Home / Blogs

A Single Price Index for Domain Names Is Misleading

Alex Tajirian

Having a single price index for the domain name industry would be worse than useless. Such an index is presented in a recent study by Thies Lindenthal. The index is intended to be a benchmark for domain owners and investors. But it's out of line with other studies and the common sense of how a market operates. A much better barometer to follow is average prices for groups of domain names with similar characteristics.

Remember that the Dow, the S&P500, and NASDAQ sometimes move in opposite directions. A single index for stocks couldn't contain that movement, and a single index for domain names couldn't contain the differences in the amount and direction of price fluctuations among various domain name groups with common characteristics

Using a single index would have dire consequences. Basing transactions on such values would cause some owners to overpay, others to underpay. The resulting prices would be worse than imprecise, they'd be wrong — contrary to fair market value. Inflated prices can cause owners to resist selling and thus, forego lucrative deals. In the early days of the domain industry, wild price predictions by some owners not only put the industry on the map but created so much interest that investment returns rocketed above those of other industries. However, it is hard to determine if on average the sales have been over- or undervalued.

Other problems with Lindenthal's single price index:

1. It's based on information taken from just one market, Sedo. But other studies reveal that sale prices for similar names vary across marketplaces. Moreover, Lindenthal's model ignores the ask prices of listed domain names. That's relevant data if the aim is more precise estimates.

2. The reported index-price correlation with the NASDAQ 100 is grossly misleading. Why would prices of domain names of large companies move in synch with those of small and backdoor domains! One would expect that different groups of domain names would be correlated with different indexes depending on their risk characteristics.

3. The methodology is not robust to the presence of outliers in domain name price data.

4. The model does not identify risk factors driving domain name prices, such as the number of a domain's key word searches, the cost per click (CPC) of the key word, the popularity of the key word as shown by uses of it on Web sites, etc.

5. The model ignores evidence of a nonlinear relationship between prices and the explanatory variables.

Yes, there are domain name pundits who dismiss quantitative modeling and listen to the predictions of high-flying "experts." This approach all but guarantees more misses than hits. Studies confirm that "people who make wild bets but successful are remembered for those hits — but on average are the worst predictors." (See Jerker Denrell's "Experts" Who Beat the Odds Are Probably Just Lucky.) Tom Cruise could use sheer instinct to fly a jet in Top Gun, but in real life smart pilots use their instruments.

We need models based on rigorous analytics if we want to track the price movements of domain names. Of the models available, the single price index doesn't measure up.

By Alex Tajirian, CEO at DomainMart
Follow CircleID on
Related topics: Domain Names

If you are pressed for time ...

... this is for you. More and more professionals are choosing to publish critical posts on CircleID from all corners of the Internet industry. If you find it hard to keep up daily, consider subscribing to our weekly digest. We will provide you a convenient summary report once a week sent directly to your inbox. It's a quick and easy read.

I make a point of reading CircleID. There is no getting around the utility of knowing what thoughtful people are thinking and saying about our industry.

Vinton Cerf, Co-designer of the TCP/IP Protocols & the Architecture of the Internet

Share your comments

I agree, domain prices are in the eye of the beholder Chris McElroy  –  Aug 15, 2013 11:56 AM PDT

For instance, a little self-promoting, but only to make a point. I have SearchEngineOptimizationSchool.com on Sedo for sale at 10,000 dollars as a buy now option.

If you use standard appraisals, the domain is probably only worth about $500.

However, I do training for people who want to learn about marketing and SEO. I have a current business doing content marketing and SEO that keeps me really busy. I have a business plan for that domain name. I think it's a good one that'll make me a lot of money.

But do I have the time to start a new business? Not right now. But I do have an app I want to build for my current clients and could use that money to produce it.

So if someone buys it, great. I move forward with my app. If they don't buy it, then I will get around to building it down the road.

So, whatever a domain name appraiser or standard index would say the domain is worth doesn't matter in the least. That's what the domain is worth to me right now and maybe to someone else with a good idea for using it.

Domain appraisals have always been faulty. Great Domains inflated domain name appraisals for years. An example; Wanteds.com was listed there at one time and Great Domains valued it in the high six figures range.

I suppose if someone really had a good plan for that domain and a large budget to promote it, they could recoup their investment, but wanteds isn't even a real word, so I would have valued it at $500. Again, though, that price is subjective. It's what I see as the value. Great Domains disagreed. Did they have a vision? . . . or a hallucination?

Your opinion would be just as valuable as theirs.

A domain name is as valuable as the owner or purchaser sees it for whatever reasons they have.

