Gringotts DAO: A Sneak Peek Inside the Vaults

It’s been a while since we formally announced Gringotts DAO in early December. In just 3 months, we have fostered a culture of execution and experimentation. With all the distractions of crypto, it’s nice to feel part of a group that is quietly paying attention and getting work done.

What’s Gringotts? Gringotts DAO is a community-led decentralized organization that is focused on alternative finance opportunities. As we continue to evolve, our aim is to still be the one-stop-shop for NFT holders to access liquidity.

This article summarizes what we’ve accomplished since our last update, covering:

• Contributor curation and onboarding
• Market research into the current NFT-backed lending landscape
• Strategy formulation (automatic, manual, DAO 2 DAO)
• Community activation & proprietary tools
• What lies ahead

In aggregate we have:

• Onboarded +50 members including 25 capital contributors
• Raised 200 ETH, with no person controlling >13% of funds
• Deployed 113 ETH in loans at a total APR of 60% with no defaults, through a combination of automated and manual strategies
• Built an analytics dashboard, that is integrated with discord bots
• Established a DAO 2 DAO lending strategy and made proposals

Onboarding Contributors

In early December we decided to open our doors to those wanting to join Gringotts. We were fortunate enough to screen over 200 high-quality applicants. Our focus then, and today, is centered around right-sizing the community. That said, onboarding people with a diverse set of experiences have been essential in defining the Gringotts strategy. These Goblins have provided unique perspectives around NFT valuation, borrower credit evaluation, automated and manual lending strategies which have collectively strengthened Gringott’s capabilities in underwriting NFT loans.

We have onboarded members that have contributed to, or are from:

• Fingerprints DAO, POAP, 1kx, Mirror, Ready Player DAO, Rarible, NFTfi, Squiggles DAO, NFTBank, and more…

Establishing a Lay of the Land

After iterative discussions with the community, we knew we wanted to be the one-stop-shop for NFT collectors to access liquidity. With the help of many data-driven community members, we started to collect copious amounts of data. We were able to collect trading platform data, social data, and on-chain data, which have helped us define our lending strategies while allowing us to remain nimble during times of market volatility.

NFTfi was the first P2Peer lending platform we started collecting data on. High-level takeaways can be seen within the dune analytics dashboard created by one of our members, and include:

• Transaction volume: NFTfi transaction volume is >$70 million (from >4k loans), with +40% of the volume was accruing in 1Q22
• Asset denomination: ~80% of loans are in wETH / ~20% are in DAI
• Lenders: The top 5 lenders account for ~50% of the volume
• Borrowers: The top 5 borrowers account for ~25% of the volume

Some of the more interesting findings are centered around NFT market caps, defaults, and APRs. Of note, some of this analysis leverages multiple data sources.

• Asset market caps: We were able to identify patterns within the market caps of different NFT projects. We found that for NFT projects following a power-law-like distribution, 30–50% of the market cap is made up by NFTs close to the floor price. Mid-tier NFTs (defined as 1.4 to 2.5 times the floor price) make up 10%–20% of the market cap, while the remaining 20%–40% of the market cap is made up by the 5%–10% of top tier NFTs.
• APRs: Average APRs range from 42% to 431% over 7–90 day loan durations. We found that if you had 1 ETH to lend you would get more bang for your buck if you chose 7-day or 90-day loan durations (earning 9–10% return on initial investment), while 14-day & 30-day loan durations range between 4.0% to 5.5%. We note that when optimized for capital efficiency, deploying shorter-duration loans becomes more attractive.

Below we highlight the relative return on the initial investment (not including repayments and defaults).

• Defaults: We analyzed default data both at the market level and within our NFT whitelist and found default rates tend to jump once LTV breaches ~65%. We also learned that data curation at the top of the funnel allowed us to drastically outperform the market (minimizing liquidations across our lend-to-earn lending strategy).

Gringotts curated whitelist outperforms market liquidations, with assets defaulting less than 1% of the time when LTVs are below 60%.

