Advertising, R&D expenditures and the market value of the firm

Advertising, R&D expenditures and the market value of the firm – Mergers and Acquisitions

Keith W. Chauvin

This paper provides evidence that advertising and research and development (R&D) expenditures have large, positive and consistent influences on the market value of the firm. Like information on current cash flows, data on advertising and R&D spending appear to help investors form appropriate expectations concerning the size and variability of future cash flows. As a result, spending on advertising and R&D can be viewed as a form of investment in intangible assets with predictably positive effects on future cash flows.

While the significant market value effects of advertising and R&D are generally apparent for all COMPUSTAT firms, such aggregate evidence has the potential to obscure meaningful differences across firm size classes and industry groups. Consistent with the fact that only a handful of firms are responsible for substantial advertising and R&D spending, the valuation effects of advertising and R&D are most uniformly evident in the case of large firms. In fact, the valuation effects of advertising and R&D are typically greater for larger as opposed to smaller firms in both manufacturing and nonmanufacturing sectors. These findings suggest that size advantages make advertising and R&D relatively more profitable for larger firms. Nevertheless, smaller firms do not appear to be precluded from making profitable investments in advertising and R&D. Some evidence emerges to suggest that the well-targeted advertising and R&D efforts of the smallest firms in the economy can be highly profitable. I. Investment Aspects of Advertising and R&D Expenditures

An emerging body of research considers the market value effects of a wide range of corporate investment decisions. McConnell and Muscarella |15~, for example, find that unanticipated increases in planned capital expenditures have a positive effect on the market value of the firm and that unanticipated decreases have a negative impact. Statman and Sepe |20~ report a correspondingly positive market reaction to the termination of investment projects that have poor prospects, while Brickley and Van Drunen |4~ find a negative reaction to operating unit liquidations. Blackwell, Marr, and Spivey |2~ find a significantly negative stock market reaction to plant-closing announcements. Chan, Martin, and Kensinger |5~ and Doukas and Switzer |7~ report interesting evidence on the stock market reaction to a very specific type of corporate investment decision: announcements of increased R&D spending. Chan, Martin, and Kensinger |5~ and Doukas and Switzer |7~ find statistically significant R&D announcement day returns, especially in the case of large “high-tech” industrial firms that devote substantial resources to R&D. The positive share-price reaction to announcements of increased R&D spending discovered in these studies can be taken as evidence of a strong link between R&D spending and the market value of the firm, a link that is especially robust in the case of large firms. To further investigate this relation, our paper considers the cross-sectional influences of both advertising and R&D expenditures on the market value of the firm. This paper considers how the traditionally recognized valuation effects of current cash flow, growth, risk and market share are augmented when both advertising and R&D are considered as potentially important sources of intangible capital. To the extent that advertising and R&D expenditures represent a type of investment expenditure that gives rise to economic benefits lasting more than one year, a market value influence can be anticipated. Of course, current cash flows may also reflect, at least in part, the positive effects of previous investments in advertising and R&D. Once the valuation effects attributable to current cash flows are controlled, any incremental valuation effects of current advertising and R&D expenditures represent evidence of intangible capital or asset-like influences.(1)

This paper builds on related research by Hirschey |10~, Jose, Nichols, and Stevens |12~, Lustgarten and Thomadakis |14~, Morck, Shleifer, and Vishny |17~, and Morck and Young |18~. Each of these studies reports positive market value effects of advertising and R&D expenditures that are consistent with a forward-looking perspective of stock market investors. Unfortunately, each of these studies is restricted by data considerations to fairly limited samples of large firms. Data is now available that makes a much more complete analysis possible, including consideration of the differences in the effectiveness of advertising and R&D across firm size classes and between manufacturing and nonmanufacturing industry groups.

II. Economic Characteristics of Advertising and R&D

A. Advertising and R&D Spending by Industry Group

Exhibit 1 shows the distribution of advertising and R&D expenditures by COMPUSTAT firms organized according to broad two-digit industry groups.(2) These data are firm averages over the 1988-1990 period. The purpose of Exhibit 1 is to illustrate that the distribution of advertising and R&D is fairly skewed. The distribution of advertising intensity and R&D intensity across all industries is clearly not uniform. While both advertising and R&D are important methods of intangible capital investment, it is relatively rare to find firms and industries that employ both methods in tandem.

