Valuable Words: Pricing Internet Domain Names

Thies Lindenthal thies@lindenthal.eu July 13, 2011

Abstract
This paper estimates the rst constant quality price index for internet domain names. The index can serve as a benchmark for domain name traders and investors looking for information on price trends, historical returns and fundamental risk of internet domain names. It thereby increases transparency in the market for this newly emerged asset class. Furthermore, it can serve as a fever curve for the well-being of the internet economy, since the demand for domains is linked to online business opportunities in general. This proxy also includes small enterprises that are excluded by traditional stock-price indices. The quality of the domain names is controlled for by hedonic repeat sales regressions. The empirical work is based on a large dataset of real domain transactions spanning the years 2005-2011.

Acknowledgements
The author thanks Tim Schumacher and Ulrich Essmann for granting research access to the unique database of domain transactions facilitated by Sedo.com, the largest market place for domains. David Geltner is thanked for generously sharing his advice and feedback.

Introduction

Internet domain names bring back the location to the otherwise location-less internet economy. A domain name provides a virtual street address for any website or service on the internet. It is comparable to a tract of land on which a business or just a private home (page) can be built on. Understanding the World Wide Web as an agglomeration of interlinked network locations is not uncommon. This is visualized by many terms related to the internet exhibiting a spatial connotation. Labels for technical network addresses, for instance, are called domains, users are visitors, internet browsers have been baptized Navigator or Explorer, websites are home pages the list can be easily extended. Understanding domains as a novel form of land oers the opportunity to transfer established theoretical and empirical frameworks for the pricing of land into virtual space. Theoretically, AlonsoMuth-Mills models of urban layouts (Alonso, 1964; Mills, 1972; Muth, 1969) explain dierences in the

INTRODUCTION

rents and prices of land by dierences in the distance to jobs or amenities. Applying this reasoning to domains, the price of a domain is hypothesized to depend on its proximity to potential users. Since a voyage on the world wide web usually begins with the user entering the domain name of the desired website into her web browser, distance to the user can be seen as the eort a user is required to make to correctly remember and type a domain name. An appealing domain name like Apple.com is easy to recall and quickly entered. In this sense, an intuitive domain name is like a convenient down-town address linked to excellent transportation systems. Long or cryptic domain names are more burdensome, which is comparable to a longer commute to a location somewhere in the outskirts. Dierences in location fuel a heated race for the shortest and most memorizable domain names that sprung up since the very rst domain was created in March 1985. By now, more than 200 million unique domain names are registered (Verisign, 2011) with no end of growth in total numbers in sight. Furthermore, an active secondary market for domains proves that good domain names carry value. Exclusive domains oftentimes trade for 5 or 6 gure dollar amounts, and some for even more. The current record in reported sales prices is the 13 million US dollar transaction of sex.com in 2010. Trading of and investing into domains has quickly evolved from a geeky pastime of a few to the serious bread and butter industry feeding hundreds of professional traders today. The focus of this paper is to estimate the rst constant quality price index for internet domain names. It adapts an empirical framework borrowed from real estate research which is suitable for valuing infrequently traded and non-standardized assets like houses, antiques, pieces of art or, by analogy, virtual locations. The linguistic nature of Domains causes substantial heterogeneitiy in their quality, which makes domain names very comparable to traditional asset classes where the intrinsic value of an asset is not directly observable as well.1 The index spans more than 5 years and will be updated on a monthly basis in the future. Despite all rapid growth in the last decade, domain names are still a relatively small investment class which lacks any information on its inherent risk and return prole.2 Market participants and investors simply do not know whether domain names are a good investment. Did domain holdings deliver positive returns in the last years at all? How big were any returns? Rational investors evaluate returns and risk of their holdings simultaneously. The market index is the best proxy fundamental risk of domain names. The price volatility of a market portfolio hypothetically containing all domain names cancels out any domain-specic volatility. The fundamental or market risk of domains can be compared to the risk of an investors portfolio, which puts any return on this portfolio into a risk-adjusted perspective. Furthermore, the fundamental risk of domains can be compared to the risk of other investment classes like stocks, bonds or real estate.
1 Practitioners rely on coarse classications to describe the quality of a domain name.

