733 15th Street, NW . Suite 700 . Washington, DC 20005 . Tel.: 202-393-1348 . Fax: 202-393-1349

www.capitalresearchers.com

HOME

 

WEBSITE TRAFFIC ("EYEBALLS"), FINANCIAL PERFORMANCE,

AND INTERNET STOCK PRICES: A SYNTHESIS OF THE EMERGING

EMPIRICAL EVIDENCE

By Michael J. Isimbabi, Ph.D.

Principal

misimbabi@capitalresearchers.com

This Draft: December 2000

Comments Welcome

I. Introduction and Summary

A popular area of financial research is the analysis of the relation between firms’ stock prices or returns and economic, industry, and firm-specific factors. A sub-area of this type of research seeks to determine the extent to which various financial and non-financial firm characteristics are relevant in equity valuation. Obviously, the determination of such "value-relevance" is of particular interest with respect to Internet firms—until the Spring 2000 shakeout, the stock prices of most Internet firms skyrocketed to "bubble" levels even though most were unprofitable or only marginally profitable. Of course, the relatively high valuations – and over-valuations in some cases -- of Internet stocks have largely been based on the expected high growth and performance of the sector.

This paper synthesizes the findings of empirical studies that examine the relation between Internet stock prices and financial and nonfinancial variables. [The studies and their URLs are listed in the Appendix.] In particular, in light of the lack of profitability of most Internet firms, the studies examine the value-relevance of nonfinancial factors – especially Website Traffic ("eyeballs") and its underlying determinants – relative to the conventional financial "value-drivers" such as earnings, book value of equity, cashflow, etc.

This review is written for a general business audience and, therefore, to the extent possible, avoids unnecessary technical jargon.

In general, the studies follow the conventional methodological approach of undertaking multivariate regressions to analyze cross-sectional differences in valuation among Internet firms. Typically, firms' stock prices or market values (or variants such as market-to-book ratios, market value/asset ratios, stock price/sales ratios, and stock returns) -- the dependent variable -- are regressed on financial and nonfinancial variables (earnings, book value, Website Traffic, etc. and corresponding variants) as explanatory variables. However, the studies’ regression model specifications, sample firms, data sources, and variable proxies differ, which makes direct comparisons of results difficult and may partly explain some conflicting results. (Brief descriptions of data samples and sources are provided in the Appendix.)

Furthermore, because the Internet sector is relatively young, the empirical studies suffer from data limitations, inadequacy of the proxies for some factors, etc. For example, there are questions about the accuracy of web traffic data, and much of the traffic data available does not include corporate Web-users. Thus, as is usually the case with pioneering work, the findings of the studies reviewed here are preliminary and inconclusive. As the industry matures, additional research that builds on these studies using larger and "cleaner" datasets and more refined methodologies will, of course, provide better insights on the subject.

Notwithstanding the foregoing caveat, the findings of the studies largely support some of the prevailing beliefs about the factors that investors have taken into account in their valuation of Internet firms. However, it should be noted that all but one of the studies cover only the period prior to the Spring 2000 Internet shakeout. The only study that examines the post-bubble period reports some changes in investor perceptions and valuation approaches after the bubble burst.

The following is a summary of the key findings (the reader should note that the studies examine different combinations of financial and nonfinancial variables and, therefore, a particular finding may be that of only one or two studies in certain cases):

• Consistent with the conventional wisdom, the studies' findings indicate that investors did not place much emphasis on bottom-line earnings in their valuation of Internet firms. Some of the studies find a negative relation between net income and Internet stock prices, notably for some groups of B2C and negative-income firms. That is, it appears that the larger an Internet firm’s losses, the higher the value that investors accorded the firm’s stock. This is apparently due to the high valuation placed on large expenditures by Internet firms for technology acquisition, product development, and customer acquisition, which were considered critical to success. Investors evidently believed that such expenditures, while immediately expensed in financial statements (therefore resulting in little or negative income), were intangible investments that would generate high profitability in the future. In this regard, some of the studies found a strongly positive relation between Internet stock prices and research and development (R&D) and sales and marketing expenses. However, this appears to have been the case only during the Internet bubble period, i.e., prior to the Spring 2000 shakeout. The only study that examined data for the "post-bubble" period [Demers and Lev (D&L), 2000] reports evidence indicating that, after the bubble burst, investors apparently became more skeptical about the usefulness of the expenditures and no longer ascribed high valuations to them.

