Original author:  Lawrence Lee, Mint Ventures Researcher

Introduction

Lybra Finance is a new project of LSDFi that was born in April this year. It has been controversial since its inception, with disputes over the source of IDO funds, contract issues, and implications of its relationship with Lido on social media. The lack of anchoring ability in its stablecoin design has also been criticized by DeFi players.

Nevertheless, Lybra's TVL has already occupied a prominent position during the so-called LSDFi summer.

Source: https://dune.com/defimochi/LSDFi-summer

The price of its governance token LBR has also increased nearly 20 times in just over 20 days in May, with a current fully circulating market value of over $150 million.

Source: https://www.coingecko.com/en/coins/lybra-finance

In this article, we hope to answer the following questions by reviewing the mechanism of Lybra Finance:

• How did Lybra achieve the aforementioned results?

• What challenges is Lybra currently facing?

• How will Lybra's V2 address these issues?

Our aim is to provide readers with a more comprehensive understanding of Lybra Finance and its stablecoin eUSD.

The following article represents the author's perspective at the time of publication and may contain factual errors and biases. It is provided for discussion purposes only and should not be considered as investment advice. Corrections and feedback from other industry professionals are welcomed.

1. Basic Information and Business Data

Lybra Finance is a stablecoin protocol, with its stablecoin being eUSD and its governance token being LBR.

Lybra Finance has a relatively short history. They launched their testnet on April 11th and the product officially went live on April 24th. It has been just over 3 months since then.

Lybra Finance is an anonymous team and did not engage in private fundraising. Its only financing round was a public IDO on April 20, 2023, where 5,000,000 LBR tokens (5% of the total supply) were sold at an estimated valuation of 1 ETH = 20,000 LBR, equivalent to a valuation of 5,000 ETH. Based on the price of ETH at 2000 USD, Lybra Finance had a valuation of 10 million USD and raised a total of 500,000 USD.

Since its launch, Lybra's TVL has grown rapidly, reaching 100 million USD in TVL within the first month. Currently, its TVL is close to 400 million USD, making it the 18th-ranked protocol in terms of TVL on the Ethereum network.

Source: https://defillama.com/protocol/lybra-finance

Based on the scale of stablecoin issuance, eUSD with a circulation of nearly 200 million ranks 11th among stablecoins. Especially in the category of decentralized stablecoins, the circulation of eUSD is only second to DAI, FRAX, and LUSD, and it is also not lacking well-established decentralized stablecoins like MIM and alUSD. eUSD has become an undeniable new force in decentralized stablecoins.

Source: https://defillama.com/stablecoins

At present, there is a total liquidity of $25.5 million in the eUSD liquidity pool on Curve. However, the price of eUSD is unstable and has long been in a premium state. On June 16th, the price of eUSD once rose to $1.36 due to a whale buying $900,000 worth of eUSD. Although arbitrageurs quickly brought the price of eUSD back to $1, the issue of eUSD premium is indeed a major problem for eUSD at present, which we will discuss in detail below.

Source: https://curve.fi/#/ethereum/pools/factory-crypto-246/deposit

However, at present, the number of addresses holding eUSD is relatively small, only 829. In addition to the Curve liquidity pool, all the top ten addresses holding eUSD are personal addresses, and the overall distribution of holding addresses is relatively scattered. eUSD has very few use cases on the chain, and most holders hold it in their accounts to earn interest. Of course, this is also related to the poor composability of the current eUSD design (which will be described in detail later).

Top 10 eUSD Holding Addresses Source: etherscan

2. Stablecoin Core Mechanism

In the current V1 version, eUSD can only be generated by over-collateralizing with stETH (Lybra also allows users to deposit ETH, but in actual business operations, Lybra will deposit users' ETH into Lido Finance to obtain stETH for further business processes, and users can only withdraw stETH after repaying all debts, so we believe that eUSD can only be generated by over-collateralizing with stETH). Lybra requires a minimum collateralization ratio of 150%, meaning that for every 1 eUSD generated, there must be at least 1.5 U worth of stETH as collateral.

