How recursive royalties will save NFTs

A fully decentralised, fair and robust solution to NFT royalty payments

UPDATE: Many thanks to Tyler Cowen for highlighting this article at Marginal Revolution. Follow me on twitter @XTZtcg.

A lauded feature of NFTs is their ability to allow payments to be made to the original creator of the token upon each sale, giving an artist exposure to the value of their work on the secondary market.

But NFT royalties are doomed.

Contrary to expectations and popular belief, royalties are not enforced by an uncompromising golem on the blockchain. Royalties are voluntarily enforced by marketplaces.

The launch of Yawww’s zero royalty marketplace in early 2021 triggered a race to the bottom, and more ‘zero-fee’ marketplaces soon followed. Royalty-optional brokers have been making serious inroads into the market share of encumbents. On October 15th, a major marketplace joined them:

Magic Eden 🪄 @MagicEden

🧵After some difficult reflection and discussions with many creators, we’ve decided to move to optional royalties on @MagicEden. Effective later today, we will also begin running a promotion to waive our platform fees.

As marketplaces which honour royalties will be outcompeted by those which do not, openSea, Rarible, Objkt and others will be forced to follow the path of Magic Eden, unless a solution is found. I will be presenting one such solution in this article, and comparing the various proposals emerging in upcoming posts.

We can rescue royalties - but should we?

Yes. Fairly rewarding creatives for making art consumers love is (shock) a real-world use case for crypto. Hell, for an person without special insights into finance or blockchain (i.e a normal person) it may be the only social benefit from crypto that they can see.

Royalties are a path to a decentralised future for art and music - where more people can pursue a creative career, where more creative niches are explored, new genres discovered and artists at all levels of success are more richly rewarded. This will be a world where crypto is creating social value, not merely moving it. Royalties are great for creators, for collectors and for marketplaces.

Until now, royalty payments were protected by a social convention. We could be cynical such conventions are so fragile, or cheered this one lasted so long. Either way - shouldn’t blockchain technology allow us to find a ‘trustless’ solution to this problem. Isn’t this exactly the sort of problem blockchain should allow us to solve?

Again: yes! I am working on a solution which I will outline in this article. I will share further progress here over the coming weeks and months, informed I hope by your comment and criticism.

The Problem

Very broadly, I see two issues with existing NFT royalties, (henceforth: legacy royalties) a policy problem and a technical challenge.

1. They are not fair

2. They are not enforcable

1. Legacy royalties are not fair

Compare these two cases of royalties under the legacy regime:

Creator 2 recieves greater royalties, even though Creator 1’s artwork is valued equally on the market. This is unfair on Creator 1.

Meanwhile, Rebecca, Higgins and Keeley have made a shared royalty contribution of $750 for one artwork valued at $5000 in Case 2, whereas Ted and Nathan have pay $250 less for an artwork of the same value. This is unfair on the collectors in Case 2.

Consider another example: an NFT is dropping in price as successive collectors try to get it off their books - say at $100 then $90, $80, $70 and finally $60? At 10%, the total royalties paid will be $40 total to the creator for this depreciating work! Meanwhile, an artist who sold a single piece for $300 to a collector who loved it so dearly they never wish to part with it will make only $30!

With these incentives, creators may devise ways that encourage collectors to resell NFTs, even if this means reducing the value of their NFTs currently in circulation¹.

On the topic of fairness, it is worth noting that the current system is fair in one important way. Although some collectors may resent paying royalties that seem to arbritrarily accrue to NFTs more often transacted, collectors do share the burden of paying creators in a fair(ish) way. For a given NFT, each collector who owns it over time pays royalties proportionate to her valuation (for instance, in figure 1, Rebecca values Creator 2’s artwork twice as highly as Higgins, and she contributes twice as much to the total royalties paid as Higgins).² This is an excellent quality of the current system, and one we should seek to preserve.

2. Legacy royalties are not enforceable

A good overview of some of the technical problems that make NFT royalty payments difficult to enforce on chain is given here by Arthur Breitman (co-founder of Tezos blockchain):

Quick summary: it is hard to know when an NFT sale has occured and at what price.

There is a lot of great material in this video beyond the discussion of technical difficulties, especially on the Harberger Tax which my design riffs on.³ We will explore this further in an upcoming post.