• Sales Revenue was found to be value-relevant after controlling for other financial and non-financial factors. This finding is consistent with reports that, absent earnings, some investors use Sales Revenue as an indicator of Internet firms’ performance and prospects. However, Sales Revenue does not appear to have been a dominant factor relative to the other financial and nonfinancial variables.

• In general, the studies find that various measures of "Website Traffic" were highly value-relevant for Internet stocks, particularly those of B2C firms. [The Traffic measures differ across studies but each used variants of one or more of the following: (i) the number of unique visitors to a Web site as a percentage of the Web-user population; (ii) the number of pages viewed by visitors; (iii) time spent at a site per visit; and (iv) the number of visits to the site per unique visitor per period. The Appendix provides definitions of the Traffic measures used in each study.] However, there are differences across the studies with respect to the degrees to which each of the measures was value-relevant. This may be due to differences in the computations of the measures, data manipulation techniques, and the types of firms in the samples. Some of the studies find that Website Traffic measures had stronger explanatory power for market values than financial variables such as net income, book value, and revenues.

• Website Traffic measures remained value-relevant for Internet stocks even during the post-bubble period in 2000, which contradicts some analysts' claims that Website Traffic was no longer an important factor in the valuation of Internet firms after the bubble burst. [D&L finding]

• The market evidently adjudged that the much-hyped strategic alliances of business-to-consumer (B2C) firms with top traffic-generating sites did not provide the anticipated benefits to Internet firms. [D&L]

• Internet firms’ ability to sustain their current rate of "cash burn" was found to be value-relevant. This supports the view that the market had a negative view of unprofitable Internet firms that were quickly depleting their cash holdings. [D&L]

• The only study that compares e-tailers and portal/content/community (P/C) firms (Trueman et al) finds that there were differences in the extent to which investors considered some factors to be value-relevant across the two groups of B2C firms. For example, the association between Internet stock prices and net income was negative for the e-tailers but positive for the P/C firms. Also, for the P/C firms, the incremental explanatory powers of "Pageviews" and of "Unique Visitors" were about the same, while, for the e-tailers, the incremental explanatory power of "Pageviews" was much greater than that of "Unique Visitors."

II. Financial and Non-Financial Variables and Value-Relevance

In general, previous studies of firms in established industries have found a significant relation between stock prices and financial variables such as earnings, book values, and cash flows. However, some recent studies have also found that the degrees of value-relevance of these financial variables have changed over time. In particular, there is evidence that there has been a significant decline in the value-relevance of earnings in recent years.¹

One explanation advanced for this decline is that investors have become increasingly less dependent on financial statement information for valuation because of the rapid change and uncertainty – as a result of innovation, deregulation, and increased competition – that have characterized the business and economic environment over the last two decades. This is particularly so in new, rapidly growing, or fast-changing sectors such as information technology and telecommunications. For such sectors, because "intangible investments" in areas critical to future performance -- such as research and development (R&D) -- are immediately expensed rather than capitalized, accounting information such as net income becomes less useful in valuation. Empirical evidence also indicates that an increased propensity of firms to report transitory income or nonrecurring items (which often tend to result in losses) over time has made the market place less importance on bottom-line earnings.

The nature of the value-relevance of financial variables varies across types of firms and industries. In this regard, there is empirical evidence that, in certain sectors, nonfinancial information may have greater value-relevance than financial information such as earnings and book value. For example, in the telecommunications sector, population size may be an indicator of growth potential for a firm that has been granted a license to provide services in a geographic area, and would therefore be more value-relevant than current net income. Similarly, the granting of a major patent to a pharmaceutical company portends expected future performance that may make this factor more value-relevant compared to accounting information.

