Pac 2.0 Whitepaper
Last updated
Last updated
The overwhelming preference for overcollateralized loans in DeFi is due to two core characteristics of cryptocurrencies: trustlessness and automation. In a fully anonymous and unregulated environment, such as the smart contract blockchain of Ethereum, it is impossible to enforce repayment in the event of default. Thus, the only reasonable recourse for lenders is to require borrowers to provide collateral exceeding the loan value. When the value of the collateral falls relative to the loan value, smart contracts and preset economic parameters allow for timely liquidation, ensuring that lenders (in most cases) can recover their funds.
Overcollateralized loans are not without risk. On the contrary, they can be highly risky and have historically been one of the most exploited primitives in DeFi. Overcollateralized loans rely on two key assumptions: ample liquidity and accurate price feeds. Liquidity is required when liquidating collateral, and accurate price feeds are essential to determine whether a loan remains solvent. The failure of either can lead to bad debts or even the complete collapse of the money market. Different design choices made to mitigate these risks have created fundamentally different money market protocols, leading us to a hard truth: DeFi money markets are broken.
The money markets we are most familiar with, such as Aave and Compound, have popularized a fully cross-collateralized and rehypothecated model, where every asset can be borrowed against any asset on the platform. This provides the highest capital efficiency and flexibility for users. However, it also means that every asset is exposed to the risks associated with all other assets. The most common attack vector is oracle/price manipulation, where attackers make the price of an asset ridiculously high, use it as collateral to borrow all available assets on the platform (since the collateral value is ridiculously high), and then walk away.
This forces money markets to only include the largest assets whose prices are the least manipulatable, limiting user choices.
Aave V3 has a main asset pool in the style of Aave V2, but also includes isolation mode (where risky assets can be used as collateral to borrow a limited amount of stablecoins) and eMode (which allows higher borrowing power when collateral assets are highly correlated with debt assets, such as stETH and ETH). This can be understood as inserting silos into a shared pool and controlling risks through active gatekeeping and parameter settings. Compound V3 takes a more conservative approach: only USDC can be borrowed, thus it is cross-collateralized but not rehypothecated.
Pac Finance is the first self-repaying lending and margin trading protocol on Blast, featuring unique one-click loop and one-click long/short functions for users. Pac Finance 1.0 quickly became one of the largest lending protocols on Blast, with a peak TVL of $400 million. Pac Finance has always been exploring the balance between capital efficiency and risk, and now we have the answer. It is time to take the next step.
Pac Finance 2.0 uses Relend to activate and unlock capital efficiency. Relend adopts a tiered market structure design where each layer operates independently and uses the receipt tokens from the lower layer as collateral. For example, if we consider Aave as T0, a T1 market will accept Aave's aTokens (receipt tokens users receive after supplying assets, such as aETH and aUSDC) as collateral and allow them to be borrowed by other users.
This is a simple but powerful primitive. With Relend, we can now enter a higher-dimensional design space and have some very neat properties:
Liquidity in the Relend protocol can be shared across all markets and all assets. There is no fragmentation of liquidity or forced choices for users. Users can freely supply their receipt tokens across T0, T1, and other market tiers without worrying about liquidity restrictions. The liquidity in T0 remains in the base layer after the receipt tokens are supplied to higher tiers (T1, T2, etc.) until they are borrowed, and the borrower withdraws the underlying assets with the receipt tokens. For example, a borrower can borrow aETH from a T1 market and use it to withdraw ETH from Aave.
It is easy to see that in Relend, all liquidity is aggregated in the T0 market and shared among all markets.
From the user's perspective, lenders no longer need to choose between low-yield-low-risk and high-yield-high-risk markets. If they wish, they can supply assets and earn interest from all available markets.
In Relend, risks are isolated between tiers. A key observation is that T0 is unaware of T1. From T0's perspective, T1 is just another asset supplier. All lending and borrowing activities in T1 are entirely insulated from T0.
In the example above, T0 issues aETH as the receipt token for ETH suppliers, and T1 issues aaETH as the receipt token for aETH suppliers. Even if T1 is exploited and becomes only half solvent, T0 remains unaffected because the bad debt is confined to the aaETH-aETH level and fully borne by users who chose to supply aETH to T1.
In simpler terms, Relend allows users with higher risk tolerance to achieve higher yields with higher risks without affecting those with lower risk tolerance, while still letting everyone share the same liquidity pool.
The fact that Relend markets are stacked means that the interest rate in T0 becomes the "base rate" for T1. For suppliers in T1, they at least receive the interest rate of T0 even if there is zero borrowing in T1, and continue to enjoy that base rate on top of the local borrowing rate.
For example, in Pac Finance 1.0, the ETH pool has a current utilization rate of 62.89%, with liquidity of 17,376 ETH, a borrowing rate of 20.53%, and a supply rate of 16.46%. By creating a T1 market on Pac Finance, aETH can be borrowed, and 20% of the total aETH supply is added to the T1 market. Since T1 serves riskier assets, its interest rate curve is 2x steeper than Aave's, resulting in a 2x higher interest rate at the same utilization rate. When T1 reaches an 80% utilization rate, the borrowing rate for aETH is 52.23%. Finally, for T1 borrowers, the effective rate (or composite borrowing rate) is 20.53% + 52.23% = 72.76%.
Now we understand the interest rate transmission rules in Relend:
T1 depositors receive interest from both T0 and T1.
T1 borrowers pay interest to both T0 and T1.
All T0 depositors receive higher interest rates when there is borrowing activity in T1.
Interest rate transmission between tiers shows that Relend markets are more closely connected than existing money markets. This is a solid step toward a more unified on-chain rate environment. Even for non-participants, Relend can bring higher yields without increased risk exposure.
One major criticism of established lending protocols like AAVE is the limited support for assets, typically only mainstream ERC-20 tokens, which limits the utilization of mid-tier assets. Pac Finance 2.0 can support various novel and diverse assets.
This offers significant application potential for the global money market. By reducing the need for isolated markets, our protocol promotes a unified liquidity pool, enhancing the overall efficiency of the DeFi ecosystem.
Pac Finance 2.0 aims to drive the integration and development of the DeFi ecosystem. With Relend, we can, for the first time, have a global DeFi money market accommodating the newest assets and the most innovative experiments. Through risk-isolated liquidity sharing, we provide a more efficient and safer money market solution.
Pac Finance 2.0 is committed to delivering a better financial experience for users and driving the progress of the entire industry.
