ARC: Add support for LUSD (Liquity)

Add LUSD as a borrowable asset on Aave


Paragraph Summary

LUSD is stablecoin minted directly by the protocol users by opening a Trove and depositing ETH as collateral. Loans are interest-free, with a 0.5% base initiation fee. All the contracts of the Liquity protocol are immutable, which makes LUSD the most trustless stablecoin currently available on the Ethereum mainnet.

LUSD is overcollateralized by ETH with two main mechanisms helping it keep its peg: first, LUSD can be redeemed for ETH at face value (i.e., 1 LUSD for $1 of ETH). Secondly, there is a minimal collateral ratio of 110%. Both mechanisms respectively create a price floor and ceiling through arbitrage opportunities.

Since the protocol is immutable, no additional developments are planned, yet the Liquity team is working on another project that will prove synergetic with LUSD.

LUSD’s resilience makes it a highly sought-after stablecoin for DAOs and protocols looking to diversify their treasury and users looking for cost-effective leverage on ETH. The current focus is to grow the DeFi ecosystem around LUSD to enable more diverse use cases.

This proposal aims to add LUSD as a borrowable asset on Aave v2 mainnet (0% LTV). If the current proposal is met with enthusiasm, other proposals will follow suit.


AIP author status

I, TokenBrice, author of this ARC recently joined the Liquity team to help with DeFi strategy.

Overview of the project and token

Liquity is a protocol offering cost-efficient leverage using ETH as collateral. While its base design can be described as similar to Maker/DAI, there are several key differences:

  • Only ETH is accepted as collateral to maximize the protocol’s trustlessness and resilience.
  • There is no interest fee on Liquity. Instead, the borrowers pay a predictable initiation fee starting at 0,5% — and ramping up algorithmically when needed to protect the protocol.
  • Liquidations are close to instant, and enabled by a protocol-level backstop: the stability pool, a reserve of LUSD deposited by users.
  • Just like Aave, Liquity uses ChainLink as the main price feed. However there is a fallback system in place, switching to Tellor, if the ChainLink price feed data is deemed incoherent.
  • Since the contracts are immutable, there is no governance. There is still a token associated with the protocol – LQTY, but it’s used for incentivization and fee-sharing purposes only.
  • The minimal required collateralization ratio is lower – at 110%, increasing the capital efficiency of the protocol. This is possible thanks to the stability pool and instant liquidation mechanism which also helps maintain the peg by acting as a liquidity buffer. A failsafe mechanism, called the Recovery Mode, is triggered when the total collateralization ratio of the protocol falls below 150%. Being an exceptional measure, the Recovery Mode enforces more substantial restrictions on the Troves, such as temporary minimal collateralization of 150%, until the total collateralization ratio of the protocol recovers. So far, it was only triggered once in the early days of the protocol.

Positioning within Aave Ecosystem

LUSD is a stablecoin offering some genuinely unique features, the leading one being the trustlessness of the whole Liquity protocol. In that sense, LUSD is one of the most resilient stablecoin and the most decentralized. Similar to the addition of RAI on Aave, having LUSD will help Aave diversify from the current dominance of centralized stablecoins such as USDC used for borrowing.

Furthermore, as mentioned above, this proposal is the first step. Several others could follow to nurture the relationship between Aave and Liquity further:

  1. Enablement of LUSD as collateral (LTV>0%)
  2. Enablement of efficiency mode against selected stablecoins once v3 mainnet is live.
  3. Addition of LUSD on Aave/Optimism + potential enablement as collateral + again enablement of the efficiency mode.

I believe that LUSD as a borrowable asset will be met with a strong demand: it will be the first time such a resilient stablecoin could be borrowed on Aave and the first protocol where LUSD is borrowable.

History of the project and its components

The Liquity protocol was launched on the 5th of April, 2021. As the code is immutable, there has been no update since then. However, the team is still actively working on the protocol’ DeFi integrations, such as supporting LUSD liquidity on Curve.

