- Competition is ramping up in the NFT lending and fractionalization space with a high degree of similar product features and roadmaps being offered
- Market is still in its nascency with untapped potential for new products and features to unlock greater value in NFTs
- Lending products are hard to scale currently due to a lack of reliable pricing mechanisms
- As the “Airbnb of the metaverse”, NFT rental marketplace has big potential and is a vertical that merits a close watch
- NFT pricing oracles are key to the success of most financialization primitives
Aggregators can capture value more effectively with the use of gas saves and price fetching functions
Thinking about the Financial Layer of NFTs
In 2021, we witnessed tremendous growth in volume, reach and new users of NFTs and their applications.
The proliferation of NFT technology in the industries of Art and Collectibles, Games, Entertainment and Fashion, as well as use cases for community engagement, content monetization, social interaction have created deep and meaningful implications for all stakeholders of the NFT ecosystem — users, creators, and founders alike.
As new brands are being built, game experiences are being developed and dozens of metaverses are set to roll-out, marking a very exciting time for the community to participate and be immersed in the space.
That being said, with the rapid development and ambitious roadmaps published by NFT brands, games and metaverses, we are also seeing bottlenecks emerging in the current NFT technology stack. As builders create ever more novel applications for NFTs, we have yet to see corresponding development in the financial layer of NFTs to support such accelerated growth.
Critical issues to be addressed include:
- The difficulty of accessing liquidity for blue-chip NFT holders (i.e., Can I generate interest or take out loans against my CryptoPunks and Bored Apes?)
- Problems with maintaining community growth without raising barriers to participation
- Scaling issues with gaming guilds as they continue to grow their scholar base, as well as incubating new DAOs
- Limitations to yield generation on NFT assets owned by Metaverses landowners and developers
This report by Matrixport presents the latest development in NFT Financialization to facilitate a better understanding of the landscape, as well as the key developments within. We will also identify protocols and solutions that could potentially plug the gaps in the sector and fuel growth.
Three key areas will be discussed:
- The current NFT financialization landscape
- Pros and cons of the existing NFT protocols – Fractionalization, Lending, Rental, Pricing and Aggregators
- Key areas for product innovation
Matrixport is Asia’s largest crypto financial services platform with ~$10 billion worth of crypto under custody and management. We offer our users one-stop crypto financial services across lending, trading, asset management and structured products.
If you are building in this space, we would love to talk to you and see how we can contribute.
The NFT Financialization Landscape
Looking at the current landscape, NFT financialization is distinctively divided into the five verticals of fractionalization, rental, pricing, lending and aggregation. In order for liquidity and yields to be generated from NFTs, these key primitives have to be built for the financial layer to function efficiently and scale. Taking a closer look at the critical areas for development for each segment:
- Indexing: Building a floor price index by fractionalizing similarly priced NFTs from the same collection into fungible tokens
- Collection: Creating an “index fund” like product that adds different NFTs into the fund (see Bridgesplit’s roadmap on Curated Index)
- Peer-to-Peer (P2P) Lending / Crowd-Lending: For borrowers and sellers to negotiate on loan terms and durations
- Liquidity Pool-based Lending: Building a lending model that generally allows NFT holders to deposit their NFTs into the liquidity pool and take out permissionless loans, and lenders will be able to earn liquidity provider fees by providing liquidity
- DAO-to-DAO lending: For gaming guilds to lend or lease their assets to emerging guilds
- Collateral-backed Rental: Renters will deposit assets as collaterals, which will be managed by the protocol
- Wrapped NFT + Custody by Marketplace: Renters pay rent and receive a wrapped NFT, while the lenders lock their NFT into the protocol and receive rent
- Expirable Dual-Role NFT: Renters will be assigned the role of user and the NFT will expire automatically when the rental period ends
- Machine Learning-based Pricing: Using statistical analysis and machine learning to derive a value range for any NFTs based on historical trading data (e.g. Upshot in 2022)
- Time-weighted Average Price (TWAP) from Chainlink or Sushiswap: JPEG’d has built their customized floor TWAP with Chainlink
- Peer-to-Peer Pricing: Leveraging a curator committee incentivized by rewards by the final pricing session output (Abacus, Taker Protocol and Upshot in early 2021)
- Pricing using Financial Instruments: Using various financial derivatives and auction mechanisms to enable real-time pricing. Paradigm’s Dave White wrote a piece on floor price perpetuals, and Abacus Spot is developing their liquidity pool-based pricing tool
- Genie.xyz and Gem.xyz: Enabling rapid fetching of NFT price data across marketplaces and bundles multiple NFT buy orders into one transaction
There are two critical aspects that shape the framework for effective NFT fractionalisation:
- infrastructural capabilities to generate deep liquidity for NFT holders and create NFT exposure to ERC-20 holders;
- novel price discovery mechanisms that create new consumer experiences.
