Overview

Lending protocols are a critical component within the DeFi ecosystem. However, different lending protocols have introduced various architectures with different governance mechanisms in mind. This article will primarily focus on the architectures of lending protocols from the perspective of governance. Specifically, it will explore the relationships between Collaterals and Liabilities, and discuss the configuration of key parameters in lending protocols.

Collaterals and Liabilities

Monolithic

The Monolithic architecture remains the most widely adopted structure for lending protocols. In this model, users can use multiple assets as Collaterals to borrow other assets. Additionally, the Collaterals deposited by users can be borrowed by other users, allowing the Collaterals to earn interest.

In the Monolithic architecture, to mitigate risks, certain tokens may only be borrowed but cannot act as Collaterals. For example, as shown in the diagram below, users can deposit wstETH and USDC as Collaterals to borrow LINK tokens, but LINK tokens cannot be used as Collaterals to borrow WBTC.

The most well-known example of a Monolithic lending protocol is AAVE V3, the largest lending protocol on Ethereum. In AAVE V3, most assets can serve both as Collaterals and Liabilities (as shown in the “Normal” case in the diagram below). AAVE V3 also features an Isolation Mode, where certain assets can only be used as Collaterals to borrow stablecoins but cannot be used as Collaterals to borrow other tokens.

The Monolithic architecture offers relatively high capital efficiency. For typical users, it allows them to directly deposit multiple tokens as Collaterals and borrow the required assets, with Collaterals also earning interest. However, the major issue with this architecture is the inability to incorporate most tokens as Collaterals. Introducing volatile tokens as Collaterals can lead to significant losses for the lending protocol.

AAVE V3, for example, has stringent token audits, and currently supports only around 30 tokens as collateral assets. Many of these tokens are in Isolation Mode, meaning they cannot be used as general-purpose Collaterals.

Isolated Pairs

Isolated Pairs is another popular lending protocol structure in use today. In this model, the Collaterals that users deposit can only be used to borrow a specific type of token. Additionally, the deposited Collaterals are not lent out again, so users don’t earn interest on their Collaterals.

The protocol using this model is Morpho Blue, the third-largest lending protocol on Ethereum. In this lending protocol, there is a series of discrete markets, where users can choose to deposit Collaterals into one market and borrow the corresponding Liabilities tokens. However, since the Collaterals are not borrowed again, users cannot earn interest on the Collaterals. The following diagram shows the market for WBTC and USDC. The WBTC deposited by users can only be used to borrow USDC, and the WBTC deposited by users will not earn interest.

Therefore, there is another group of users in Morpho Blue who provide USDC to the lending market to earn interest. For example, users might inject USDC into the market where WBTC is used as Collateral to earn interest. So, in the Isolated Pairs system, users are divided into two categories:

1. Deposit Providers: These users provide the Liabilities tokens to the market for others to borrow, earning interest from the lending.
2. Borrowers: These users provide Collaterals to borrow tokens but forgo the interest on the Collaterals.

For deposit providers in Morpho Blue, the same token might be available to borrow in multiple markets. For example, there could be a market where WBTC is used as Collateral to borrow USDC, and another market where wstETH is used to borrow USDC. The interest rates paid to USDC in these markets are different. The following diagram shows the interest rates for different markets where USDC can be borrowed using different Collaterals.

Therefore, if users deposit USDC and wish to earn a higher interest rate, they need to allocate their USDC to different markets. This is not a simple task because some markets may choose highly volatile Collateral tokens. If users provide USDC to such a market, they may suffer losses from extreme price fluctuations of the token, potentially leading to liquidation. As a result, a third type of market participant often appears in Isolated Pairs markets — the Curator. Curators are usually professional financial institutions that analyze the risks and returns of each market and then design their own capital allocation plans. Users can entrust their USDC to Curators to manage according to their distribution plans and return conditions.

Another advantage of Isolated Pairs is that it completely avoids the governance issues seen in the Monolithic architecture. Any user can create their own lending market without needing governance approval. For instance, USD0++, as a new type of yield token, might find it difficult to enter a protocol like AAVE, but it can be directly incorporated into Morpho Blue by establishing its own market.

From a developer’s perspective, Isolated Pairs architecture has another significant advantage: because it does not involve the secondary lending of Collaterals, there is no need for code that tracks interest on Collaterals. The overall implementation is simpler. As a result, Morpho is currently one of the most sought-after lending protocols by developers.

Isolated Groups

The Isolated Groups architecture is currently used only by Compound V3 within lending protocols. In this model, different Liabilities are paired with different Collaterals.

