An Empirical Investigation What Happened to Internet

by Damir Tokic

a peer-reviewed articleDamir Tokic is a doctoral candidate at College of Business Administration at the University of Texas - Pan American and a part-time member of the faculty of that institution's Department of Finance and Economics. He is the co-founder of the New Finance Research Institute, a non-profit research organization whose mission is to explore the idea of "the new finance for the new economy." 

ABSTRACT: This article (below) points out that the investment opportunities approach is able to theoretically support the practical explanations of the recent decline in the prices of Internet companies' stocks. First, this approach shows that owners/managers had the incentive to over-spend on R&D and advertising to increase the underlying volatility of their company's stock and, thereby, maximize their personal wealth. Second, it suggests that financial analysts contributed to the crises by ignoring the negative profitability in their target value projections. This study reported on in this article empirically proves both that underlying volatility increases as managerial ownership increases and that managers over-spent on R&D and advertising expenses.                                                    




What caused the recent Internet stock valuation downfall? The financial media has offered several practical explanations. The management of Internet firms, they say, lacked business savvy but had an abundance of technological expertise. They were not able to administer a viable long-term business model and ignored a lack of short-term profitability. Financial analysts also ignored negative short-term profitability and contributed to the crises by constantly raising target stock values based solely on future expectations.

Although there is some empirical evidence that supports these practical explanations, the financial literature has not been able to provide a theory that explains the how Internet stocks are valued. This study provides a theoretical explanation for the causes of the recent devaluation of  Internet stocks. It also provides the empirical evidence that managers and financial analysts must share the blame because both used a hazardous valuation technique based on option pricing theory to irrationally overprice Internet stocks. 

Black and Scholes (1973) developed a model to price the European Call options. A buyer of call option, optimistic about future prospects of a firm, acquires the right to buy shares of a firm in the future at the contract (strike) price. As the value of shares increases over the strike (plus the price of option), the investor exercises (buys at contract price and sells at the market price) the option for profits. A fundamental feature of the option pricing model is a positive relation between the option value and volatility. The higher the underlying volatility, the higher is the chance that the value of shares will increase over the strike price (plus the price of option), and, thereby, the higher will be the value of a call option. Traditional valuation models (DCF) assume a negative relation between volatility (discount rate) and net present value.

It is very important to understand the proper theoretical valuation technique for several reasons. First, it provides the answer to the "What went wrong?" question. It also gives managers guidance for strategic planning and investors the basis for the evaluation of managers and firms’ overall performance. Finally, it provides directions for future research by isolating critical valuation issues.


The starting point for explaining the valuation process is to select the proper valuation methodology. Miller and Modigliani (1961) propose that the investment opportunities approach is the most appropriate method to value "growth shares". A growth firm is a firm that has the opportunity to invest capital in new assets at a rate of return higher than normal. If the assumption that Internet firms have the opportunity to realize abnormal returns holds, at least in the short-term future, than the investment opportunities approach is the appropriate model to value Internet stocks.

The model predicts that a manager buys stock in a firm that has the potential for earning an abnormal (higher than normal) rate of return in order to maximize his or her personal wealth. The price that a manager pays for a business equals the sum of the value of current operations and the value of growth. (See the Appendix for a mathematical formula and detailed explanation.)

Value of a Growth Firm = Present Value of Current Operations + Value of Growth Opportunity

The present value of current operations is a function of earnings (or operational cash flows) that present assets can produce during the forecasted period, discounted a normal rate of return. The value of growth opportunity is a function of (a) the presence of rate of return that is higher than normal (abnormal) and (b) the investments in new assets (projects) to take advantage of an abnormal (higher) rate of return. The opportunity for abnormal returns can arise due to a competitive advantage, which in this case is the development of a new technology, and brand recognition in the growth stage of the industry. Firms that invest more in R&D and advertising have higher probability of dominating the industry in the future. However, investments in intangible assets such as R&D and advertising are highly volatile. This implies that the positive relationship between the volatility of earnings and value of Internet firms' stocks. 

