The Abstract Narrative of NFTFi: The Liquidity Dilemma, Introduction of New Roles, Risk Substitution and Its Extensions

Author: Kylo, Zonff Partners

From the explosive rise of BendDAO that put the NFTFi narrative back on track, to Sudoswap introducing a new paradigm for NFT trading, NFTFi has developed for nearly a year. Looking back at the opportunities for the rise of BendDAO and Sudoswap, some were inevitable while others were coincidental. The inevitability lies in the shift of user demand, while the coincidence was the emergence of a permissionless product at the right moment.

BendDAO is not the starting point of NFTFi; many NFT Lending Protocols have come and gone, seizing the early opportunities in NFT lending but failing to capture the market. Sudoswap is also not the first project to engage in homogenized NFT trading; NFTX was at least a year ahead of it.

So what is the takeaway? An underexplored narrative has actually existed in past stories; it just had a different cover and was forgotten in a corner. When people suddenly remember that this story can be told in a different way, it becomes fresh again, but the core remains the same narrative.

Starting with BendDAO

BendDAO is a story of borrowing demand. The airdrop of BAYC mobilized NFT holders' demand to expand their NFT exposure, and borrowing ETH against NFTs can maximize holders' interests. The advantage of BendDAO's permissionless collateral P2Pool lending model over the previous P2P Lending is that the borrowing experience is seamless for borrowers. As long as BendDAO can solve the funding side issues, the subsequent lending process will become natural.

The core of BendDAO is a lending protocol in DeFi, but the difference lies in that for traditional funding pool lending protocols, the user's collateral position is part of the lending supply funds. This means that borrowers are also providing excess liquidity for the lending protocol's funding pool. Here, excess liquidity means that the funds deposited into the lending protocol by users exceed the funds borrowed by users, without considering repeated borrowing. Under this mechanism, the TVL of the lending protocol will be positively correlated with the demand from borrowers. However, NFT lending differs in that the collateral itself does not possess strong capital attributes, and its commercial attributes cannot directly support the TVL of the lending protocol. This requires the protocol to additionally utilize its tokens as liquidity incentives to attract a large amount of lendable assets into the NFT lending protocol's funding pool, leading to an imbalance between the supply and demand of the protocol's tokens.

The issue of tokens is secondary and can be addressed by raising the benchmark lending rate to reduce liquidity incentives. The main problems with BendDAO focus on two areas: the liquidation mechanism and floor price manipulation. Currently, trading solutions for NFTs still primarily rely on order books, where sellers place orders and wait for buyers to act. This trading model is unsuitable for a liquidation mechanism that has high liquidity requirements. When the health factor of a certain NFT collateral lending position in BendDAO falls too low, the NFT collateral position will auction that NFT in the secondary market. Due to the lack of existing liquidity for that NFT in the market, the number of NFTs needing to be auctioned accumulates rapidly, leading to further declines in the NFT floor price and panic among depositors; floor price manipulation is very similar to oracle attacks in DeFi. Due to the lack of liquidity in NFT trading, whales can manipulate the NFT floor price by maliciously placing low-priced orders on NFT marketplaces and executing self-deals, effectively clearing targeted NFT lending positions.

Ultimately, the two problems with BendDAO are essentially liquidity issues with NFTs, where insufficient immediate liquidity leads to the accumulation of NFTs being liquidated, and the floor price is easily affected. However, the liquidity issue of NFTs has always been a "characteristic" problem of NFTs. New trading paradigms like Sudoswap have contributed to solving the liquidity problem of NFTs, primarily by constructing instant liquidity pools that allow NFTs to be liquidated around the floor price without a lengthy waiting process. Currently, the market share of Sudoswap-like AMM NFT DEXs is still too small, and the heterogeneous characteristics of PFP NFTs mean that the mainstream solution remains order book-based.

What are the solutions?

In the face of the currently unsolvable NFT liquidity problem, NFTFi projects can actually choose other ways to address the issues of the liquidation mechanism and floor price manipulation, mainly by transforming collective risk into individual risk and establishing a separate NFT pricing mechanism.

