The Market Valuation of Corporate Reputation

Transcription

1 Corporate Reputation Review Volume 3 Number 1 The Market Valuation of Corporate Reputation Ervin L. Black and Thomas A. Carnes University of Arkansas, Fayetteville Vernon J. Richardson University of Kansas, Lawrence ABSTRACT We provide evidence that corporate reputation has value relevance, as measured by its ability to explain the rm's market value of equity at the end of the scal period. Corporate reputation is assessed using a summary measure from the Fortune survey of `America's most admired companies.' The Fortune measure serves as a proxy for intangible assets, such as internally generated goodwill, customer service, and intellectual capital. We demonstrate that this summary measure of non- nancial information adds to market value, even after controlling for the nancial performance `halo' e ects on the Fortune ratings. INTRODUCTION Measurements of corporate reputation appear nowhere in the nancial statements, yet it is apparent to the most casual of observers that corporations expend vast amounts of time and money to burnish their reputations. Such expenditures, of course, are undertaken with the expectation of resultant bene ts to the corporation. The bene ts may not always be nancial in a direct sense (eg, corporate sponsorship of a community event may be a `good neighbor' gesture that does not translate into increased sales). But it has been shown that favorable reputations have rm-speci c nancial bene ts to corporations by reducing the mobility of industry rivals (Caves and Porter, 1977; Wilson, 1985); by allowing rms to charge premium prices (Milgrom and Roberts, 1986); or by enhancing rm access to capital markets (Beatty and Ritter, 1986). Firms appear to be `involved in a competitive market for reputational status in which, because of information asymmetries, rms signal their key characteristics to constituents' (Fombrun and Shanley, 1990). Corporate reputation therefore meets the customary accounting de nition of an intangible asset, though it is not one that is speci cally identi able (in contrast to a patent or trademark). Instead, it is a qualitative asset, and the determination and timing of its future bene ts to the rm are extremely di cult to quantify, thereby posing serious valuation problems. Edvinsson and Malone (1997) refer to such assets as `invisible assets'. Despite the di culty in quantifying the worth of favorable corporate reputation, it should have value to the investor since it results in nancial bene ts to the corporation. As Edvinsson and Malone (1997) write, `Somehow, if only by hunches and intuitions, the market is putting a value on invisible assets. And some of these qualitative assets seem to hover in the ether almost inde nitely, converting to line items on the balance sheet years after the market has accounted for them.' There is evidence regarding the value-relevance of corporate reputation: Fortune reports that a 1988 investment in its ten most admired companies would have grown to nearly three times as much by 1998 as would have a comparable investment in the Standard Corporate Reputation Review, Vol. 3, No. 1, 2000, pp. 31±42 # Henry Stewart Publications, 1363±3589 Page 31

