The NFT lending market is inevitable, and any DeFi user would have predicted its rise after the 2021 NFT boom. This is a story of the NFT lending market — its start, challenges, and the path forward.
NFT lending expansion depends on fairer loans, support for other NFT types, and composable debt.
The path to NFT lending is challenging. The design patterns that make the fungible market work don’t apply to the non-fungible side — added to the fact that NFTs carry an intrinsic off-chain value that fungibles don’t. Each piece means something more to the holder. There is no standard, nor will it ever be easy to create a standard to value an NFT. Digital Art, PFPs, and community access tokens can have a floor price, but that does not consider what the NFT means outside of its traits and utility. It is never going to be easy to price a subjective market objectively.
Metrics like floor price, moving averages, or any other “pricing” methodology can work, but these are blunt instruments that are not designed for the market they operate in. Breaking the NFT lending space down, you are effectively asking if a borrower will pay back the loan terms, and if not, will the piece be worth the outstanding debt at the time of liquidation?
Peer-to-peer
Before we jump into the new design space, let’s walk through the options available today. The first is a throwback to the barter system, where NFT lending is done through an OTC peer-to-peer process. Borrowers and lenders agree to terms, and a contract trustlessly handles the transacting. This process has the distinct advantage of allowing users to match terms in an isolated trade. The peer-to-peer model will continue to exist in the space as there will always be a need for a high-touch market for unique pieces.
The constraint to the peer-to-peer model is that both borrowers and lenders must agree to terms. A borrower will generally overvalue their piece, meaning they are more to the holder than a simple floor price. A lender will place a premium on their risk if they are unaware of the borrower’s credentials or concerned about the volatility of the piece’s value. This mismatch between the aggregate borrower and the lender’s sentiment leads to a lower clearing between the two parties.
Peer-to-pool
If peer-to-peer is one method, the other alternative is to use a peer-to-pool model. While peer-to-pool has only recently started taking traction, it has distinct advantages over peer-to-peer — and a few drawbacks.
The advantage of peer-to-pool is that lenders no longer have to lend to each individual and instead can lend to a pool that handles the liquidity management. This offers a far less negotiation-driven process offering a better user experience for both borrowers and lenders.
The obvious next question is how the vaults know what terms to write for a specific piece. The answer is to find a particular valuation methodology. In the case of BendDAO, which uses appraisal services and bases terms on the appraisals. These methodologies can change over time.
Peer-to-pool also adds a collective risk to all LPs in a pool. Where peer-to-peer places an isolated risk on each loan, a peer-to-pool model naturally pools that risk.
While the risk of peer-to-pool is concentrated, it offers a far more advantageous scaling path. LPs earn yields similarly to the fungible market–add money to a pool and earn. Borrowers can access instant liquidity, and the tedious negotiation process is removed. Liquidity provider and borrower demand can scale as the market scales.
What’s wrong with peer-to-pool Highlighting the biggest flaw in peer-to-pool modeling: the valuation methodology. Any oracle-based price calculation method (like floor prices) can be manipulated. This is the hardest nut to crack in the NFT lending space.
The alternative is to get specialized appraisers. This removes the “flat” collection pricing issue and offers a more nuanced pricing methodology. Appraisers (like Upshot, NFTBank, NFTValuations) appraise based on their in-house subjective pricing algorithms. Their algorithms are typically machine learning based and give an expected value of what an NFT is worth even within each collection.
The appraisal system does not guarantee any piece’s price and is just as subjective to the model. Appraisers can value characteristics like traits, previous owner history, and a myriad of other data points to estimate a piece’s value.
At launch, Astaria will have competitive dynamics for each appraiser to ensure that their valuations’ terms balance fair terms for borrowers with the risk-return profile for lenders.
Summing up peer-to-pool vs peer-to-peer
The Astaria protocol makes the conjecture that peer-to-pool will win out as the winning protocol design. I suspect both peer-to-pool and peer-to-peer will have distinct roles depending on users’ needs. Ultimately the scalability of peer-to-peer will always limit its scope to bespoke deals, while peer-to-pool, if designed correctly, stands a chance to be the protocol design that will unlock the future of NFT lending.