Advertising expenditures and advertising intensity are especially high among toy companies included in COMPUSTAT’s miscellaneous manufacturing industry (such as Tonka Corp., Mattel, Inc., and Hasbro, Inc.), motion picture companies, and firms that provide amusement and recreation products. Advertising intensity is also high among firms providing educational services, food and kindred products, tobacco, and leather products. Little, if any, advertising is noted for firms in heavy manufacturing TABULAR DATA OMITTED sectors, such as construction, mining, petroleum refining, and so on.

R&D spending is even more concentrated than advertising; a few high-tech sectors account for the overwhelming share of R&D activity. R&D spending is highest in industries such as industrial machinery and computing equipment, measuring instruments, photography, electronic equipment, and chemicals and allied products. Little R&D activity takes place in most business and consumer service industries, the financial sector, and retailing. In fact, COMPUSTAT reports zero R&D activity tot firms in 31 out of 63 two-digit industry groups. By way of contrast, at least some advertising is reported for firms in 47 out of 63 industry groups.

It is relatively rare to find companies that report high levels of advertising and R&D activity. Exceptions to this rule are provided by firms that offer educational services, chemicals and allied products, industrial machinery and computer equipment, electronic equipment, measuring instruments, and transportation equipment. It is clearly more typical to find firms that employ advertising or R&D, but not both, as an effective means of product differentiation.

The most obvious implication of these data is that advertising and R&D must be considered as investments in alternative means of product differentiation that have the potential to differ in their effectiveness across industry groups. In particular, it is relevant to ask whether or not the valuation effects of advertising and R&D differ for firms in manufacturing (20 |is less than or equal to~ SIC |is less than~ 40) versus nonmanufacturing (SIC |is less than~ 20 or SIC |is greater than or equal to~ 40) industries. Based on spending patterns, it is clear that R&D activity is concentrated among manufacturers, whereas advertising promotion tends to be more broadly dispersed across both manufacturing and nonmanufacturing sectors. These data alone suggest that R&D spending may be an effective means of new product development and product differentiation for manufacturers, whereas advertising may be a broadly effective means of product differentiation for both manufacturing and nonmanufacturing firms. B. Firm Size, Advertising and R&D Spending

Exhibit 2 shows advertising and R&D expenditures by the 20 largest firms ranked in terms of aggregate advertising and R&D expenditures. Average advertising expenditures of $1.2 billion by the 20 largest firms ranked by advertising expenditures are roughly 43 times greater than the $27.7 million dollar average advertising expenditure reported for all sample firms in Exhibit 1. R&D spending is even more highly concentrated among large firms than is advertising. Average R&D expenditures of $2.1 billion by the 20 largest firms ranked by R&D expenditures are roughly 91 times greater than the $23.1 million dollar average R&D expenditure for the entire sample of firms. Over the 1988-1990 period, the 20 largest firms accounted for an average of $24.6 billion in advertising, or 41.8% of the $58.9 billion spent on advertising by sample firms. At the same time, the 20 largest firms account for $42.3 billion in R&D expenditures, or 45.9% of the $92.2 billion spent on R&D by all firms included in the sample. Not only do the 20 largest firms consistently account for a substantial share of advertising and privately financed R&D, roughly the same group of large firms is found at the top of the advertising and R&D expenditure lists year after year. Unilever, Phillip Morris, General Motors and Procter & Gamble are perennial leaders in advertising expenditures. Similarly, General Motors, IBM, Ford and AT&T regularly appear as leaders in terms of R&D spending.

