Is the NFT-Fi Summer led by Sudoswap here?

Author: Bryan, IOSG Ventures

Is the NFT-Fi Summer Led by Sudoswap Here?

This article is original content from IOSG and is intended for industry learning and exchange purposes only. It does not constitute any investment advice. Please cite the source if you wish to quote it, and contact the IOSG team for authorization and reprint guidelines.

This article will discuss some interesting sub-tracks/protocols around NFT collateral lending and derivatives in the NFT-FI direction, as well as the current bottlenecks and potential future developments.

NFT AMM

The application scenarios of NFT AMM are mainly divided into blue-chip NFTs and long-tail NFTs. For blue-chip NFTs, a step before entering the AMM is to fragment them.

NFTx

NFTx is a well-established NFT fragmentation protocol that mainly references the old DeFi gameplay, which means that using the AMM mechanism can provide liquidity mining after fragmentation. At the same time, Floor Dao provides a certain liquidity solution for fragmented NFTs, allowing users to exchange ERC-20 NFTs for floor tokens, with high APY as rewards. The basic logic is consistent with the Olympus DAO of DeFi 2.0. It indeed increases the use case of ERC-20 NFTs, but it is not sustainable.

Fragmented NFTs will facilitate price discovery, for example, the floor price in the secondary market for NFTs aligns with the price of NFT fragments circulating in the AMM. If the trading price of fragmented NFTs is lower than the secondary trading market, such as OpenSea, arbitrageurs can buy a complete NFT fragment from the AMM and redeem the NFT from the treasury, then sell it in the secondary trading market to achieve risk-free arbitrage.

Sudoswap

Since the trading directly involves the entire NFT, its positioning is closer to secondary markets like OpenSea and SuperRare. The distinguishing point is that it can provide instant liquidity that secondary markets do not have, ensuring that traders can know the predetermined trading profits through linear/exponential bonding curve settings (more curves matching different trading needs will be introduced in the future) and have an instant trading experience. The pricing methods for the same NFT series can vary, allowing for risk-free arbitrage within the protocol.

Currently, the traction is quite good, with the platform reaching 100,000 NFT trading volumes and showing a good upward trend in TVL. One concern is mainly UX, as it is permissionless, and currently, the same NFT series can have up to 200+ pools, with varying depths, some having only single-digit depths. This liquidity dispersion may be a potential bottleneck hindering the further development of the protocol. Additionally, most NFTs traded on the platform are long-tail assets, with lower popularity for blue-chip series that have higher entry barriers.

Comments:

We believe that the main application scenario for NFT fragmentation protocols is in blue-chip NFTs with relatively high floor prices, where fragmentation lowers the user entry threshold and increases liquidity. For long-tail non-blue-chip NFTs, their floor prices are relatively lower, and there is no need for further fragmentation; they can be traded directly. Although both trading models are AMM, the target NFTs are different.

Is constant product AMM a suitable trading model? Constant product (x*y=k) is a common model in ERC-20 swaps (Uniswap, Bancor), but LP's impermanent loss is proportional to the volatility of the assets they provide liquidity for. This means that the greater the price volatility of the assets, the higher the LP's impermanent loss. For fragmented NFTs, their price volatility is proportional to that of spot NFTs, so impermanent loss is also a significant risk exposure for LPs. (NFTx has designed a new staking model to reduce impermanent loss in this scenario—inventory staking, but it does not fundamentally improve the existing impermanent loss issue).

Currently, the mainstream innovation in the market is bonding curves like Sudoswap, which ensure that the price fluctuations within the same pool are independent of the number of NFTs in the pool, fixing price slippage and promoting a better trading experience.

The author's personal view is that AMM is one of the ways to improve liquidity/capital efficiency (mainly to facilitate price discovery for NFTs near the floor price), but AMM is not a suitable trading model.

NFT as Collateral Lending

In addition to directly trading NFTs, lending protocols using NFTs as collateral are also one of the means to promote NFT liquidity. NFT lending mainly falls into two models: P2P/P2Pool. P2P refers to peer-to-peer trading, where one borrower interacts with one lender. The P2Pool model involves multiple lenders providing assets, not targeting a single borrower.

P2P: Unique Play in P2P Lending.

Flowty:
A lending protocol on the Flow chain, it focuses on PFP NFTs as collateral, mainly featuring NBA Top Shot and Ballerz. Co-founders are early participants in NBA Top Shot, and this platform is more focused on lending projects in the sports NFT segment. One distinguishing point is its collateral—NBA Top Shot's brand effectiveness/events, such as regular challenges held by NBA Top Shot, where participants can use their NFTs to participate in challenges and earn rewards. Flowty can help participants unlock additional funds to purchase the NFTs they need.