Since the demand for LUSD is directly dependent on the appetite for leverage on ETH, the LUSD supply had varied quite a bit since the launch, with the highest point around 1.5B LUSD when ETH reached a peak price of ~$4K. The following correction on ETH, with the lowest point around $2K, led to a drastic reduction in the LUSD supply – this episode is the first and so far unique time when the Recovery Mode was triggered. Since this episode, the supply has stabilized more and is now around 325M LUSD. This episode was the first massive stress test for the Liquity Protocol, which handled it gracefully – a recap is available here:

Since the protocol launch, various DeFi integrations have been built on top of the Liquity Protocol, here are some of the most notable:

  • A Curve pool (+integration on Convex and Yearn) was created against the 3Pool and became the main liquidity source. It’s amongst the stablecoin pools with the highest base APR due to the trading volume it processes.
  • DeFiSaver offers some automation features to streamline Troves management.
  • B.Protocol delivered an auto-compounder for depositors of the Stability Pool who wish to keep a 100% LUSD exposure: B.Protocol
  • Some protocols, such as Olympus or Fei Protocol diversified part of their treasury to LUSD to shield themselves from volatility with an unstoppable stablecoin. Synthetix is also using LUSD on Optimism to stabilize the sUSD price.

LUSD usage

A sizeable share of the LUSD supply (~66%) is deposited into the Stability Pool, the liquidity backstop for liquidations: Liquity: Stability Pool | Address 0x66017d22b0f8556afdd19fc67041899eb65a21bb | Etherscan

The top LUSD holders also include other protocols, such as Olympus, DXdao or some treasuries.

LUSD is also present on Optimism, where Synthetix provided a wrapper contract to help stabilize the sUSD price (~45M LUSD) + to provide liquidity on Velodrome (the pool is ~9M TVL, against USDC).

LUSD Emission

The only way to mint LUSD is to open a Trove, while LUSD is burned through debt repayment, either directly (by the trove owner) or through redemptions (by an arbitrageur).

Token & Protocol Permissions

The LUSD token as well as the whole Liquity protocol are fully permissionless. There is no multisig and no governance, and no-one can interact with the contract underlying the protocol in an administrative manner: all users are the same in front of the contract, just like with Uniswap v2 for instance.

Market Data

As of June 03, the total LUSD supply is around 325M. There is about $90M of total liquidity in pools where LUSD is involved.

The main liquidity source for LUSD is the Curve pool with ~$65M TVL:

There are also liquidity sources against FRAX, USDC and DAI on Uniswap V3: Uniswap Info

There is a 4-tokens pool on Saddle Finance (alUSD, FEI, FRAX, LUSD) with $25M liquidity: Saddle

Finally, on Optimism, LUSD is the largest stablecoin pool with around ~$9M against USDC

The observed volume can vary depending on the market’s activity (massive ETH price movements are usually followed by days with large volume), with a baseline around $3M daily volume (mainnet only).

Social channel data

The main channel for the community to gather is the Liquity Discord server, with around 6.5K members. For outward-facing communications, the main channel used is the @LiquityProtocol Twitter account with over 28.5K followers. We also regularly publish newsletter updates to subscribers (2.5K) and have recently become more active on LinkedIn (<1K).

Contract Related Information

Deployment of the LUSD contract - 05 April 2021: Ethereum Transaction Hash (Txhash) Details | Etherscan

The first minting of LUSD - 05 April 2021: Ethereum Transaction Hash (Txhash) Details | Etherscan

Number of transactions on LUSD: 197,461 (+ ~2000 on Optimism)

Number of token holders on LUSD: 4758 (+ ~100 on Optimism)

Risk Assessment

We followed Aave’s risk assessment methodology to provide all the necessary metrics:

I’m not sure what the proper rating is for a stablecoin with no governance and immutable code should be when it comes to the “Permission” category, so I went with an A to reflect that in that regard, LUSD is almost as safe as ETH. Feel free to challenge me if you think that is not appropriate.

(Volume data from CoinGecko)

Smart Contract Risk

Maturity Transaction
429 197990
B+ C+

Counterparty Risks

Holders Permission
4763 Permissionless and Immutable

Market Risk

Market Cap Average Volume Normalized Volatility
. 1M 3M 1M 3M 6M
337989637 12244118.23 7406951.186 0.002836979528 0.002166153429 0.004069270527
B C C A+ A+ A+

Implementation - Oracle

Since Aave v2 uses ETH-based ChainLink price feed, and LUSD’s feed is USD based, an adapter contract was needed, similarly to what ENS did.