Based on spending patterns, it is clear that advertising and R&D spending tends to be relatively concentrated among large firms. By itself, the fact that a relative handful of firms is responsible for a substantial portion of advertising and privately financed R&D suggests important size advantages (economies of scale) or economies of scope in advertising and R&D. This evidence provides support for the hypothesis that advertising and R&D spending is likely to be generally more effective for relatively larger firms. Exhibit 2 also shows advertising and R&D expenditures by the 20 largest firms ranked by advertising intensity and R&D intensity. Advertising expenditures claim a notable share of revenues for smaller firms seeking to penetrate consumer products industries dominated by large, diversified competitors. The leading firms, as ranked by advertising intensity, devote an average of 20.44% of sales to advertising, or roughly 13 times the COMPUSTAT average of 1.6%. Even more impressive is the share of revenues devoted to R&D by small biotechnology, pharmaceutical, and software companies. It is simply astounding that the 20 largest firms, as ranked by R&D intensity, devote 41.6% of sales to R&D, or roughly 52 times the COMPUSTAT average of 0.8%. These data suggest the presence of meaningful threshold levels of advertising and R&D spending that must be reached for effective competition in such sectors. Alternatively, meaningful economies of scope may be present in these industries. In either case, significant large-firm advantages to advertising and R&D are suggested once again.

TABULAR DATA OMITTED

III. Methodology and Data

A. Sources of Capitalized Market Value

Of primary concern to this study is the role played by advertising and R&D as determinants of the current market value of the firm. To isolate such influences. the effects of other factors that might affect current market value must be constrained, including cash flow, growth, risk, and market share considerations. Current cash flow is taken as the best available indicator of a firm’s ability to generate cash flows during future periods. Cash flow is measured by operating income before depreciation minus interest expense, taxes, preferred dividends and common dividends. Miller and Modigliani |16~ argue that growth has a positive effect on market values if future investments are expected to cam above-normal rates of return, and if growth is an important determinant of these returns. Growth is measured by the least-squares estimate of the three-year rate of growth in sales for each firm. While growth affects the magnitude of anticipated excess returns, a valuation influence may also be associated with the degree of stability such returns exhibit. Risk is measured in this paper using the logarithm of the ratio of the 52-week high and low stock price for each firm, an index that is proportional to the Garman and Klass |8~ “ideal” volatility estimator. With an increase in risk, the market value of expected returns will fall (Thomadakis |22~). Thomadakis |22~, among others, considers the valuation effects of market share data as a potentially important indicator of the firm’s ability to earn economic rents tied to market power or superior efficiency. Market share is defined as total revenue in the firm’s primary-product four-digit SIC industry expressed as a percentage of industry sales.

B. Measuring Capitalized Market Value

Alternative cross-sectional measures of the market value of the firm can be employed to learn whether investors recognize long-term or asset-like characteristics of advertising and R&D expenditures. Among the most familiar of these are Tobin’s “q ratio” (Brainard and Tobin |3~), defined as the market value of the firm normalized by the replacement cost of tangible assets; and Thomadakis’ |22~ “relative excess valuation” (EV/S), measured as the market value of common minus the book value of stockholders’ equity, all normalized by sales. Both are attractive variants of the ratio of the market value of the firm divided by the book value of tangible assets studied by Miller and Modigliani |16~.

While q and EV/S provide appealing market-based views of investor expectations concerning the firm’s future profit potential, neither is completely free from the use and/or abuse of accounting conventions. Both q and EV/S are market rum accounting-based measures that are subject to measurement errors in the valuation of tangible assets. The q data are subject to error to the extent that flaws persist in accounting replacement cost data, just as EV/S is subject to error given that accounting book values measure imperfectly the economic value of tangible assets. While it is commonly presumed that accounting replacement costs provide a measure of tangible assets that is superior to traditional book value data, this is not necessarily true. Indeed, Watts and Zimmerman |23~ argue, from a market-based perspective, that accounting replacement cost numbers are irrelevant for security pricing purposes. It follows that replacement cost adjustments to book value data have the potential to obscure, rather than make more precise, the level and determinants of the capitalized value of the firm (Landsman and Magliolo |13~). In addition to their obvious exposure to accounting measurement error, both q and EV/S are subject to accounting bias as well. As Schwert |19~ argues, such measures are likely to reflect. at least in part, the effects of accounting policy decisions. If firms capitalize nonproductive assets so as to smooth or hide monopoly profits, both q and EV/S will be correspondingly biased. To provide an unbiased framework for analysis, this study considers the effects of advertising and R&D on the market value of common equity without any accounting-based adjustments. Both q and EV/S can be viewed as accounting-adjusted and size-adjusted valuation measures, where the replacement value of tangible assets and sales, respectively, are employed for size normalization purposes. Therefore, this study focuses on the market value implications of advertising and R&D in a manner that minimizes the potential for accounting error or bias.