Key players in the NFT fractionalization space include NFTX, Fractional.art, NFT20 and unic.ly and this report will focus on NFTX (blue-chip, floor-price fractionalization) and fractional.art (rare NFTs and any NFT fractionalization).
NFTX have pioneered the fractionalization of NFTs by launching NFT vaults such as Punks, Hashmasks and Squiggles. With 133 punks and over $20 million Total Value Locked (TVL) in the CryptoPunk vault, they have created deep liquidity for blue-chip NFTs and enabled users to gain exposure to Punk’s price through the PUNK-WETH token (refer to Figure 1).
Figure 1: NFTX Token Pool Rewards
Their vault method also enables buying, selling and swapping of Punks by Punk holders, and for ERC-20 holders to participate in the market making process and earn both trading fees on Sushiswap and vault tokens. The team also launched their Olympus DAO-style market-making protocol FloorDAO.
On the other hand, fractional.art focuses on the fractionalization of any NFTs. One of the most famous examples being the tokenization of the original Doge NFT by pleasrDAO and the subsequent auction of $DOG token on Sushiswap. This resulted in a whopping $225 million implied valuation of the original doge NFT and showcased the amazing power combo of NFT and tokenized communities (refer to Figure 2).
Figure 2: The Doge NFT Profile
Since then, fractional.art have seen an influx of influencers and collectors / curators fractionalizing their own NFT collections including Bored Apes and ArtBlocks as well as rare 1-out-1 NFT arts on Foundation. Such creations helped to further onboard new users to crypto and NFTs as well as providing 1-out-of 1 NFT asset exposures to their communities.
With the rise of metaverses and blockchain games, we believe there are opportunities for innovation which will enable an enhanced user experience. These are namely:
- Fractionalizing the NFT utility over potential airdrops and access to key in-game features and economic output
- Building a financial derivatives layer on top of the floor price indexes to enable granularity in pricing
- Introducing debt instruments that can be raised against the acquisition of blue-chip NFT to further improve on some of the buyout mechanisms of rare NFTs
We can break down the NFT lending protocols into 2 segments. The first, is a P2P lending model where borrowers and lenders negotiate loan terms such as duration, interest rates and loan-to-value ratios. The other segment focuses on the creation of liquidity pools around floor prices and enables permissionless loans to be taken out against NFTs. There are also players such as MetaStreet that act as liquidity providers for NFT lending protocols to enable structured loan products.
For NFT lending, there are several criteria in evaluating such protocols. These include:
- Risk Management: Presence of a robust risk management system in place in case of large-scale liquidation, sudden NFT floor price collapses and bad actors
- Protocol Design: Accessibility to sustainable liquidity and yield generation, whether it’s P2P or liquidity pool-based protocols
- NFT Pricing: Implementation of a decent NFT pricing mechanism, and the effectiveness of loan term generation
- Liquidation Process: Creation of incentives for liquidated NFTs to be acquired
- Token Economics: Proper design of token economics to encourage both lenders and borrowers to participate in lending and liquidation
As case studies, we will explore NFTfi and Pawnfi’s P2P lending product designs and Bridgesplit’s liquidity-pool mechanisms.