Domain market places like Sedo.com or auctions.godaddy.com, for instance, classify their listings inter alia by extension (TLD), language, domain length, number of words, price range, industry sector, trac, special characters, listing type. 2 A rst glance at the price trends in domain sales are oered by industry sources like Sedo.com (2011).

A PRIMER ON INTERNET DOMAIN NAMES

Finally, any cautious economist will surely ask: Are domains names for real or just another fad? An entertaining gamble? A return of the dot-com-bubble? Or does an economic rationale justify the prices paid? Are domain name prices linked to the performance of the general economy or more specic to stock prices of internet companies? Or are domains a totally detached new economic phenomenon? Comparing the price path of the market index to other asset classes gives rst insights into the economic foundations of domain names. The remainder of this paper rst presents a primer on internet domain names, including a brief introduction on the nature of domain markets. Then the choice of an adequate index estimation methodology is motivated, followed by an overview of the data the econometric analysis relies on. Section 5 discusses the empirical results leading to general conclusions.

2 A Primer on Internet Domain Names

2.1 The Internet Domain Name System

Domain names make the internet navigable for humans. All servers, computers and other devices linked to the internet use a unique Internet Protocol (IP) number as their address within the network. Typing 74.125.39.106 into the address bar of a web browser will lead to the website of a well known search engine from California. Since IP numbers are hard to remember for internet users, a Domain Name System (DNS) provides shortcuts, assigning more memorizable domain names to the underlying IP numbers. Each domain name consists out of at least two components: Top Level Domains (TLD) structure the overall namespace into a limited number of subsets that either have a global scope (gTLD) like COM, NET, ORG or that are country specic (ccTLD) like DE, FR, NL.3 Each TLD is subdivided into Second Level Domains (SLD) that can consist out of letters, numbers and hyphens only. Spaces are not allowed. The combination of a TLD and SLD, separated by a colon, is commonly referred to as an internet domain name. It cannot exceed 67 characters in total.4 From a lawyers perspective, it is not resolved yet whether a domain name is an intangible asset or just a contractual right (Burshtein, 2005) in practice, the case is clear: domains are treated and traded like assets.

2.2

Primary Market for Internet Domain Names

The primary market for domain names resembles the land rushs in 19th century America. The general rule is: Whoever les a registration for a domain rst receives this domain. Some country spe3 Some

countries subdivide their namespaces again, like CO.UK for commercial UK domains and the AC.UK namespace reserved for academic usage. 4 More hierarchical layers like mail.SLD.TLD or www.SLD.TLD can be added if needed.

METHODOLOGY

cic TLDs allow registration only by citizens of their countries. No costs but a modest registration fee to the TLD registry are incurred. As a result of this liberal rst-come-rst-serve approach, the number of domain names grew rapidly in the last decade. Figure 1 visualizes the growth in the total number of registered .COM/.NET/.ORG domains since 1998. == insert Figure 1 about here == A domain must be renewed in periodic intervals, again incurring a fee. Domains that do not get renewed become available for registration again. About 27 percent of .COM/.NET domains were not continued in 2010 (Verisign, 2011) but handed in again. As long as the renewal fees are taken care of, the domain owner enjoys full ownership benets, including the possibility to rent out the domain or to sell it at the secondary market.

2.3

Secondary Market for Internet Domain Names

Domain names are actively traded in secondary markets. Estimating the size of these markets is difcult as no central registry explicitly keeps track of domain sales. The Internet Corporation For Assigned Names and Numbers (ICANN) reports 8 million transfers of .COM/.NET domains for January 2011, which is the upper bound of total sales as this number contains many not-for-sale transfers. The number of domain transfers across whole-sale registration companies (registrars5 ) could serve as a better proxy for domain transactions. In January 2011, 0.65 million .COM/.NET domains were transferred from one registrar to another (ICANN Internet Corporation For Assigned Names and Numbers, 2011). Domains are privately traded and no central market place exists. No trading licenses or other formal requirements restrict access to the market. Several trading platforms oer marketing and transaction handling services to buyers and sellers.