Based on these considerations, a general model is suggested:

Market Value = |a.sub.0~ + |b.sub.i~Cash Flows + |b.sub.2~Growth + |b.sub.3~Risk + |b.sub.4~Market Share + |b.sub.5~Advertising + |b.sub.6~R&D + e. (1) Size normalization is accomplished in this study by deflating (dividing) both dependent and independent variables for each firm i by total sales revenue raised to the 1.5 power, |Mathematical Expression Omitted~. This method of size normalization was chosen on an empirical basis; |Mathematical Expression Omitted~ is the log likelihood ratio-minimizing deflator.

C. The Data

This study focuses on cross-sectional samples of COMPUSTAT firms over the three-year period from 1988 through 1990, inclusive. An average of roughly 1,500 firms per year are considered. Firm-level rather than industry-level data are analyzed to ensure an exact match between advertising, R&D and market value figures. This approach allows for a detailed consideration of the role played by firm size and industry conditions and has the advantage of avoiding industry classification errors for widely diversified firms. Such an analysis seems worthwhile in light of findings reported by Acs and Audretsch |1~ who study the large-firm and small-firm share of innovation in 247 four-digit SIC industries. While not directly tested in their analysis, Acs and Audretsch |1~ suggest that meaningful size-based differences in the effectiveness of R&D spending may be observed for relatively small firms. The potential for size-based differences in the valuation effects of advertising is also of interest; it lies at the heart of suggestions that high advertising intensity creates barriers to entry for new rivals and barriers to mobility for established nonleading firms. Of similar interest is the potential for industry characteristics to influence the valuation effects of both advertising and R&D. For example, the basic research typical of high-tech firms is quite different from the applied research or developmental work typical of the R&D effort of firms in low-tech sectors. Similarly, the valuation effects of price and product promotion can also vary depending upon the potential and ability of advertisers to build durable brand loyalty. To capture such influences, industry effects can be broadly indicated using intercept dummy variable interactions for two-digit SIC industry group classifications. Such two-digit SIC dummy variables have the effect of isolating the market value effects of advertising and R&D from other industry-specific valuation considerations. In addition, the basic valuation model can be analyzed over samples of advertising-intensive and R&D-intensive industries to learn the extent to which the valuation effects of advertising and R&D expenditures are mitigated by substantial promotional and innovative activity by competitors. By analyzing the overall sample of firms, in addition to a simple two-part breakdown for manufacturing versus nonmanufacturing firms, it becomes possible to learn the extent to which expenditures on advertising and R&D have broad rather than narrow implications for the value of the firm. By considering the market value implications of a three-part sample partition according to firm size (measured by sales revenue), the extent to which firm size plays a role in determining the market value effects of advertising and R&D can also be learned. III. Empirical Estimation

A. Comparative Results for the Manufacturing and Nonmanufacturing Sectors

After allowing for potentially important simultaneous influences among relations describing market values and other important elements of market structure, Connolly and Hirschey |6~ find no significant endogenous influences. Therefore, the model described previously is estimated through straightforward application of ordinary least squares (OLS). The model is estimated for each year within the 1988-1990 period for both manufacturing and nonmanufacturing sectors, as well as over pooled cross-sectional samples of all available firms for the 1988-1990 period. In each annual regression, two-digit SIC code intercept dummy variables are employed to control for industry-related differences in the market value of the firm. In each pooled cross-sectional sample regression, these two-digit SIC code intercept dummy variables are supplemented with annual dummy variables to control for transitory influences related to overall stock market conditions. Using an F-test, the hypotheses of no industry-related or time-related effects can be easily rejected at the |Alpha~ = 0.01 level for each sample analyzed. Using an F-test, the hypothesis of no difference in estimation results for manufacturing versus nonmanufacturing industries can also be rejected at the |Alpha~ = 0.01 level. Since the annual and pooled sample regressions are uniform in terms of the inferences they provide, only pooled sample results are reported in the interest of saving space.

