While it’s easy to say with recent events that DeFi is the antidote – the question remains: what’s stopping these markets from fully being brought over to DeFi?

Although Aave and Compound (and increasingly Euler), have solidified their place within DeFi – it’s important to step back and look at what is the current set of major issues with lending markets.

For new protocols to innovate within lending, there’s a few open questions that they need to address:

• Rehypothecation is highly dangerous

One major source of complexity is with the rehypothecation of collateral. This is when supplied collateral is borrowable by other users – usually for the purpose of shorting.

A hypothetical attack on Aave, illustrated by the highly profitable trader Avi Eisenberg, is enabled by rehypothecation, which in turn enables recursive borrowing.

Newer protocols have begun to recognize this – Fraxlend, Compound v3, and Silo only allow the borrowing of one base asset.

This removes the rehypothecation in addition to removing cross asset borrows – drastically simplifying risk management, even if net borrow rates are marginally higher due to the decreased capital efficiency.

• Money Markets don’t work with long tail collateral

Lending markets are notoriously fickle to add new assets or materially change LTVs – and for good reason. Outside of the majors, the lack of onchain liquidity means that if the asset isn’t supported by Chainlink, it’s highly susceptible to manipulation. DeFi is littered with many such cases – the most recent being Quickswap Lend.

This is to say nothing of pricing for derivatives such as LP tokens, or structured products – which don’t necessarily rely on spot markets for pricing. Tarun Chitra of Gauntlet gave a great talk at SBC on how xSushi put Aave at risk – highlighting this exact issue.

There is work being done to address this – isolation mode as introduced in Aave v3, along with Euler’s permissionless isolated markets, but the lack of adoption in addition to the low LTVs indicate that maybe this isn’t the right approach.

Notably, of the 6 most borrowed assets on Euler, only WSTETH, and EUL are not borrowable on Aave.

• Lending protocols are difficult to compose with

This ties hand in hand with the previous issues around complexity of these lending protocols. DeFi’s main strength lies within composability, but existing protocols are difficult to integrate with.

While it’s trivial to use the interest bearing assets within protocols – nearly every other aspect of the protocol is under lock by governance or constrained due to the protocol design.

• What does taking a 99% LTV loan against ETH look like?

• How can money markets integrate credit scoring?

• How could Aave modify liquidations to be more efficient?

These are all difficult to answer with existing money markets – they’re usually built to be self reliant for risk management. This is also why innovation within lending is more slower moving than spot trading or derivatives.

Fin

With all these drawbacks said - money markets work well. They’ve processed billions in loan volume, and are a great repo market for liquid tokens.

The key realization is that we need to move on from money markets in order to address these issues.

There’s good indication that P2P markets could be an exciting area of innovation here.

Arguably, Euler’s innovation of permissionless isolated markets is a P2P market, given how concentrated the borrowers and lenders are.

NFT finance, where oracles are not a given, is an area where we’ve seen the most amount of innovation within P2P lending, with volumes on NFTfi, the leading NFT P2P lending platform, staying robust despite the downturn within NFTs.

Metastreet, Goblin Sax, and others have innovated around lending on NFTfi – driving efficiencies in loan pricing and creating mechanisms that allow for fills of loan requests with a UX that closely resembles P2Pool than the traditional P2P model.

We’ll have more to share soon, but this is a small teaser.

Follow our Twitter: @stardustfi