At the same time, there are a series of third-party platforms on the Flow chain for reference, such as the aforementioned ranking system Moment RANKS, and other analytical platforms like Own the Moment (which includes historical sales data, price predictions, trends), etc., helping lenders make better judgments on whether to provide loans to borrowers. Overall, there are many gameplay options, but scalability is limited, and the ceiling depends on the Flow/NBA Top Shot ecosystem.

NFTfi:

The pioneer of NFT P2P lending, and currently the platform with the highest depth and best user experience for PFP series NFT P2P lending on the market. In April this year, the open loan positions reached over 2k, far exceeding other similar competitors. Some interesting findings: high-value NFTs, such as BAYC/Punk, generally have longer durations—ranging from 30 to 90 days. The LTV of these high-value NFT series is also relatively high—(past month: BAYC-77%, Punk-53%, MAYC-58%, with collateral prices based on the average transaction price over the past 30 days). For a typical pawn shop with 25-30% LTV, it can be concluded that blue-chip NFTs have relatively high bargaining power in the lending market. A lender-friendly aspect is that the borrower's lending history will be displayed (such as the amount previously borrowed and whether there have been any defaults), and other reference data includes the floor price of the NFT series and the value estimates provided by Upshot/NFTBank, which help lenders make reasonable judgments.

Comments Generated While Exploring P2P Lending Models:

Basic terms of P2P: The standard LTV for NFT P2P lending is close to that of pawn shops in real life, fluctuating between [25-50%]. Moreover, the longer the duration, the higher the risk for lenders, so short-term + low LTV is a more suitable choice for NFT P2P.
Two interesting findings:

a. Overall, the default rate is relatively low, around 10%. There are also hierarchical differences in default rates among blue-chip NFTs. For example, the highest borrowing volumes for BAYC, Punk, and MAYC have default rates of approximately 1%, 3%, and 5%, respectively. However, blue-chip NFTs with greater price volatility, such as Azuki, have a default rate of 30%. This more indicates that for blue-chip NFT lending, borrowers tend to hold rather than default (i.e., cashing out using the platform).

b. The APY and duration data can also explain the differences in perceived value retention among different blue-chip NFT series in the eyes of lenders. For example, Punk's APR is relatively lower compared to other NFTs (averaging 14%, even during recent volatility, it remained around 10%, while BAYC's APY reached 40%), and the duration utilized (actual repayment time/scheduled repayment time) indicates that lenders and borrowers have a higher recognition of Punk's value retention—despite BAYC's historical floor price being more stable than Punk's.

Development Bottlenecks:

a. Terms—The current inefficiency of P2P protocols is due to the process of reaching a deal involving multiple rounds of bidding and negotiation (the validity period of an offer is 7 days). How to formulate terms more accurately and quickly will be a consideration for P2P protocols.

b. For some high-value NFTs, due to their inherent high risk, it is necessary to rely on liquidity solution providers like Metastreet to complete lending.

P2Pool:

BendDAO: A mainstream NFT P2Pool lending protocol on the market. The gameplay is straightforward: lenders lend ETH, and borrowers collateralize blue-chip NFTs. It has developed very quickly, with 830 borrowers (NFT-Fi has over 1200), considering that this number is just a few months after its launch. The rapid development can be attributed to the higher liquidity of P2Pool compared to P2P, with collateralized NFTs close to the floor price, and the protocol initially provided cold start subsidies: 6% deposit APR and -16% borrowing APR. The collateral data is also good—MAYC/BAYC has stabilized at over 300+ and 250+ in the past two months. Currently, the protocol has not seen any liquidations of NFTs other than Azuki. Recently, there has been a significant risk of liquidation for many BAYC NFTs.

Innovative Protocols:

Defrag: A lending protocol attempting risk control.

When a borrower collateralizes an NFT, they will purchase a put option (if the collateral price drops, the underwriting pool needs to lock a certain amount of ETH to redeem this NFT). The ETH in the underwriting pool can be used for: 1. Converting to UDTC to lend to users 2. Redeeming NFTs. The profit of the underwriting pool comes from the premium fees of the put options.

If the put option strike price is 100 and its current value is 80, then this put option is worth 20. The liquidation price is 70, and the user has not been liquidated yet, so the put option will not be executed (liquidation means executing the option). If the price drops to 60, then it executes, the underwriting pool buys back the NFT, and the user receives 100-60=40 + borrowing amount. The underwriter essentially combines the roles of lender and option seller, but in reality, the two roles are not the same—while the lender is shorting the collateral as the counterparty to the borrower, as the option seller (short put), they are long on the collateral, which creates a conflicting setup.