Exhibit 3 offers broad-based evidence of long-lived benefits to advertising and R&D by showing their valuation influences for size-based samples of firms taken from manufacturing and nonmanufacturing sectors. According to census classification criteria, manufacturing firms fail within two-digit SIC code industry groups between 20 (food and kindred products) and 39 (miscellaneous manufacturing industries). From Exhibit 1, it is clear that R&D activity is concentrated among firms from within the manufacturing sector, whereas advertising expenditures are more broadly dispersed among both manufacturing TABULAR DATA OMITTED and nonmanufacturing firms. By itself, the relatively concentrated nature of R&D activity suggests the potential for a greater level of R&D effectiveness in the manufacturing sector. On the other hand, it is likely that advertising expenditures give rise to effective product differentiation in a broad range of industries.

Despite some differences in explanatory power between the manufacturing and nonmanufacturing sectors, it is clear from Exhibit 3 that a simple model can describe a meaningful share of the firm-by-firm variation in market values. In the manufacturing sector, more than one-half of the variation in market values can be attributed to variation in cash flows, growth, market share, advertising and R&D. In the nonmanufacturing sector, roughly one-third of the variation in market values can be similarly attributed; the only exception being the replacement of growth with risk as an important consideration. Given the relative consistency of valuation effects due to advertising and R&D in the manufacturing and nonmanufacturing sectors, it appears safe to argue that both types of expenditures give rise to intangible capital. From an economic perspective, both advertising and R&D constitute alternative forms of investment in profitable intangible assets. An economically appropriate accounting treatment would therefore involve a capitalization and amortization treatment typical of other forms of capital expenditures. Public policy that allows an immediate expense treatment of advertising and R&D expenditures for tax purposes, rather than a more economically appropriate capitalization and amortization treatment, results in an implicit tax subsidy for advertising and R&D intensive firms (see Hirschey and Weygandt |11~). Therefore, this finding of long-lived benefits for advertising and R&D expenditures is of practical interest to both firms and public policy-makers.

And finally, in contrast to some popular assumptions that stock market investors are myopic in their focus on short-run performance, these findings suggest that investors evaluate the advertising and R&D effort of firms within a long-term perspective.

B. Firm Size Effects

Acs and Audretsch |1~ find higher levels of innovative activity in industries that feature relatively large firms, but that the elevated level of creative endeavor emanates from smaller rather than larger competitors. Grilliches |9~ reports that small (less than $2 million R&D) firms appear more “efficient” than their larger competitors in that they are able to generate a relatively larger number of patents per dollar of R&D expenditure. As suggested by Scherer |19~, the most favorable climate for rapid technological progress may be an environment that includes large firms with the capacity to undertake ambitious developments complemented by a large number of small (below $500 million sales) technologically oriented competitors.

These findings suggest the potential for interesting differences in the effectiveness of R&D expenditures according to firm size. Similarly, it is worth investigating the potential for firm size effects on the effectiveness of advertising expenditures. To the extent that advertising and R&D effectiveness is influenced by firm size considerations, the potential exists for the market value impact of advertising and R&D to differ according to firm size. To the extent that economies of scale or other size advantages in advertising and R&D are present, the market value effect of a dollar in advertising and/or R&D expenditures will be greater for larger as opposed to smaller firms. Conversely, to the extent that diseconomies of scale or other size disadvantages in advertising and R&D are present, the market value effect of a dollar of advertising and/or R&D expenditures will be moderated for relatively larger firms.