For example, in the mainnet version of Compound V3, if we want to borrow ETH, we can provide specific types of collateral:

Unlike the Isolated Pairs model, Isolated Groups allow users to offer multiple types of collateral for a single asset. This setup theoretically allows developers to freely create lending markets and configure the relationships between Collaterals and Liabilities. However, the only protocol currently using this model, Compound V3, has implemented a governance system based on COMP tokens, meaning the creation of markets and changes to Collaterals require votes from COMP token holders.

There has been some controversy surrounding Compound due to two instances where vulnerabilities in its code were discovered. These issues are detailed in the Rekt article “Compound Errors and Overcompensated.” Fortunately, these errors did not impact the assets in the lending protocol, which is why Compound remains the 5th largest lending protocol on Ethereum.

Isolated Groups with Mutual Collateral

The Isolated Groups with Mutual Collateral architecture is also less commonly used. The protocol Silo Finance, which currently ranks 18th on DeFiLlama’s lending protocol TVL leaderboard, utilizes this model.

The image above illustrates the lending relationships within the Silo Arbitrum version of Silo token markets. In the Isolated Groups with Mutual Collateral setup, tokens are interlinked and serve as collateral for each other. In the Silo Arbitrum version, every lending market includes both ETH and USDC tokens, along with a third token chosen by the market creator, which in this case is the Silo token.

According to the Silo documentation, ETH and USDC are called bridge assets, while the third token selected by the market creator is referred to as the base asset. This naming convention makes sense because users can create chain-based lending paths using ETH and USDC to meet their borrowing needs. For example, users might use Silo tokens as collateral to borrow ETH, then use the ETH as collateral to borrow ARB. This creates a pathway that allows Silo tokens to indirectly facilitate borrowing ARB.

One advantage of the Isolated Groups with Mutual Collateral structure is that it combines the risk isolation features of Isolated Groups, giving users the flexibility to choose lending markets that align with their risk preferences. It also helps mitigate liquidity fragmentation by enabling users to set up chain-based lending paths using bridge assets. Additionally, because of the Mutual Collateral relationship, users’ deposited collateral can generate interest.

However, Silo also provides an option to configure the Collateral Status. By choosing this option, users can prevent their collateral from being borrowed, thus ensuring its safety, though this comes at the cost of forgoing any interest income from the collateral.

Parameters

In the second part of this article, we will discuss the issue of parameters within lending protocols, focusing primarily on the parameter Collateral Factor, which is most relevant to the capital efficiency of lending protocols. This parameter measures the discounting of collateral within the lending protocol. The higher the Collateral Factor, the lower the discount of the asset within the lending protocol. Generally speaking, stable assets like ETH have higher Collateral Factors, while low-market-cap assets tend to have lower Collateral Factors.

Global Paternalism

Global Paternalism is the most common Collateral Factor parameter configuration pattern within lending protocols. In simple terms, it means that the protocol operator directly sets and modifies the Collateral Factor parameters. For example, in AAVE, the AAVE DAO organization proposes and modifies the parameters related to certain collaterals. The image below shows the configuration of parameters for ezETH as collateral in AAVE governance.

For both the protocol and users, Global Paternalism is the simplest solution.

The Invisibal Hand

Currently, lending protocols represented by Morpho adhere to not introducing a governance system, so Global Paternalism cannot be executed within the Morpho protocol. Therefore, Morpho uses another approach. In this system, any user can create markets with arbitrary Collateral Factor parameters, and depositors will store their assets in the markets they believe are the most reliable.

For example, in Morpho, we can see that for the USD0++ / USDC market, there are markets with different parameters, and users will deposit their USDC assets into different markets based on their own risk preferences.

Conclusion

This article focuses on the common architectures of DeFi lending protocols. Simply put, different lending protocols have emerged based on how much they rely on governance. We can categorize lending protocols into two main types:

1. Governance-Dependent Protocols: These protocols require strict governance processes to set parameters and choose collateral assets.
2. Governance-Free Protocols: These allow anyone to create markets with their own parameters, with users making decisions based on their preferences.

For governance-dependent protocols, the governance mechanism is crucial. Both Compound and its forks have faced issues with governance, leading to problems within the protocol. Currently, AAVE is considered to have the best governance system. A significant benefit of governance-dependent protocols is that users only need to deposit assets without needing to take any additional action, as long as they trust the governance structure. This is particularly appealing to organizations with a foundation-based structure, such as the Ethereum Foundation.

On the other hand, governance-free protocols leave everything to the market, which puts more responsibility on participants. Both depositors and borrowers need to track market movements and frequently shift funds between different markets to get the best interest rates. This model can be difficult for foundation users who cannot regularly adjust their funds due to the inherent limitations. Of course, users can delegate their funds to third parties, but currently, no third-party solution within Morpho is as trustworthy as AAVE.

Disclaimer：

1. This article is reprinted from [TechFlow]. All copyrights belong to the original author [gccofficial]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.