Managers/owners may be tempted to over-invest in R&D and advertising in order to maximize their personal wealth. However, this produces negative earnings and lack of earnings visibility, which reduces the present value of cash flows. Similarly, as industry enters its mature growth stage, the relative amount invested in R&D and advertising should decrease because the opportunity for abnormal returns diminishes, and the source of value shifts to earnings. The next section of this article presents a detailed theoretical explanation of the recent collapse of the prices of  Internet stocks and establishes the foundation for the development of explanatory  hypotheses.


A careful analysis of fundamental relations between the value of a share of a firm's stock and its earnings, investments in new assets, and the opportunity for abnormal returns as predicted by the investment opportunities approach suggests the existence of the following stages in the collapse of Internet stocks. (See Exhibit 1 below.)

Exhibit 1

What went wrong with Internet stocks according to the investment opportunities approach.

Opportunity to invest new capital at rate of return higher than normal exists.

1. ®

Industry experts start new firms to manage them.


2. ®

Firms compete in R&D spending and advertising campaigns.

3. ®

High uncertainty increases the volatility.


4. ¯

Negative profitability causes the sharp decline in stock prices.



Excess spending resulted in negative profitability.



Managers overspent on R&D and Advertising.



High volatility increases the stock prices.



1. Opportunity to invest new capital at rate of return higher than normal exists

In 1995, there were only 22 million people in U.S. on-line, which was less than 10 percent of its total population. At that time expectations were that there would be a 150 million Americans on-line by the year 2000. Firms that sell over the Internet or provide a service or support (hardware, software, security, hosting, etc.) had a great opportunity to capture the market leadership in its early growth stage and convert it into a profitable venture once this market matured. The opportunity to earn an abnormal rate of return on a share of Internet stock was a product of  a fast growing customer base, new high-tech products, and a new location (virtual market place).

2. Industry experts start new firms to manage them

There were high barriers to entry into the Internet industry. The major one is the necessary technological expertise. The typical business person was not capable of  managing or starting a firm whose success or failure depended on technological innovations targeted at a growing customer base. So, it was only natural that high-tech experts started their own firms. This way they were able to issue an initial public offering (IPO) and maximize their personal wealth by maximizing the value of the firm's shares. In support of this assumption, Schultz (2001) finds that the ownership of the stock of  Internet firms by their managers is greater than in other firms; that managers/owners are generally optimistic about the future of their firm; and they are less likely to diversify their personal stock portfolios than are managers of other types of  firms.

3. Firms compete in R&D spending and advertising campaigns

The strategy when this industry began was to develop a superior product and sell it to the growing target market. Therefore, Internet firms started competing on the basis of spending on R&D and advertising. A firm that was able to deliver a superior product or service and successfully market it was the prime candidate for market leadership and financial rewards once this market matured.

4. High uncertainty increases the volatility

Outside investors understood that there is an opportunity to earn high returns by investing in Internet stocks. All dot.coms invested heavily in R&D and advertised extensively, however only few would become market leaders. Investors' challenge was to pick the survivor. Firms that invested the most in R&D would probably have the greatest chance of becoming the market leader. However, because of their lack of expertise in this field, an outside investor did not have the knowledge needed to evaluate R&D projects. As a result, overall volatility increased across the board in the Internet industry.

5. High volatility increases the stock prices

In conformance to option pricing theory, firms with higher volatility have higher values. In other words, firms that do not spend on R&D and advertise (less volatile) have less of a chance to become market leaders. There is also a "psychological" side to this story. By engaging in extensive R&D and advertising, managers of Internet firms inflated investors' hopes, which in turn increased their valuation of Internet stocks, as they hoped that all those R&D efforts and dollars would produce a superior product.

6. Managers overspent on R&D and Advertising

The managers of Internet firms were industry experts with technological know how. However, they lacked essential business training, including fundamental financial management and strategic planning skills. Tokic (2001) finds that a selected sample of twenty Internet firms spent on average 75 percent of revenue on R&D and advertising during fiscal 1999. This is a clear example of overspending. Shama (2000) calls it a lack of vision or of business savvy, while Schultz (2001) calls it over optimism or over confidence. Either way, the mismanagement of Internet firms set the stage for the bubble to burst.

