An end to domain squatting

TL:DR; Domain name speculation causes economic harm, and the easiest way to stop it is by increasing prices to $100 a year.

Domain speculation is value-destroying

I do not oppose asset speculation in general. In the financial markets speculators provide liquidity and reduce spreads. If you wanted to buy 10 shares of GOOG, you can do it instantly and cheaply in a stock market with lots of speculators but you will struggle to find a counterparty if only long-term investors were allowed to trade. In short, speculators in financial markets generally provide positive economic value.

Domain name speculation is fundamentally different. Currencies, commodities and vanilla securities are fungible — the ¥1 you may buy from one guy is identical to the ¥1 you might buy from someone else. This allows a competitive market to exist.

Buyers in the secondary market for domain names, on the other hand, are there because they need a particular name and simply do not have any other choice. It is inevitably an uncompetitive, monopoly-seller situation.

Obtaining a domain that’s being squatted on involves significant dead-weight losses: the considerable effort and expense of negotiation, escrow and legal proceedings in addition to losses from delays. Even more value is destroyed when no deal is reached — the domain name remains in the hands of a squatter rather than someone who could put it to good use.

Even when deals happen, the practice results in misallocation of resources. Wealth is transferred from productive businesses and individuals to unproductive squatters.

Domain name speculation is an unashamedly rent-seeking, value-destroying activity. It must be stopped.

Making domain speculation unviable

It’s hard, if not impossible, to draft objective criteria that distinguish between speculators and legitimate domain name users; adding layers of regulation and oversight will only cause costs to explode and corruption to run rife. The answer lies in understanding and disrupting the speculators’ business model.

Speculators buy up tens or hundreds of thousands of domain names containing generic words, combinations of words or misspellings of trademarks at an average price (for .coms) of about $8 per name per year. They “park” these domains, serving a single page of ads and malware to anyone who types the URL by mistake. A tiny fraction of these domains (typically 0.1%) are sold each year.

My (very rough) estimates are that parking revenue averages about $12 per domain per year, and revenue from sales are around $2000 per domain sold (or about $2 per domain squatted on). Both revenue streams have very tail heavy distributions (a small fraction of the domains provide a majority of the revenue).

The average revenue per domain with this business model is around $14 per domain per year. Increasing the registration cost of a domain to, say, $20 per year would make this model as practiced today unviable.

However, because of the tail-heavy skew in revenues, $20 is not enough to kill domain speculation completely. Speculators will release their less profitable domains but continue to hold the memorable ones. There is still room for the majority of squatted domains to be unprofitable but still make money on the aggregate.

$100, on the other hand, will make profitable domains too few and mistakes too expensive for the squatters’ business model to be viable. The .io top-level domain, which charges $96 and sees remarkably little squatting, is a good example.

Registries (the monopoly authority controlling a top level domain, like .com) are currently under strict price regulation through their contract with ICANN. Below them is a highly competitive industry of registrars and resellers, which means that domains are often sold at very close to the registry’s regulated price. Implementing this increase will therefore be easy — just one change (to Section 7.1d) of the ICANN-Verisign agreement.


Impact on legitimate domain owners: Domain name registration represents a a neglible fraction of total cost incurred by businesses, individuals and other productive domain owners. Even at $100 a year, it will be dwarfed by the costs of building and hosting a website, marketing it, and creating content for it.

Impact on registries: If implemented naively, registries like Verisign will make (undeserved) windfall gains. Perhaps an intermediate measure might be to require them to donate a portion of this to a nonprofit for recovering domain speculators.

Apply to existing 2+ year registrations: The increased pricing should apply for any annual auto-renewal that happens after the increase is announced; For example, if a domain has been registered for 10 years at the old pricing one day before the increase, the first year will still cost $8 while the second year and beyond will be considered to cost $100. The extra $72 paid beyond the first year’s cost will be rolled over — the domain owner must pay $28 more for the second year and $100 per year after that.

This is to prevent squatters holding on to their domains via 10-year registrations.


Now read this

Proportional representation

The AAP manifesto mentioned something about “supplementing the existing first-past-the-post system with proportional representation.” This was a pleasant surprise — it mirrored something I’d discussed with friends in 2010. In response to... Continue →