In the Internet sector, high initial expenditures for technology acquisition, product development, and customer acquisition have been considered critical to B2C firms' ability to achieve critical mass and attain competitiveness and leadership in their markets. Thus, in such an "intangible-intensive" sector, these (uncapitalized) expenses that are expected to generate future benefits may have greater value-relevance than earnings and book value. For B2C firms in particular, such large expenditures are directed largely at generating traffic to Web sites. The studies reviewed next therefore examine the relative value-relevance of financial information and nonfinancial factors (particularly Website Traffic).

III. Financial Performance and Internet Stock Prices

Following the conventional approach of prior research on firms in other industries, the studies examine the relation between Internet stock prices and financial statement information (earnings, book value of equity, expenses, etc.) and then the additional explanatory power of nonfinancial variables.

Trueman et al (TWZ) find a negative and statistically significant relation between net income and Internet stock prices for a sample consisting of only firms that they judged to be primarily portals, content/community providers, or e-tailers (i.e., B2C firms). Upon separating their sample into e-tailers (revenues from product purchases by site visitors) and the portal and content/community (P/C) firms (revenues from advertising), the negative association observed for the whole sample is observed only for the e-tailers, while a positive association is observed for the P&C firms. Hand(A) reports a similar negative association, but only for firms with negative net income: "larger losses cross-sectionally correlate with higher, not lower, market values." For firms with positive net income, the relation is positive and significant. Rajgopal et al (RKV) also find a negative association between net income and stock prices but the relation is statistically insignificant. [As discussed shortly, these different and seemingly conflicting results may be due to the differences in the data samples used by the studies. For instance, TWZ and D&L’s samples consist of only B2C firms, while RKV and Hand’s samples consist of both B2C and business-to-business (B2B) firms (see Appendix).] Consistent with most previous studies on other industries, TWZ, RKV, and Hand(A) all report that, beyond net income, book value of equity is positively related to, and has significant incremental explanatory power for, Internet stock prices.

The observed negative correlation between net income and market values is consistent with the view that, at least during the period covered by these studies, the market ascribed high valuations to Internet firms that recorded large losses on the basis that such losses were due to high expenditures on R&D and sales and marketing, which, though expensed, are intangible investments expected to generate high sales and profitability in the future.

Absent earnings, some market participants use Internet firms’ sales revenue figures as an indicator of performance and prospects. This is based on the notion that revenues reflect a firm’s market share and market acceptance. In this context, some analysts and investors use the price-to-sales ratio in lieu of the price-earnings ratio in valuation, since the latter ratio is not applicable for most Internet firms. If investors use revenues as a key factor in valuation, then Sales Revenue should be value-relevant in the regressions.

To examine the implications of expenses and revenues for value-relevance, some of the studies "decompose" net income into the various income statement items – revenues, gross profits, R&D expenses, sales and marketing expenses -- in their regressions.

Hand(A) finds revenues to be only "weakly positively priced", after controlling for book value of equity and expenses, and therefore rejects the hypothesis that revenues dominate the pricing of Internet stocks. RKV find that Sales Revenue is positively and significantly related to market value after accounting for book value, earnings, and Website Traffic (discussed in the next section).

TWZ report that the significance of book value disappeared when net income was replaced by its income statement components -- Gross Profit, Sales & Marketing Expenses (SME), R&D Expenses, and Other Expenses -- in their regression model. The model's explanatory power increased with this decomposition. Also, Gross Profit has a positive and significant association with market value, in contrast to the negative relation observed for net income. When Gross Profit is further decomposed into Revenues and Cost of Revenues, TWZ find that the relations between market values and the two variables are positive and negative respectively, as would be expected. The decomposition had no effect on the significance of the other financial statement variables.

TWZ opine that the positive association between market value and Gross Profit is due to the fact that the latter "is viewed by investors as more permanent in nature, and as a less noisy measure of current operating performance than is bottom-line net income…(which may include) large transitory items (such as merger-related costs), upon which investors likely place less weight in valuation, as well as items that might be considered in some firms to be investments rather than expenses (such as SME and R&D)." However, they find neither R&D nor SME to be significantly associated with market values, either in a positive or a negative direction. They opine that this is because either (i) investors considered these expenses to be of little use in valuing Internet firms or (ii) investors viewed them as normal expenses for some firms and as investments for others.