In terms of the price stabilization mechanism for eUSD, the rigid redemption mechanism plays a major role. The rigid redemption mechanism means that any user can use 1 eUSD to rigidly redeem stETH worth 1 USD in the system after paying a fee of 0.5% (modifiable by Lybra DAO) at any time. Since the protocol itself is over-collateralized, as long as the protocol mechanism works properly, the combination of over-collateralization and rigid redemption mechanism can provide a lower limit of $0.995 for the price of eUSD. Stablecoin protocols that have adopted this design in the past (such as Liquity) have been able to effectively maintain the lower limit of the stablecoin price, which constitutes a core component of the current price stability mechanism of eUSD.

Regarding the upper limit of the price, in the current V1 version, Lybra Finance lacks an effective mechanism to return eUSD in a positive premium state to 1 U. This is a significant flaw in its current product design, which we will discuss in detail later.

The design of the liquidation module for stablecoins that adopt the over-collateralization model is a crucial factor in protocol security, and a robust liquidation mechanism is also necessary as a complement to an effective price stabilization mechanism.

In terms of liquidation, Lybra introduces two roles: Liquidator and Keeper. Liquidator provides liquidation funds (eUSD), and Keeper triggers the liquidation operation. They can receive rewards of 9% and 1% respectively. To better protect the collateral of the pledgers, Lybra adopts the mechanism of forced partial liquidation: the maximum liquidation ratio for a person being liquidated is 50%.

At the same time, when the overall system's collateralization ratio is below 150%, all users with a collateralization ratio below 125% can be completely liquidated in order to quickly increase the overall collateralization ratio of the system.

We can see that Lybra Finance mainly refers to the core mechanisms of stablecoins (minting/redemption/liquidation/price stability), such as the stablecoin protocol Liquity (readers can refer to our previous articles about Liquity: Liquity: the emerging star of the stablecoin market and Liquity: an overview of the leading decentralized stablecoin), but it is not a mechanical copy of Liquity. They retain Liquity's redemption and recovery modes, remove the stability pool (thus eliminating some token incentives), and instead have the Liquidator provide liquidation funds.

Liquity's core mechanism has been validated over a period of more than 2 years and has good stability. We can see that several main stablecoin projects on LSDFi, such as Lybra, Raft, Gravita, and Prisma, have drawn inspiration from Liquity to varying degrees. Based on Liquity, Lybra Finance has been operating for over 3 months and the state of operation is relatively good. However, whether it can maintain this in the future still needs time to test.

3. Stablecoin Interest Mechanism

Simply holding the stablecoin eUSD can earn approximately 8% APR, which is an important selling point for Lybra Finance to attract users. It is also an innovative feature in the stablecoin mechanism that I believe is rare recently. Let's delve into the interest mechanism of eUSD.

Source: Lybra Finance Official Website

We know that Lido Finance's goal in issuing LSD stETH is to maintain a 1-to-1 peg with ETH, and that the ETH pledged by users continues to generate income on the Ethereum consensus layer. This requires the mechanism of Rebase, where Lido takes a snapshot of all stETH holders and their holdings on the Ethereum mainnet every day, and proportionally distributes the increased portion of stETH to all stETH holders. In terms of user experience, this means that the amount of stETH held by users increases every day, and their income is also reflected in the increased amount of stETH. Typically, users can earn staking income from holding stETH. Most DeFi protocols that integrate stETH also proportionally distribute the income generated from stETH rebase to users.

However, Lybra has taken a completely different approach. After users deposit stETH into Lybra, the stETH generated from rebase is not distributed to stETH holders. Instead, the protocol converts stETH into its stablecoin eUSD on the secondary market, and then proportionally distributes the income from this portion of eUSD to all current eUSD holders. In this process, Lybra deducts an annualized 1.5% of eUSD based on its issuance size as protocol revenue, which is distributed to holders of $esLBR (escrowed LBR obtained through locking LBR or mining).