2 The Market Valuation of Corporate Reputation and Poor's 500. Antunovich and Laster (1999) nd that a portfolio of the top decile of Fortune's most-admired rms earns an abnormal return of 3.2 per cent in the year after the survey is published and 8.3 per cent over three years. We provide an empirical analysis of the value-relevance of one widely known measure of corporate reputation, the annual Fortune survey of America's most admired companies. We examine the value of corporate reputation through use of a model which allows us to determine if corporate reputation is a signi cant explanatory component of market value of equity (MV). We take into account the Brown and Perry (1994) nding that there is a nancial performance halo that heavily in uences the list of Fortune's most admired companies and employ a modi ed form of their methodology, thereby segregating nancial measures (MV, risk and growth) in the Fortune rankings before employing the nal regression model. The Fortune summary measure thus serves as a proxy for an aggregation of non- nancial information, such as internally generated goodwill, customer service, product quality, and intellectual capital. Consequently, our research question represents a convergence of prior studies in a variety of business-related elds such as marketing, management, economics, organizational strategy and accounting that examine corporate reputation. We do this by showing there is nancial value to a successful investment in non- nancial attributes, even though these investments in many cases are impossible to measure nancially. As discussed by Shenkar and Yuchtman- Yaar (1997), corporate reputation is important in identifying potential partners in joint ventures and potential targets in mergers and acquisitions, in attracting and retaining higher-quality employees, in improving standing with `salient publics' (Perrow 1961), and in allowing businesses to charge a premium for their goods and services. Drumwright (1996) nds that advertising campaigns with social dimensions are not particularly e ective in increasing sales or pro ts, but are highly e ective in achieving company-oriented objectives such as motivating employees or communicating the company's mission. However, few accounting studies have examined whether there is nancial value to an investment in non- nancial attributes. As discussed in Fombrun and van Riel (1997), we recognize that reputations embody appraisals of two dimensions of rm e ectiveness Ð the economic performance and the rm's social responsibilities Ð and we provide evidence of the nancial bene t to a rm e ectively meeting the latter. We nd evidence of value-relevant information provided by the Fortune survey that cannot be explained by nancial information available from nancial reporting sources. The signi cance of the non- nancial reputation component provides evidence of an invisible intangible asset that is value-relevant in explaining the market value of the rm and appears to demonstrate the validity of corporate decisions to make reputation-enhancing expenditures. The remainder of this paper is organized as follows. First, we review pertinent research on corporate reputation. We then discuss research methodology, provide results of the study, and present concluding remarks. LITERATURE REVIEW Tangible assets, as well as some classes of intangible assets (eg, patents and trademarks), can be easily transferred, thereby allowing valuation through an arms' length transaction. Corporate reputation, an intangible asset that is developed internally for the most part and that is not easily transferable to other parties, does not lend Page 32

3 Black, Carnes and Richardson itself to such easily determinable valuation. Reputation-building, though, appears to be valuable to rms. Reputations represent publics' cumulative judgments of rms over time, according to Fombrun and Shanley (1990), who conclude that these judgments `stratify industries, with potentially signi cant competitive advantages accruing to rms with higher perceived reputational status.' They theorize that conditions of incomplete and ambiguous information and heterogeneous publics make reputations relevant, and they nd that rms compete for reputation in such a market. They study the 292 rms included in the 1985 Fortune survey and nd that assessments of reputation appear to be positively related to prior accounting pro tability, advertising intensity, and size, and negatively related to prior performanceadjusted risk. The importance of nancial information, such as accounting pro tability and risk, on the Fortune measure implies that there exists a nancial performance halo in the magazine's index of corporate reputation. Since individual raters appear to be so heavily in uenced by previous nancial performance, previous attempts to determine the relationship between speci c attributes of corporate reputation and rm value may be victimized by a circularity Ð rm value improves reputation which improves rm value ad in nitum. Brown and Perry (1994) develop a statistical method to remove a signi cant portion of that halo in order to alleviate the perceptual distortion it adds to the data. After forming a halo index based upon various measures of recent nancial performance, they determine rm-speci c residuals that represent the portion of the Fortune ratings not explained by nancial performance. They apply this method to the rms in the 1992 Fortune survey and validate their results by comparing the halo-removed rating on four speci c attributes (eg, product quality or innovation) to independent evaluations of corporate performance on similar attributes. They conclude their method is especially valuable for researchers using Fortune's ratings to test `the impact of (non- nancial) attributes on corporate nancial performance or stock market returns.' Shefrin and Statman (1995, 1998) hypothesize that rms rated highly with the Fortune measure will underperform the market, since the survey typically gives high ratings to rms with high market-tobook ratios and large market capitalization Ð characteristics associated with low stock returns by Fama and French (1992), among others. However, their empirical results provide little support for this hypothesis. Antunovich and Laster (1999) sort rms into seven portfolios based upon their annual Fortune reputation measure and form stock portfolios in order to determine whether abnormal returns are associated with the rankings. They nd that the mostadmired decile signi cantly outperforms the least-admired decile for up to ve years after the survey and conclude that their ndings are an example of investors underreacting to publicly available information. Empirical investigations of limited aspects of corporate reputation have been undertaken by Ittner and Larcker (1998), Beatty and Ritter (1986), McGuire, Sundgren and Schneeweis (1988), and McGuire, Schneeweis and Branch (1990). Ittner and Larcker examine the value relevance of customer satisfaction in order to determine whether measures of satisfaction are leading indicators of accounting performance and whether the release of such measures provides new or incremental information to the stock market. Their study includes customer service measures at three levels: (1) a study of 2,491 customers buying a speci c telecommunications service from a single rm Page 33