7. Excess spending resulted in negative profitability

The chance for positive earnings was very slim if, on average, Internet firms spent 75 percent of revenue on R&D and advertising. Managers viewed profits as distraction (Shama 2000) and treated R&D and advertising as investments, rather than expenses (Schultz 2001). Surprisingly, financial analysts "fell" for managerial optimism and ignored the negative profitability of their firms for a very long time. Allegedly, there were also a conflict of interest between  financial institutions maximizing the return from their investment banking operations and providing objective financial analysis.

8. Negative profitability causes a sharp decline in Internet stock prices

Once the on-line population increased to over 50 percent of the United States' population of 155 million, the opportunity for abnormal returns diminished, and the source of value shifted to the net present value of earnings. Since Internet firms remained unprofitable, their values sharply decreased, and several bankruptcies ensued (, E-Toys, etc). This stage represents the beginning of the  maturation of this industry and accompanying  negative profitability. Most of these companies had negative profits from their initial IPO stage. Problems developed when investors realized that dot.coms did not have a viable business model that would eventually produce the expected profits. As a result of low expected earnings in the future, the net present value component of the investment opportunities approach remained a depressed value. As a result of the diminishing opportunity for growth and the low net present value of earnings, Internet share prices dropped sharply.

What went wrong? Managers and financial analysts predicted a positive relationship between value of a stock and its volatility, while ignoring the size of earnings. In order to maximize volatility, managers overspent on advertising and R&D; thereby causing  profitability to be negative. In early stages of Internet industry growth, with only a small percentage of the population on-line, it was rational to expect that firms that developed the dominant product and established a recognizable brand to get a comparative advantage over their competition and record earnings in the future. However, with the spread of Internet to over 50 percent of  the United States'. population, investors expected that some dot.coms would become profitable. However, a majority of Internet firms continued to record negative earnings, and they faced slower growth prospects and the possible loosing of an opportunity to earn abnormal returns. Therefore, according to the investment opportunities approach, the value of Internet firms became the function of earnings. Because earnings visibility is blurred and highly unpredictable even for industry leaders, this caused a sharp decline in the price of Internet stocks, .


Several studies (Bathala, Moon and Rao 1994 and Chen and Steiner 1999) find that the negative relationship between managerial ownership and volatility in supports the risk aversion hypothesis (management non-diversification problem). As managerial ownership increases, managers have an incentive to maximize personal wealth by reducing volatility, which reduces the discount rate and increases value under the traditional discounted cash flow valuation models. Similarly, as volatility increases, managers will diversify their personal portfolios and sell a portion of the shares they own.

However, if analysts use the option pricing theory to value Internet stocks, then managers will increase volatility in order to maximize personal wealth (by maximizing the value of equity). As managerial ownership increases, managers have more incentive to increase volatility by excessive spending on R&D and advertising. Similarly, higher volatility promotes more managerial ownership (and possible institutional ownership). Therefore, the author of this study expected to find a positive relationship between managerial ownership and volatility that would support the option-pricing-theory based  wealth transfer hypothesis (Jensen and Meckling 1976, Galai and Masulis 1976).

Ho1: There is a positive bi-casual relationship between the managerial ownership and volatility.

If the opportunity to invest new capital at the rate of return higher than normal is present, then spending on R&D and advertising can be defined as "investments" in intangible assets, but only up to certain levels. If a firm spends on R&D and advertising beyond the break-even point, according to the investment opportunities approach, value will decrease. If the increase in spending on R&D and advertising increases volatility, which in turn increases value, than the RDD variable (defined in Exhibit 2 below) in the volatility equation should be positive and significant. However, excessive spending on R&D and advertising will not significantly increase volatility. In turn, it will indirectly decrease value via the negative earnings,

Ho2: Managers of Internet firms over-spent on intangible assets (R&D and advertising).

The combined value of R&D and advertising expenses as a percentage of revenue (RDD) is a proxy for investment in intangible assets. Bathala, Moon and Rao (1994) used this variable to capture agency costs. Myers (1977) argues that agency costs associated with intangible assets are higher than agency costs associated with tangible assets. Further, Leland and Pyle (1977) find that high spending R&D firms have more managerial ownership due their possession of  relatively more positive, private information. Chan, Lakonishok and Sougiannis (2001) find similar market valuation effects for R&D and advertising expenses. They argue that failure to adjust PE or MV/BV ratios for intangible expenditures may result in serious mispricing of technology stocks.


