I had this idea the other week, sent it to a friend. He didn’t love it, but figured let’s post it anyway.
Over the past few weeks, everyone has been talking about NFT royalties — a percentage of the sales prices gets added as a fee at sale, or more likely subtracted from the seller’s take, and transferred to a third party, usually the collection’s creator.
If you buy a Bored Ape on OpenSea, you pay the listed price, the seller gets 95%, OpenSea gets 2.5% and the “creator” — in this case Yuga Labs — gets another 2.5%.
Some top collections like CryptoPunks, trade with 0% fees. Others like Art Blocks’ Fidenza set a 7.5% creator fee.
Many keys have been smashed over the past few weeks about whether NFTs should have royalties, how high royalties should be, and who should enforce them. Last week the Magic Eden NFT exchange hosted a Twitter Spaces about their decision to make creator royalties optional, and 19,000 people listened in.
Two of my friends and favorite writers in NFTs wrote popular articles on the subject. 0xFoobar and NFTStatistics agree on much, but draw somewhat different conclusions and emphasize different aspects of the royalty experience.
NFTStatistics stresses that many creators of great projects made their revenue on royalties — rather than the original sale, or as we call it “the mint”
0xFoobar stresses that royalties are unenforceable at the contract level, and that some of the enforcement mechanisms suggested operate against the spirit of permissionless ownership
NFTStatistics points out that Tyler Hobbs made just about all of his Fidenza revenue on royalties from secondary sales.
Fidenza by Tyler Hobbs generated 170 ETH off mint but 3,850 ETH through royalties, allowing Hobbs to profit as his work rose to fame. Both Yuga and Hobbs also generated substantial revenue through followon NFTs, however, showing that royalties are not the only path to riches for successful projects.
Pricing the Fidenza market cap conservatively at 90 ETH per item, that’s 90,000 ETH (~$121M at today’s prices). Hobbs generating 4,000 ETH in fees, <5% of the market cap, I’m sure would strike most as reasonable. More reasonable than the 170 ETH he generated from the mint. Today, the price of two Fidenzas.
0xFoobar’s point that royalties can not be in fact enforced at the contract level, while retaining decentralized ownership, is a good point, and not a trivial one. Let’s just say it would be very hard to enforce side payments to a third party creator address on every sale, and harder still to tie these payments to a percentage of consideration offered in exchange.
What happens with private sales? What about buying multiple items in a single transactions? What about trading NFTs for NFTs, like fantasy football trade szn?
These may seem small considerations contrasted with “artists getting paid” and I don’t disagree. But as we’ve seen with many examples, such as yesterday’s frenzy of NFT sales on Reddit, trading the JPEGs is part of the value, and a part of the fun. Bored Apes can be “stolen” because whomever controls the secret seed phrase controls the NFT. Part of that ownership is being able to sell and transfer the asset, or even borrow against it.
Can we imagine a way for artists to get paid in proportion to market prices, while not getting in the way of true decentralized ownership?
Ownership Claims and Indirect Ownership
We want two things
the creator participates in the upside of a successful collection
the collector has free reign to sell, stake or transfer the NFT
Unfortunately, there is no way to do this without introducing a new contract, and a bit of complexity. This is not a fully baked idea. But I think this can work, and the complexity offers positive externalities.
When you buy a pre-royalties NFT like a CryptoPunk, you own the NFT — it is fully yours. However even now, when you buy a Mint Pass — say for Tyler Hobb’s popular new QQL collection — you actually buy the right to mint in the future. In this case with your own curation step. But many drops are handled via Mint Passes — sold “pre-reveal” instead of collectors bidding on known individual items directly.
Extend this idea just a little, and you can imagine a split ownership of the NFT:
the creator owns part of the NFT
the collector owns part of the NFT
The creator can use their ownership to charge an annual tax against the asset, as well as to set the collection wide buyout schedule. The collector retains all other rights, crucially the right to sell, transfer, stake, token gate or claim drops.
At mint, the collector pays, let’s say 0.1 ETH, for the right to own 80% of the NFT. The creator keeps the other 20%.
At any point in the future, the collector can “buy out” the remaining “debt” of the NFT, at which point that NFT converts to a fully owned asset. The creator gives up all future claims, including to royalties or taxes.
Naturally the “buy out fee” will be much higher than the mint price of 0.1 ETH. I suggest that the the buyout fee should be:
subject to a max fee schedule — starting at 10x mint price, and increasing gradually over time
once market prices are established, the collector would be expected to set current buyout price based on third party price oracles — such as those we are building at DeepNFTValue
Good for the Creator
The creator gives up one valuable thing that they (sort of) have — the right to a slice of (most) future sales, forever. However, to buy out this future claim, the collector will pay either a large multiple of the mint price, or a fee proportional to current market value.