Through this process, Lybra Finance achieves "asset income distributed by liability". To better understand this process, consider the following example:

Assume on Day 1, the current ETH price is 2000 eUSD and the stETH APR is 5%:

• Adam deposited 10 stETH into Lybra and minted 7000 eUSD. Alice's CR (Collateralization Ratio) is 285.7%;

• Bob deposited 10 stETH into Lybra and minted 10000 eUSD. Bob's CR is 200%;

• Charlie deposited 10 stETH into Lybra and minted 13000 eUSD. Charlie's CR is 153.8% (only 3.8% away from the liquidation threshold of 150%);

• Adam, in need of longing for ETH, sold his 7000 eUSD to David in exchange for David's ETH. As a result, David now holds 7000 eUSD;

• Eric also holds 10 stETH, but he did not deposit it into Lybra;

There are a total of 30 stETH in the Lybra system, minting 30000 eUSD. The overall CR of the system is 200%.

On Day 2, the 30 stETH generated 0.0041 stETH at a 5% APR. Lybra will convert this 0.0041 stETH into 8.219 eUSD and distribute it to Bob, Charlie, and David, based on their respective holdings of eUSD:

• Adam, although he has created eUSD, does not currently hold any eUSD, so he did not receive any new eUSD. His APR on Day 1 is 0.

• Bob currently has 10,000 eUSD and received 2.74 eUSD based on a ratio of 1/3. His APR on Day 1 is 2.74 * 365 / (10 * 2000) = 5%. If we consider a 1.5% annual fee for Lybra, Bob's one-year earnings would be 3.5%.

• Charlie currently has 13,000 eUSD and received 3.56 eUSD based on a ratio of 13/30. His APR on Day 1 is 3.56 * 365 / (10 * 2000) = 6.5%.

• David currently holds 7,000 eUSD and received 1.92 eUSD based on a ratio of 7/30. His APR on Day 1 is 1.92 * 365 / 7000 = 10% (note that David's APR denominator is different from Bob and Charlie).

• Eric earned his stETH with a 5% return.

From the above examples, we can see that:

In the absence of considering Lybra's share, the APR of casting eUSD users in eUSD base is APR=stETH APR / personal CR * system CR. If we assume that the stETH APR and system CR are constant in the short term, the way to increase personal APR is to lower their personal CR. The lower the CR, the higher the income, but it also means that there is a greater risk of liquidation due to price fluctuations. Of course, currently, users who cast eUSD can obtain incentives for esLBR based on the ratio of casting eUSD. This part of the APR is currently around 20% (income calculated based on the amount of casting eUSD), which is the main motivation for incentivizing users to cast eUSD.

Lybra's design has obvious benefits: it provides a powerful use case for their stablecoin, eUSD - interest-bearing stablecoin. Currently, decentralized stablecoins in the DeFi world primarily serve as "yield farming tools" rather than their intended purpose as a unit of value and medium of exchange. Even MakerDAO, which has the first-mover advantage and network effects, is trying to recover its declining Total Value Locked (TVL) and stablecoin supply by offering the interest-bearing stablecoin sDAI with up to 8% APR. In comparison, eUSD currently offers a stablecoin-based annual yield of approximately 7.5% to 8%, fulfilling its role as a yield farming tool quite well.

In fact, it's not uncommon for collateral deposited by users into overcollateralized stablecoin protocols to generate returns and redistribute them. From a protocol standpoint, overcollateralization is mainly to ensure the overall security of the protocol. However, having such a large amount of high-quality collateral trapped within the protocol is a significant waste of resources. If the collateral can generate returns through a secure and reasonable method, it benefits users by providing additional yield while obtaining stablecoins. From the protocol's perspective, it is also reasonable to extract a portion of the "commission" as protocol revenue.

For example, Alchemix Finance, a project supported by DeFi "godfather" Andre Cronje that launched earlier this year, introduced alUSD, which has the feature of "automatically repaying loans". Its core logic involves depositing stablecoins from users into Yearn, and using the generated yield from Yearn to repay users' debts. Alchemix then takes a 10% cut of the earnings as protocol revenue. The subsequent logic of the alETH product is similar, relying on the interest from collateral to automatically repay debts while also meeting the demand of these interest-earning asset holders (DAI, wstETH) for asset liquidity.

Another example is MakerDAO, which uses the USDC accumulated by the Price Stability Module (PSM) to purchase various Real-World Asset (RWA) products. The resulting earnings serve as protocol funds for MakerDAO, covering daily expenses, distributing to sDAI holders, and providing liquidity for MKR (equivalent to a share buyback).