The other two studies that examine the role of expenses report findings that are more consistent with the view that investors largely considered these expenses to be investments for most firms, at least until the Spring 2000 shakeout.

Hand (A) finds that, for the sub-sample of firms with negative net income, market values were positively related to both SME and R&D expenditures: "when marketing costs are large enough to lead to reported losses, they are viewed by the market as intangible assets, not period expenses…Contrary to their immediate expensing under GAAP, large R&D costs are also priced by the market as if they are intangible assets, not period expenses."

D&L’s study, the most recent of the five studies covered in this review and the only one that examines data for the post-bubble period (i.e., the first half of 2000), compares findings for the bubble period (1999) to those of the "post-bubble" period (2000).

In their regressions, D&L utilize the price-revenue ratio concept by examining the relation between firms’ price/sales ratios and income statement expense variables -- Cost of Goods Sold, Marketing and Advertising Expenses, and R&D/Product Development Expenses – as well as measures of Website Traffic. D&L justify their use of price/sales ratios rather than market-to-book ratios as follows:

"First, price-to-sales is the financial metric that is most commonly referred to by analysts and the business press in their evaluations and discussions of Internet companies. In this sector, price-to-sales takes on the role that the price-to-earnings ratio has traditionally held in the valuation of going concern entities because most Internet companies are not yet profitable (and therefore P/E cannot be sensibly applied). A similar argument applies to the book value of equity in this sector. Book values are depressed because Internet companies have few tangible assets and their massive expenditures on the all-important intangible assets are generally expensed rather than capitalized. The market-to-book ratio therefore does not have the same economic interpretation and intuitive appeal as it does in the cross-section of more established and profitable firms. Furthermore, from a statistical perspective, the market-to-book ratios tend to "blow up" because of this small denominator problem."

Consistent with the other studies, D&L find advertising and marketing expenses and R&D/product development expenses to be significantly and positively related to price-sales (P/S) ratios in 1999. However, Cost of Goods/Services Sold (CGS) was not significant. In 2000 (the post-bubble period), CGS was negatively and significantly associated with the P/S ratio, as would be expected, while marketing and product development expenses were no longer significant, though the association remained positive.

D&L interpret these results as follows:

• During the "bubble" period, investors were apparently more optimistic about the prospects of B2C companies, and thus capitalized product development and advertising expenses as assets in their valuation of Internet firms.

• In the post-bubble period, investors apparently became more skeptical about these expenditures and stopped capitalizing them and/or began treating them as expenses, suggesting "some maturation of the industry." And, since the relation between price/sales ratio and these expenses remained positive though not significant, this suggests that "the market is not, on average, treating the marketing and development expenses as reductions to firm value."

D&L also find that Internet firms’ ability to sustain their current rate of "cash burn" was significantly associated with their price-to-sales in both the 1999 and 2000 periods covered by the study. This supports the view that the market negatively viewed unprofitable Internet firms that were quickly depleting their cash holdings as likely to face liquidity problems that would threaten their survival.

IV. The Value-Relevance of Website Traffic and Other Non-Financial Variables

In general, the studies determine the value-relevance of Website Traffic and other nonfinancial variables by measuring the additional explanatory power of the nonfinancials relative to the respective sets of financial variables (discussed above) in the regression models. Their findings indicate that, after accounting for financial factors, various Website Traffic measures were value-relevant, notably for B2C Internet stocks (whose revenues depend on traffic volume), even in the aftermath of the shakeout.

The key findings of each study are summarized below:

Trueman et al (TWZ)

Measures of Website Traffic

UNIQUE VISITORS: the estimated number of different individuals who visit the firm’s web site(s) during a particular month.

PAGEVIEWS: the estimated number of pages viewed by those individuals visiting the firm’s web site(s) during the month.

Key Findings

• "Unique Visitors" and "Pageviews" had significant explanatory power for Internet stock prices after accounting for financial statement information such as book value of equity and earnings.

• However, the Traffic measures and individual income statement line items (especially Gross Profits) appear to capture some of the same information: "Once the information conveyed by the components of net income is taken into account, the informational role of Internet usage appears to be considerably diminished."
• "Pageviews" had more explanatory power for stock prices than "Unique Visitors," suggesting that investors placed greater emphasis on the former as for valuation.