Methodology

Domain names are unique: By denition, it is not possible to have the same combination of top level domain (TLD) and second level domain (SLD) twice. When estimating the value of a domain name, one cannot draw a direct comparison to recent transactions of exactly the same (or at least very similar domains), as it is common for pricing standardized goods like stocks or bonds. Taking the average of transaction prices gives a rst indication of price developments in the market for domains but does not control for quality dierences in the domains sold. If there is a time
5 A registrar is a company facilitating the domain registration process on a whole-sale basis, as retail customers cannot register

domains with the registry directly.

METHODOLOGY

period in which many high-quality domains are sold, the average transaction price will increase, regardless of any true trend in prices. Median based indices are less sensitive to extreme values but suer from the same systematic shortcoming of disregarding the characteristics of the underlying transactions. Putting it dierently, one is comparing apples to oranges.

3.1

Hedonic Regression

Two general methodologies are commonly used to price non-standardized assets like real estate or art. The rst one, the so-called hedonic regression analysis, explains the price of an asset by a set of quality variables that describe the characteristics (hedonics) of that asset. For example when investigating the sales prices of cars, hedonic characteristics one might think of are the mileage of the car, the year of production, the manufacturer etc. Dierences in quality will be captured in the regression coecients of the hedonic variables while a general price trend can be separated from potential changes in the underlying quality of the traded assets. The overall explanatory power of this approach depends on how well the hedonic variables capture the price-relevant characteristics of the asset. Therefore, this method is only feasible, if sucient characteristics on the individual observations are known.

3.2

Repeat Sales and Hedonic Repeat Sales Regression

Alternatively, the repeat sales methodology (Bailey et al. , 1963) controls for quality by tracing individual assets in time, comparing each transaction to previous transactions of the very same asset. It assumes that the quality of the asset does not change between the two sales. While this is probably the most direct form of making sure to compare apples to apples, it disregards all information on transactions that are sold only once. In practice, this approach is suitable for samples of non-standardized goods that provide sucient numbers of repeat sales. This paper follows the Hedonic Repeat Sales (HRS) methodology, rst suggested by Shiller (1993). HRS combines the advantages of having individual dummies for each SLD (which are notoriously suering from omitted variables in a pure-bred hedonic framework) and a hedonic classication of the highly standardized TLD. Including the hedonic characterization extends the number of observations entering the regression signicantly. To give an example: A repeat sales regression only considers repeat transactions of exactly the same domain (like XYZ.COM) while a hedonic repeat sales approach include single sales that share the SLD but have dierent TLDs (like XYZ.COM and XYZ.NET). Conceptually, the sales price P of domain i is split into three components: Pi,t , = SLDi + T LDi + Dt , (1)

METHODOLOGY

where SLD and T LD capture the quality of each SLD and TLD. Time dynamics enter the equation through Dt . Dierencing transactions that share the same SLD cancels out the SLD component, which leads to an equation that can be empirically tested in a regression: ln(Pi,t2 ) ln(Pj,t1 ) =
k=K k=1

k T LDCombik,i,j +

t=T t=2

t Dt,i,j + t,i,j ,

(2)

where the dierence in the natural logarithm of prices from transaction i and j is the dependent variable, explained by a set of time dummies Dt and TLD combination dummies. For each pairwise combination k of TLDs in our sample, T LDCombik is dened to be one if T LDi equals the rst element and T LDj equals the second element of k, and zero otherwise. K is the squared number of uniques TLDs in the sample. The time dummy Dt is dened in a Bryan and Colwell (1982) way, as described in Geltner (1997). For each time period between t1 and t2, Dt is set to 1. When only parts of a given time period fall between the two sales dates, the value for Dt is scaled down accordingly. If t1, for instance, is June 30, 2009, the annual dummy for 2009 will be set to 0.5, for November 31st it will be adjusted to 1/12. The default value is 0. This methodology provides end-of-time-period return estimates and avoids averaging within each time period (Geltner, 1997). The k and t are regression coecients. The relative price dierences between TLDs are kept constant in time. For each pair of TLDs (a, b), the coecients k=(a,b) are restricted to be k=(b,a) . The error term t,i,j is assumed to be independently and identical distributed. The goal of this paper is the estimation of an index that can be updated frequently and that still provides reliable index gures not subject to excess volatility. When estimating a high frequency index, naturally the number of observations per time period will be low. This causes the resulting index to be sensitive to noise, resulting in excessively volatile index estimates with low signal/noise ratios. Goetzmann (1993) further shows that repeat sales regressions suer from spurious negative autocorrelation in the estimated return series and an excess return volatility, especially at the beginning and at the end of the series where the data thin out. Imposing a time structure on the time coecients t reduces the eect of transaction price noise in thin markets. He therefore suggests a Bayesian shrinkage techniques where the log levels are modeled as a random walk with drift process. Francke (2010) generalizes this approach and extends the random walk with drift model to a structural time series model. This paper follows a novel frequency conversion approach suggested by Bokhari and Geltner (2010). In a rst step, lower frequency indices are estimated staggered in time. For the domain name index, I estimate twelve annual indices based on (2), each starting in a dierent month.6 Applying a general6 Case