LTVs between 40% and 60% default 5%-10% of the time, without a curated NFT whitelist.

Formulating a Strategy

After we established the lay of the land, it was time to hone our lending strategies. Lending against NFTs comes with a variety of challenges, including the need to understand and define market supply/demand, collateral value, capital-efficient LTVs and APRs, asset liquidations, and borrower evaluations to name a few. With the help of our contributors, we were able to collect, manipulate and visualize transaction, wallet, and social data to cut through the noise and start implementing a capital-efficient NFT collateralized lending strategy.

Our rigorous valuation process has allowed us to deploy >113 ETH across 35 transactions (~3 ETH average transaction size), while achieving a 60% APR with no defaults. The bulk of capital deployed was through our automated lending strategy.

Here are a few things we observed:

• Our manual lending strategy came with significant operational overhead that we weren’t quite ready for, therefore we doubled down on our automated strategy for Season 0, feeding our aggregate valuation capabilities into the loan bot.
• There is the perception that NFT valuation is non-uniform across projects. While this may hold true for individual pieces, we at Gringotts found that is not entirely true. Top-tier NFTs of a project can sequester up to 40% of that project’s market cap, with approximately 50% of all lower-tier NFTs staying close to the market floor price.
• Implementation of trade data was crucial in defining risk. For instance, we looked at bid/ask spreads, active buyer/seller counts, avg. transaction frequency, transaction volumes, etc…
• We can effectively evaluate NFTs valuation on any piece. This has allowed us to automate our parameters to base our discounted valuation through 30-day moving averages and the volatility during that period. Sometimes this would result in a 20%-40% valuation discount.

In all, we aim to improve the stability of the NFT ecosystem by valuing NFTs based on historical data and not current price, reducing the risk of negative equity and allowing us to continue providing a service indefinitely rather than a blip in the time continuum.

Activating the Community

We augmented these quantitative data points with expert knowledge sourced from our contributors! Contextual knowledge has allowed us to make key strategy alterations in real-time. Working as a community is and will continue to be our edge! Our ultimate goal is to include all of our internal sector knowledge — whether in specific projects, categories (generative art, pfps) — into our lending models to obtain maximum capital efficiency.

We onboarded plenty of non-technical members to Gringotts, so making the data we collect easy to understand and act upon became crucial. We worked hard to translate and display our data. While this is of no importance to our automated strategies, giving members the power to act autonomously is a differentiator.

For instance, members can easily look at our portfolio (real-time) and download data within our proprietary dashboard, which includes a variety of portfolio and asset-level data.

Members can look at collection stats, including average floor, price standard deviation, total volume, and # of trades. We have also pulled asset-level stats, including rarity distribution, charting functionalities, and the ability to download CSV data.

The team also built some cool discord bots which allow members to input simple (well communicated) commands to look at assets, borrowers, and our portfolio.

Discord valuation bot

Borrower evaluation bot

Member autonomy will continue to remain one of our core tenants as we navigate the future of NFT financialization.

Looking to the Future

Over the past few months, we experimented with various lending strategies and were able to find a product-market fit. We cultivated a tight-knit community, with Season 0 members displaying a bias for action and fast experimentation. It’s now time to get our vaults in order.

Thus…

We are formally announcing the close of Season 0 and are heading into our first offseason.

We have a lot planned during the offseason, with a core focus on formalizing Season 1 objectives and structure. A few things we aim to accomplish over the coming weeks, include:

• Identify new lending strategies and products to bring to market with a roadmap
• Establish a full-time and part-time paid contributor team — to be compensated through an operational fundraise
• Create clear and concise onboarding documentation so new DAO members can hit the ground running
• Design a roadmap to develop further DAO autonomy (path to tokenization)
• 5 pages of other stuff… :)

For those interested in joining the DAO, keep your eyes glued to our Twitter. We will be opening up the doors to new members (still a permissioned application) soon!

Cheers,

Fellow Goblin