• For the P/C firms, the incremental explanatory powers of "Pageviews" and of "Unique Visitors" were about the same, while, for the e-tailers, the incremental explanatory power of "Pageviews" was much greater than that of "Unique Visitors."

Rajgopal et al (RKV)

Measures of Website Traffic

REACH: Number of unique visitors as a percentage of total estimated population of Internet users during the time period.

UNIVIS: the number of web-active individuals who visited a particular site or web company within a given time frame.

Key Findings

• Consistent with TWZ, either of the measures, i.e., REACH or UNIVIS, was a significant factor reflected in Internet stock valuation after controlling for the influence of book value of equity and earnings. [The correlation between the two variables was 0.97, thus the results obtained using either variable in the regressions were similar.]

• REACH remained a strong explanatory factor even with the inclusion of Sales Revenue in the regression model, and the relation between traffic and revenues or sales growth was "weak and insignificant". Thus, REACH and Sales Revenue "appear to capture different aspects of market values of Internet firms." RKV conjecture that this implies that Traffic reflects expectations of future performance that are not necessarily reflected in current sales. They argue that such expectations are based on the "network effects" of attracting a critical mass of customers, building customer relationships, and collection of data about the buying and surfing behavior of site visitors. They cite, as an example, Excite@home’s acquisition of Bluemountainarts.com, an electronic greetings card website, for $780 million in cash and stock--even though the company had virtually no revenues and profits, it ranked behind only AOL and Amazon.com in terms of traffic.

• Analysis of 42 acquisitions of Internet companies during the year ended September 30, 1999 showed a strong relationship between the acquisition prices paid by acquiring firms and the acquired firms’ traffic numbers (UNIVIS), which is additional evidence of a strong association between market values and expectations of future growth based on traffic volume.

• Several factors which RKV hypothesized as affecting Website Traffic collectively explained a "significant variation (62%)" in REACH: Strategic Alliance with America Online; Use of Affiliate Programs; Media Visibility (measured as the total number of articles published about each firm in LEXIS/NEXIS’s "Major Newspapers" database); Marketing expenditures; and Cash Availability (for "chasing traffic"). Specifically, Media Visibility and Cash Availability were the key determinants of Traffic. Strategic Alliance and Affiliate Programs were positively related to Traffic but were only weakly significant. RKV opine that the value-relevance of Traffic apparently stems from these underlying factors that drive traffic.

Demers and Lev (D&L)

Measures of Website Traffic

D&L use factor analysis to extract from several Nielsen/Netratings Web metrics (listed in the Appendix) three "parsimonious factors...that capture the most relevant dimensions of Website performance":

REACH: the extent to which an Internet firm is able to attract unique visitors--defined as the number of unique visitors to a web site, as a percentage of the (total or active) web surfing population.

STICKINESS: a measure of how long visitors stay at the site once they are there—based largely on the average time spent at the site per visit and the average number of pages viewed per visit.

CUSTOMER LOYALTY: based primarily on the average number of visits to the site per unique visitor per period.

Key Findings

• Beyond the income statement expense factors (discussed in the previous section), all three Website Traffic measures were value-relevant for Internet stocks in both the 1999 and 2000 periods examined. D&L opine that their findings of significance of these measures for the 2000 post-bubble period contradict the claims of some analysts that Website Traffic measures were no longer important in the valuation of Internet firms after the bubble burst: "absent meaningful financial results, investors continue to rely heavily on web traffic measures."

• The Internet stocks that were relatively over-valued prior to the shakeout -- as indicated by D&L's empirical model -- experienced relatively larger declines in their price-to-sales ratios when the bubble burst: "the excesses of the early 1998-1999 valuations have been systematically eliminated."
• "Firms that went public later in the Internet boom experienced more negative changes in their price-to-sales ratios when the bubble burst."

• To examine the value-relevance of Strategic Alliances, D&L use three proxy variables to capture the nature of a firm's alliances with strategic partners:

1. Whether or not a firm has announced a strategic alliance with AOL
2. Whether or not a firm has announced a strategic alliance with one or more of the other "top 10" Internet traffic-generating companies: Lycos, Amazon, Yahoo!, MSN/Microsoft, Excite@Home, Alta Vista, GO Network, Go2Net, Time Warner (prior to the merger with AOL), and C/NET.
3. The cumulative number of strategic alliances that the company has announced itself to have entered into.