Estimating a genral trend is very different from valuing individual domain names Thies Lindenthal  –  Aug 17, 2013 1:12 PM PDT

Indices provide information on the market in general, not on individual assets. A financial investor will look at e.g. the S&P;500 for a general understanding on stock markets trends - but not for a valuation of an individual company. 

The arguments presented in this blog post unfortunately miss this fine but important distinction.

It can never be worse to have more information than less.  The purpose of IDNX is to estimate a fair trend of overall re-sales prices, which it does well based on tested economic techniques. I've addressed some of Mr. Tajirian concerns below, but for more detail I'd welcome a closer look at my paper on IDNX.

Also, other big market places beside Sedo see the value in the index and have offered to open up their sales data.  I am in the process of signing a very interesting data sharing agreement - stay tuned.

More specifically:

1) IDNX is based on massive data (~ 250,000 sales) ranging from very high value to low prices ranges. They contain sales from auctions, fixed price domains and brokerage. I will further extend the data basis in the coming weeks. IDNX is only looking at real sales - ask prices, as suggested by Mr. Tajirian, are a very poor metric to look at since they only show the valuation by the seller. We all know how often sellers reach for the stars with their asking price. The only "true" price is an actual sales price.

2) Common sense and statistics both show that domain markets boom when the economy is doing well and vice versa. Why shouldn't there be a link between general economic indicators like the NASDAQ 100 index and IDNX?

3) The two stage estimation technique behind IDNX deals with outliers very well. Actually, it was developed at the Massachusetts Institute of Technology as a technique to reduce noise from outliers in similar instances.

4) IDNX circumvents the need to explicitly find and any "risk factors". Like the big housing indices, it compares transactions of the same domain in time. This is a better way to make sure to compare domains of exactly the same quality than trying to describe the quality of a domain with a few general factors.

5) IDNX does consider non-linearities in the best way possible. By comparing the sales prices of domains with the same SLD in time, non-linearities are accounted for. Better than a factor model ever could.

Thanks Dr. Lindenthal for taking the time Alex Tajirian  –  Aug 19, 2013 9:40 AM PDT

Thanks Dr. Lindenthal for taking the time to provide in-depth comments.

Following a private conversation with Dr. Lindenthal, I would like to outline the following points of agreement:

1.  A single price index can be better than no index at all. As such, the IDNX’s functionality, not computation, can be compared to broad indices like the Wilshire 5000 for the stock market, which represents the market value of all stocks actively traded in the United States. An index tracking the overall market trend might deviate from more narrow measures of groups of US stocks such as the DJIA, the S&P500;, or the NASDAQ 100. It is well known that these indices can move in opposite directions and represent different risk/return trade-offs. Nevertheless, a general index like the Wilshire 5000 can provide an unbiased overall direction of the market. For investors with specialized domain portfolios, however, it would be desirable to have sub-indices that match the risk-return profile of the investor's portfolio more closely. Unfortunately, such specialized indices are not available yet.

2.  It is possible to derive multi-indices using the methodology adopted by DomainMart for valuing individual domains. The methodology is robust and can handle noise in the data.

3.  When estimating an index, only sale prices should be used. However, when estimating prices for individual domains, ask prices can provide valuable information.

4.  Using transaction data from multi sources improves the accuracy of the price estimates.

5.  Statistical analyses of large amounts of relevant data yields superior estimates to heuristics.

Okay Chris McElroy  –  Aug 21, 2013 8:52 AM PDT

I agree with everything said by Dr. Lindenthal and Alex, IF you are looking at domain names being bought and sold as speculators do. If you're looking to buy and sell domain names in the same way you buy stocks, your points are truly valid.

But, if you sell domain names to business owners that plan to use them for business, then the index is useless as a guide. The value of the domain name is how valuable it is to your business. Can you get a return for the specific investment? That's all that matters.

If I own the last house on the block that you are trying to buy for a development you are about to build, average house sale prices are useless. How much is my little house worth to you so you can get started on your development? That's all that counts.

So, as I said, if you're "investing" in domain names, then the index is a good tool. If you're buying a domain name for your business, then what it is worth to you is all that counts. You may want to use the index to negotiate, but if the seller also has a use for the domain name and has applied a value to that, then they, or in my case, I wouldn't consider the index valuation as important at all.

To post comments, please login or create an account.



New TLDs

Sponsored byAfilias

DNS Security

Sponsored byAfilias

IP Addressing

Sponsored byAvenue4 LLC

Domain Names

Sponsored byVerisign


Sponsored byVerisign