Lybra Finance's approach is different from Alchemix and MakerDAO. Lybra focuses on the interest-bearing assets held by users and redistributes the interest generated by those assets. Critics argue that all the profits obtained by eUSD come from the original profits of stETH depositors. Depositing stETH into Lybra does not generate any additional income, but instead incurs a fee of 1.5% charged by Lybra. This creates a completely zero-sum game among all minters. While this may be true if we disregard other incentive factors, as we have seen in the current decentralized stablecoin market, the circulation of any stablecoin is essentially sustained by opportunities for various yield farming. Ultimately, the mines being dug are sourced from the stablecoin project's own governance token (although stablecoins like DAI, which are in a leading position, may also mine other tokens rewarded by other projects for promotional purposes). With the addition of protocol token incentives, the entire system can operate to a certain extent, just like the current state of Lybra.

For stablecoin holders (users who hold stablecoins without minting and holding the stablecoins), this design clearly allows them to obtain more "organic" profits: the profits are directly in the stablecoin itself, and there are no lock-up clauses or various complex tokens to deal with. They just need to hold eUSD every day, and the profits will be automatically credited to their accounts.

Lybra's idea may be as follows: by providing stablecoin holders with a better interest-bearing stablecoin, stimulating the demand side of stablecoins through mechanisms rather than incentives, and then combining early incentives for the supply side to stimulate stETH holders to mint and form a closed loop.

Of course, this design also brings a series of problems, such as the characteristic of eUSD rebase, which makes it difficult to be integrated into other DeFi protocols and inconvenient for cross-chain transactions, greatly affecting the composability of eUSD.

Furthermore, to some extent, Lybra puts all eUSD minters in a "prisoner's dilemma".

Since everyone's APR = stETH APR / personal CR * system CR, and stETH APR is an external input parameter that cannot be adjusted, individuals mainly focus on: 1 reducing personal CR, and 2 increasing system CR in order to increase their APR.

In the above example, Charlie obtained a mortgage rate of 200% for the average system yield (i.e., excluding Lybra's commission) for stETH. Since Lybra's liquidation CR is 150%, it means that Charlie bears a maximum stETH depreciation risk of no more than 25% by obtaining stETH base yield through Lybra. If the stETH depreciation exceeds 25%, Charlie may lose their stETH.

If the overall system CR is 300%, Charlie only needs to set their CR to 300% in order to obtain stETH base yield. This way, Charlie can bear a stETH depreciation of 50%. Both methods can generate stETH base yield, but the latter can tolerate more risk, making it clearly better than the former.

From a macro perspective, if all the minters can maintain a higher CR, their yield rates can be consistent and they can tolerate more risk. From a micro perspective, each minter has an incentive to reduce their CR in order to increase their yield. As a result, minters collectively reduce their CR, reducing the overall risk-bearing capacity of all participants without changing the overall yield.

Of course, compared to the two issues mentioned above, the problem of eUSD premium could be more severe in the short term. In the current design of Lybra V 1, the daily income generated by all collateralized assets is used by the protocol to purchase eUSD from the secondary market and then distribute it to all holders of eUSD. This means that there is a fixed buy side for eUSD every day. The forced redemption mechanism of eUSD mainly solves the problem of eUSD deviating below its anchor. However, when it deviates above the anchor, the protocol has no means to push the price of eUSD back to 1, which means that eUSD remains slightly deviated upwards in the long term. Although the degree of deviation is not high, the risk-reward ratio of an investment project with a target annual return of 7.5% and a premium purchase of 3% is not high, which also restricts the development of eUSD from a mechanism perspective.

eUSD history price source Geckoterminal

4. Token Model

The governance token of Lybra Finance is LBR, with a total supply of 100 million tokens. Of these:

• 60% of the tokens are allocated for mining incentives across the protocol. This includes incentives for minting eUSD, incentives for the eUSD-USDC LP, and incentives for the LBR-ETH LP.

Incentives for LBR can be found at: https://lybra.finance/earn

• 8.5% of the tokens are allocated for the team, which will be linearly released over 2 years starting 6 months after the TGE.