The critical task for this study is to prove that there is a bi-casual positive relationship between managerial ownership and volatility. The proper methodology to test this hypothesis is the two-stage, least square method (2-SLS) with two simultaneous equations having managerial ownership and volatility as dependent variables in respectable equations. The 2-SLS is more appropriate than OLS because OLS assumes that the error term is uncorrelated with each of regressors. This assumption leads to biased and inconsistent parameter estimates. Both managerial ownership and volatility (as a function of R&D and advertising expenses) are endogenous to the system. Therefore, it is unreasonable to assume that the error term in managerial ownership equation is uncorrelated with the volatility posing as an independent variable. The two stage, least square method is able to show how volatility affects managerial ownership separate from how managerial ownership affects volatility.

Managerial Ownership Equation

Volatility Equation

Exhibit 2

Definition of Variables

MGTOWN – Managerial ownership--Found on proxy statement--Defined as total percentage of shares held by directors and executives

INSTOWN – percentage of shares outstanding held by institutional investor (pension funds, mutual funds, other corporations)

VOL – Standard Deviation of Daily Stock Returns--Daily closing stock prices for the period from April 1, 1999 – April 1, 2000 are collected from Yahoo. This data is exported to Microsoft Excel, where the standard deviation of daily stock prices movements is calculated.

RDD – Research and Development expense plus the Advertising expense as a percentage of revenue

GROWTH – One-year revenue growth rate

DEB/EQ – Total Debt to Equity ratio

SIZE – Log of total market capitalization

Other variables include the institutional ownership, which is supposed to provide an external, substitute monitoring effect (Shleifer and Vishny 1986) and reduce volatility according to the prudent man behavior hypothesis (Badrinath, Gay, and Kale 1989), as well as the debt/equity ratio, growth, and size, consistent with the similar models in Bathala, Moon and Rao (1994) and Chen and Steiner (1999).


The sample includes 50 companies from the Internet Information Provider and Service Industry listed on stock evaluator. These companies range from on-line retailers ( and auctioneers (E-bay) to e-business application software (Vignette) and Internet security (ISSX). The Internet Information Provider and Service Industry on includes over 100 companies, however, there was missing or unreported information for many of the smaller start-ups. Exhibit 2 (above) defines each variable. The sampling frame for institutional ownership, one-year growth rate, debt-to-equity ratio, and size variables is the stock evaluator. Managerial ownership for each company was found on proxy statements filed with the SEC using the Lexis/Nexis index. Daily one-year stock return volatility was calculated using daily closing prices adjusted for stock splits that was found on Yahoo's historical stock returns. R&D and advertising expenses as a percentage of revenue were found in the management discussion section of annual reports.

Table 1 

Descriptive Statistics




Standard Deviation













































The descriptive statistics in Table 1 (above) show combined managerial and institutional ownership of 77 percent on average in Internet firms, compared to 55 percent in cross-section of firms found by Bathala, Moon and Rao in their 1994 study. Also, Internet firms on average spent 72 percent of revenue on R&D and advertising, compared to 3.5 percent in cross-section sample found by Bathala, Moon and Rao in their1994 study. Similarly, growth rates are higher, volatility is higher, and debt ratios are much lower. The broad cross-section of firms is dominated with "old economy," mature firms competing in industries with insignificant growth opportunities (no opportunity for permanent, abnormal returns) whose values were function of the net present value of earnings that naturally spend only small percentage of their revenue on R&D and advertising. In contrast, Internet firms were start-ups that competed in an industry promising permanent, abnormal returns for a firm that brought new technology, shifting value to growth opportunity. Therefore, these firms spent heavily on R&D and advertising. Similarly, managerial ownership was higher due to managerial optimism and confidence in their ability to utilize technological know-how to develop a dominant product. These fundamental differences between "new economy" and "old economy" firms provide the basis for expecting a positive, bi-causal relation between managerial ownership and volatility in the Internet firms sampled, in contrast to previous studies that used a broad cross-section dominated by old economy firms (Bathala, Moon and Rao (1994).