4 The Market Valuation of Corporate Reputation (2) a business-unit analysis of customer satisfaction from 73 retail branch banks of a major US nancial services provider (3) rm-level data from the American Customer Satisfaction Index. They nd the relationships between customer satisfaction and future accounting performance generally are positive, but some are non-linear, showing evidence of diminishing bene ts at high levels of satisfaction. They also nd that public release of these measures is statistically associated with excess stock market returns. Beatty and Ritter (1986) examine the underpricing of initial public o erings and nd that it is related to the reputational capital of investment bankers. They conclude that investment bankers have nonsalvageable reputational capital at stake, and this enforces the underpricing equilibrium, as they show the market penalizes underwriters who cheat on this equilibrium by underpricing too much or too little. McGuire, Sundgren and Schneeweis (1988), using Fortune's ratings as a proxy for corporate social responsibility, nd that perceptions of social responsibility are more closely associated with prior nancial performance than with subsequent nancial performance. These results may re ect the `halo e ect' referred to by Brown and Perry (1994), as the authors fail to control for the in uence nancial variables have upon the Fortune reputation measure. McGuire, Schneeweis and Branch (1990) extend these results to all eight dimensions of perceived rm performance contained in the Fortune measure and nd that high returns (both return on assets and market returns) are highly correlated with subsequently high rm image Ð the `halo e ect' of Brown and Perry (1994). However, they nd that the Fortune metric `has little value as a forecaster of future rm nancial performance.' Our research extends the existing literature by undertaking an explicit search for market e ects of non nancial factors a ecting corporate reputation. We employ the Fortune `America's most admired corporations' summary reputation measure, thereby investigating a broader de nition of reputation than the speci c aspect studied by Ittner and Larcker (1998). By applying a variation of the model developed by Brown and Perry (1994), we remove the `halo e ect' of nancial performance and isolate the other in uences upon the Fortune rankings in order to extend the research of McGuire, Sundgren and Schneeweis (1988), McGuire, Schneeweis and Branch (1990), Shefrin and Statman (1995, 1998) and Antunovich and Laster (1999). METHODOLOGY The market value of equity (MV) is the expectation of the expected future cash ows (CF) accruing to stockholders discounted at the appropriate risk-adjusted rate r:? MV t = ~(1 + r t ) ±1 E t [CF t ] (1) t=0 MV can be disaggregated into the market value of assets and liabilities. A commonly used proxy for the market value of assets and liabilities is the book value of assets and liabilities. However, due to conservatism, historical costs, and the exclusion of most intangible assets, accounting book values are unlikely to capture completely the implications of their market value counterparts. Therefore, Equation 1 becomes: MV t = Assets t ± Liabilities t + Intangible Assets t (2) where intangible assets represent, for example, the discounted cash ows accruing to shareholders due to reputation Page 34

