Spending associated with one type of intangible asset, research and development (R&D), and the problems it poses for stock valuation, have lately been the subject of much attention. In part, the concern reflects the growth of science- and knowledge-based industries. The technology sector, for example, accounted for about 5 percent of the total value of the market in 1990, but it has jumped to almost 20 percent of value by 1998. More strikingly, the amount of R&D spending is at least as large as the sector’s earnings.

The rise in the importance of such companies raises the question of whether their stock market values reflect their large investments in R&D. In the idealized case that economists call an "efficient market," investors make the appropriate corrections to stated accounting earnings. As a consequence, the stock price includes an unbiased estimate of the value of a firm’s intangible assets (such as the capital value of its R&D investments). On the other hand, several factors make setting the value of such R&D-intensive firms especially challenging. Many such firms have few tangible assets; instead, their prospects are tied to the success of new and untested technologies and, hence, are extremely unpredictable. Large expenditures are usually required at the outset, and the outcome of many research projects is far from assured. The benefits, if any, are likely to materialize only much later, while the life-cycles of any resulting products may be quite short. Finally, detailed financial information about a firm’s R&D activity is not generally available to investors. Firms have some leeway in identifying what counts as an R&D expense. All such spending is reported as one aggregate item, so there may be very little information about the kind of R&D involved. More important, U.S. accounting principles require a firm to deduct all R&D expenditures as a current expense against income — even though the benefits of R&D are long-lived, and, therefore, at least part of the spending might be more appropriately treated as a capital expenditure. As a result, some yardsticks commonly used by investors, such as price-earnings ratios and price–to-book ratios, may not be particularly useful in looking at high R&D firms.

In light of these difficulties, the link between a firm’s R&D spending and its stock price is not clear. Many market commentators have argued that R&D-intensive firms are likely to benefit from technological innovation and, hence, represent superior investments. In light of the uncertainties inherent in R&D, however, it is also possible that the market may simply ignore any future benefits and accept firms’ financial statements at face value, without adjusting for the long-term benefits of R&D. If, for example, investors value a firm at a fixed multiple of current earnings, the effects of R&D expensing may lead to mispricing. Similarly, some have argued that stock market investors fail to reward businesses for long-term investments. In contrast, many observers suggest that the valuations of R&D-intensive technology stocks are excessive and may reflect investors’ over-optimism about the effect of R&D on future profits.

Recent research by Louis K. C. Chan and Josef Lakonishok of the Department of Finance and Theodore Sougiannis of the Department of Accountancy investigates whether the stock market appropriately accounts for investment by firms in R&D. To do so, they examined all traded U.S. companies from 1975 onwards, measured the relative amount each firm spent on R&D, and whether this was related to subsequent stock returns.

Contrary to the argument that R&D-intensive technology stocks represent superior investments, the evidence does not indicate that the prices of firms who invest in R&D do better than those of firms with no R&D. Instead, the average return is comparable. Put another way, historically, the stock of an unexciting cement company did just as well, on average, as a highly publicized technology stock. The absence of any differences is consistent with the idea that the market price, on average, currently takes into account the intangible capital associated with R&D. It may also be the case that in many industries continued R&D investment, in the form of technological enhancements and new product development, is as much a basic input as labor and capital. Thus, continued R&D expenditures are vital to maintaining competitive position and so are not associated with extraordinary stock price performance.

Although, on average, the stock market attaches fair values to the intangible R&D capital of firms, a closer examination of the set of stocks active in R&D reveals further interesting patterns. First, glamour stocks with high price-to-book ratios that are very active in R&D earn higher average returns than other glamour stocks. Consequently, an investor choosing among glamour stocks might prefer a stock that does a lot of R&D such as Merck, over a glamour stock such as Starbucks, which does little or no R&D. Based on conventional valuation measures such as price-earnings or price-to-book ratios, both firms look equally expensive. However, it appears that the market tends to underestimate the future opportunities of Merck arising from its R&D investments, relative to those of Starbucks.

In addition, firms with high levels of R&D relative to their equity market value tend to have higher average future returns. These firms tend to have turned in poor performance in the past. Because of their poor track record such firms would be facing strong pressures to cut costs and improve earnings. Instead, some firms continue to devote a substantial part of their earnings to R&D. The willingness of the managers of these firms to forgo current earnings should send a strong signal to potential investors that management feels that prospects are likely to improve. The evidence indicates, however, that the market tends to overlook such signals so that the stocks tend to be relatively under-priced.

American Firms Gain as Targets of Foreign Takeovers

Foreign takeovers of American firms have surged in the last decade. The 1997 Mergerstat Review reported that such acquisitions increased from 154 in 1982 to 333 in 1996, and their value increased from $5.1 billion to $73.5 billion. In 1996 cross-border takeovers accounted for 15 percent of all acquisitions of U.S. firms.