The high standard deviation in the values of each variable raises concerns about the normality of the series and potential outliers, which would make the parameter estimates biased. The Jarque-Bera test (Table 2 below) shows that it is not possible to reject the null hypothesis of  a normal distribution at 0.01 level. The test statistic measures the difference of the skewness and kurtosis of the series with those from the normal distribution. A small probability value leads to rejection of the null hypothesis of a normal distribution.

Table 2 

Jarque-Bera Normality Test



























Table 3 (below) presents empirical evidence in support of null hypothesis1, stating a positive, bi-casual relationship between managerial ownership and volatility. In the managerial ownership equation the volatility coefficient is positive and significant. (Also, the institutional ownership coefficient is negative and significant, which supports the external, substitute monitoring hypothesis.) Similarly, in the volatility equation, both managerial ownership and institutional ownership coefficients are positive and significant. These findings contradict  Bathala, Moon and Rao (1994) and Chen and Steiner (1999), who find the negative relationship between managerial ownership and volatility predicted by traditional valuation methods. (As volatility increases, managers diversify personal portfolios.) This finding is contradictory, because managers and financial analysts based their valuations on the option pricing theory, predicting, as explained earlier, a positive relationship between the volatility and value.

Table 3 

Nonlinear two-stage least square simultaneous equation regressions.










































*** Significant at the 1% level.

** Significant at the 5% level.

* Significant at the 10% level.

The RDD coefficient in the managerial ownership equation is positive and significant, implying that, as spending on  R&D and advertising increases, managerial ownership increases. This is consistent with Schultz’s (2001) managerial optimism and over-confidence in future prospects of a firm and their ability to manage it. The RDD coefficient is insignificant in the volatility equation. This implies that volatility did not significantly increase with heavy spending on R&D and advertising. Therefore, managerial efforts to maximize value by maximizing volatility via R&D and advertising spending were not fruitful. Volatility increases as R&D and advertising spending fall within certain levels where a firm can at least break even. Once a firm becomes un-profitable, any additional spending insignificantly increases volatility and indirectly decreases value. Therefore, there is a support for hypothesis 2, which states that managers over-spent on R&D and advertising expenses.

The future success of Internet firms is a function of managerial strategic planning. Managers must understand the importance of earnings in the valuation process. In the early stage of growth, managers should invest heavily in R&D and advertising, but only at the break-even point because negative profitability decreases the value. This would maximize volatility, which in-turn will maximize value. At the more mature growth stage where the opportunity for abnormal returns diminishes, managers should decrease the relative amount of spending on R&D and advertising and attempt to increase profitability.


The investment opportunities approach provides the theoretical foundation for valuation of Internet stocks. It suggests that managers are tempted to increase volatility in order to maximize value. Therefore, the major reason for recent Internet stock devaluation was over-investment in R&D and advertising that caused negative profitability. Financial analysts must share the blame because they ignored losses in their projections and target values.

There are several issues that need to be studied in the future. First, it is important to develop an optimum investment policy in R&D and advertising that will maximize volatility at different stages of growth. Secondly, since R&D expenditures may be more critical for some firms, while advertising expenditures may be more important for other firms, future research should attempt to evaluate the individual effect of each component of intangible assets on managerial ownership and volatility, as well as on the overall valuation process. Also, we need to understand better the duration of the period during which abnormal returns can be earned.



According to Miller and Modigliani (1961), the value of a growth firm is a function of: (a) normal rate of return that an investor can earn by investing in other securities; (b) present value of cash flows from existing operations; and (c) opportunities to make above-average returns on additional investments in a business. The equation below (1) illustrates the investment opportunities approach:


where, X(0) represents the earnings from assets-in place, "" is the normal rate of return, "" is the rate higher then normal that can be obtained by investing additional capital "I" at time "(t)" in new real assets.

The first expression resembles traditional discounted cash flow (DCF) models. The "normal" rate of return discounts forecasted earnings from assets-in-place to the present value. As this rate increases (), the present value decreases (), illustrating in equation below (2) the negative risk-value relationship fundamental to all DCF models.