NFTfi is currently the category leader in the P2P NFT lending space. The protocol has done more than $80 million in cumulative loan volumes with 5,000+ loans underwritten since its inception in 2020.
NFTfi focuses on the lending and borrowing of blue-chip NFTs such as Punk, Apes, ArtBlocks and CloneX. Borrowers and lenders will fill in expected loan value, duration and interest rates, and loans can be underwritten once the interests are matched. Typical loan durations are 7,14, 30, 90 days and the annualized yields are ranging from 43% to 426% based on Dune Analytics dashboard (refer to Figure 3).
However, calculating the actual borrowing cost must take loan durations into consideration as well, so the true loan cost will be around 4.1% to 10.6%.
Figure 3: Average APR on NFTfi by Gideontay
In the event of a liquidation, NFTfi’s current protocol design is to implement a non-recourse loan where the borrowers’ own assets will not be affected. In other words, the lenders will waive their claims on the outstanding loan amount.It then begs the question on the liquidity of such NFTs given that the typical cause of default would be that the floor price of the NFT becomes lower than the loan value, and it is no longer in the borrower’s interest to pay back the loan amount.
NFTfi’s roadmap in 2022Q1 includes issuing loans against ERC1155 NFTs (i.e. 1/150 NFTs such as Shibuya.xyz’s producer passes) and ERC998 NFT bundles (different NFTs bundled into one); early repayments, unlimited duration loans, loan extensions, NFTs as loan receipts; and offering granularity in loan term listings that can be automatically accepted by lenders.
The P2P model has strengths in risk mitigation since most of the transactions are negotiated between parties and defaults only happen locally and in separate cases. However, this model also has given rise to the question of liquidity and scalability as lenders and borrowers will have to wait for agreement on loan terms before they can access the NFT or loan.
To fill the gap, NFTfi is partnering with MetaStreet as a liquidity provider to enable structured credits and facilitate large borrowing demands. To date, they have executed the largest NFT loan so far – a 10% APR, 30-day duration $8 million loan collateralized by 101 CryptoPunks.
MetaStreet is also working with Arcade.xyz and other P2P NFT loan platforms to improve liquidity and facilitate transactions and are in the process of developing an automated NFT pricing and underwriting system (i.e. max loan-to-value ratio suggestions). In the future, participants of MetaStreet can expect to put their stablecoins and ETH into the MetaStreet DAO and earn interests on loans underwritten by the MetaStreet platform.
The other approach in this segment is crowd-lending developed by Pawnfi. Similar to syndication loans in traditional finance, it supports a single user to borrow against multiple NFTs from multiple lenders, with an initial focus on blue-chip NFTs and whitelisted assets.
Borrowers can set loan terms based on their own liquidity needs. If borrowers need quicker access to liquidity, they can choose lower loan-to-value ratios with higher APYs, or vice versa. Lenders can decide whether to participate and how much they want to contribute to the syndicated loans. If a lender is willing to take 100% allocation, it becomes the case of P2P lending.
Pawnfi also provides instalment loan options for borrowers. This option splits repayments into different terms and increases the durations of loans and possibility of lenders’ acceptance of higher loan value. This is because using an instalment loan usually helps borrowers better manage their cash positions, so the theoretical default risk is lower.
In the case of liquidation, a duction auction will be initiated and crowd loan lenders have the first priority to bid for the liquidated assets (where auction stands). If lenders are not interested in acquiring the asset, the NFT will flow to the market and other interested liquidators will be able to make purchases. This auction process allows the lenders to decide whether they want to deal with the collateral themselves or have it liquidated by market participants.
The crowd loan approach has also enabled potential product innovation such as credit default swaps and structured debt with multiple creditors.
