Pricing Internet Stocks: Traditional Methods of Stock Valuation Can Give Very Conflicting Results in E-Business
Estrada, Javier, European Business Forum
Terra Networks, the crown jewel of Spanish Internet companies, went public on November 17, 1999 at [euro]11.87. It closed its first day of trading at [euro]37, and quickly went on to reach a high of [euro]139.75 on February 25, 2000, for a gain of over 1,000%. It was all downhill thereafter; by late August the stock was trading at around [euro]40, thus losing over 70% of its value. Trying to explain these huge changes in valuation with fundamental analysis is difficult, to say the least. And Terra is not the exception but the rule in the Internet world.
To be sure, there are many disagreements about how to value "old-economy" companies; but when it comes to Internet stocks, most analysts are often shooting in the dark. Is Terra nowadays fairly valued, undervalued or outrageously overvalued? There are market participants who support all three positions--a situation which may stem from the wide range of different valuation models being used. In this article, I will first summarize some methods of relative valuation, and then some "traditional" methods of absolute valuation which, I will suggest, should not be forgotten when pricing Internet stocks.
Methods of Relative Valuation
Many of the methods of relative valuation currently used illustrate the analysts' ongoing struggle with finding some measure (any measure!) that helps them assess the value of Internet companies. Some of these methods, and their pros and cons, are described below.
Price-earnings ratios, or P/E ratios for short, are defined simply as a company's current stock price divided by its (past or expected) earnings per share. This method, used to assess the value of both new-economy and old-economy companies, simply consists of comparing a company's P/E with an "appropriate" P/E, in order to determine whether the company is undervalued, fairly valued, or overvalued. The issue, of course, is what is the "appropriate" P/E?
There are at least three possibilities: historical, industry-wide, and theoretical benchmarks. A historical benchmark involves estimating an average P/E based on the company's past P/Es; an industry-wide benchmark involves estimating an average P/E based on the P/Es of companies in a given sector; and a theoretical benchmark involves building a model based on some predetermined variables that yields equilibrium P/Es.
A simpler alternative is to make a straightforward comparison of the P/E of two companies at a given point in time. To illustrate, by the end of July, 2000, the P/Es of Yahoo and Lycos stood at 386 and 309, respectively, thus implying that Yahoo is expensive relative to Lycos.
Pros P/Es are widely used for their simplicity. In addition, they have a theoretical underpinning, for they can be expressed in terms of a company's fundamental variables.
Cons Deciding what is the "appropriate" or benchmark P/E is a mix of art and science (hence, the use of more than one benchmark may be desirable). In the Internet world, historical benchmarks are usually misleading, for most companies have very short trading histories, and hence no reliable "equilibrium" historical P/E. Industry-wide benchmarks in principle solve this problem but, if the whole industry is overvalued, then an industry-wide P/E will lead to overvalue the shares of individual companies. Theoretical benchmarks beg the question of what are the right variables to include in the model to estimate equilibrium P/Es and, to the best of my knowledge, nobody has built such model. Finally, perhaps the most important objection to this method: Many (if not most) internet companies have no earnings, anyway!
Another widely-used measure of valuation is a company's price-to-book ratio, which is obtained by dividing a company's stock price by its book value per share. This measure, just as the P/E ratio, also needs a benchmark to determine relative valuation. …