• 5% of the tokens are allocated to advisors, which will be released linearly over 1 year after TGE.

• 10% of the tokens are allocated as ecological incentives. 2% will be unlocked at TGE, and the remaining portion will be released linearly over two years.

• 10% of the tokens are allocated to the protocol treasury and will be released linearly over two years.

• 5% of the tokens are allocated for IDO, raising a total of $500,000.

• 0.5% of the tokens are allocated for whitelist rewards of the IDO.

According to Coingecko data, the current circulating supply of LBR is 11.78% with a total of 11.78 million tokens.

Source: https://www.coingecko.com/en/coins/lybra-finance

The use case of LBR is mainly achieved through esLBR (escrowed LBR), and the rewards for protocol mining are also distributed in the form of esLBR.

esLBR cannot be traded or transferred, but it can share protocol revenue (a 1.5% annualized eUSD scale deduction). Users can unlock esLBR into LBR through a linear release over 30 days, or lock LBR to obtain esLBR for mining speed bonuses and sharing protocol revenue. In addition, esLBR also has the function to participate in protocol governance.

In V 2, there are also significant adjustments to the Lybra reward distribution module. We will now provide detailed explanations.

5. Lybra V 2

Lybra V 2 is currently live on the testnet and the documentation has been released. It is currently undergoing an audit by Halborn and is expected to go live no earlier than the end of August.

Regarding the scope of V 2, Lybra has provided a detailed description in the following image:

In summary, the changes in V 2 are as follows:

First, Lybra V 2 will introduce a new stablecoin called peUSD and will also support more Liquid Staking Tokens (LST).

Currently, there are two types of LST (Language Specific Tags) in the market, one is stETH and sETH issued by Stakewise with rebase features, the characteristics of which we have already introduced in previous texts. Lybra can support this type of LST very well. The other type is "value-accumulation" LSTs, such as rETH from Rocket Pool, cbETH from Coinbase, wBETH from Binance, swETH from Swell, and wstETH from Lido. The characteristic of this type of LST is that the value of LST increases relative to ETH exchange rate. For example, when users hold rETH, the quantity of rETH does not change, but the amount of ETH that can be exchanged for each rETH will continue to increase. Therefore, we call it value-accumulation LST. Lybra's current eUSD interest mechanism can only be applied to rebase LSTs and does not apply to value-accumulation LSTs. In order to solve this problem, Lybra has launched a new stablecoin called peUSD (pegged eUSD), which can be directly minted by value-accumulation LSTs. peUSD and eUSD are similar in price stability, liquidation, and fee mechanisms. The main difference is that peUSD generated by value-accumulation LSTs is not an interest-earning stablecoin. Holding peUSD does not automatically generate income (because the accumulated value of the collateral still belongs to the user). Of course, peUSD can also be wrapped by eUSD, and peUSD wrapped by eUSD can obtain the income generated by eUSD's rebase.

Through peUSD, the composability of Lybra can be greatly improved: on one hand, Lybra can obtain a more stable stablecoin peUSD, which is beneficial for its expansion across multiple chains and integration with other DeFi protocols; on the other hand, through the design of peUSD, Lybra can expand its protocol collateral to value-accruing LST that were previously unsupported, achieving comprehensive coverage of LST. Additionally, the eUSD stored in peUSD can be used to provide flash loan services to generate additional income for eUSD holders according to Lybra's plan. However, issuing two stablecoins in a stablecoin protocol is relatively uncommon and will raise the user's cognitive threshold. Although peUSD and eUSD are somewhat related, their mechanisms are quite different, which to some extent affects their overall expansion in the C-end. Moreover, if Lybra's composability is achieved through peUSD instead of eUSD, it may be awkward for eUSD's positioning: the sole purpose of holding eUSD for users is to obtain a stablecoin yield of 7.5% (other uses such as leverage require holding peUSD), and this yield is only derived from the token incentives of LBR, which could potentially turn eUSD into a typical Ponzi scheme that is sustained by governance token incentives.

Secondly, Lybra V2 has made significant adjustments to the acquisition of esLBR token rewards, proposing two bounty programs: Advanced Vesting Bounty and DLP Bounty. These two adjustments are mainly based on the V2 version of the lending project Radiant on Arbiturm.