The overall NFT P2P lending model has 2 potential areas for improvement:
- Potential loopholes for bad actors to exploit the P2P lending mechanism;
- Liquidity provision for both loan underwriting and the liquidation auction process.
In this case, risk management systems (i.e., indicative loan-to-value ratio and NFT floor price suggestions), on-chain activities proof and curator DAOs (pricing 1-out-1 NFTs or rare NFTs) would be key for the platform to scale and control default rates. Possible dark-pool order book building mechanisms similar to options market-making in the traditional finance industry can also be useful to facilitate better liquidity in the auction process.
Moving on to liquidity-pool based NFT lending protocols, we find Bridgesplit to be particularly interesting to look at.
Bridgesplit is working on NFT liquidity and lending solutions on the Solana blockchain where there is a growing number of Gamefi and NFT projects seeking new ways to unlock more utility and yield-earning potential.
A glance at their process, Bridgesplit starts off by building fractionalized NFT liquidity pools based on Serum and Raydium (individual NFT pools are called “vaults”), and then moves on to collaborate with projects and collector DAOs to build floor price indexes and what they call “Curated Indexes”.
After building tractions with their indexes, the plan is to launch NFT-collateralized loans and asset rental based on the liquidity generated in various pools of assets. Users currently can earn Liquidity Provider tokens and share of trading fees by depositing their NFTs into the floor index such as the Boryoku Dragonz series. NFT holders can also gain access to liquidity by trading their $DRGNZ token to $SOL after they have deposited their NFT.
Bridgesplit is also working on their “automated indexes,” which is integral to their NFT lending solutions. The automated indexes will be based on a NFT pricing oracle that leverages on Bridgesplit’s engineering capability and outside solutions such as Abacus and Upshot. Users can expect to deposit and redeem their NFTs automatically based on the pricing provided by the NFT oracle (when they deposit the NFT) and the liquidity condition of the indexes after depositing (when they take out the NFT).
The collateral value of the NFT will then be determined based on the real-time share prices of the fractionalized NFT pool (vaults), instead of the pricing provided by the NFT oracle. Automated indexes will also execute liquidation once the traded value of the NFT falls below the collateral loan value and exchange the NFT vault tokens back to SOL, ETH or USDC tokens to keep enough liquidity in the index. However, with the pricing of NFTs being executed by floor price oracles, this information will only be more useful when it comes to blue-chip NFT collections such as CryptoPunks, Bored Apes and Azukis.
Other liquidity pool-based protocols such as Pine Loans and JPEG’d are in the process of creating similar automated lending processes where lenders deposit money into the protocol and set terms; and borrowers can take out permissionless loans if they find such terms attractive.
While Bridgesplit’s approach towards providing NFT liquidity and lending solutions provides a good example of innovation, the sustainability of liquidity on the liquidity pool-based lending protocols, as well as their ability to counter potential NFT oracle attacks remain to be tested by the market.
We are keeping a close eye on liquidity pool-based lending protocols and continue to think about NFT pricing, sustainability of liquidity and risk management of such models. We think that such models give flexibility to lenders and borrowers and have better scalability on blue-chip NFTs with large volume and more active price feeds.
The NFT rental market remains top of mind because we believe the large-scale adoption of blockchain games, guilds and metaverses is imminent. Rapid development of such virtual experiences will result in large amounts of idle NFTs held by users, guilds, and projects. Therefore, it is crucial to have an active rental market available to bring down cost to participate and increase existing users’ engagement through rental income.
Our thesis in this vertical is that a collateral-free rental model with a product design to separate ownership and utility of an NFT would prevail in this market.