5 Black, Carnes and Richardson e ects, competitive advantages, anticipated growth opportunities, or positive net present value projects. Fortune annually provides a reputation summary measure of America's largest corporations in its `America's most admired corporations' issue. Brown and Perry (1994) nd that this rating, which is a measure of the rm's competitive advantage and reputation, is heavily in uenced by the rm's previous nancial results, thus creating a `halo e ect.' This e ect must be removed, or controlled for, before these qualitative reputational measures can be used appropriately in this study to measure the value relevance of the intangible asset unexplained by nancial reporting data. Brown and Perry nd several nancial factors that a ect the Fortune summary measure (SCORE): return on assets (ROA), market-to-book value (MKTBV), size, growth, and risk. Adding in dummy variables for each year to control for other economic e ects yields Equation 3: 96 SCORE t = a 0 + ~ DMY y + a 1 ROA t + y=82 a 2 MKTBV t + a 3 SIZE t + a 4 BETA t + a 5 GROWTH t + e t (3) where: SCORE t = Corporate reputation Score in year t taken from Fortune magazine annual reputation survey. ROA t = (Income Before Extraordinary Items De ated by Total Assets) *100 in year MKTBV t t. = Market Value of Common Equity De ated By Book Value of Common Equity at the end of year t. SIZE t = the natural log of the Market Value of Common Equity in year t. BETA t = Market Model Beta computed as 5-year (60- month) time period, ending at the end of year t (as computed by Compustat). GROWTH t = (Sales t-5 ± Sales t )/ Sales t. The residuals and estimated coe cients from this model are used to test the valuerelevance of the non nancial factors (the invisible assets) represented by the SCORE. The estimated coe cients are used to calculate a predicted reputation score which measures the portion of the Fortune rating score that nancial statement users can obtain from nancial performance data. The residuals from estimating Equation 3 represent the intangible assets related to reputation e ects that a rm has which cannot be obtained through the analysis of the nancial statements and market data. We use a valuation model (used by Burgstahler and Dichev 1997 and Barth, Beaver and Landsman, 1998) that expresses, in general form, market value of equity (MV) for rm i in year t, as a linear function of recognized net assets in place (BV) and unrecognized net assets, UNA: MV it =a 1 BV it +a 2 UNA it (4) If book values of recognized assets equal their fair values and fair values are wellde ned as in a setting economically equivalent to perfect and complete markets, UNA equals the present value of incremental cash ows of unrecognized net assets and a 1 and a 2 each equal one. If book values do not equal their fair values, then UNA also includes the di erence between fair and book values of recognized net assets; ie, it would include the di erence in the fair and book values of assets and the value of the intangibles, such as market reputation. Thus, in the more realistic Page 35

6 The Market Valuation of Corporate Reputation setting of imperfect and incomplete markets, UNA re ects the di erence between assets' values-in-use over entry or exit values, and a 1 and a 2 need not equal one. Because revenues and expenses relating to unrecognized net assets, including any excess of values-in-use over entry or exit values, can be re ected in net income (NI), net income is a proxy for UNA (see Bernard 1994; Barth and Landsman 1995, and Ohlson 1995). We modify Equation 4 to control for annual e ects and to include a measure of intangible assets by including the non- nancial factors estimated from Equation 3. This measure of the market reputation of the company that is beyond what can be determined by nancial factors is a proxy for the value of corporate reputation not measured in the accounting measure of the book value of net assets or common equity. Thus, our estimating equation becomes: 96 MV t = b 0 + ~ DMY y + b 1 BV t + y=82 b 2 NI t + b 3 NonFREP t±1 + SIZE t + n t (5) where, MV t = Market Value of Common Equity at year t DMY y = Annual year dummy variables set equal for 1 if t = y; 0 otherwise. BV t = Book Value of Common Equity at year t NI t = Net Income Before Extraordinary Items at year t NonFREP t±1 = Non- nancial Component of Corporate Reputation Ð derived by taking the residual value from the estimation of rm reputation in Equation 3. SIZE t = A control for size, measured as total assets at year t. The value-relevance of the non nancial component of corporate reputation, NonFREP, is estimated in Equation 5. We use market value of equity at the end of the year t as the dependent variable. It is expected that the market will have impounded all current and past information into this value. The Fortune survey results are published early in year t, but apply to the survey taken during t-1. Thus, if the intangible asset component of the survey results is value-relevant, NonFREP should be signi cant. From the previous discussion we expect this to be true; therefore, our main hypothesis becomes: H: We expect that the non- nancial component of the Fortune summary measure is value relevant, ie, that the coe cient on NonFREP will be signi cant and positive. Sample selection The sample consists of all of the rms rated by Fortune magazine in the 1982 through 1996 `America's most admired corporations' rating published early each of the following years. For each year included in the sample, Fortune sought ratings of corporate excellence from several thousand top executives, outside directors, and securities analysts. These experts were asked to rate the ten largest rms (measured by sales) in their own industry or economic sector, comparing them to competitors with respect to eight key attributes of reputation: innovativeness, quality of management, employee talent, quality of products/services, long-term investment value, nancial soundness, social responsibility, and use of corporate assets. The ratings are on a zero (poor) to ten (excellent) scale. Fortune typically gets a response rate of about 50 per cent to its survey, and it publishes the results every March. In addition to the Fortune rating, we obtained other nancial data from Compustat PCPlus. These sample selection criteria Page 36

