Liquity USD Versus USDC Loans Where Your ETH Collateral Costs Less
Interest isn't paid when borrowing against Liquity USD, the recurring expense that typically accompanies any loan or borrowing product. That one sentence is the fitting summary of every liquity price conversation since 2021. It's also why any argument comparing Liquity to USDC-denominated lending protocols has something of an uphill battle before it even begins. Liquity's native stablecoin LUSD stablecoin is minted when ETH is deposited into a Liquity "Trove" (smart contract), which holds debt against the collateral at a minimum ratio of 110%. Compare this to taking out a loan in USDC from Compound or Aave, where the borrower has to pay an interest rate (dependent on market demand) that can average under 3% APR in quieter markets but can rise above 10% when these platforms are heavily utilized.
Sound too good to be true? There's a caveat. Liquity vs Compound vs Aave isn't as simple as "free versus you pay fees." Liquity's APR of 0% comes with trade-offs in upfront costs, liquidation mechanics, and flexibility that can make centralized stablecoin lending preferable for the right user, use case, and risk appetite. Time to break down the variables.
The Fee Structures Borrowers Pay
Liquity V1 charges a one-time 0.5-5% of the crypto borrowed in LUSD as a borrowing fee. The exact percentage depends on what the base rate of the protocol is when the borrow occurs. The base rate goes up when others redeem (trade LUSD back to underlying ETH collateral) and decays back down to zero over time. During quiet times where there are few or no redemptions, a borrower can expect to pay a fee of approximately 0.5% of loan amount upfront and never have to pay fees again. A $10,000 loan equals a $50 fee at the lowest possible rate. No monthly payments. No compound interest.
Cumulative borrowing cost over 365 days on a $10,000 loan. Sources: Liquity Protocol documentation, Aave V3 historical APR data.
Compare that to lending USDC through Aave V3 on Ethereum mainnet today. Variable borrow APRs have ranged from 4-8% average over the last year. Using 5% APR as an example, that same $10,000 USDC loan costs $500 interest per year. In two months the borrower has already paid more than Liquity's minimum fee. After 6 months the difference is around $200 versus maybe $50.
This math starts to change when borrowing for shorter periods of time. Someone who needs liquidity for two weeks and will repay their loan soon after opening their position is now at a point where Liquity's 0.5% fee is equal to or more expensive than the interest paid on a USDC loan with 5% APR (approximately 0.19% for two weeks). Any user who plans to make a Liquity USD price arbitrage trade over a couple of days has paid less in fees through Aave. Once a position runs above 30 days, Liquity's pricing advantage gets super linear due to the nonlinear nature of time versus a flat fee. One other fee to note on Liquity is a 200 LUSD liquidation reserve which is paid when a Trove is opened. This gets refunded when the position is closed, so it's not really a cost unless liquidation occurs. There is no equivalent "deposit" for USDC loans on Aave.
What An ETH Crash Does To Each Position
Liquidation mechanics is the field where the two systems diverge most wildly. Study them. Comprehend them. They may be the deciding factor between losing 10% of principal, or losing it all. Liquity forces liquidations of Troves when their collateral ratio dips below 110%. If a user had locked up $11,000 worth of ETH as collateral, borrowed 10,000 LUSD against it, a decline in ETH price of greater than 10% should trigger liquidation. Said debt is consumed by the protocol's Stability Pool (read: users who have deposited LUSD into the pool in exchange for collecting liquidation rewards). Depositors of the Stability Pool receive the collateral. The borrower loses their ETH but keeps the LUSD they have already withdrawn. There is no such thing as partial liquidation in Liquity V1. Once it's happening, the Trove is liquidated 100%.
Aave differs here. Once the health factor dips below 1.0, liquidation is possible, but only a percentage (up to 50% of the debt) of the position is liquidated at a time. There is a 5% liquidation penalty on ETH-collateralized positions. This means a borrower can essentially "ride out" a large drawdown if they have enough collateral left over after the partial liquidation occurs. They can then deposit more ETH into their position to "recapitalize" it.
An example helps clarify. Imagine ETH crashes 40% in one day. Both systems are under stress. The Liquity borrower who previously opened a Trove with a 150% collateral ratio now finds themselves with a ratio around 90% after the price decline. Well below Liquity's 110% safety buffer. Their position will be liquidated completely. The Aave borrower also opened a position with $11,000 worth of ETH collateral and borrowed 10,000 worth of any token. After ETH's price decline, a series of partial liquidations will occur, with a penalty of 5% applied each time. Because the liquidations are partial, the Aave borrower has more control over how the unwind affects them. Incentivization built into Liquity's smart contract pushes borrowers to always over-collateralize their loans, even past the absolute minimum requirement of 110%. Most loans should strive for ratios over 150%, according to Liquity's own protocol documentation. That recommendation means something, because unlike Aave there is no progressive warning system. A Trove is silently liquidated with no chance of recovery.
Recovery Mode: Liquity's Hidden Liquidation Trigger
There exists an always-on circuit breaker within the LUSD stablecoin protocol which most borrowers are completely unaware of until it's too late. Known as Recovery Mode, it basically invalidates all existing borrow positions which are not sufficiently overcollateralized. Recovery Mode is a special protocol state which is entered when the total collateral ratio of all Troves on Liquity drops below 150%. Under Recovery Mode, any Trove with a collateral ratio below 150% (note NOT 110%) is subject to liquidation. This is extremely important to understand. Say a borrower is at a collateral ratio of 130%. The position is safe, right? Wrong. If the total collateral ratio of the system dips below 150% as a result of changing market conditions, that perfectly healthy 130% position can now be liquidated even though it was not directly undercollateralized at all.