Exhibit 3 illustrates the influence of firm size on the market value effects of advertising and R&D expenditures using a simple three-part breakdown of each sample according to sales revenues. For example, over the entire COMPUSTAT sample, the valuation model is estimated for subsamples of large-size (|greater than or equal to~ $1,424.2 million in sales, $7,782.3 million average), medium-size ($384.8 million |is less than~ middle |is less than~ $1,424.2 million, $764.5 million average), and small-size (|less than or equal to~ $384.0 million, $184.6 million average) firms. Samples of manufacturing and nonmanufacturing firms are divided in a similar fashion. From Exhibit 3, it is clear that the valuation effects of advertising and R&D are consistently positive and of a comparable order of magnitude for large firms. By way of contrast, some variability is present in terms of the valuation effects of advertising and R&D sponsored by smaller firms. Interestingly, in the nonmanufacturing sector, no significant valuation effects are associated with advertising by medium-size firms nor with small-firm R&D. After controlling for other important valuation effects, evidence reported in Exhibit 3 suggests that the market value influence of advertising does indeed depend upon firm size considerations. Using a t-test of the difference in coefficients, the superiority of large-size firm over medium-size firm advertising is statistically significant in the manufacturing sector (t = 4.76), nonmanufacturing sector (t = 4.10), and for all COMPUSTAT firms (t = 4.46). Similarly, the superiority of small-size firm over medium-size firm advertising is evident in the manufacturing sector (t = 6.45), nonmanufacturing sector (t = 3.54), and for all COMPUSTAT firms (t = 5.33). Taken together with data reported in Exhibit 2, these findings suggest that size advantages are relevant in determining the valuation effects of advertising. At the same time, it does not appear that large-size firm advantages preclude effective advertising by smaller, specialized firms. Indeed, the advertising of large firms is more effective than small-size firm advertising only in the manufacturing sector (t = 2.19).

The influence of firm size on the stock market valuation of R&D expenditures is also evident in the findings reported in Exhibit 3. Again, using a t-test of the difference in coefficients, the superiority of large-size over small-size firm R&D is statistically significant in the manufacturing sector (t = 1.36), nonmanufacturing sector (t = 5.99), and for all COMPUSTAT firms (t = 4.02). The superiority of large-size firm over medium-size firm R&D is also evident in the manufacturing sector (t = 3.13). Therefore, especially when the R&D-intensive manufacturing sector is considered, large firm advantages in R&D are unmistakable. The superiority of medium-size over small-size firm R&D is less evident, but significant in the nonmanufacturing sector (t = 2.07), and for all COMPUSTAT firms (t = 1.85). The ability of smaller firms to conduct effective R&D should not be overlooked, however, as evidenced by the fact that superior small-size firm over medium-size firm R&D is evident in the manufacturing sector (t = 3.67).

Therefore, based upon a market value perspective, the advertising and R&D activity of larger firms appears to be relatively more effective than that of smaller companies. These results are consistent with the previously discussed event-study results on R&D reported by Chan, Martin, and Kensinger |5~ and Doukas and Switzer |7~. Just as the market’s response to R&D announcements is more favorable in the case of larger spending increases reported by large firms, the market capitalization of advertising and R&D seems to generally increase with firm size.

C. Industry Effects

Lustgarten and Thomadakis |14~, building upon earlier research by Thomadakis |22~, find that the cross-sectional relation of Tobin’s q to firm characteristics can depend upon market structure influences. Thus, the potential exists for the market value implications of firm-specific features, such as market share, advertising and R&D, to vary with industry conditions. In support of this hypothesis, Doukas and Switzer |7~ suggest that the benefits accruing from R&D expenditures may be greater for firms operating in industries where the share of industry sales concentrated among the four largest firms is relatively high.