7 Black, Carnes and Richardson result in a sample size of 2,905 rm-year observations for Equation 3 and 2,769 rm-year observations for estimation of Equation 5. RESULTS We rst estimate Equation 3 in order to control for the `halo e ect' described by Brown and Perry (1994). Each of the variables in Equation 3 is statistically signi cant, with signs the same as found by Brown and Perry. Descriptive statistics for the variables used in estimating Equation 3 are found in Table 1. The residuals from Equation 3, which represent the non nancial reputation component, are saved for use in Equation 5. Coe cients for the estimation of Equation 2 are found in Table 2. The mean of these residuals, as expected, is equal to 0. 1 We then estimate Equation 5 in order to determine whether the non- nancial reputation component of the Fortune survey is incrementally value-relevant once we control for the nancial information. In the rst column of Table 4, results of the estimation of Equation 5 are reported. Non- nancial components of the Fortune reputation score are incrementally valuerelevant at p=.001, as are BV and income. The signi cant positive coe cient on the non- nancial components (NonFREP) provides evidence that information provided by the Fortune survey is value relevant beyond what is provided by nancial information available from nancial reporting sources. This non- nancial reputation component is evidence of an invisible intangible asset that is valuerelevant in explaining the market value of the rm. Table 1: Descriptive statistics for rms used in Equation 3 Variable Mean Median Std. Deviation SCORE t ROA t MKTBV t SIZE t BETA t GROWTH t N = 2,905 Firm-year observations De nition of Variables: SCORE t = Corporate reputation Score in year t taken from Fortune Magazine Annual Reputation Survey. ROA t = (Income Before Extraordinary Items De ated by Total Assets) *100 in year t. MKTBV t = Market Value of Common Equity De ated By Book Value of Common Equity in year t. SIZE t = the natural log of the Market Value of Common Equity in year t. BETA t = Market Model Beta computed as 5-year (60-month) time period, ending at the end of year t (as computed by Compustat). GROWTH t = (Sales t-5 Ð Sales t )/Sales t. Page 37

8 The Market Valuation of Corporate Reputation Table 2: Estimation of fortune reputation score Score 0 = a 0 + a 1 ROA t + a 2 MKTBV t + a 3 SIZE t + a 4 BETA t + a 5 GROWTH t + e t (3) Variable Coe cient T-statistic Constant ROA t MKTBV t SIZE t BETA t ± GROWTH t Adjusted R 2 : 39.82% F= Note: Statistically signi cant annual dummy variables are estimated (not reported) over pooled cross-section samples of rm data. De nition of Regression Variables: SCORE t = Corporate reputation Score in year t taken from Fortune Magazine Annual Reputation Survey. ROA t = (Income Before Extraordinary Items De ated by Total Assets) *100 in year t. MKTBV t = Market Value of Common Equity De ated By Book Value of Common Equity in year t. SIZE t = the natural log of the Market Value of Common Equity in year t. BETA t = Market Model Beta computed as 5-year (60-month) time period, ending at the end of year t (as computed by Compustat). GROWTH t = (Sales t-5 Ð Sales t )/Sales t. As a form of sensitivity analysis, we test Equation 5 using various combinations of the independent variables. These results also are reported in Table 4. The second column contains the results of the basic model (Equation 4), omitting any variables related to reputation. As expected, both book value and income have signi cant positive coe cients. When the reputation score from Fortune is included in the model, its coe cient is signi cant and positive (see column 3, Table 4), indicating that the score is incrementally value-relevant beyond income and book value. We also nd that the non- nancial component of reputation, NonFREP, continues to be positive and signi cant (see column 4 of Table 4) even when the nancial component of reputation is included. 2 We also note the increase in the adjusted R 2 from 74.9 per cent to 76.2 per cent when the reputation score, SCORE t-1, is disaggregated between the non- nancial and nancial components (see columns 3 and 4 of Table 4). 3 It is also interesting to note the di erent coe cients and valuations given to a one-point change in score ($2.052bn in column 3) versus a one-point change in the non- nancial component and nancial components ($1.015bn and $4.137bn, respectively, in column 4). Page 38