You did everything right. Borrowed less than you were allowed and minted tokens against it, but YOUR POSITION got rekt anyway because the system as a whole was under extreme stress. There is no equivalent to this on USDC lending platforms like Aave. When borrowing USDC from Aave, the liquidation price is solely dependent on individual collateralization ratio. There is no way for a position to become riskier based on actions of other borrowers. Each position's health is siloed from each other. This is great if you are a paranoid borrower.
Recovery Mode can and has been triggered a few times since Liquity's inception. It's a non-zero risk that liquity crypto borrowers have to be aware of. For those exploring liquity price prediction models, the recommendation is to bake in some catastrophic system-wide black swan stress events. This is always relevant but particularly during a time where an ETH-only selloff could happen and all Troves lose collateral value at the same time. The two things to understand: once crypto is deposited into Liquity it can potentially be lost to forced liquidation, and the smart contracts that define how LUSD works are immutable. Recovery Mode thresholds can never be made less severe (no soft-fork option) even if all stakeholders agreed to it on some imaginary governance call. Liquity has no governance.
Why Zero Interest Isn't The Same As Zero Cost
Liquity can make these loans interest-free because the protocol charges fees instead of using an interest rate system. The protocol's revenue stream comes from one-time fees (borrowing fee plus redemption fee). This creates an incentive for borrowers to hold their loans long term instead of short term. A borrower who opens a Trove could hold it for two years and pay the exact same total fee as someone who holds a Trove for two months.
The LUSD token itself comes with an added cost a USDC borrower does not deal with: deviation from the $1 peg. At the time of writing, LUSD is trading right around $1.00 but has traded down to $0.8967 and up to $1.16 over its entire history. During times of high redemption demand or low liquidity, LUSD may trade below its $1 peg. A user that mints LUSD at $1.00 and sells it for other assets may later have to purchase LUSD back at over $1.00 per token to close out their Trove. USDC's peg has been much more stable (outside of March 2023) and it also has much deeper liquidity on exchanges, making par purchases for large buy orders much easier.
Then there is capital efficiency. Liquity has a 110% minimum collateral ratio, one of the lowest in DeFi lending. Essentially this means more liquidity for every dollar of ETH that gets locked up as collateral. For ETH, Aave has much higher collateral ratios. The minimums posted are often lower than users actually use, but borrowers want to play it safe so the effective minimum is closer to 130%. If a borrower had $15,000 worth of ETH, they could borrow approximately $13,636 in LUSD from Liquity. By comparison, they could borrow $11,538 in USDC from Aave. That extra $2,098 may mean the world to someone who wants as much capital as possible. Liquity's price stability mechanism backed by direct ETH redemptions is what allows this low of a ratio.
When Centralized Stablecoin Lending Wins
Even though Liquity's fees are far lower on longer duration loans compared to other platforms, there are still a few common use cases where USDC lending on Aave (or even just a centralized alternative) is more attractive. Short-term borrowing under 30 days. If borrowing for only a few days or weeks, then the amortized interest rate on a USDC loan will be cheaper than Liquity's upfront fee.
Multi-asset collateral. Liquity only supports ETH as collateral. Aave supports ETH, wBTC, stablecoins, and a dozen other tokens. Borrowers who hold multiple assets and don't want to sell their non-ETH assets to free up ETH collateral will find Aave more convenient. Partial liquidation protection. Borrowers who want to weather volatility without having their entire position liquidated find Aave's system of incremental liquidations gives more time.
Ultra-deep liquidity matters for institutional and whale borrowers. USDC trades with exponentially higher liquidity than LUSD crypto on every exchange. To buy or sell $500,000+ of an asset to move in or out of a loan position, USDC is much easier to trade at par. Regulation. USDC is issued by Circle, a regulated U.S. entity. For borrowers who need to prove compliance or otherwise work inside the traditional finance system, USDC may be the only choice. LUSD has no issuer. Whether that's a feature or a dealbreaker is up to the borrower.
Matching The Loan To The Borrower
The data here paints a pretty straightforward picture. Liquity USD is the lowest cost option for borrowers who plan to maintain an ETH-collateralized position over 30 days, maintain a collateral ratio over 200% to avoid both normal liquidation and Recovery Mode liquidation, and don't mind the rare thin liquidity conditions of the LUSD stablecoin. For borrowers who care about the fact that this protocol cannot be altered (no admin keys, no governance) through either a DAO proposal or corporate intervention, the Liquity USD token may be the best option.
Borrowing USDC from Aave (or other protocols) makes a lot of sense for borrowers with use cases that include the need for short-term capital, partial liquidation protection, multi-asset collateral, or loan sizes that benefit from higher liquidity. Cost is inevitable. Flexibility is too. Neither Liquity nor Aave V2 are ideal in all aspects. Liquity USD trades flexibility for decentralization and cost certainty. Aave V2 USDC trades latency for flexibility.
Both systems are deployed. Liquity's V2 is LIVE. BOLD is already rated A- from Bluechip, better than USDC's B+. Nobody is questioning whether Liquity is technically competent. They just offer different things. Pick what works best.