The second expression represents the value of growth opportunities. Expecting returns higher than the normal, the owner invests in new real assets (). Increased investments induce high uncertainty regarding the commercial success of these investments. The uncertainty translates into increase in volatility (), a measure of risk associated with the opportunity for abnormal returns. As illustrated in the equation below (3), the increase in investments and the underlying uncertainty positively affect the value () of the growth component in the investment opportunities approach. It is important to point the positive volatility-value relationship, a feature fundamental to option pricing models (Black and Scholes, 1973), opposite to the traditional DCF predictions.


Traditional DCF models assume that there is no opportunity to consistently produce above normal returns, (). This assumption is supported with the semi-strong version of Efficient Market hypothesis which says that it is impossible to produce abnormal returns based on all publicly available information, including historical price behavior. Many argue that traditional DCF models undervalue Internet stocks because they ignore the value of growth opportunities in the absence of significant earnings.

The growth component of investment opportunities approach is similar to the real option pricing method in several ways. First, it assumes the positive value-volatility relationship. Second, it recognizes the importance of managerial project selection in investments in new assets. Many argue (Boers 1998, Jagle 1999) that the real option-pricing model is applicable to value Internet stocks because of these assumptions. However, it tends to overvalue Internet stock because it ignores the negative effect of losses on the value of earnings.


Badrinath, S.G., G.D. Gay, and J.R Kale, "Patterns of Institutional Investment, Prudence, and the Managerial ‘Safety Net’ Hypothesis," The Journal of Risk and Insurance, LVI (December 1989), pp. 605-625.

Bathala, C.T., K.P. Moon, and R.P. Ramesh, "Managerial Ownership, Debt Policy, and the Impact of Institutional Holdings: An Agency Perspective," Financial Management, Vol. 23, No. 3, Autumn 1994, pp 38-50.

Black, F., M. Scholes, "The Pricing of Options and Corporate Liabilities", Journal of Political Economy, 1973, pp. 637-654.

Boer, P.F., "Traps, Pitfalls and Snares in the Valuation of Technology," Research Technology Management, (1998), pp. 45-53.

Chan, L. K.C., J. Lakonishok, and T. Sougiannis, "The stock market valuation of research and development expenditures," Journal of Finance 56, (2001), pp. 2431 - 2456.

Chen, C.R., T.L. Steiner, "Managerial Ownership and Agency Conflicts: A Nonlinear Simultaneous Equation Analysis of Managerial Ownership, Risk Taking, Debt Policy, and Dividend Policy," The Financial Review 34 (1999), pp. 119-136.

Galai, D. and R.W. Masulis, "The Option Pricing Model and the Risk factor of Stock," Journal of Financial Economics 3, (1976), pp. 53-81.

Jagle, J.J., "Shareholder Value, Real Options, and Innovations in Technology-Intensive Companies," R&D Management 29, 3 (1999), pp. 271-287.

Jensen, M.C. and W.H. Meckling, "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure," Journal of Financial Economics, (October 1976), pp. 305-360.

Leland, H.E. and D.H. Pyle, "Information Asymmetries, Financial Structure, and Financial Intermediation," Journal of Finance (May 1977), pp. 371-387.

Miller, H. M., F. Modigliani, "Dividend Policy, Growth and the Valuation of Shares," The Journal of Business, (1961), pp. 411-433.

Myers, S.C., "Determinants of Corporate Borrowing," Journal of Financial Economics (March 1977), pp. 147-175.

Myers, S.C., "Finance Theory and Financial Strategy," Interfaces, 1, 14 (1984) pp. 126-137.

Schultz, P., "Do the Individuals Closest to the Internet Firms Believe they are Overvaluated," Journal of Financial Economics, 59 ( 2001), pp. 54-78.

Shama, A., "For Dot.Coms It’s the Vision Thing," Barrons, September 2000, p.46.

Shleifer, A and R.W. Vishny, "Large Shareholders and Corporate Control," Journal of Political Economy 94, (1986), pp. 461-488.

Tokic, D., "Do Internet Firms Overspend on R&D and Advertising," Working paper, 2001.


(Business Quest) 

A journal of applied topics in business and economics

Click to go to


Click to go to