Hybrid Mechanism

Collectivism describes the P2Pool model, where the common creditor of NFT lending positions is the ETH funding pool, and in the event of default, the asset returned will also be ETH. Therefore, the P2Pool model must undergo an NFT-ETH asset conversion process, which relies on good NFT liquidity to achieve. However, the P2P model differs in that the final liquidation essentially does not involve an auction process; in the event of default, it can be realized simply through the transfer of the NFT, thus bypassing the liquidity issue of NFTs. Yet, the P2P model has an inefficiency problem, so NFT lending projects can adopt a hybrid model, using P2P as the core while wrapping it in a P2Pool facade.

A possible manifestation is to first attract ETH lenders to establish NFT Lending Pools for different NFTs and set conditions for borrowing in that Lending Pool, such as interest rates, supported NFTs, and loan terms. Those who meet the conditions can borrow unconditionally. The primary responsible party for the NFT Lending Pool is the creator of that Lending Pool. This Lending Pool can accept third-party deposits and act as a direct creditor to the Lending Pool creator. Thus, the above mechanism constructs a hierarchical structure:

Third-party depositors ---> Lending Pool creator ---> NFT collateral borrowers

The Lending Pool creator acts as the direct creditor to the NFT collateral borrower. When a default occurs, the NFT collateral will be deposited into a multi-signature address, and then the third-party depositors, as creditors of the Lending Pool creator, will seek to recover their ETH deposits. At this point, the Lending Pool creator can choose to sell the NFT or retain it and repay the debt owed to the third-party depositors. If the Lending Pool creator fails to repay the debt within a certain timeframe, they effectively forfeit ownership of the NFT and are penalized with a portion of their funds, after which the protocol executes the liquidation of the NFT and returns the auction proceeds to the third-party depositors.

The above mechanism is one possible form of a hybrid mechanism, with the advantage of utilizing a hierarchical structure that grants the primary creditors the right to dispose of the NFT, allowing them to choose between auctioning or retaining it. More options and extendable auction times also help smooth the price curve of NFTs.

Moreover, since the above mechanism generally involves fixed-term borrowing, there are no issues of default and liquidation during the borrowing period; defaults and liquidations only occur after the term ends, thus avoiding the situation of clustered liquidations due to a declining market environment. This mechanism retains the option for liquidation, which can quickly integrate NFT liquidation pools once the liquidity issues of NFTs are resolved, making the hybrid mechanism more open and future-oriented compared to the closed nature of a purely P2P model.

Individual NFT Pricing

In NFTFi, the delineation of the liquidation line is generally directly related to the floor price of the NFT collection. In the current context of insufficient NFT liquidity, relying the core of the entire NFTFi stabilization mechanism on a pricing mechanism that is easily manipulated poses risks to the entire system. One possible solution is to price each NFT individually and set a unified LTV. The main methods include machine learning pricing and asset class substitution.

Machine Learning Pricing Method

The machine learning pricing method is one of the least sensible approaches; it estimates prices for NFTs with different characteristics based on empirical data combined with certain curves. Due to the scarcity of trading data for certain characteristics of NFTs, such price estimates are often very inaccurate. Another issue with this pricing model is its lack of scalability and ecological integration. Because of its inaccurate estimates for NFT pricing, this data, as a non-fair price, is difficult to be utilized by NFTFi applications, making it more of a price reference. However, the problem arises that machine learning models have weak interpretability, and without accurate data feeding, it is impossible to compare the effectiveness of models trained under such conditions, leaving one unable to determine which toy model is more effective.