In V2, the unlock time of esLBR will be extended from 30 days to 90 days, but users are allowed to unlock it early, albeit with a cost ranging from 25% to 95% depending on the length of time remaining until full unlock. The cost of unlocking esLBR early will become the advanced vesting bounty.

The DLP (Dynamic Liquidity Provisioning) bounty requires eUSD users to maintain a minimum of 5% LBR/ETH liquidity provisioning in order to receive their esLBR rewards. If a forging user is unable to maintain more than 5% LBR/ETH liquidity, their esLBR will be converted to DLP bounty.

The unlocked esLBR from the advanced unlocking bounty and the DLP bounty allows users to purchase LBR/eUSD at a discounted price. The received LBR will be completely destroyed, while the received eUSD will enter the stability fund (detailed below).

By unlocking the bounties in advance, Lybra V2 aims to establish a longer-term connection between protocol incentives and the development of the protocol itself. However, from another perspective, such high friction may impact users' willingness to participate in Lybra's ecosystem mining.

Another important improvement in Lybra V2 focuses on the price stability of eUSD. As we analyzed above, the current mechanism for eUSD's interest accrual is the daily purchase of stETH on the secondary market, which is then distributed to all eUSD holders, creating continuous purchasing power for eUSD. In V2, Lybra introduces the following two measures to address the premium of eUSD:

• Firstly, a premium suppression mechanism is designed: When the exchange rate of eUSD/USDC exceeds 1.005 (i.e., a premium of more than 0.5%), the protocol will change the daily stETH yield to purchase USDC on the secondary market and distribute it to eUSD holders. This transfer of continuous purchasing power to USDC resolves the long-term root cause of eUSD premium;

• Next is the stable fund. The stable fund is accumulated by users using eUSD to purchase esLBR at a discount. The eUSD reserved in the stable fund can also be used to control the price of eUSD when it is overpriced. By using the premium suppression mechanism and the eUSD stored in the stable fund, the overpricing issue of eUSD should be solved. Combined with the rigid redemption mechanism that provides a price floor for eUSD, it is possible to achieve a stable anchor for the price of eUSD.

Overall, Lybra V2 addresses the issues exposed in the previous version, such as poor composability of eUSD, inability to extend the product model to value-accumulating LST, and difficulty in recovering from overpriced eUSD. It is more targeted and attempts to closely link the incentive token LBR with the long-term development of the protocol.

Of course, these changes will greatly increase the complexity of the protocol, and having two sets of stablecoins with significant design differences may also affect user adoption. At the same time, the advantage of composability of peUSD may make the positioning of eUSD somewhat awkward.

Summary

In the concurrent LSDFi project, Lybra has a relatively poor investment background (no institutional investment) and the lowest fundraising amount of $500,000. There have also been ongoing concerns and FUD about Lybra. However, in the current LSDFi project, Lybra has the highest TVL and token market value, and the fastest progress. Not only has V1 been successfully launched and running, but V2 has also entered the testnet phase. This reflects the operational and BD capabilities of the Lybra team on one hand, and the strength of their product on the other hand.

Compared to other LSDFi stablecoin protocols that make minor improvements on existing stablecoin protocols, Lybra has truly brought innovative ideas to the distribution of LSD's own earnings. By providing users with a stablecoin with an APR of around 8%, a relatively stable demand scenario is constructed. On the supply side, appropriate token incentives are applied, enabling the protocol to develop rapidly. From the perspective of stablecoin holders, this type of APR is more sustainable than stablecoin APR relying on protocol token incentives, which also forms the foundation for the long-term development of Lybra.

But as a stablecoin that already has a scale of 200 million and is striving for breakthroughs, the above is far from enough. Expanding use cases is currently the top priority for all decentralized stablecoins, and compared to FRAX, LUSD, and even smaller-scale alUSD and MIM, eUSD obviously has much fewer use cases. If peUSD cannot build a more diverse use case in future development, Lybra remains a mining game - although Lybra's mining game design is more sophisticated and further complexified in V2.

However, regardless, Lybra Finance has already become a cornerstone of LSDFi, and we look forward to its future development.