and Shiller (1987) expect the variation in the error terms of a repeat sales regression to increase in the time between sales. Any heteroscedasticity in the error terms is a violation of the assumptions underlying OLS regressions. They therefore suggest to rst run an auxiliary regression, following the methodology of (Bailey et al. , 1963). In a second step, they regress the

DATA

ized inverse estimator converts these annual indices to a monthly return series. The Bokhari-Geltner procedure relies on sympathetically few assumptions as a structure on the time coecients does not need to be formalized. It reduces noise in the index estimates without introducing a time lag. Bokhari and Geltner (2010) provide a very detailed step-by-step explanation of the methodology.

Data

This study relies on a dataset of real domain transactions traded on the trading platform Sedo.com, which is the largest domain market place in the world based on completed transactions. The sample comprises of 200,170 transactions for the period from January 2005 through June 2011. Figure 2 presents the number of observations per quarter. The sample is not uniformly distributed in time, with more transactions taking place in later quarters. The growth rate of numbers of sales in the sample exceeds the growth rate in total registered domains, suggesting that the sales sample covers a larger market share in later quarters. == insert Figure 2 about here == Completed transactions are a small subset of all domains oered for sale. In 2010, about 14 million domains were listed at Sedo.com in contrast to just 44,000 sales. At this point, one can only speculate why the sales rate is so low. Domains that were sold could be of higher quality than the shelf warmers. Alternatively, the low initial registration fees of just a few dollars per domains and the potentially high returns in case of a successful sale could warrant a lottery-like market where domains were registered and put up for sale on a large scale.7 This perfectly rational calculation of expected values is contrasted by a more behavioral explanation of too optimistic sellers. They over-estimate the probability of a sale at an extremely high price and reject bids that are perceived as too low. If that explanation was true, I would expect dierences in domain turnover across seller types, as professional domain traders are expected to be less prone to the big-catch fallacy.8 Finally, Sedo.coms fee structure is likely to induce substantial cheating as sellers rst use the platform for marketing purposes but close the deal privately in order to avoid the fees due when domains change hands through the platforms ocial channels. Future research is needed to verify which factors determine a successful sale.
error terms derived from the rst step and regress them on the time between sales. Based of tted values from this regression, they construct weights for a nal re-estimation of step one by Generalized Least Squares (GLS). The annual indices do not exhibit signicant coecients in the second step, which indicates that the Case/Shiller correction is not needed for this data. 7 A back-of-an-envelope calculation illustrates that domain trading is protable at the aggregate level: Multiplying the median transaction price in 2010 (500 USD) by the number of transactions (44,000) gives a total turnover that is several times bigger the sum of all registration and renewal fees (6-10 USD per year and domain). 8 A reliable domain name index could help to close the gap in bid and ask prices, leading to more successful transactions and higher liquidity in the market.

DATA

In May 2011, 28 percent of all sales are closed at a xed price set by the seller in advance while for 42 percent of trades the price is set in bilateral price negotiations, where the buyer has to give a rst price quote. 13 percent of transactions are initiated by professional brokers aliated with Sedo.com. 17 percent of sales were closed in auctions. The selling mechanism is not a true hedonic characteristic since it is not necessarily linked to the quality of the domain name. Nevertheless, including this information into the hedonic framework is expected to improve the model t: Bulow and Klemperer (1996) show that auctions lead to higher transaction prices than negotiations with one less bidder. All transactions that include a website or other content beside the naked domain name are excluded from the sample, since disentangling the value of the structure from the land is impossible. The sample does not contain any information on buyers and sellers.