Asset Class Substitution

The basic logic of asset class substitution is that a third party locks additional liquidity to form an asset package to back the value of a single NFT, and then uses that asset package as collateral for borrowing. This approach essentially creates a nested structure, where the NFT holder pays interest twice for their borrowing, converting the liquidated assets from NFTs to ERC-20 assets. The specific structure is as follows:

NFTHolder ---> ERC-20 asset package ---> DeFi collateral lending

The first interest payment occurs when obtaining the ERC-20 asset package from the C-end user, which requires a bribe. There are two points to note about the bribery process: first, this mechanism generally targets high-rarity NFTs; since mainstream NFT lending protocols price based on floor prices, high-rarity NFTs are often excluded from the NFTFi ecosystem. Through this pricing model, NFT holders can have their NFTs participate in NFTFi at a value close to their true worth after paying a certain fee, with the bribe being merely a cost in this process. Second, the C-end users who assess the price of the NFT "must bear more costs once they receive the money." Each price assessment has a fixed term, during which the ERC-20 assets paid by the C-end user and the NFT of the NFT holder are held in custody by the pricing protocol. NFT holders can utilize the ERC-20 asset package to participate in DeFi collateral lending and pay interest, with the amount that can be borrowed being LTV x $(ERC-20 asset package). Therefore, the main risk faced by C-end users when assessing the price of the NFT is that when the price assessment period expires, if the borrower believes the current value of the NFT is < LTV x $(ERC-20 asset package), they have the incentive not to repay the borrowed assets from DeFi, leading to the corresponding ERC-20 asset package being liquidated, and the C-end price assessor ultimately only receiving the NFT.

The essence of the above mechanism is a game process. First, the NFT holder, after incurring a certain cost, transfers the risk of the NFT to the counterparty through P2P and obtains the ERC-20 asset package, then uses that asset package to participate in DeFi lending, converting the final liquidation into ERC-20 liquidation rather than NFT liquidation, thus directly avoiding the liquidity issues of NFTs. This NFTFi-DeFi substitution model is theoretically feasible, and compared to the machine learning pricing method, it is more reasonable in terms of complexity and logic.

Current Common Trading Paradigms for Solving NFT Liquidity

Order books, aggregators, fragmented protocols, and AMMs constitute the core of NFT trading. Currently, NFT trading volume is primarily concentrated in order books and aggregators, while fragmented protocols and AMMs serve merely as supplements for instant liquidity in NFTs. The four paradigms have been iterating since their inception, reflecting their respective problems, development directions, and improvement trends.

Order Book NFT Marketplace

NFT marketplaces can be divided into corporate and community models. Corporate models, represented by Opensea, aim to provide NFT trading as a service to customers and earn transaction fees; while community-based NFT marketplaces aim to return profits to community users through reasonable token models, represented by X2Y2 and LooksRare.

Currently, the leading order book NFT marketplace remains the corporate Opensea. According to the rational agent hypothesis, trading on the X2Y2 platform incurs the lowest transaction fees and also allows users to earn some trading mining income, providing an incentive for NFT traders to switch to platforms like X2Y2. However, in reality, this conversion has not occurred.

The main reason may relate to the profile of NFT traders. Most NFT speculators do not come from token speculators; their wealth appreciation primarily comes from NFT premiums. Therefore, for them, trading is mainly a service, and compared to higher quality, faster, and smoother trading services and more and better NFTs, a small amount of trading rebate is negligible. Thus, for community-based NFT marketplaces, it is essential to fully understand the disconnect between NFT speculative users and token speculative users and to design economic models considering this disconnect. In fact, mainstream community-based NFT trading platforms that have already issued tokens have not carefully considered the separation of interests between these two groups in their tokenomics design.

The main issues include:

• Governance tokens do not provide trading conveniences or privileges for NFT traders;
• Since NFT traders have no incentive to purchase governance tokens, they do not retain the token rewards from trading mining, leading to significant selling pressure;
• Governance token holders can earn a share of platform fees through staking, but the potential for significant selling pressure reduces the motivation for token holders to stake;
• NFT traders and governance token holders essentially have conflicting interests, which may lead to irreconcilable conflicts in community governance;

In summary, the tokenomics design of NFT marketplaces that have already issued tokens has not bound the interests of the community, tokens, and NFT traders to any extent, and the core of solving the problem lies in how to encourage NFT traders to hold tokens. The demand for holding tokens should align with the essential needs of NFT traders, which include two main points: better access to whitelist NFTs and better trading of NFTs. Better access to NFTs requires a community-oriented approach, while better trading of NFTs needs to consider product, liquidity, and pricing aspects.

