V God: Break the monopoly of ENS domain names and set normal fees based on demand

Domains on ENS are cheap right now. The cost to register and maintain a five-letter domain is just $5 per year. From the perspective of someone trying to register a single domain, this sounds reasonable, but the global picture is very different: in the early days of ENS, someone could register a random five-letter word domain and prepay for a hundred years of ownership. Today, almost all of these word domains are already taken, waiting for someone to acquire the domain from them at a higher price. OpenSea data shows that about 40% of domains are for sale or are only sold on the platform.

Is this really the best way to distribute domains? The ENS DAO is earning far less than it could, which limits its ability to improve the ecosystem. The status quo is also not conducive to fairness in the ecosystem. Being able to buy domains cheaply was good for people in 2017 and is still feasible in 2022, but the consequences may seriously hinder the ecosystem in 2050. Considering that the actual cost of purchasing a domain ranges from 0.1 to 500 ETH, the nominally cheap registration price does not actually save users money. In fact, relying on the secondary market makes domains more expensive than high-quality in-protocol mechanisms.

Can we distribute ownership of domains in a better way? Is there a way to generate more revenue for the ENS DAO and better ensure that domains are owned by those who can make the most of them, while maintaining the credible neutrality and accessibility that makes ENS valuable with strong guarantees of long-term ownership?

Problem 1: There is a fundamental trade-off between the power of ownership and fairness

Suppose there are N "high-value domain names", and after P years, no one can obtain a high-value domain name again.

The unlimited time allocation of finite resources is incompatible with long-term fairness, and this is true for land. This is why there have been so many land reforms throughout history, and it is also a big reason why many people today advocate for a land tax. The same is true for domain names, although the problem of the traditional domain name field was temporarily alleviated by "forced dilution" of early .com domain name holders through the massive introduction of .io, .me, .network, and many other domain names. ENS has already paid lip service to not adding new top-level domains to avoid undermining its chances of eventually integrating with the mainstream DNS, so such dilution is not an option.

Fortunately, ENS charges not only a one-time fee to register a domain, but also a recurring annual fee to maintain the domain. Not all decentralized domain name systems have had the foresight to implement this. Not so with Unstoppable Domains, which favors short-term consumer appeal (never having to renew) over long-term sustainability. The normalcy fees of ENS and traditional DNS are a healthy way to mitigate the excesses of a truly unlimited “one-time ownership” model. At the very least, the normalcy fees mean that no one can accidentally lock into a domain forever because they forget or are careless. But that may not be enough. It’s still possible to spend $500 to lock in an ENS domain for a century, and there are certainly some types of domains that are in high enough demand that they are vastly underpriced.

Problem 2: Speculators don’t actually create efficient markets

Because of the "first come, first served" model with a low fixed fee, many domains will be bought by speculators, but speculation is natural and a good thing. This is a free market mechanism, and speculators truly want to maximize their profits. They are incentivized to resell the domain in such a way that it goes to anyone who can make good use of the domain, and their excess returns are simply compensation for this service.

But it turns out that a profit-maximizing auctioneer does not actually maximize social welfare. This is shown in the figure below.

Maximizing revenue for the seller almost always means accepting the possibility that the domain will not be sold at all, leaving it completely unused. Auctions that maximize revenue for the seller are least efficient when there is only one potential buyer (or at least one buyer who values ​​the domain much higher than the others), and this efficiency decreases rapidly once there are many competing potential buyers. But for a large number of domains, the first category is exactly the situation they are in. For example, there are people, projects, or companies that buy domains with the same name as them. So if a speculator buys such a domain, they will of course set the price very high, accepting a high probability that a transaction will never be completed, in order to maximize their revenue in the event that a transaction occurs.

Therefore, we cannot say that speculators taking a large portion of the domain allocation revenue is simply compensation for their efforts to ensure an efficient market. On the contrary, speculators can easily make the market worse than if the protocol had well-designed mechanisms that encouraged domains to be sold directly at a fair price.

Setting stricter standards for stable domain ownership

The monopoly problem of overly strict ownership of non-fungible assets has long been known. Solving this problem in a market-based way was the original goal of the Harberger tax: require the owner of each backed asset to set a price at which they are willing to sell it to others, and charge an annual fee based on this price. For example, 0.5% of the sales price could be charged each year. Holders would be incentivized to leave the asset available for purchase at a reasonable price, and "lazy" holders who refuse to sell would lose money every year.

But the risk of being forced to sell something at any time can carry huge financial and psychological costs, which is why advocates of Harberger taxes generally focus on its application to ownership by established market participants.