Advertising and R&D coefficient estimates in a valuation model of this type, that differ according to firm size, offer evidence of size-based differences in advertising and R&D effectiveness. Moderately reduced effects of advertising and R&D on the market value of smaller firms may reflect the fact that such efforts are often the focus of rivalry and imitation by bigger and better-financed rivals. Alternatively, market value differences across smaller firms may simply be tied more closely to the fortunes of specific industries, whereas large diversified firms are more broadly positioned to take advantage of scale advantages in advertising, and both chance and intended discoveries. To further investigate the extent to which the basic valuation model results may be influenced by industry-related considerations, Exhibit 4 reports the market value effects of variations in the levels of advertising and R&D expenditures for a variety of advertising-intensive and R&D-intensive (or “high-tech”) industries. In both instances, industries were arbitrarily chosen on the basis of having advertising intensity or R&D intensity of at least two percent of sales as shown in Exhibit 1, plus a sufficient number of observations for meaningful analysis. Based upon these criteria, six industries were chosen as representative of the advertising-intensive and R&D-intensive (“high-tech”) sectors. Chemicals and allied products (SIC 28) has the distinction of appearing in both categories; it is the sole industry group that displays high levels of advertising intensity and R&D intensity. In addition to industry-specific results, findings are also reported for pooled cross-sectional samples of all firms included within each of the advertising-intensive and high-tech categories. And finally, results are reported for all COMPUSTAT firms that devote more than two percent of sales to advertising (“all advertising-intensive firms”) or more than two percent of sales to R&D (“all high-tech firms”). When samples of firms are selected on the basis of having high levels of advertising intensity or R&D intensity, as is the case here, then the amount of sample variability in both advertising intensity and R&D intensity is necessarily constrained. This makes it relatively difficult to find statistically significant market value effects of advertising in samples of advertising-intensive industries. Similarly, it is relatively difficult to find statistically significant market value effects of R&D in samples of R&D-intensive industries. It is therefore not surprising to find some industry-by-industry variability in the market value effects of advertising and R&D, as shown in Exhibit 4. Nevertheless, despite some variability within various advertising-intensive industries, advertising intensity has generally positive market value effects across all advertising-intensive industries and advertising-intensive firms. Similarly, despite some variability within individual high-tech industries, R&D intensity has strongly positive market value effects across all high-tech industries and high-tech firms.(3)

TABULAR DATA OMITTED

IV. Conclusions

Positive effects of advertising and R&D expenditures on the market value of the firm are illustrated in this paper for broadly representative samples of COMPUSTAT firms. Like current cash flow, growth, risk, and market share, advertising and R&D expenditures constitute key determinants of the market value of the firm. The market value effects of advertising and R&D are broadly operative throughout both manufacturing and nonmanufacturing sectors. As such, advertising and R&D appear as attractive alternative means of investment in valuable intangible capital that have differing degrees of relevance in different economic sectors.

Findings reported here also suggest the potential for interesting differences in the effectiveness of advertising and R&D expenditures according to firm size. Consistent with R&D event-study results reported by Chan, Martin, and Kensinger |5~ and Doukas and Switzer |7~, size advantages exist in advertising and R&D activity; the market value effect of a dollar in these types of expenditures tends to be greater for relatively larger firms. Still, the evidence also suggests that advertising and R&D expenditures by smaller, specialized firms can be highly effective. While there are obviously important differences between advertising and R&D, these two types of expenditures can be regarded as alternative forms of investment in intangible capital that contribute to shareholder value.

1 Hirschey and Weygandt |11~ show that if economic amortization of the exponential decay type can be assumed, along with constant percentage rates of growth in advertising and R&D expenditures, then the magnitude of intangible capital equals annual expenditures on advertising or R&D multiplied by a constant. Given these assumptions, the stock of intangible advertising and R&D capital is strictly proportional to the flow of advertising and R&D expenditures, and the valuation effects of each type of expenditure can be taken as indicative of intangible capital influences.

2 COMPUSTAT annual data items are taken from the full coverage version of COMPUSTAT PC Plus for the 1988-1990 period. The overall sample is limited to those firms having nonegative book values and at least $100 million in market capitalization (share price times the number of shares outstanding, price-adjusted to December 31, 1990). The market value screen is used to avoid market value estimation problems for firms that are inactively traded in thin markets. The COMPUSTAT firms studied in this paper include those responsible for an aggregate average $58.9 billion in advertising expenditures per year, or 77.5% of the $75.9 billion aggregate ave rage for all COMPUSTAT firms over the 1988-1990 period. Firms included are responsible for an aggregate average $92.2 billion in R&D expenditures per year, or 78.8% of the $117 billion aggregate average for all COMPUSTAT firms over the 1988-1990 period.

3 Valuation effects of advertising and R&D may be influenced by factors that affect the degree of congruence between managerial and stockholder interests. To study the possibility that the market value effects of advertising and R&D may be sensitive to agency problem considerations, the basic valuation model was tested for both independent and interactive effects of the percentage of closely held shares, measured by the share of stock held by insiders and large outside interests. However, no consistent and material influence of the percentage of closely held shares on the valuation effects of advertising and R&D was noted (at the 0.05 level). This suggests that the role played by agency problems in modeling the valuation effects of advertising and R&D is somewhat tangential to the main thrust of this research.