It can be found here: LSUDUsdToLUSDEth | Address 0x60c0b047133f696334a2b7f68af0b49d2f3d4f72 | Etherscan


LUSD would add more diversity in the stablecoin reserves within the Aave Protocol and further would empower the support for decentralized stablecoins. LUSD stablecoin mechanisms are also relatively simple (only ETH collateral) and is in line with the risk averse ethos of the Aave Community. Fully supporting LUSD, only as a borrowable asset for now.

Moreover, congrats @TokenBrice for leading the DeFi strategy at Liquity.


Thank you @TokenBrice for this well-written proposal.

I’m personally a strong supporter of stablecoin diversity and especially in favor of decentralized stablecoin diversity and LUSD is a good candidate.

As pointed out by @stani and according to historical market data, stablecoin usage in DeFi is mainly used as borrowed currency and we witness little to no usage as a collateral asset (I guess no one wants to use the future of France to short the future of France? :upside_down_face:)

In the context of the upcoming V3 upgrade that will introduce on L1 both isolation mode and emode (LUSD used as collateral most promising use case framework)

I’m in favor of a 0% LTV onboarding to let the Liquity & Aave fam meet inside the protocol and after the V3 upgrade reconsidering the LTV to allow emode to participate in LUSD & other stablecoin stability.

A 0% LTV onboarding minimizes as much as possible the risk for the protocol while opening the doors to an increasingly successful synergistic relationship in the future.


Seems to have support from the larger Aave team and community.

Personally have never used Liquity but am aware of its respect within the larger ecosystem.

@TokenBrice yours as well - acting as a large delegate across a myriad of DeFi protocols.

0% LTV feels like a safe starting point.


Thanks to @stani @MarcZeller and @fig for the support and the kind words, I’m happy to put my knowledge of DeFi to good use by helping promote the adoption of an immutable stablecoin. :slight_smile:

Indeed, I think the bulk of the demand for LUSD as collateral will probably manifest itself once the efficiency mode is enabled against USDC and other stablecoins, as it would provide some neat arbitrage opportunities against the Curve pool. But we have a few more steps to go through before we can get there: one at a time!

So LUSD on mainnet with 0% LTV is indeed a great starting point that will help us better assess the next steps and ensure a safe & risk-minimized onboarding for Aave.

Happy to answer any questions about Liquity if there are some left unanswered with the post.


Thanks for this comprehensive ARC @TokenBrice

Could you share some dashboards on liquidations and Liquity’s market metrics please?

1 Like

Hello @Alex_BertoG, happy to!

The main Liquity Dune Analytics Dashboard is a great place to follow all the metrics related to the protocol, including the latest liquidations:

TokenTerminal also supports Liquity and provides some neat metrics:

Finally, we saw what I think is the biggest liquidation to date a few days ago: a Trove with a 60.7M LUSD debt got liquidated. It was massive but everything worked just fine!

Let me know if you need something else.


The signaling proposal vote is now live on Snapshot:

1 Like

Hi @TokenBrice,

If Liquidity is looking for some help in listing LUSD on the Aave v2 mainnet market, please drop us a DM and we can help with that. :slight_smile:

Matthew Graham


Voted in support of this proposal.
current risk parameters limit risks and having a diverse pool of decentralized stablecoins is healthy for the Aave procotol.


Enabling an asset as collateral can be one of the riskiest phases for a new asset coming onto Aave. As such, we strongly recommend listing assets without enabling them as collateral to observe how usage evolves for supplies and borrows and to derisk any mechanism design issues that could pose an outsized risk to Aave. It is great to see that is the approach being taken for LUSD. Once an asset has sufficient time in the market (O(weeks)), we are comfortable enabling them as collateral as part of the biweekly parameter update process, provided there is sufficient liquidity and community buy-in.

We recommend the below initial parameters for LUSD. We recommend setting the reserve factor as the same as other stablecoins on the platform:

Max LTV: 0%
Liquidation threshold: 0%
Liquidation bonus: 0%
Reserve factor: 10%


I’ve updated the main post to clarify the situation regarding the oracle. Since Aave v2 uses ETH-based ChainLink price feed, and LUSD’s feed is USD based, an adapter contract was needed, similarly to what ENS did.

It can be found here: LSUDUsdToLUSDEth | Address 0x60c0b047133f696334a2b7f68af0b49d2f3d4f72 | Etherscan