4.1

Summary statistics

The median transaction price is 500 USD. The sales price distribution has an enormous right tail, with 12 domains exceeding one million USD. The biggest sh in the sample is sex.com which has been sold for 13 million USD in November 2010. The average SLD consists out of 10 characters, 2.1 percent of the samples SLDs contain diacritic characters or non-ascii characters, 6.2 percent consist of at least one digit, while 9.6 percent are split by dashes. == insert Figure 3 about here == The most senior .COM/.NET domain in the sample was registered in early 1989. The vintage of the domains slowly deteriorates throughout the years. In 2005, the average year of rst registration was 2003 compared to 2005 in 2010. == insert Table 1 about here == Domains that are sold at least twice (repeat sales) are on average more expensive than the universe of all sold domains. This dierence is statistically signicant. The number of characters in the domain or number of words contained in the domain do not dier in the subsample. In total, 156 dierent TLDs are present in the database. Figure 4 compares the share of the 10 most frequent TLDs within the sample with their share in the universe of registered domains. COM accounts for almost half of all domains, followed by DE, NET, ORG, and UK. The weights in the sample do not divert much from the true weights for most TLDs. The country-code top level domain DE and the European TLD EU, however, are over-represented, probably due to Sedos strong roots in the German market. Less popular domains that are subsumed under the label other are less prominent in the sales sample. == insert Figure 4 about here ==

RESULTS

Results

Figure 5 shows 12 annual indices each starting at a dierent month of the year. The solid black line represents the resulting aggregate price trend at a monthly frequency, which shows a general robust upward slope in domain prices. Overall, the domain price index increased by 83 percent since January 2006. The average of historic returns is one percent per month, with a standard deviation of 3.2 percent. In the last 6 years returns ranged from -6.3 percent up to 14.4 percent per month. Internet domain names gain in value in 2006 through 2007, with prices peaking in November 2007 (+76 percent) before falling by 34 percent in the subsequent 5 quarters. Since then, domain prices regained their strength, climbing to an all-time high in May 2011 (Figure 6). This pronounced cycle demonstrates that domain names oer attractive investment opportunities to the skilled (or lucky) investor that can identify boom and bust phases in advance. This study does not investigate any predictability in domain prices, so the question of market eciency which rules out predictable and thereby risk-free investment opportunities is not answered yet. Table 5 presents the Internet Domain Name Index (IDNX) estimates on a monthly frequency for the years 2006 through 2011. == insert Figure 5 about here == The results, however, clearly show that domain names are risky investments the observed bust phase in 2008 wiped out a large share of market value before bouncing back. In addition, liquidity in the market dries up just when it is needed most. Investors trying to liquidate their domain name holdings in 2008 had to nd a buyer in relatively thin markets, indicated by reduced total transaction numbers during the bust period (Figure 2). The regression coecients conrm a domain traders mantra: COM is the most valuable TLD. Table 2 shows pairwise price dierences for identical SLDs under the 10 most frequently traded TLDs. DE domains trade for 24 percent less than COM domains (for identical SLDs), the discount for NET is 73 percent, ORG is worth less than a quarter of COM, BIZ less than a tenth (please refer to Table 2 for all TLD combinatios). == insert Table 2 about here == == insert Figure 5 about here == == insert Figure 6 about here == Domain prices have an economic foundation. They are not detached from the economy in general. On the contrary, Figure 6 suggests that domain price changes are very similar to changes in the NASDAQ 100 index, which covers the 100 largest technology companies listed on the NASDAQ stock market. The estimated correlation on monthly changes in both indices is 0.25, while the monthly correlation estimate in levels is 0.79. When the internet economy is expanding, domain names as a production factor are in high demand as well. 9