They find that an alliance with AOL was positively associated with the price-to-sales ratios of B2C stocks in 1999, which is consistent with RKV's finding of a weak positive association between alliances with AOL and their Website Traffic measure during 1999. However, in 2000, AOL alliances ceased to be value-relevant. Also, the cumulative sum of alliances entered into was significantly negatively associated with the price-to-sales ratios of Internet stocks in both 1999 and 2000. Thus, the findings support the view that the market adjudged that the much-hyped strategic alliances had not provided the anticipated benefits to Internet firms.

Hand(B)

Measures of Website Traffic

UNIQUE VISITORS: Number of visitors to a particular web property (combination of web sites and pages owned by the same company).

PAGE VIEWS: Total pages viewed by visitors for a specific web property.

HOURS VIEWED: Total hours that a visitor spends on a specific web property.

Key Findings

Hand's findings appear to contradict those of the other studies with respect to the relevance of Website Traffic in Internet stock valuation. In Hand(A), which did not directly examine the relevance of Website Traffic, he opines that financial variables -- book value, revenues, and expenses -- explained such a large proportion of the cross-sectional variation in the market values of Internet firms that the proportion available to be uniquely explained by non-financial data was quite low. He concludes that "the strength of basic accounting data and the lack of room it leaves for non-financial data thus runs opposite to the claims of many analysts that non-financial information is the central factor in the pricing of Net stocks."

Hand(B) extends the analysis in Hand(A) by examining the value-relevance of Website Traffic, supply and demand for shares, and forecasted financial performance:

• Current book equity, forecasted one-year-ahead earnings, and forecasted long-run growth in earnings -- which he collectively terms "Economic Fundamentals" -- "overwhelmingly dominate the pricing of both Internet and non-Internet companies." Negative forecasted earnings are negatively priced, while book equity and positive forecasted earnings are positively priced.

• "Incremental to 'Economic Fundamentals', Internet firms’ equity market values are only marginally related to one of the three web traffic measures examined--the number of unique visitors to the web site." None of the other Traffic variables -- the number of page views, the number of hours spent on the page views, the average income, age or gender of web site visitors -- had any significant association with Internet firms’ stock prices.

• Supply and demand factors [proxied by: the percentage of outstanding shares publicly traded (public float); the percentage of public float sold short; and the degree of institutional ownership] also did not have value-relevance for B2C firms after accounting for the Economic Fundamentals and Website Traffic variables. However, supply and demand factors were value-relevant for B2B Internet firms, i.e., those with little or no B2C traffic.

The apparent conflict between Hand's results and other studies' with respect to the value-relevance of Website Traffic may be partly because his sample includes B2B firms, for which investors do not consider Website Traffic to be of much relevance to current or future performance. D&L, for example, note that a direct comparison of their results with those of Hand(B) is not possible because of differences in their samples of Internet firms, the proxies for Traffic factors, and statistical models.

Furthermore, Hand (B)’s finding of only "marginal" or insignificant value-relevance of Website Traffic may be because forecasted earnings data already capture traffic measures to a substantial degree--in the absence of historical financial and other information that would normally be factored into forecasts, investors/analysts' projections of a B2C Internet firm's future earnings may have been based largely on traffic data.

V. Further Research

As the authors of the studies generally acknowledge, much of the empirical evidence thus far is preliminary, and they suggest various directions for further research. As the industry matures and more data become available and less prone to measurement problems, more extensive studies (including updates of the pioneering ones reviewed here) will provide better insights. For example, ongoing and future empirical work will examine, among other issues, the following:

• Changes in the set of factors that investors and analysts take into account in valuing individual Internet stocks in the aftermath of the shakeout, e.g.: the impact of recent questions about the accuracy of Website Traffic data and their efficacy as indicators of future performance; the relevance of nonfinancial factors other than Website Traffic; changes in the value-relevance of financial statement information relative to nonfinancial variables.