However, since there’s no universal minting standards built around NFTs regarding the separation of ownership and utility, NFT rental protocols in this space would have to work closely with projects to build reputation and trust around their services. Projects such as Double Protocol, reNFT, IQ Protocol and Pine Loans are all trying to develop solutions to this problem using various designs such as:
- Building on ERC-721 token standard’s extension dual-role setting that distinguishes owner and user, which is used by Decentraland ( Double Protocol’s expirable NFT EIP-4907 proposal and Split-and-merge design that enables multiple use cases in different games; Landworks / EnterDAO’s EIP-4400 proposal)
- Assigning utility to renter while holding NFTs in escrow (see reNFT’s design and IQ Protocol)
- Separating utility and ownership through a designated wallet. The renter can use NFTs while the ability to transfer NFTs is disabled (see Pine Loan’s Wallet)
The preceding approaches all have their own benefits and limits, but for NFT rental protocol to become successful, the go-to-market strategy towards guilds, games and NFT projects remains essential. If protocols can collaborate with high potential games to drive adoption with a massive group of users, then it is possible to create Ethereum Improvement Proposals (EIPs) to solidify the approach towards the NFT rental and mortgage market.
We can already see early signs of such approaches in NFT profile picture projects (PFP) such as Azuki, which utilizes the ERC721A improvement to reduce mining costs. This was quickly followed and adopted by many upcoming PFP projects.
On top of the separation of ownership and utility of NFTs, we can expect new uses cases such reserving or subletting NFTs (i.e. the Airbnb model in NFTs), instalment loans, and buy-now-pay-later options to be developed, which will unlock a new array of applications for games, guilds, metaverses and NFT projects for users to utilize.
NFTs are hard to price for a few reasons. Its non-fungible nature, lack of liquidity, volatile price movement and features such as rarities and traits (NFT-specific attributes) make it hard to generate continuous price information.
Projects in this space have developed roughly 4 ways approaches to value NFTs. The first approach is to acquire on-chain Time Weighted Average Prices (TWAP) on NFT floor prices through Chainlink or Sushiswap. This is commonly used by NFT lending projects that are attempting to build permissionless loan functions.
Using TWAP for NFT pricing has a few potential disadvantages. TWAP is only applicable to NFT collections with an active market and large transaction volumes, of which are prone to oracle attacks and potential manipulation. Additionally, applying TWAP in NFT lending primitives can limit the level of loan-to-value ratios and amount of loan borrowers can take out against their NFTs, resulting in lower capital efficiency.
The second approach is to use statistical methods and machine learning techniques to extrapolate the NFT’s possible valuation range based on past sales data and corresponding NFT traits. Upshot, after pivoting from a peer-appraisal model in November 2021, has been actively exploring and experimenting with the approach, as well as building NFT pricing APIs. Using Shapley Value, a method which measures specific features’ impact on historical data, it is possible to conduct analysis on each individual NFT’s traits and the relationship with its historical sales price (refer to Figure 4).
Through the application of machine learning methods, Upshot has built prediction variables to conduct principal component analysis (filtering out the traits that have the most impact on pricing) and refined their prediction model on CryptoPunks, Bored Apes and Cool Cats.
The possibility to arrive at a valuation range on NFTs can be very useful for both NFT lending and rental primitives. Through the suggested valuation range, market participants can strike loan and rental terms more efficiently. However, because of the adoption of machine learning methods which feed historical data into the model, this pricing can have potential lags and might be subject to potential attacks.
The third approach is a peer appraisal model, where NFT collectors and collector DAOs can participate in either direct appraisal for NFT values, or enter a prediction-market style valuation session where the participants can be rewarded and punished by the session results. It is obvious that aside from a few selected super-rare NFTs that would work well with this method, pricing NFT using the appraisal model would produce less efficient and accurate outcomes.
Lastly, there are financial derivatives-driven primitives such as Abacus Spot. With Abacus Spot, participants can open pools, deposit ETH to speculate on NFT prices and track down NFT owners to activate emission rewards in the form of Abacus’ token. Holders of NFT can demand bribes from liquidity pool traders for an earlier spot in the pool, which in turn will give the traders an advantage to profit in the auction process.