Design Criteria

While we are tinkering with NFT royalties, we may as well attempt to design something fundamentally better from first principles. Here are the criteria I have sought to meet:

There are two ultimate goals for a new royalties regime:

1. It must be fair

2. It must be enforceable

I have 8 criteria on the next level down:

1. Fair

   1. Creators should recieve royalties proportionate to the market value of their work (not the number of times it changes hands).

   2. Total royalties paid to the creator should be spread among the collectors who own the NFT throughout its lifetime, proportionate to how they valued the NFT and how long they held it.

   3. Creators should be able to set royalties at the rate they choose.

   4. Creators and collectors should have the greatest possible degree of ownership and control over their assets.

2. Enforceable

   1. Royalty collection should be automatically enforced and fully decentralised.

   2. Paying royalties should be preferable to not (e.g via social or financial incentives or penalties built into the solution).

   3. Royalty collection should be robust against known workarounds (e.g wrapping the NFT in a new contract, use of a custodian). Note that if criteria 4 is met fully this criteria is redundant.

   4. The solution should be “backward compatible” with existing NFTs operating under the legacy royalty standard. It should be possible to bring existing NFTs under the new royalty regime.

I have developed a design that I believe meets all these criteria. This is a work in progress! I invite comment and criticism. Please share your constructive criticism and general abuse in the comments. Feel free to suggest further criteria and changes to those above.

Proposed Solution: Recursive Royalties

I propose packaging NFTs in smart contracts called “Recursive Royalties”

The design involves four main conceptual elements:

• Tidemark royalties: the royalty is only charged on the amount above the tidemark, or highest previous transaction price. This is paid by the buyer.

• Tidemark Auctions: The NFT will be automatically entered into a tidemark auction - perhaps once every three years (unless the current owner triggers this earlier). The owner can choose to either sell the NFT to the highest bidder or keep the NFT by paying royalties on the amount above the tidemark. This also raises the tidemark to equal the winning bid.

• Recursive Contracts: The practical implementation could be achieved through a smart contract that recursively generates a new copy of the contract with an updated tidemark, a new tidemark auction deadline and a record of the new owner (or rather ‘last legitimate purchaser’). The old contract gets burned.

• Royalty burden redistribution: Each time the contract is recursed (usually a sale), the burden of the royalty payments are redistributed fairly among previous owners. Previous owners get a small payment from the current seller. This is deducted from the payment to the seller (only when this is necessary). If a current owner is choosing to keep the NFT following a tidemark auction, both the redistribution and royalty payments must be paid.

Please note that the creator NEVER refunds anyone any part of the royalty they have been paid and that collectors NEVER make payments to future owners, they can only recieve payments once they have sold the NFT.

The goal is to get these contracts, or something better, implemented on chains. Please share your thoughts, especially on approaches to implementation and anticipated challenges. Do subscribe to the newsletter and join us on this journey!

Future Posts

In upcoming newsletters I will be diving into the details of how the proposal will work, looking at each of the three elements above.

Upcoming topics will include:

• A deep dive into the details of the Recursive Royalties proposal, including worked examples and comparisons with the status quo legacy royalty system.

• How marketplaces and creators will be able to work together to enforce royalties on already minted NFTs.

• The expected impacts of tidemark royalties on the wider NFT market.

• How to redistribute royalty burdens fairly through a smart contract.

• More on the “Last Legitimate Purchaser” field in smart contracts. This can support or act as an alternative to tidemark auctions. This option trades off autonomous enforcement for complete ownership rights for collectors.

• Also: Emergents TCG content:

  • A look at the whitepaper

  • Decks and strategy

  • Analysis and predictions on Emergents NFT price movements

• There’s gonna be tables and diagrams people.

My profile picture is a screenshot from the animated art of ‘The Mad Confectionist’, an EmergentTCG NFT trading card. Art: VOLTA - Jennyson Rosero || Color: VOLTA || Animation: VOLTA - Gustavo Guimarães

1. Up to a point.

2. Note that this does not consider the length of time which Rebecca and Higgins each held the artwork for, which (spoilers) it should!

3. On the Harbenger tax, Breitman references this medium article by Simon De La Rouviere who in turn references recent work from Posner and Weyl.