[IDEA] Isolated Lending Markets


To expand its market scope and onboard more assets, Venus needs the ability to separate more risky assets and allow users to control in which markets they take risk.



The current Venus collateral pool is a common pool, which means all the assets are vulnerable to bankruptcies in a single asset. This is good for the capital efficiency of the included tokens, but requires a very conservative risk profile of chosen assets. More risky assets do not have the opportunity for lending because the net risk of the included assets cannot exceed the most conservative lender in the target market.

However, if the collateral is separated into independent lending environments, lenders and traders can choose to participate based on their personal risk preferences. Venus should add Isolated Lending Markets with custom pools that can be created by any user at any time as long as the assets have adequate oracle sources.


To increase the XVS Market Cap / TVL ratio, Venus needs to expand into newer DeFi feature categories. With lending pools of varying risk, Venus expands beyond the typical risk-conservative Financial Primitive customer base, and changes its brand narrative into one that can participate in the latest innovative protocol tokens.

DeFi Protocol Market Cap ($MM) TVL ($MM) Market Cap / TVL Ratio ($MM)
Rari Fuse (Current*) 280 785 0.36
Sushi (Current*) 830 4,950 0.17
Venus (Current*) 155 1,830 0.08
Venus (Improved Ratio) 240-335 2,500+ 0.17-0.36
Venus Gains +55-115%

*Updated Jan 10, 2022


Several isolated lending models already exist in the DeFi ecosystem, and Venus does not need a custom built solution. Imitating an existing model is the quickest route to launch of the feature.

Sushi’s Kashi Model

Sushi’s Kashi Isolated Lending model would be a very different experience for Venus users. A single collateral token is paired with a lending token, and each pair is custom-created by users, like Uniswap AMM pairs.

This system is easy to understand and isolates risk into highly customized risk levels, but it means that capital is separated across pools, reducing capital efficiency. This also means that general use is much lower as compared to larger pool lending systems. Isolated Lending Pairs like Kashi measure TVL in the millions, vs. billions for other systems.

Sushi’s Kashi system also comes with additional built-in benefits like the BentoBox model that is used as the base fund management package for lending pairs. BentoBox allows idle funds to be allocated to various farming strategies to increase net capital efficiency for users.

However, the BentoBox capital management system does not need to remain unique to the Kashi Isolated Lending model, and Venus can choose to implement this type of Account Vault structure for users’ funds in the future, even without Kashi Lending included.

Rari’s Fuse Model

Rari Capital uses a model very similar to Venus’ current lending structure, but they offer the ability for users to create custom pools of collateral for lending and borrowing rather than a single common pool for all users. This Isolated Lending model would be very familiar to current Venus users, and would also give users the ability to create pools with more risk not otherwise possible.

Upon creation, each isolated lending pool can receive custom parameters. Only if the market is chosen to be “upgradeable” can the assigned admins update parameters after the initialization process is complete.

Implementation of the core of this model would be simple due to the familiarity with the current Venus codebase, but because each pool can include more than two tokens (like Sushi’s Kashi model), risk assessment for each pool is much more difficult. A single risky token in a custom pool can create high levels of risk for the entire pool, but an inexperienced user glancing at a pool’s assets might not understand this risk. Therefore, Venus would need to include a custom pool risk rating system. Developing and testing this requirement with Venus’ risk team would extend the launch of a Fuse-style lending model.

Pool Risk

Pool Risk would be analyzed based on common token metrics for each asset from exchange data, as well as typical lending market metrics for the asset. For example:

  • Exchange Metrics
    • Market Cap
    • Liquidity
    • Volatility
    • Swap Count
  • Lending Market Metrics
    • Collateral Factor
    • Reserve Factor
    • Utilization Rate
    • Debt Ceiling
    • Jump Rate

Integrate Venus’ Protocol Pool

Because the Fuse model is so similar to Venus’ current common pool model, the interface for Venus’ current pool (protocol pool) could be integrated with the custom pools, but retain the Venus governance system for the parameters of this main protocol pool. This would make the overall Venus experience more seamless and allow users to easily compare the performance and risk of the protocol pool with custom pools.


Which Isolated Lending Model would be best for the current Venus community and draw new users looking for more risky assets to lend and borrow?


I think when it comes about community driven markets, capital efficiency isn’t the number one goal. The goal is to empower people and do partnerships. Higher TVL should come from institutional offerings, liquid staking and so on etc.

So I bet on isolated market not on common pool.

1 Like

I like this proposal. It’s different from other big competitors such as Aave. Aave has its permissioned pools, while Venus will have customized pools for users can earn greater rewards for higher risk tokens. I would like to hope that the generalized pools will be retained alongside the customized ones.

So long as the protocol and team can necessarily provide the infrastructure for these pools, I am for it.

1 Like

Partnerships are definitely possible with either model, but your intuition that partnerships might be simpler with pair-type isolated lending pools is definitely accurate. Great point.


After reading this proposal more thoroughly, as well as listen to the first community call that took place a few weeks ago, I would support the type of isolated lending that Rari Capital has. As was already stated, it provides seamless user experience and it provides easier integration with Venus’ interface. Users can assess better how much risk they would like be exposed to, not to mention that it can attract much higher TVL than Sushi’s Kashi Model. The latter seems to be a bit confusing for users, especially any new user who are just starting out. This is not to mention that would make pools capital inefficient since users can only borrow a particular token from that pool and no other token. For example, if somene wants to borrow deposit their Tron token as collateral for BUSD, but there isn’t any pool, nor the demand for it, then that user is constrained by what he can borrow.