This would enable real-time NFT pricing and provide a pricing and liquidation mechanism for NFT lending protocols (refer to Figures 5 and 6). Lending protocols can simply issue loans using the ETH locked in the pool as collateral, and enable the auction process if the holder defaults on the loan (refer to Figure 7).
Expanding further, Gradient Protocol is working with Abacus Spot to enable permissionless NFT loans and yield-generation on NFTs. Their product is currently in closed beta test and will be released on testnet soon.
We also believe that the NFT aggregator will be an important piece in the NFT financialization vertical. The valuation propositions to users are clear: lower gas fees by combining multiple buys into one transaction; and better prices and liquidity through the aggregation of multiple marketplaces.
Gem.xyz and Genie.xyz have been leading the race in transaction volumes for NFT aggregators. Genie initially had more users and transaction volume, but Gem has surpassed Genie since February (refer to Figure 8).
Figure 8: Gem.xyz and Genie.xyz Trading Volumes on Dune Analytics
WeMeta has also built a specialized metaverse land NFT aggregator using Gem.xyz’s API with Gem.xyz to attract users that are interested in sweeping the floor and bulk-buy metaverse lands. Other upcoming NFT aggregators include Flip.xyz, co-founded by the UpOnly host Brian Krogsgard (Ledger Status) and Hyperspace on Solana blockchain (previously Solanlaysis).
It is highly likely that NFT aggregators will be able to capture value as more specialized NFT marketplaces and “Opensea challengers” (Looksrare, Universe.xyz) gain more traction this year. Games and metaverses NFT aggregators can also be promising applications as items become interoperable and increasingly financialized.
Potential Areas of Product Innovation
With the accelerated developments of each vertical, coupled with the evolving needs of NFT holders, the potential for product innovation is immense.
For P2P NFT lending protocols, a dark-pool order book auction mechanism that enables periodic display of bids on the liquidated NFT can improve liquidity and facilitate liquidation. It would also prove interesting to introduce Dutch auction mechanisms to find the best supply-demand matches at different interest rates and Loan-to-Value ratios.
In order to solve the NFT liquidity issues, protocols can also introduce Keeper Bots to fetch liquidity from various marketplaces. For liquidity pool-based lending protocols, the combination of multiple NFT pricing approaches to enhance risk management and increase Loan-to-Value ratios would be key for the protocol to grow TVL sustainably.
On the premise of a secure, adaptable NFT rental protocol, buy-now-pay-later (BNPL) or Instalment loans can drastically help to reduce the entry barrier for NFT brands, games and metaverses.
NFT aggregators can also build arbitrage bots that increase liquidity across various P2P NFT loan protocols by arbitraging NFTs, and enhance value to its token holders similar to 1inch Earn.
Across the protocols Matrixport has researched and tested, we are seeing intense competition in the lending and fractionalization verticals where protocols are offering similar product features and roadmaps. Adding to that, lending products are finding it hard to scale since there are no proven, reliable NFT pricing protocols to work with. We will continue to watch closely for high quality projects within these two spaces with strong product and execution capabilities.
We believe that NFT pricing oracles would be the key to unlock the growth of NFT financialization primitives and plug the critical gaps in scalability and accessibility. At the same time, we should be paying close attention to two emerging verticals: the “Airbnb of the metaverse” – NFT rental marketplaces – and NFT aggregators. The former holding promise given the high adoption rates across games and mateverses at present, and the latter stepping up to offer effective value capture through gas saves and liquidity aggregation across marketplaces.
As the financialization of NFTs evolves, the entry of institutional participants such as Genesis Trading and Nexo can potentially catalyze this ecosystem to the next level with their expertise in the fungible token lending space.
If you are building in this space, we would love to connect with you and see how we can collectively contribute to the growth of this space.
Reach us on:
- Twitter: @looksrare_eth
- Email: email@example.com; firstname.lastname@example.org
- Website: matrixport.ventures
Download the full report here