On the whole, foreign firms do not appear to gain as much value when they acquire U. S. firms as do American companies, and U.S. firms targeted by foreign companies appear to gain more than those taken over by American. firms. Perhaps, despite the fears about foreign takeovers of U.S. firms expressed by some in this country, international acquisitions should be welcomed as a source of wealth for American stockholders.

One market where the opposite phenomenon occurs is the New York Stock Exchange (NYSE) where traders placing larger orders pay higher prices. The standard explanation offered by financial theory are that when a trader seeks to make a larger stock purchase this suggests to the dealer on the other side of the transaction that the buyer has developed information that the stocks’ true value is higher than their current price. Consequently, the dealer might lose potential profits on each share, and the more shares he sells at this price the greater his potential loss. Therefore, he charges a higher price for larger orders to compensate for the risk. Accepting a large order may also unbalance the dealer’s portfolio, reducing its diversity and increasing his risk, so that he may also seek compensation for this risk.

Most financial theorists have focused on trying to explain why large orders are met with higher prices since, historically, the best data on stock prices has come from the New York Stock Exchange and other markets where this is so. However, recent research has shown that on many other financial markets, for example, NASDAQ, the London Stock Exchange (LSE), and foreign exchange markets, large orders receive more favorable prices than small ones.

This raises a fundamental question: What accounts for the very different pricing on these markets? Recent research by Illinois Professor of Economics Dan Bernhardt and colleagues Vladimir Dvoracek of Simon Frazer University and Eric Hughson of the University of Utah has attempted to identify the differences in the structure of the markets that can explain why larger orders are charged higher prices on the NYSE but lower prices on NASDAQ and some other markets.

One fundamental difference in the design of the exchanges is in the timing of competition for a stock. When a trader places a buy order for stock on the NYSE, everyone willing to fill the order bids simultaneously for the business; and the best price prevails. The simultaneous bidding spurs competition and ensures that the party that most values the order fills it.

In contrast, on NASDAQ, the LSE, or foreign exchange markets, while dealers post price quotes (good only for very small orders) on electronic computer screens, the trader placing the buy order must then pick a dealer and telephone him. The actual price is negotiated one-on-one over the telephone.

There are no other bidders to spur competition. In fact, practicing dealers report that a trader with a large order may not even call the dealer with the best quote listed. The result is that for a single order there is essentially no price competition between market makers. Rather than call the dealer with the lowest listed price, the trader may call one who has a better reputation based on the prices he offers his clients and the volume he is willing to handle at those prices. Such a dealer needs to handle amounts of stocks over the quote size posted without charging a premium. Also, dealers with good reputations often discount the posted quotes during negotiations.

If the trader placing the order is unhappy with the price offered, he can either accept it (but switch dealers in the future), or call another dealer and try to negotiate a better price. Calling another dealer takes time—and there is also the risk that information about the stock and the order can leak to others in the market. The consequence of this system is that competition for orders takes place over a period of time. What determines the price for a buy order placed is the relative value that the dealer places on retaining the trader’s business. If the dealer offers a good enough price, he will continue to receive the trader’s orders. If he offers a bad price, he can earn immediate profits on the order; but he is likely to lose the trader’s business in the future. The price discount that the dealer will offer the trader, then, depends on how the dealer values the immediate profits from gouging the trader against the value of offering a better price and maintaining an enduring trading relationship. According to Bernhardt and his colleagues, dealers will offer a greater discount to traders who submit orders more frequently—the value of their future business is greater. They also show that traders who submit orders more frequently also tend to submit larger orders. Thus, larger orders get discounts.

The reason that frequency of ordering is more of a factor in a favorable price than size of the order is that there may be some large orders from traders who either have not yet established enduring relationships, or do not intend to. As a consequence, traders who receive greater breaks on quoted prices are more likely to stay with their customary dealers than to follow the best price quoted. It is to be expected that dealers who trade infrequently will be much more conscious of the quoted price, because they will receive a smaller discount during negotiations.

Berhardt and his colleagues then explored the related consequences of the NASDAQ market design. One implication is that the dealer who most values the order is less likely to get it—a trader will prefer to submit his order to his "standard" dealer to help maintain the ongoing relationship and receive better prices in the future. Consequently, there is more repeat business among dealers and traders on NASDAQ, as well as the London Stock Exchange and foreign exchange markets. The Securities and Exchange Commission has recently targeted NASDAQ in several investigations, and several lawsuits have been filed alleging collusion among dealers. Currently, Bernhardt and fellow Illinois economics professor Charles Kahn are in the process of investigating how the design of NASDAQ can make anti-competitive collusion easier. The basic argument is that the incentives to offer better price quotes on NASDAQ are weaker because, since many traders channel their orders to their usual dealers, the quotes draw only a limited number of additional orders.

Dan Bernhardt, a professor of economics, pursues research interests including industrial organization of financial markets, competition among financial intermediaries, and human capital aquisition. His co-researchers teach at Simon Fraser University and the University of Utah. For further information on this article or other research, contact Dan Bernhardt, danber@cba.uiuc.edu.