Image source: https://twitter.com/DefragFinance/status/1461770614881067009/photo/1

Astaria:

Just announced an $8M financing, a project for NFT P2Pool lending led by the former CTO of Sushi, with no specific protocol design yet. According to existing information, its model is similar to Maple Finance, with an intermediary (delegate/appraiser) setting the pool's terms—price/amount/rate, etc., and having a last resort function.

Comments on P2Pool Protocol:

Currently, the high TVL NFT P2Pool protocols lack innovation compared to DeFi's ERC-20 P2Pool. We expect more interesting protocol designs to emerge in this field.

The distinguishing points among different P2Pool protocols mainly include: liquidation mechanism, LTV, and risk control.

The liquidation mechanisms mainly include two types: time-based and price-based. Time-based is more borrower-friendly, allowing them to repay within a specified time without facing liquidation risk, while LTV is potentially a trade-off.

Moreover, in P2Pool, liquidators play a key role in ensuring the protocol's risk control, but they are also relatively independent participants from the protocol itself and the lending parties. Their existence and requirements are: 1. Profitability 2. A certain understanding of collateral pricing.

Regarding profitability, in the case of insufficient liquidity for NFTs, especially in extreme market conditions, the rapid decline in collateral prices may lead these independent liquidators to choose not to participate in liquidations, ensuring that their pricing understanding prevents them from selling NFTs at irrational prices. This requires more reasonable pricing mechanisms. (The third-party valuation systems are not mature enough, and even blue-chip NFTs can have significant price discrepancies for the same NFT).

Risk control mainly concerns the protocol mechanism for timely repayment to lenders in the event of borrower defaults (such as black swan events).

Image source: https://blockcrunch.substack.com/p/NFTfi-understanding-the-NFT-lending

Closing Thoughts

Many people say NFTs are like real estate, but I don't quite agree. NFTs are a product of the blockchain, inherently possessing liquidity attributes that real estate does not have. Although current NFT protocols mainly focus on NFTs as collateral (similar to the low liquidity of real estate), an interesting future direction is whether there will be lending/borrowing protocols where the underlying assets are all NFTs (for example, collateralizing BAYC to borrow Crypto Punk). A positive signal in this direction is the Seaport protocol launched by OpenSea, which opened the door for NFT trades of NFTs. It is not hard to imagine the future emergence of situations where NFTs are collateralized and borrowed.

In an efficient capital market, such as the BTC lending market, the forces of long and short sides are evenly matched. However, current NFT protocols are all focused on protocols where NFTs are used as collateral (i.e., Long NFTs), essentially promoting the liquidity of ERC-20 (what is borrowed is ERC-20). At the same time, the current development of NFTs is still mainly based on PFPs, which do not inherently have income-generating characteristics, so the demand from borrowers is very vague. This is also a very urgent issue that needs to be considered in this direction. Perhaps NFTs with practical landing scenarios/production activity needs, such as gaming NFTs, are more suitable.

Regardless, the next cycle of NFT lending will definitely be closely related to the broader crypto/NFT market's bull/bear trends. In a bull market, such lending protocols serve as a natural leverage tool, greatly promoting the development of the protocols.

P2P or P2Pool?

Logically speaking, P2P is more suitable for NFTs with net values relatively higher than the floor price, while P2Pool is more suitable for NFTs with relatively lower net values, close to the floor price. P2P emphasizes the borrower's experience (customized terms), while P2Pool emphasizes the lender's experience (multiple lenders disperse risk while lowering the entry threshold).

P2P does not have the liquidation risk of collateral price fluctuations; it only requires timely repayment. In P2Pool, borrowers may frequently receive margin calls.

P2P LTV has a bar (even for BAYC, it averages around 80%), but P2Pool's LTV is relatively high—BendDAO's maximum LTV can reach 90%. The capital utilization rate is higher in P2Pool.

What is Next?

DeFi lending giant Aave expressed interest in entering the NFT as collateral lending market about a year ago but has not made any new progress. Compared to traditional finance, NFT lending remains a high-risk activity, requiring risk hedging measures to diversify risks, including but not limited to NFT insurance, futures, options, and even structured products.

NFT Futures

NFTperp is an NFT futures trading market based on the vAMM model, which was designed by Perpetual Protocol. Compared to AMM, the vAMM model requires LPs to inject liquidity and has pool size limitations, with a funding rate mechanism ensuring that futures prices converge with the underlying asset prices—the vAMM price is analogous to the perp price, and its price feed from oracles serves as the basis for calculating the funding rate. However, like the AMM mechanism, this vAMM mechanism requires trading volume for cold starts, and without LP rewards, early project launches may encounter some resistance.

