Discrete Time-Based Liquidation
Introduction
The entire NFT ecosystem, including NFT Finance, has expanded at breakneck speed in recent years. Although many people feel that NFTs are useless JPEGs (except perhaps for their use as ‘digital bling’ for social status signaling), there are many builders and crypto enthusiasts who, like us, see limitless future NFT use cases and are working hard to turn that dream into reality by adding more utility to NFTs, such as being able to use NFTs as collateral to borrow crypto and enhance capital efficiency.
We believe NFT lending will play a pivotal role in the decentralized Future of Finance. Billions of VC investment dollars are currently being channeled into building out the Metaverse and GameFi niches, and undoubtedly ‘Metabanks’ will thrive in the Metaverse (just like ‘Megabanks’ are invaluable IRL). More broadly, NFTs will play an integral part in the Metaverse future, as virtual land and other rare gamified assets will be required to own, buy, sell, trade and interact with other players/avatars.
Some NFT lending platforms, like NFTfi, have adopted a peer-to-peer approach for NFT lending, and some, like BendDAO, are trying to solve the problem by pursuing the peer-to-pool path. These up and coming NFT x DeFi protocols possess unique advantages and disadvantages, and all are instrumental in promoting the development of the nascent NFT lending niche.
This article will focus on introducing and explaining an innovative Discrete Time-Based Liquidation (“DTBL”) mechanism that is effective for both peer-to-peer and peer-to-pool NFT lending protocols. It not only solves the problem of cascading liquidation risk, but also greatly improves the user experience (UX) for borrowers.
DTBL User Examples
Alice, our NFT-loving heroine, wants to access liquidity (i.e. borrow some ETH) using her NFTs as collateral to add a hot new NFT to her growing NFT collection, pay her daughter’s tuition in the real world, and/or take care of some other pressing financial need. She turns to an NFT lending platform that allows her to borrow immediately based on the collection floor price. If the platform utilizes a Price-Based Liquidation mechanism like BendDAO, she must constantly pay attention to the NFT floor price 24/7/365 as the crypto market literally never sleeps. Unfortunately, Alice has a real life outside of the wildly gyrating crypto markets and needs her beauty sleep every night and definitely doesn’t want to wake up one morning or after returning from a night out with friends to find her prized NFT has been liquidated and lost forever.
One may say that Aave justifies the Price-Based Liquidation model by successfully surviving several bull and bear markets. However, NFTs are not FTs. In other words, liquid fungible tokens can be partially liquidated with minimal risk to lenders and the crypto lending protocols (like Compound and AAVE) even in highly volatile market routs, such as the Terra/UST depeg debacle in Summer 2022. However, we must remember the indivisible property of NFTs, which means that once the price drops and triggers a liquidation, Alice will lose her entire NFT, not just a portion based on the liquidation fee logic used by fungible token lending protocols.
To make matters worse, imagine if there were not just 1 but 1,000 Alices and Bobs who received NFT-backed loans from a price-based liquidation NFT lending protocol. If a Black Swan Event occurs, and they somehow seem to occur quite regularly in the Wild West of crypto markets, it means that 1,000 NFT loans are likely to face liquidation at the same time, which may require 100 or more days of forced selling to fully auction off NFTs with inherently low liquidity, which will undoubtedly have a negative impact on NFT market prices and trading activity, which in turn could induce further declines in NFT floor prices, which is the dreaded cascade liquidity crisis that everyone in the NFT space incessantly frets about.
We do not believe that a NFT lending platform should be designed in such a way that a single factor (e.g. NFT floor price) determines the life and death of both the protocol in question and perhaps the entire NFT ecosystem. Because there is a significant non-zero probability that this can, at some point, happen, even if it seems safe enough at the moment. For this reason, we propose a new clearing mechanism that is better suited for NFT lending: we call it Discrete Time-Based Liquidation (“DTBL”).
DTBL Mechanism
Back to Alice’s story, she decides to try an NFT lending platform for an instant loan. This platform employs a discrete time-based liquidation mechanism. So what does this mean exactly?