It turns out that domain names don’t hodl very well. Domain owners are often immature, the cost of changing a domain is often high, and the negative externalities of changing a domain wrong can be large. The highest value owner of coinbase.eth might not be Coinbase, but a scammer might grab the domain and then immediately fake a charity or ICO, claiming it’s run by Coinbase, and get people to send money to the address. For these reasons, Harberger’s tax on domain names is not a good idea.

Alternative Solution 1: Demand-Based Normalized Pricing

Today, maintaining ownership of an ENS domain requires a normal fee. For most domains, this is a simple and very low fee of $5 per year. But what if we charged a fee based on the actual level of market demand for the domain?

This does not require that the domain name be immediately available for sale at a specific price. Instead, the pricing process will fall to the bidders. Anyone can bid on a particular domain name, and if they keep an open bid long enough (e.g. 4 weeks), the domain name's valuation will rise to that level. The annual fee for the domain name will be proportional to the valuation. It could be set at 0.5% of the valuation). If there are no bids, the fee may decrease at a constant rate.

When a bidder sends their bid amount to a smart contract to bid, the owner has two options: they can accept the bid, or they can reject it, although they may have to start paying a higher price. If the bidder bids more than the domain is actually worth, the domain owner can sell the domain to them, causing the bidder to lose a lot of money.

This property is important because it means that "compromising" a domain name holder is risky and expensive, and may even end up benefiting the victim. If you own a domain name, and a powerful person wants to harass or censor you, they may try to bid very high for the domain name to greatly increase your annual fee. But if they do, you can sell it to them directly and charge a huge amount.

This already provides more stability to ENS domain prices, and is friendlier to new buyers than the aforementioned Harberger tax. Domain owners don't need to constantly worry about whether they set the price too low. Instead, they can just pay the annual fee, and if someone makes an offer, they can take 4 weeks to make a decision, either to sell the domain or to hold on to it and accept the higher fee. But even this may not provide enough stability. To go further, we need a compromise within a compromise.

Alternative Solution 2: Normalized demand-based pricing caps

We can modify the above scheme to provide stronger protection for domain name holders. We can try to provide the following attributes:

Strong time-limited title guarantees: For any fixed number of years, it is always possible to calculate a fixed amount that you can pay in advance that unconditionally guarantees title for at least that number of years.

As shown in the figure:

Please note that the numbers in the table are only the theoretical maximums required to guarantee holding a domain name for a certain number of years. In reality, few domain names will have high bids, so the price that holders of almost all domain names will eventually pay is much lower than the maximum price.

Some versions of an “annual fee cap” would be strictly more favorable to existing domain name holders than the status quo. We could imagine a system where domains that received no bids paid no annual fees, and bids could increase the annual fee to a maximum of $5 per year.

Demand from external bidding clearly provides some signal about the value of a domain (and therefore the extent to which the owner is excluding other buyers by maintaining control of the domain). Therefore, I think it should be possible to find some attractive parameter choices for demand-based fees.

A superlinear f(N) value with a maximum annual fee that increases over time is a good idea. First, paying more for long-term security is a general feature of the entire economy. Fixed-rate mortgages generally have higher interest rates than variable-rate mortgages. You can earn higher interest by providing a deposit that is locked in for a long time. This is the compensation the bank pays you for providing the bank with long-term security. Similarly, longer-term government bonds generally have higher yields. Second, the annual fee should be able to adjust eventually based on the market value of the domain, we just don't want it to happen too quickly.

Two possible solutions

1. Democratic legitimacy: Proposing a compromise that is truly adequate, satisfying more people, and perhaps even resulting in higher returns for some existing domain name holders (not just potential domain name holders).

   For example, we could charge an annual fee based on demand, with a maximum annual fee of $640 for domains up to 8 letters and a maximum annual fee of $5 for the longest domains, and if no one bids, the domain holder does not have to pay anything. Many ordinary users can save money under such a proposal.

2. Market Legitimacy: Legitimacy is gained by creating a new institution (or subsystem) that avoids overturning people’s expectations under the existing institution.

   One possible compromise would be for the ENS DAO to hand over single-letter domains individually to projects that run some other type of trusted neutral marketplace for their subdomains, as long as they hand over at least 50% of their revenue to the ENS DAO.

Summarize

If changing ENS domain pricing is not feasible, then market-based approaches that explicitly encourage different rules for subdomains should be considered.

For me, the crypto space is not just about money. Instead, my interest in the space is more about credible neutrality and providing a high degree of assurance of ownership, but at the same time increasing the cost of domain name squatting, generating more revenue for the ENS DAO, and increasing the chances of getting a domain name for those who don’t have the domain name they want. A high degree of assurance of ownership is very important for the proper functioning of the domain name system.