CONCLUSION

== insert Table 3 about here == The most important source of income for the majority of internet enterprises is revenue from online advertisements. This dependency is visible in the correlation coecient of 0.76 between the domain names index to a time series on the total revenues from online advertisement for the US, which exceeds the estimated correlation coecients for NASDAQ 100 (0.29) index and Googles share prices (0.38) versus advertisement revenues. The correlation of quarterly changes in online ad revenues and IDNX is 0.34, while it is negative for the NASDAQ 100 share price index. == insert Table 4 about here == From a theoretical point of view, prices for internet domain names should be the discounted future cash ows that can be generated by owning this domain. Domain prices are therefore forward looking, giving a indication of not only the current prots but include expectations about future opportunities as well. These expectations cover small and large internet ventures as long as they share similar business models. This suggests that the domain name index can serve as a fever curve for the well-being of the internet economy, considering also small enterprises that are currently excluded by stock-price based indices.

Conclusion

This paper estimates the rst constant quality price index for internet domain names. The aggregate price path reveals the fundamental risk/return prole for domain names as an investment class. The quality of the domain names is controlled for in a hedonic repeat sales regression framework. The strong correlations of domain prices with the high tech economy and online advertisement revenues are re-assuring in two ways. First, they show that domain name buyers and sellers make economically motivated price decisions. Domain markets are not a cloud-cuckoo-land where dreamers trade esoteric goods at imaginary prices. Second, nding a link that is expected to be present based on economic reasoning conrms that the domain name index makes sense. Summing it up, the domain name index put forward in this paper can serve as a benchmark for domain name traders and investors looking for information on price trends, returns and fundamental risk of internet domain names. It thereby increases transparency in the market for this newly emerged asset class. == insert Table 5 about here ==

10

REFERENCES

References
Alonso, W. (1964). Location and land use: toward a general theory of land rent. Harvard University Press. Bailey, M. , Muth, R. , and Nourse, H. (1963). A regression method for real estate price index construction. Journal of the American Statistical Association, 58(304):933942. Bokhari, S. and Geltner, D. (2010). Estimating real estate price movements for high frequency tradable indexes in a scarce data environment. The Journal of Real Estate Finance and Economics, pages 122. Bryan, T. and Colwell, P. (1982). Housing Price Indices. In Sirmans, C. F. , editor, Research in Real Estate, volume 2. JAI Press. Bulow, J. and Klemperer, P. (1996). Auctions versus negotiations. The American Economic Review, 86(1):180194. Burshtein, S. (2005). Is a domain name property? Journal of Intellectual Property Law & Practice, 1(1):59. Case, K. E. and Shiller, R. J. (1987). Prices of single-family homes since 1970: New indexes for four cities. New Eng. Econ. Rev., September/October:4556. Denic (2011). Available from: http://www.denic.de/ hintergrund/statistiken/internationale-domainstatistik.html [cited 2011-06-03]. Domainzahlenvergleich international.

Francke, M. (2010). Repeat Sales Index for Thin Markets: A Structural Time Series Approach. The Journal of Real Estate Finance and Economics, 41(1):2452. Geltner, D. (1997). Bias and precision of estimates of housing investment risk based on repeat-sales indices: A simulation analysis. The Journal of Real Estate Finance and Economics, 14(1):155171. Goetzmann, W. (1992). The accuracy of real estate indices: Repeat sale estimators. The Journal of Real Estate Finance and Economics, 5(1):553. Goetzmann, W. (1993). Accounting for taste: Art and the nancial markets over three centuries. The American Economic Review, 83(5):13701376. IAB (2011). IAB Internet Advertising Revenue Report. Available from:
pdf [cited 2010-5-17]. http://www.pwc.com/en_US/

us/industry/entertainment-media/assets/IAB-internet-advertising-revenue-report-2010.

ICANN Internet Corporation For Assigned Names and Numbers. Registry Operators Monthly Report .COM/.NET January 2011 [online]. (2011). Available from: http://www.icann.org/en/tlds/
monthly-reports/.

11

REFERENCES

Mills, E. (1972). Studies in the Structure of the Urban Economy. The Johns Hopkins Press, Baltimore, Maryland. Muth, R. (1969). Cities and housing: the spatial pattern of urban residential land use. University of Chicago Press. Schumacher, T. , Ernstschneider, T. , and Wiehager, A. (2002). Domain-Namen im Internet. Ein Wegweiser fr Namensstrategien. Springer: Berlin. Sedo.com (2011). Available from: http://www.sedo.com/uk/resources/ market-trends/?tracked=&partnerid=&language=e [cited 2011-5-8]. Market trends.