She selects one of her whitelisted NFTs and decides to stake it on the protocol as collateral for an instant loan. The smart contract calculates some key data for her loan: how much she can borrow (Borrow Limit), what fixed interest rate she must pay for the use of the funds (Borrow APR), and how long she has to repay or extend the loan (Collection Term). Alice decides to borrow 50 ETH and after accepting and confirming the loan terms, she will immediately see the 50 ETH in her wallet and can now use the funds as she wishes.
Before the loan expires, she can extend the loan and the smart contract will recalculate the borrow limit and borrow APR. If the new borrow limit is lower than the current loan balance, she’ll have to top up her loan plus interest to ensure the security of the lending system and keep her loan in good health. If the new borrow limit is higher than the current loan balance, she can borrow more crypto if needed.
In addition, she can repay her loan at any time, and when the loan is fully repaid, her pledged NFT will be immediately transferred back to her crypto wallet.
If for any reason, Alice doesn’t repay the loan by the due date, she will have a 2-day Grace Period in which she can extend or repay the loan along with a required late penalty fee. However, if she fails to take action, the NFT will be in default and will no longer belong to her.
Now, as in our previous example, imagine there are 1000 Alices and Bobs who receive loans from the discrete time-based liquidation (DTBL) lending protocol. Since each borrower borrows at different times, in the extreme case where the floor price at the time of repayment is lower than the borrow balance and many borrowers simultaneously decide not to repay their loans, the overall market effect will be muted because only a fraction of the total loans will be at risk of liquidation on any given day.
That is to say, this elegant but powerful time-based liquidation mechanism discretely diversifies the liquidation risk across time, greatly reducing the cascading liquidation risk.
A bit more detail on the DTBL mechanism
DTBL is applicable for both peer-to-pool and peer-to-peer platforms where the borrower can choose to extend the loan for an additional fixed term (e.g. one month) within the current fixed term or repay it fully to claim their NFT and avoid the loan late fee. DTBL is based on overcollateralization, and we believe that the possibility of continuously experiencing severe NFT floor price drops during a given fixed term period is extremely low. For example, taking BAYC’s floor price as an example, the probability of the floor price decreasing by 60% in a 28-day period is quite small. As can be seen from this chart, this hasn’t happened. That is not to say that it will not happen, but the probability of it occurring is low, and the probability of it occurring continuously for a sustained period is even lower.
The advantages of this DTBL mechanism are: 1) The time factor changed the liquidation trigger mechanism from a single factor (minimum floor price) that can occur at any time during the entire term of the loan to comparing the borrow amount to the floor price during a discrete time window on only a fraction of the outstanding loans during any discrete time window. In this way, the cascading liquidation risk problem has likely been solved. 2) This mechanism is very borrower-friendly, as the borrower no longer needs to pay attention to the collection floor price 24/7, but only needs to determine when the loan is due (perhaps with a calendar reminder) just like paying a credit card or mortgage payment in the real world.
Risks
We believe that any mechanism, whether in the real world or in the crypto world, is subject to a number of hazards and tests, but the basic principle is that the platform cannot collapse because of a single risk factor. In layman’s terms, eggs should not be kept in a single basket because there is no remedy in case of an unfortunate and sudden drop of said basket.
On the contrary, we believe that all financial markets, TradFi, DeFi, CeFi, etc, will continue to experience bull and bear market cycles, but the risk of a total loss with little chance of protocol recovery must be reduced as much as possible. Of course, there is no easy way to completely eliminate all risks, and this applies to any investment or asset. For example, a black swan in one of the prominent NFT collections could potentially cause the collapse of the entire NFT market. This is a major risk for the platform, but even then it meets the discrete characteristics, as is highly unlikely for all NFTs to collapse all at once.
Of course, only one new risk management method is proposed in this article. In the future, some new risk mitigation systems may be developed to further reduce the risk to lenders. For example, Dutch auctions may be used to conduct quick and efficient auctions for defaulted NFTs. Also, a LiquidationDAO with a related investment treasury can be established to ‘pre-fund’ liquidations and further protect the viability of the lending protocol and, by extension, the NFT market as a whole.
If you’re interested in this risk reduction mechanism, please reach out to us at @eurisNFT on Twitter. We’d be thrilled to hear from you and further discuss our plans and thoughts.