Synfutures launched NFTures, which is also an NFT futures contract trading market, currently supporting contract trading for four NFTs—PUNK, UJENNY, NFD, The Doge NFT. The price feed for the underlying assets comes from Unicly/NFTx. Synfutures' sAMM is also battle-tested, with trading volumes comparable to Perpetual Protocol/DYDX.

For NFT futures, there are several bottlenecks:

The price of the underlying asset. In comparison to BTC futures, such as DYDX/Perpetual Protocol, the price of the underlying asset is fed through oracles for spot prices (DYDX uses MakerDAO's oracle, while Perpetual Protocol uses Chainlink). However, for NFTs, there is a lack of a suitable and reliable price feed—essentially still a liquidity issue. Perhaps when trading platforms that can handle higher NFT trading volumes, like Blur/Sudoswap, achieve a certain level of liquidity, their trading prices can serve as price feed options.

Demand. If we take BTC futures as an example, one of the demands is as a hedging tool for BTC miners, while many NFT holders do not have such needs. This presents an opportunity for gaming NFTs—gamers continuously obtain NFTs during the process of leveling up in games, creating a hedging demand.

NFT Options

The main direction in NFT options is still from the perspective of NFT holders, which is to hedge the risks of holding NFTs by purchasing put options.

Putty is a platform providing NFT put options, with currently 11 orders executed (including 3 BAYC put options). The platform is relatively early in its development and lacks activity. It supports customizable options (users wishing to issue put options provide NFTs, premiums, strike prices, durations), but the liquidity/transaction volume of such options is inherently low.

Nifty Options is also a platform providing NFT put options, where NFT holders can lock their NFTs on this platform or other collateral lending protocols like NFTfi while initiating a put option. The counterparty needs to lock ETH equal to the strike price, and the option issuer can choose to exercise/cancel at any time. Exercising means selling the NFT for the strike price, while canceling returns the original NFT, and the counterparty will receive a certain interest reward regardless. The interest rate here needs to be higher than that of mainstream lending protocols like Aave/Compound, as the counterparty bears more risk.

In terms of traction, it is similar to Putty, still in early exploration with little usage. A similar scheme to Defrag is to buy this option while collateralizing the NFT, ensuring that even in liquidation, the lender can still obtain assets of certain value. Compared to Defrag, the option seller and lender are not the same group, making this design more reasonable.

Jpex.finance is an option platform different from the above two, where NFT holders sell call options rather than buy put options. This is equivalent to a covered call strategy. However, liquidity is an issue—1. The counterparty must pay to enter the pool 2. European options only settle on the delivery date, and the counterparty cannot collect interest while in-the-money.

For NFT options, there are several bottlenecks:

Low liquidity. Whether it is FTX or Binance, their option setting mechanisms are quite simple (strike price, expiration date, premium)—FTX has fixed choices for strike prices, while Binance directly uses the futures price as the strike price, and both automatically calculate the premium. For NFT options, everything is filled out by the NFT holder, and the premiums are often arbitrarily set, leading to a lack of willing trading counterparties. Additionally, the UI/UX is significantly inferior compared to professional trading platforms.

Option pricing. Option pricing is mainly determined by two factors—intrinsic value and time value. The time value is the reason NFT options are priced higher—due to high volatility (historical & implied). One method to consider for lowering premium prices is to shorten the exercise time window, promoting speculation on NFT prices within smaller time frames.

The number of NFT collections. Especially in bull markets, there is a higher demand for calls, while the sellers of calls are relatively few (compared to BTC holders), which cannot adequately meet the leverage demand in bull markets.

NFT Insurance

For some high-value/high-scarcity NFT holders, in addition to relatively indirect hedging measures, a more direct approach is to obtain insurance. This may increase borrowing costs, but some risks are also transferred. Even when users choose P2P/P2Pool platforms, if the protocol has internal insurance or collaborates with third-party insurance, this could likely become an attractive point for users.

NFT Structured Products

NFT20 has an NFT index that tracks the floor prices of a series of NFTs, similar to the S&P 500 in traditional finance.
In the metaverse, land as an income-generating asset NFT (for example, users need to pay rent to conduct activities on the land) can refer to real-life mortgages as a type of financial derivative, such as CDS.

Some cash flow-based structured products, similar to Element Finance/Solv in DeFi, can allow lenders to sell their future cash flows in the form of bonds. Metastreet is doing something similar. Additionally, there are some buy-now-pay-later products, similar to mortgages, with regular repayments, such as Cyan/Ape Now, which have potential arbitrage opportunities—staking in NFTx and other NFT fragmentation protocols to pay off regular repayment rates, and similar promissory notes are also a form of play.