Shiller, R. J. (1993). Measuring Asset Values for Cash Settlement in Derivative Markets: Hedonic Repeated Measures Indices and Perpetual Futures. The Journal of Finance, 48(3). Verisign. The Domain Name Industry Brief / February 2011 [online]. (2011) [cited 2011-4-26]. Available from: http://www.verisigninc.com/assets/domain-name-report-feb-2011.pdf.

12

TABLES AND FIGURES

Tables and Figures

Figure 1: Number of registered domain names for selected Top Level Domains

number of registered domains (in millions)

.com .net .org

20

40

60

80

2000

2002

2004 time

2006

2008

2010

Notes: A heated race for the best domain names sprung up since the rst domain name in history, symbolics.com, was created in March 1985. By now more than 200 million unique domain names are registered (Verisign, 2011) with no end of growth in total numbers in sight. Data from Zooknic (http://www.zooknic.com).

13

TABLES AND FIGURES

Table 1: Summary statistics

1st Qu. all sales (N = 188,432) price # characters in domain # words in domain repeat sales (N = 10,581) price # characters in domain # words in domain 232 7 1 295 5 1 Median 500 9 1 701 8 1 Mean 2,199 10 1.4 3,538 9 1.3 3rd Qu. 1,287 13 2 2,000 11 2 Max. 13,000,000 70 14 1,600,000 32 14

Notes: Repeat sales are on average more expensive than the universe of all sold domains. This dierence is statistically significant. The number of characters in the domain or number of words contained in the domain do not dier in the subsample.

Figure 2: Distribution of sales per quarter (2005Q12011Q2)

12000 all transactions repeat sales

Frequency

0 20051 20053 20061 20063 20071 20073 20081 20083 20091 20093 20101 20103 20111
Notes: The sample is not uniformly distributed in time, with more transactions taking place in later quarters. The growth rate of numbers of sales in the sample exceeds the growth rate in total registered domains, suggesting that the sales sample covers a larger market share in later quarters. Liquidity in the market dries up just when it is needed most. Investors trying to liquidate their domain name holdings in 2008 had to nd buyers in relatively thin markets, indicated by reduced total transaction numbers during this period of declining domain prices. Data for 2011Q2 not complete yet.

2000

4000

6000

8000

10000

14

TABLES AND FIGURES

Figure 3: Distribution of transaction prices

40000 Frequency 0 0 10000 20000 30000

1000

2000

3000 Price (USD)

4000

5000

6000

Notes: The top 5 percent of transaction prices is not shown. The maximum transaction price in the databases is 13 million USD.

Table 2: Pairwise price comparison of selected Top Level Domains (TLD), in %

com de net eu co.uk info org mobi es biz -24 -73 -74 -69 -87 -79 -93 -75 -91 de -52 -66 55 -76 -68 -65 -59 -80 net eu co.uk info org mobi es

-34 76 -59 -35 -79 -46 -79

88 -27 -5 -57 -1 -64

-82 -61 -85 -69 -81

72 -28 -3 -55

-58 -42 -65

53 -10

-35

Notes: The matrix shows the estimated price dierences (in percent) across TLDs based on the regression coecients from Table ??. A .DE domain, for instance, is estimated to be 24 percent more aordable than the .COM equivalent, controlling for time trends and the Second Level Domain (SLD). In the current econometric setup, the TLD dierences are assumed to be constant in time. Furthermore, the language of the SLD is not considered. This might lead to lower estimated discounts for ccTLDs from non-English speaking countries vs. COM whenever the SLD carries some meaning in the local language.

15

TABLES AND FIGURES

Figure 4: Distribution of Top Level Domains (TLD)

all domains sales in sample

Share in %

10

20

30

40

com

de

net

eu

uk

info

org

mobi

es

biz

other

Top Level Domain

Notes: The red bars represent the weight of all domains under a TLD versus the total universe of registered domains. The blue bars show the weight of a TLD in the sample. In total, 156 dierent TLDs are present in the database (top 10 shown above). The country-code top level domain .DE is over-represented, probably due to SEDOs strong roots in the German market. Data on total domain counts from Denic (2011).

