to negative carry. Separate native yield, lending yield, token incentives, and borrowing incentives before deciding whether the spread is durable.

Manual looping versus one-transaction leverage

Early DeFi loops required several transactions: deposit, borrow, swap, deposit again, then repeat. That process creates operational friction. Each step can incur gas, slippage, approval risk, and execution delay. A user can also stop halfway through with a position that differs from the intended target.
Modern lending interfaces increasingly compress the workflow.

Morpho's Bundler3 documentation describes one-click leverage flows that combine collateral conversion, deposit, and borrowing actions in a single transaction. Euler's Multiply interface automates the repeated loop process.
Automation improves execution, but it does not change the economics. A 4× position is still a 4× position whether it took twelve manual transactions or one bundled transaction to build.
That distinction matters. Better user experience can reduce operational errors while making high leverage easier to access. The risk analysis still starts with the final collateral, debt, LTV, and liquidation threshold.

How to evaluate a looping strategy

A looping position can be reduced to a small set of questions.

1. Where does the collateral yield come from?
An 8% APY generated from Treasury income has different risk than an 8% APY generated from token incentives or leveraged basis trading. Identify the source before multiplying it.
2. What is the current cost of debt?
Check the borrow rate, utilization, and the shape of the rate model. If a market is already near a high-utilization region, the current borrow rate may be a poor estimate of the rate after a large new position enters.
3. What leverage are you targeting?
Do not start with the market maximum. Start with the downside you are willing to absorb and work backward toward a target LTV.
At 2× leverage, a $10,000 account has $20,000 of collateral and $10,000 of debt. At 5×, it has $50,000 of collateral and $40,000 of debt. Those are materially different risk profiles even if the assets are the same.
4. What breaks the price relationship?
For ETH collateral against USDC debt, a decline in ETH is the obvious stress. For a stablecoin pair, model a depeg. For an LST pair, model the LST-to-underlying exchange rate. For a vault share, model a strategy loss or redemption discount.
5. How do you exit?
Identify the unwind path before entering. Check DEX liquidity, redemption delays, withdrawal limits, debt liquidity, and whether the position can be deleveraged atomically.
6. What happens when incentives disappear?
Recalculate the loop without token rewards. If the strategy becomes unattractive, its economics depend on the subsidy lasting.
7. Can you monitor it?
A loop with variable debt is an actively financed position. If the allocator cannot monitor rates, health, liquidity, and collateral behavior, lower leverage or no leverage may be the better structure.

How Curvance supports looped positions

Curvance supports leveraged position management for selected collateral and debt pairs. Its current PT-AUSD research states that a user can size a leveraged position in a single transaction rather than manually repeating the deposit-borrow cycle. Its savUSD strategy article also lays out the standard loop: deposit savUSD, borrow USDC, convert the USDC into more savUSD, and redeposit.

The protocol's broader lending design matters for looped positions in three areas:

Productive collateral. Curvance supports assets that can continue accruing value while they are posted as collateral. That allows the asset-side yield to remain part of the loop's economics rather than disappearing when the position is financed.
Isolated lending markets. Curvance separates markets and calibrates risk parameters by collateral and debt pair. Its published research describes independent LTVs, liquidation thresholds, rate curves, collateral caps, and oracle configurations for isolated markets. A loop therefore has to be evaluated at the specific market level rather than against one protocol-wide risk number. See The Architecture of Isolated Lending Markets for the broader risk framework.
Rate and liquidation awareness. The current Curvance savUSD article tells users to monitor collateral yield, borrowing cost, utilization, LTV, and liquidation buffer throughout the position. That is the right framework for any leveraged loop, regardless of the asset pair.

Curvance's lending contracts use cToken-style deposit positions in which supplied assets become available for borrowers and lender value accrues through the relationship between cToken shares and underlying assets.
The interface can make constructing the position simpler, and the underlying financing risk remains with the user.

When DeFi looping matters

Looping can make sense for users who already understand the collateral they want to own and are comfortable managing debt against it. A fund holding a yield-bearing asset may use a loop to increase exposure to a positive financing spread. A DAO treasury may use a conservative stablecoin loop to increase return on a reserve asset while keeping a defined LTV buffer. A trader may use a volatile-asset loop for directional leverage.

The strategy is a poor fit for capital that requires a fixed return, immediate liquidity under all conditions, or no liquidation exposure. There is also a simpler alternative that deserves to stay in the comparison: hold the underlying asset without leverage.

If a yield-bearing asset earns 8% on its own, a loop should earn enough incremental return to compensate for the additional debt, monitoring, liquidation exposure, and protocol dependencies. A projected 9% looped APY may not clear that bar. A projected 14% might, depending on how the position behaves under stress.
For a live application of the mechanics, see Curvance's Avant savUSD Yield: Looping With USDC strategy research. For the collateral side of the equation, see Productive Collateral in DeFi. For the market-risk structure underneath the loan, see The Architecture of Isolated Lending Markets.

Before opening any loop, write down four numbers: collateral yield, borrow rate, target LTV, and liquidation threshold. Then run the same position with a higher borrow rate and an adverse collateral move. If the strategy still makes sense under those conditions, the displayed APY has passed its first useful test.