• Trends in investors' perceptions of the risks and expected returns of Internet stocks in general.

• The relative sensitivities of Internet industry stock returns to economic factors such as interest rates, inflation, the business cycle, etc. compared to other technology sectors as well as non-technology sectors.

• Evaluation of the rationality and psychology of the pricing of Internet stocks.

APPENDIX

I. STUDIES REVIEWED

1. Elizabeth Demers (University of Rochester) and Baruch Lev (New York University). [D&L]

"A Rude Awakening: Internet Shakeout in 2000" (September 2000).

URL: http://www.ssb.rochester.edu/fac/demers/index.htm

2. Hand, John R. M. (Kenan-Flagler Business School, University of North Carolina, Chapel Hill)

A. "Profits, Losses and the Non-linear Pricing of Internet Stocks" (January 2000). [Hand (A)]

URL: http://papers2.ssrn.com/paper.taf?ABSTRACT_ID=204875

and

B. "The Role of Economic Fundamentals, Web Traffic, and Supply and Demand in the Pricing of U.S. Internet Stocks" (April, 2000). [Hand (B)]

URL: http://papers2.ssrn.com/paper.taf?ABSTRACT_ID=221232

3. Rajgopal, Shivaram (University of Washington), Suresh Kotha (University of Washington), Mohan Venkatachalam (Stanford University). [RKV]

"The Relevance of Web Traffic for Internet Stock Prices" (January 2000).

URL: http://papers.ssrn.com/paper.taf?ABSTRACT_ID=207989

4. Trueman, Brett, M.H. Franco Wong, Xiao-Jun Zhang (all of Haas School of Business, University of California, Berkeley). [TWZ]

"The Eyeballs Have It: Searching for the Value in Internet Stocks" (January 2000).

URL: http://papers.ssrn.com/paper.taf?ABSTRACT_ID=206648

II. DATA: SAMPLE FIRMS & WEBSITE TRAFFIC

1. Demers & Lev [D&L]

Sample Firms

84 publicly-traded "B2C" Internet companies: InternetStockList™ (www.internet.com).

Sample consists of only e-tail, content/communities, financial news/services, portal, and services firms.

Data period: 1999 and first half of 2000.

Website Traffic Data:

From the Nielsen/Netratings "Audience Measurement" database.

Three "factors" -- "Reach", "Stickiness", and "Customer Loyalty" -- extracted using factor analysis from several web traffic measures:

• unique audience (the number of unique web surfers who have viewed the site during the month),

• active reach (the percentage of active web surfers who viewed the site during the month),

• universal reach (the estimated percentage of the universe of web surfers who have viewed the site during the month),

• rank by unique audience for the month,

• the total number of pages viewed by web surfers during the month,

• the number of visits to the site per unique visitor during the month,

• the average time spent at the site per person, and

• the percentage of pages that were viewed from browser cache during the month.

2. Hand (A&B)

Sample Firms

Publicly-traded Internet firms listed on www.internet.com’s InternetStockList

Web Traffic Data

From PC DATA Online.

UNIQUE VISITORS: Number of web-active PC Data Online participants who visited a particular web property in January 2000.

PAGE VIEWS: Total pages viewed by web-active PC Data Online participants for a

specific web property.

HOURS VIEWED: Total hours that a specific web property is viewed by web-active PC Data Online participants.

3. Rajgopal, Kotha and Venkatachalam [RKV]

Sample Firms

86 publicly traded Internet firms.

Selected from (1) listing of top-50 Internet firms in 1998 and 1999 in Internet World magazine and (2) the list of publicly traded Internet firms in Business 2.0 magazine (September 1999).

21 e-commerce firms, 17 content providers, 10 portals, 10 access providers, 7 software vendors, 3 infrastructure providers, 6 advertizing agencies, 6 security firms, 4 business-to-business firms, and 2 "miscellaneous" firms.

Web Traffic Data

From PC DATA Online. Data period: February to June 1999

REACH: Number of unique visitors as percentage of total estimated population of Internet users during the time period (U.S. and Canadian households only—corporate users not included).