Table 3: Correlations of Monthly Changes in IDNX, NASDAQ 100, and Google Shares
IDNX IDNX NASDAQ 100 Google Shares 1 0.25 0.22 NASDAQ 100 1 0.72 Google Shares

16

TABLES AND FIGURES

Figure 5: Staggered annual indices for all TLD and resulting monthly index

annual indices (1-12) monthly index (converted from annual) 200 2006 = 100% 100 2006 150

2007

2008

2009 Year

2010

2011

Notes: Twelve annual indices based on (2) are estimated, each starting in a dierent month. Applying a generalized inverse estimator converts these annual indices to a monthly return series (black line).

Table 4: Correlations of Quarterly Changes in IDNX, NASDAQ 100, Google Shares, and US Online Advertisement Revenues
IDNX IDNX NASDAQ 100 Google Shares Ad Revenues 1 0.40 0.45 0.34 NASDAQ 100 1 0.80 -0.30 Google Shares Ad Revenues

1 -0.21

17

TABLES AND FIGURES

Figure 6: Price dynamics of internet domain names compared to NASDAQ 100

IDNX NASDAQ 100 US online advertising revenues

Dommain prices (2006-1 = 100)

80 2006-1

100

120

140

160

180

200

2007-1

2008-1

2009-1 Time

2010-1

2011-1

Notes: Domain prices have an economic foundation. This gure suggests that domain prices changes are very similar to changes in the NASDAQ 100 index, which covers the 100 largest technology companies listed on the NASDAQ stock market. The estimated correlation on monthly changes in both indices is 0.26, while the monthly correlation estimate in levels is 0.79. When the internet economy is expanding, domain names as a production factor are in high demand as well. The most important source of income for the majority of internet enterprises is revenue from online advertisements (time series based on IAB, 2011). This dependency is visible in the correlation coecient of 0.76 between the domain names index to a time series on the total revenues from online advertisement for the US. The NASDAQ 100 index is rescaled for a better visualization. Scaling it back to 100 in 2006 will not aect any of the correlation estimates.

18

TABLES AND FIGURES

Table 5: Internet Domain Name Index (IDNX) 20062011

Month 1/1/06 1/31/06 2/28/06 3/31/06 4/30/06 5/31/06 6/30/06 7/31/06 8/31/06 9/30/06 10/31/06 11/30/06 12/31/06 1/31/07 2/28/07 3/31/07 4/30/07 5/31/07 6/30/07 7/31/07 8/31/07 9/30/07 10/31/07 11/30/07 12/31/07 1/31/08 2/29/08 3/31/08 4/30/08 5/31/08 6/30/08 7/31/08 8/31/08 9/30/08 10/31/08 11/30/08 12/31/08 1/31/09 2/28/09 3/31/09 4/30/09 5/31/09 6/30/09 7/31/09 8/31/09 9/30/09 10/31/09 11/30/09 12/31/09 IDNX 100 101.7 103.3 103.2 104.8 106.2 108.1 105.5 108.6 109.4 111.4 111.8 127.9 131.8 135.1 139.7 143.6 149.2 157.3 157.7 163.1 168.6 173.5 176.4 168.5 168.3 167.3 164.6 160.1 155.1 151.3 143.7 138 133.2 128.4 125.6 121.5 118.9 118.4 116.8 116.7 117.1 120.2 128.7 133.8 136.7 140.3 142 142.5 Month 1/31/10 2/28/10 3/31/10 4/30/10 5/31/10 6/30/10 7/31/10 8/31/10 9/30/10 10/31/10 11/30/10 12/31/10 1/31/11 2/28/11 3/31/11 4/30/11 5/31/11 6/30/11 IDNX 144 146.6 147.5 151 152.2 159 167.8 166.2 170.2 172.8 179.8 178.2 181.2 181.5 189.2 191.7 195.6 183.2

Notes: All index gures can be downloaded at http://www.idnx.com/idnx.csv

19