A simple heuristic for NFT lending
Rule of thumb for lending to NFTs
How many loans is it safe to issue for one collection ?
Is there a magical formula for knowing when debt becomes unsafe to issue ?
As it turns out, yes there is !
With the recent cascading liquidations of certain NFT collections, lenders are paying closer attention to the fineprint of their lending agreements.
Quickly understanding the risks associated to your NFT loans has become essential for profitable lenders to understand.
While it mostly serves P2P lenders (who are now discovering the greater risks involved) it also benefits peer-to-pool lenders, as they pick and choose which collections are worth their liquidity.
How it works ?
There has been a lot of great advice already…
Solami is right. Avoiding collections with lots of underwater loans is a great start. Or atleast, you should expect a much higher interest rate on those collections.
But this is a bit of a lagging indicator.
By the time NFTs are underwater, it’s probably too late, as the downwards cycle of liquidations is either immeninet or has already begun.
It can also be hard to find without proper data analytics tools like Dune Analytics or Lender Labs. So, not the simple heuristic we’re looking for.
Additionally, I recommend paying close attention to the outstanding debt of a collection1, but not as a function of the collection's size.
People will often warn that “this collection has too much leverage” and point towards 2%-10% of the collection being collateralised. In isolation, this figure doesn’t mean anything. What matters is volume.
Each collection has a certain amount of outstanding debt, which is extremely easy to find.
On Frakt, this is called “Borrowed”, on Sharky this is called “Active loans” and on Honey, you substract the “Supplied” liquidity with “Available” liquidity to find the outstanding debt.
Once you find the outstanding debt for a given collateral, it should be weighed in proportion to it’s trading volume. The reason is simple, if that collateral is liquidated, it will usually be sold on the market. Knowing the average volume in that market allows us to know how much of our debt it can be processed and sold.
When we sell more NFTs in the market than it can handle, we’re over supplying the market, and without an equivalent rise in demand, we get a “market surplus”. All a market suplus really means is that prices need to go down for sales to happen.
You can actually predict the “underwater” loans we were discussing before, by looking at the expected surplus that the outstaniding debt will bring with liquidations.
Let’s take an example.
If Bored Apes have a weekly trading volume of 10 NFTs per day, one could reasonably assume it would take ~4 days to sell ~40 NFTs.
There are a LOT of caveats with this, but we’ll explore that later.
Let’s assume you are using a lending platform which liquidates NFTs by doing “raffles” or “auctions” after seizing it from borrowers. A practice we will appropriately call, inshallah liquidation.
If the inshallah liquidation is expected to last 48 hours, then it makes sense to compare how many NFTs are being sold to the historical average number of sales over 48h periods.
In other words, outstanding debt vs 48h trading volume.
If only 10 NFTs are sold every 48h hours, and 30 NFTs are being listed during the inshallah liquidation, you can expect that the protocol will struggle to sell so many NFTs in such short time, forcing them to reduce the asking price.
As prices fall to find bidders, more NFTs are liquidated, due to the decreasing value of the NFT collection as collateral.
There is too much debt being issued against an NFT collection once there is significantly (3x to 5x) more debt issued against it than there is trading volume.
Caveats
The comparison is not perfect for many reasons but works well enough that I feel comfortable sharing it with potential lenders to assess their own individual risk.
Firstly, buy pressure could be higher on an asset if the price were lower, and liquidations often give liquidators incentives through discounted collateral.
Volume indicates demand at the market’s equilibrium price, but we don’t truly know the demand at the liquidation price (discount of market price). We assume this is a perfectly normal market and demand goes up linearly as prices decrease.
Secondly, each collection carries a mix of intrinsic value and extrinsic value. If the intrinsic value were to evaporate, historical volume would not be useful in predicting future volume.
Yesterday’s safe loans are not necessarily tomorrow’s safe loans, and NFT collections change rapidly.
Thirdly, it’s very hard to assess volume for liquidated NFTs. Liquidation auctions are usually held away from traditional marketplaces / AMMs.
The volume on Magic Eden or Hadeswap does not perfectly reflect the potential volume on Frakt’s liquidation page.
Lastly, it’s difficult to account for wash trading, which adds noise to our equation. Try to remove platforms which enable wash trading to have a clearer picture of a collection’s volume.
That’s all :)
I hope this article was useful for you, and can help you protect your bags in the future. Lending can be extremely profitable, but with such high interest rates, you must do a bit of homework to fully enjoy the upside.
Note: None of this is financial advice ! I am only sharing how protocols measure risk for NFT lending, but your situation might be completely different as a lender. A heuristic is not fool proof, but helps you be a more educated lender.
Outstanding debt is the $$ amount that a protocol has given borrowers, and owes lenders for a given market. In NFT world, markets are collection based, so each collection has outstanding debt.
insightful article