Let’s say the creator’s post-mint share is 20% — at steady state the collectors will be buying out at 20% of the market value. Suppose the collection does well, really well. How much does the creator expect to make? Every future stream of income has a current day price.
Every future stream of income has a current day price.
If creators wants to hold on to some upside forever, they should consider keeping 10% of the total supply to themselves as well — as Larva did with CryptoPunks, and NounsDAO does today.
We appreciate that holding back supply isn’t a perfect solution, especially as most projects don’t mint perpetually over time like Nouns. If creators need to sell on the secondary to earn revenue, it does affect the market. Decisions when to sell, and what to sell become important — especially for art and PFP collections where every piece is unique. [Collections with multiple owners per item seem to support creator supply and buyback more easily.]
That said, price oracles and lending protocols mean that creators can borrow against their held back supply, without having to sell. They pay a fee, for the right to both get revenue now, and also creating a floor on their asset values. Assuming NFT lending grows and rates remain pretty reasonable, the creator gets to participate directly into the upside of their collection, without having to sell specific items into the market.
In practice, it would be nice if creators both earned fees on their collectors’ holding as those go up, and kept a bit of supply for themselves. Both seem good, in different ways.
Good (Enough) for the Collector
The biggest downside for the collector is complexity. There are now choices to make other than minting, buying and selling. Do you buy out the creator at today’s price, or wait a month? Perhaps worse, when buying an NFT, you now have to check what percentage of it are you buying…
The flip side of this complexity is leverage. You may have to now pay multiple ETH or even dozens of ETH eventually, to buy out your 0.1 ETH NFT. You wish you bought it out sooner, or perhaps not at all. But as the original minter, you have full control of buying and selling, for an asset worth much more than the price you paid.
Since theres is a max buyout schedule — starting with 10x of the mint price and rising over time — for a collection that’s doing great, there will be an obvious time to buy out the debt. And if it’s not doing great, just wait or don’t bother. Again we assume that at steady state the buyout schedule will be tied to accurate third party oracle prices such as those that we are building at DeepNFTValue.
Positive Externalities
The main reason to move from taxing transactions to taxing ownership is to remove the barrier to sales and transfers. But this partial-ownership model also leads to a few unexpected advantages:
financializeable stream of future payments for the creator
easy to collect ownership/harberger taxes
a path to repossession of lost and neglected NFTs
At every point in time, the outstanding debt on a project is known, verifiable, and can be priced in current market value. The creator may wait some time for the buyout ETH to come rolling in, but will know at all times how much that buyout is currently worth. They could even sell the claim or borrow against it — since it’s a smart contract with know dependencies, unlike royalty payments that expect the exchanges to collect on the creators’ behalf.
Harberger taxes also become simple. Instead of asking the collector what is a price he is willing to sell for, and charging himself X% of that price, the creator simply tacks on taxes to the outstanding debt on the asset, as a percentage of ownership. Similarly to how BendDAO tacks on interest to the loan principal — but without a price oracle setting liquidations…
Except there are two critical numbers on the ownership scale — 0% and 100%.
When a collector buys out the creator, he owns 100% and gets a new NFT with all trading rights and exits the original contract’s buyout schedule. But if the NFT accumulates so much debt that ownership falls to 0% — the item reverts back to the creator — who can sell it anew.
Realistically the 0% case would only happen, for a successful project, if the collector loses his private key, or otherwise forgets about the collection for a number of years. ~700 CryptoPunks have been rumored as lost. More are owned by wallets that haven’t been active in years. Perhaps it wouldn’t be so bad if these JPEGs eventually came back on the market.
Conclusion
How likely is this to happen? Not very likely. But it might be great if someone tried.
One of my favorite wrinkles is the easy way in which harberger taxes can be added on the contract level — not even requiring a price oracle. While at the same time accurate price oracles enable value to be claimed by the creator — as a small but non-trivial percentage of the collection’s market value.
Since creator debt lives with the NFT itself, there is no need to chase down buyers or sellers at transaction time — the creator’s debt simply grows by say 1% per month in proportion to the debt already owed. Eventually the debt is either paid off, or the NFT reverts back to the creator.
There is more to say, such as how low mint prices are good for everyone, provided creators can build a great allowlist. This is not a fully baked idea, and probably too complicated a solution for many. But I’ve not see anyone else write something like this elsewhere. It would be cool to see it tried, so figured I’d post about it.
What do you think?