UNIVIS: the number of web-active individuals who visited a particular site or web company within a given time frame

4. Trueman, Wong, and Zhang [TWZ]

Sample Firms

56 publicly-traded Internet firms on InternetStockList (internet.com) as of July 15, 1999 (plus Netscape, geocities, and broadcast.com, which were acquired prior to July 1, 1999, and Excite, which merged with @Home earlier in the year).

Sample consists of only portals, content/community providers, and e-tailers.

Web Traffic Data

From Media Metrix (Monthly Web Report for the months of October and December 1998, and March, June, and September 1999).

REACH: the percentage of projected individuals...that accessed the web content of a specific site or category among the total number of projected individuals using the web during the month.

UNIQUE VISITORS: the estimated number of different individuals who visit the firm’s web site(s) during a particular month.

PAGEVIEWS: the estimated number of pages viewed by those individuals visiting the firm’s web site(s) during the month. [Not directly reported by Media Metrix. Derived from three measures: (1) the number of unique visitors, (2) the average usage days per visitor in a month, and (3) the average daily unique pages viewed per visitor in a month.]

NOTES

1. The empirical evidence discussed in this section is drawn from the following studies:

Amir, E. and B. Lev. "Value-Relevance of NonFinancial Information: The Wireless Communications Industry." Journal of Accounting and Economics 22, 1996.

Blacconiere, W. G., M. F. Johnson, and M. S. Johnson. "Market Valuation and Deregulation of Electric Utilities." Journal of Accounting and Economics 29, 2000.

Collins, D. W., E. L. Maydew, and I. S. Weiss. "Changes in the Value-Relevance of Earnings and Book Values over the Past Forty Years." Journal of Accounting and Economics 24, 1997.

Deng, Z., B. Lev, and F. Narin. "Science and Technology as Predictors of Stock Performance." Financial Analysts Journal. May/June 1999.

Francis, J. and K. Schipper. "Have Financial Statements Lost Their Relevance?" Journal of Accounting Research, Autumn 1999.

Kane, G. D. "The Impact of Recession on the Value-Relevance of Accounting Ratios." Mid-Atlantic Journal of Business, December 1997.

Lev, B. and T. Sougiannis. "The Capitalization, Amortization, and Value-Relevance of R&D." Journal of Accounting and Economics 21, 1996.

Lev, B. and P. Zarowin. "The Boundaries of Financial Reporting and How to Extend Them." Journal of Accounting Research, Autumn 1999.

Zhang, G. "Accounting Information, Capital Investment Decisions, and Equity Valuation: Theory and Empirical Implications." Journal of Accounting Research, Autumn 2000.

Copyright © 2000 Capital Researchers

The Author -- Michael J. Isimbabi

Dr. Isimbabi holds a Ph.D. in Business/Finance from Temple University, Philadelphia, as well as MBA and Bachelor of Engineering (Electrical) degrees. His professional background includes extensive experience as a consultant, economist, professor, banker, and engineer.

Dr. Isimbabi's consulting, research/analysis, writing, teaching, and management experience spans several areas: Financial Markets; Investments; Banking/Financial Services; Corporate Finance; International Business & Finance; Emerging Markets; E-commerce; High Technology Business/Finance; and Electricity/Energy Markets.

His publications include Contemporary Portfolio Theory and Risk Management [Co-authored] (West Publishing, 1994) and articles in journals such as the Journal of Banking and Finance, Recent Developments in International Banking and Finance, the Atlantic Economic Journal, the SAIS Review: A Journal of International Affairs, and others. The Recent Developments... paper, "Comovements of World Securities Markets, International Portfolio Diversification, and Asset Returns: A Survey of Empirical Evidence", was extolled as "a first-class summary reading for the portfolio manager" in a Journal of Banking and Finance review.

Dr. Isimbabi's professional experience also includes the following:

• Economist/Consultant with the New York Independent System Operator (NYISO), the organization that operates New York State's competitive wholesale electricity markets.

• Finance professor (taught Financial Markets, Investments, Financial Management, and International Finance at Temple and Morgan State Universities).

• Graduate, Citibank banking/finance professional training program and banker with a foreign affiliate of J. P. Morgan and Barings.

• Electrical/electronics engineer (aircraft flight simulation, telecommunications, and other engineering systems).