References

1. Z.J. Acs and D.B. Audretsch, “Innovation in Large and Small Firms: An Empirical Analysis,” American Economic Review (September 1988), pp. 678-690.

2. D.W. Blackwell, M.W. Marr, and M. Spivey, “Plant-Closing Decisions and the Market Value of the Firm,” Journal of Financial Economics (August 1990), pp. 277-288.

3. W. Brainard and J. Tobin, “Pitfalls in Financial Model Building,” American Economic Review (May 1968), pp. 99-122.

4. J.A. Brickley and L.D. Van Drunen, “Internal Corporate Restructuring: An Empirical Analysis,” Journal of Accounting and Economics (January 1990), pp. 251-280.

5. S.H. Chan, J.D. Martin, and J.W. Kensinger, “Corporate Research and Development Expenditures and Share Value.” Journal of Financial Economics (August 1990), pp. 255-276.

6. R.A. Connolly and M. Hirschey, “Firm Size and R&D Effectiveness: A Value-Based Test,” Economics Letters (March 1990), pp. 277-281.

7. J. Doukas and L.N. Switzer, “The Stock Market’s View of R&D Spending and Market Concentration,” Journal of Economics and Business (May 1992), pp. 95-114. 8. M.B. Garman and M.J. Klass, “On the Estimation of Security Price Volatilities From Historical Data,” Journal of Business (January 1980), pp. 67-78. 9. Z. Grilliches, “Patent Statistics as Economic Indicators: A Survey,” Journal of Economic Literature (December 1990), pp. 1661-1707.

10. M. Hirschey, “Market Structure and Market Value,” Journal of Business (January 1985), pp. 89-98.

11. M. Hirschey and J.J. Weygandt, “Amortization Policy for Advertising and Research and Development Expenditures,” Journal of Accounting Research (Spring 1985), pp. 326-335.

12. M.L. Jose, L.M. Nichols, and J.L. Stevens, “Contributions of Diversification, Promotion, and R&D to the Value of Multiproduct Firms: A Tobin’s q Approach,” Financial Management (Winter 1986), pp. 33-42.

13. W. Landsman and J. Magliolo, “Cross-Sectional Capital Market Research and Model Specification,” Accounting Review (October 1988), pp. 586-604.

14. S. Lustgarten and S. Thomadakis, “Mobility Barriers and Tobin’s q,” Journal of Business (October 1987), pp, 519-537.

15. J.J. McConnell and C.J. Muscarella, “Corporate Capital Expenditure Decisions and the Market Value of the Firm.” Journal of Financial Economics (September 1985), pp. 399-422.

16. M. Miller and F. Modigliani, “Dividend Policy, Growth and the Value of Shares,” Journal of Business (October 1961), pp. 411-433.

17. R. Morck, A. Shleifer and R.W. Vishny, “Management Ownership and Market Valuation: An Empirical Analysis,” Journal of Financial Economic’s (January/March 1988), pp. 293-315.

18. R. Morck and B. Yeung, “Why Investors Value Multinationality,” Journal of Business (April 1991), pp. 165-187.

19. F.M. Scherer, Industrial Market Structure and Economic Performance, Chicago, Rand McNally College Publishing, 1980.

20. G.W. Schwert, “Using Financial Data to Measure Effects of Regulation,” Journal of Law and Economics (April 1981), pp. 121-158.

21. M. Statman and J.F. Sepe, “Project Termination Announcements and the Market Value of the Firm,” Financial Management (Winter 1989), pp. 74-81.

22. S.B. Thomadakis, “A Value-Based Test of Profitability and Market Structure,” Review of Economics and Statistics (May 1977), pp. 179-185.

23. R.L. Watts and J.L. Zimmerman, “On the Irrelevance of Replacement Cost Disclosures for Security Prices,” Journal of Accounting and Economics (August 1980), pp. 95-106.

Keith W. Chauvin is Assistant Professor of Business and Mark Hirschey is Professor of Business, both at the University of Kansas, Lawrence, Kansas.

COPYRIGHT 1993 Financial Management Association

COPYRIGHT 2004 Gale Group