Better access to NFT whitelists relies on the NFT marketplace's strong community resources, making many NFT projects willing to offer benefits to the community members of that NFT marketplace; better products, liquidity, and pricing need to fully consider the trading needs of NFT traders, whether to charge royalty fees, and the setting of trading rates.

NFT Trading Aggregators

The differentiated competitive strategies of aggregators have been a well-discussed topic since the emergence of Gem and Genie. From the initial Car, to safe mode, then gas monitoring, and now more granular secondary K-lines, aggregators have iteratively developed products towards a more professional trading experience. At the same time, various data platforms like NFTnerds and Uniswap have also begun NFT trading, leading to fierce competition in the NFT aggregator space.

Here are some basic suggestions regarding NFT trading aggregators:

First, carefully segment the user groups of NFT traders to capture the underlying needs of different types of traders. NFT traders can be divided into professional traders and ordinary traders; Blur's simple UI and sufficiently detailed K-line charts are designed for professional NFT traders, as excessive information can negatively impact their experience; Element, on the other hand, primarily targets ordinary traders, with a page rich in NFT information, making it easy for ordinary traders to get started. This is a typical example of designing products based on user groups;
Second, enhancing the product experience is beta, but discovering potential collaborations between NFT trading aggregators and other areas of NFTs is alpha.

Fragmented Protocols and AMMs

NFTX and Sudoswap both represent instant liquidity pools, treating non-fungible assets homogenously, and democratizing the identity of market makers. However, fragmented protocols have undergone multiple narrative shifts since their inception, from share fragments to trading to collective ownership. Share fragments should be considered an obsolete narrative, while the current trading banner is primarily led by NFTX, and collective ownership is a new narrative proposed by Tessera.

Regarding trading, the mechanisms of NFTX and Sudoswap are essentially indistinguishable, with the formal differences being:

• NFTX includes a fragmentation and minting process during trading;
• In NFT pricing, NFTX adopts the ERC-20UNIV2 pricing model, while Sudoswap uses price curves (linear or exponential);

In addition to meeting the needs of NFT trading, fragmented protocols like NFTX can expand their narrative:

• Like Sudoswap, they can serve as liquidity pools for NFTFi liquidation;
• They can act as an intermediary layer for NFT DeFi;

The statement about being an intermediary layer for NFT DeFi may sound abstract. Currently, there are some NFT options, NFT futures, and NFT structured products, which essentially speculate on the price fluctuations of NFTs. Similarly, based on fragmented NFTs, ERC-20 standard derivatives can be more granular than NFTs, and a relatively complete and mature system already exists for various ERC-20 standard derivatives.

However, despite this, it may be too early to create derivatives for NFTs at this stage. The underlying liquidity of NFTs has not been properly resolved, and the derivatives field, which requires higher liquidity than lending protocols, will face even more severe price manipulation and liquidity issues.

So where does the value of Sudoswap lie? The foundation of Sudo comes from the already existing NFTX, driven by retail investors' speculative demand for no royalties, fast trading, and proximity to floor prices. Therefore, as long as NFT traders' speculative demand for NFTs near floor prices does not disappear, Sudoswap will continue to exist as a reasonable trading model. Additionally, due to the homogenized trading nature of Sudoswap, NFTs like Pass Card NFTs and GameFi NFTs, which require integrity but have semi-homogeneous characteristics, are naturally suited for Sudoswap's trading model.

What might be new trading paradigms for solving NFT liquidity?

Existing trading paradigms are battling on the original continent and are gradually forming a stalemate. Breaking the current deadlock requires, on one hand, opening up new territories by introducing more categories of NFT assets, and on the other hand, introducing new participants to connect different stakeholders. Just as the increase in NFT lending products will spur the emergence of an NFT aggregated lending market, AMM DEXs like Sudoswap and royalty fees will foster the emergence of professional NFT market makers.

As for why professional market makers will emerge after Sudoswap, the main reason is that the Sudoswap protocol itself does not charge LP market-making fees and profits, while every transaction fee in order book-style marketplaces must be paid as a service fee. Additionally, the existence of royalty fees forces market makers based on Opensea to bear excessively high market-making costs, while Sudoswap and subsequent NFT marketplaces with royalty fees drastically reduce market-making costs.

Given the sharp decline in market-making profits based on ERC-20 tokens, NFTs themselves exhibit high volatility, leading to higher slippage per transaction. Under reasonable market-making conditions, the market-making profits for NFT projects may exceed those of ERC-20s, creating an incentive for original on-chain ERC-20 market makers to shift towards NFT market making based on arbitrage principles.

The introduction of NFT market makers will bring new changes to the entire NFT field. Before the emergence of NFT market makers, the value flow of on-chain NFTFi could be represented as:

Retail ETH and NFTs ---> NFTFi ---> NFT liquidation ---> NFT Marketplace

The emergence of market makers will add a link to the above path:

Retail ETH and NFTs ---> NFTFi ---> NFT liquidation ---> NFT market makers ---> NFT Marketplace

Market makers, as liquidity providers, play a crucial role in revitalizing trading across the entire NFT marketplace. To ensure that delta-neutral market makers have the incentive to collect NFT buy orders across various NFT marketplaces before taking on excess liquidity, sufficient NFT buy orders also imply the potential for NFT swaps. The core of NFT swaps lies in acquiring as many potential NFT buy orders from the market as possible and facilitating exchanges between different buy orders by providing an ETH liquidity pool. This mechanism can serve as a supplement to the current order book trading model.

The Relationship Between Game Guilds and NFT Leasing Protocols in NFTFi

Game guilds often incorporate an NFT leasing system to facilitate gold farming for their members. Their common solution involves constructing an on-chain and off-chain account system, multi-signature, or MPC co-management. Game guilds are highly localized and can attract a large number of local GameFi gold farming players. Guilds can leverage this to acquire numerous GameFi NFT gold farming assets at lower prices. Returning to the motivations for collaboration between GameFi project parties and game guilds, are GameFi project parties coveting the regional characteristics of game guilds, or the gold farming players represented by the guilds? If they are coveting the gold farming players of game guilds, how does an NFT leasing protocol with the same user base compare to game guilds?

Both NFT leasing protocols and game guilds essentially provide users with leasing functions, so for a rental system, the user base is the primary indicator determining the platform's ceiling. Certainly, game guilds can generate profits through NFT premiums, but this model is also feasible for NFT leasing protocols.

For game guilds, being highly localized is both an advantage and a disadvantage for future expansion, as the number of GameFi gold farming players in a region is limited. However, for NFT leasing protocols, their target users are all on-chain GameFi gold farming players; moreover, the GameFi game assets that guilds can provide for gold farming services need to be purchased, and they cannot buy all on-chain GameFi NFT assets for user use. In contrast, NFT leasing protocols can theoretically incentivize all NFT holders to provide NFT leasing liquidity through liquidity incentives and other means. However, game guilds and NFT leasing protocols are not entirely in competition. Currently, game guilds remain strong, and NFT leasing protocols often collaborate with them by becoming the built-in NFT leasing systems of game guilds.

Is NFTFi a Pseudo-Narrative?

Is NFTFi a real demand? If so, how can doing something based on a real demand be considered a pseudo-narrative? Currently, the market size of NFTFi is still very small, constrained not only by the liquidity of NFTs but also by the overly singular categories of NFT assets, primarily PFPs. PFP NFTs have no utility beyond their social attributes, so how can the logic of Fi achieve a closed loop?

Whether in DeFi or GameFi, the basic requirement for Fi is that its underlying assets must have certain income-generating capabilities: DeFi's underlying assets can earn interest through token swaps, and GameFi's underlying assets can realize value feedback through leasing. Therefore, I want to set aside all the factors currently limiting the development of NFTFi; the logical gap in PFP NFTFi may lie here.