9 Black, Carnes and Richardson Table 3: Descriptive statistics for variables used in market value model Variable Mean Median Std. Deviation MV t BV t INCOME t SCORE t NonFREP t FREP t N= 2,769 Firm-year Observations De nition of Variables: MV t = Market Value of Common Equity at year t. BV t = Book Value of Common Equity at year t. NI t = Net Income Before Extraordinary Items at year t. NonFREP t-1 = Non nancial Component of Corporate reputation Ð derived by taking the residual value from the estimation of rm reputation in Equation 3. FREP t-1 = Financial Component of Corporate reputation Ð derived by taking the residual value from the estimation of rm reputation in Equation 3. SCORE t-1 = Corporate reputation Score in year t taken from Fortune Magazine Annual Reputation Survey. CONCLUSIONS An important asset of the rm is its corporate reputation. Such factors as internally generated goodwill are the result of past reputation-enhancing activities and must be constantly maintained. We provide evidence of the market's ability to value this invisible intangible asset, using the annual Fortune survey of `America's most admired companies,' even after controlling for the portion of the Fortune ranking explained by nancial reporting information. Our ndings add support to existing research that internally generated intangibles not currently recognized as assets contribute to rm value and thus are viewed as assets by investors. These ndings have implications for external accounting reporting, as many critics of generally accepted accounting principles, such as Edvinsson and Malone (1997), decry the failure of nancial statements to value many intangible assets that are critically important to modern corporations. Our demonstration that one widelypublicized summary measure of such intangible assets is highly correlated with market value Ð even after controlling for nancial performance Ð provides further evidence of the relevance of non-gaap measures of rm-speci c information. This provides support for IASC Statement No. 38, `Intangible Assets' and for the position of the American Accounting Associations Financial Accounting Standards Committee (1998). We also provide evidence to support the conclusions of such authors as Fombrun and van Riel (1997) and Shenkar and Yuchtman-Yaar (1997) that reputations are valuable rm assets. A valid question for the inclusion of corporate reputation as an intangible asset Page 39

10 The Market Valuation of Corporate Reputation Table 4: Market valuation of non- nancial components of corporate reputation (Equation 5) 96 MV t =b 0 + ~ DMY y +b 1 BV t +b 2 N1 t b 3 NonFREP t±1 +b 4 SIZE t + n t y=82 Coe cient Coe cient Coe cient Coe cient (t-statistic) (t-statistic) (t-statistic) (t-statistic) Constant ± ± (4.10) (4.20) (±12.73) (±17.60) BV t (31.51) (31.74) (31.83) (30.92) NI t (26.73) (27.17) (24.63) (21.19) NonFREP t (5.91) (6.41) FREP t (19.16) SCORE t (15.67) SIZE t ±.03 ±.03 ±.02 ±.02 (±6.29) (±6.35) (±4.86) (±3.40) N F-statistic Adj. R % 72.7% 74.9% 76.2% De nition of Regression Variables: MV t = Market Value of Common Equity at year t. BV t = Book Value of Common Equity at year t. NI t = Net Income Before Extraordinary Items at year t. NonFREP t-1 = Non nancial Component of Corporate reputation Ð derived by taking the residual value from the estimation of rm reputation in Equation 3. FREP t-1 = Financial Component of Corporate reputation Ð derived by taking the residual SCORE t-1 value from the estimation of rm reputation in Equation 3. = Corporate reputation Score in year t taken from Fortune Magazine Annual Reputation Survey. SIZE t = Total assets in year t. Variables for which no coe cient was reported were not considered in that version of Equation 5. in the nancial statement continues to be the recognition criteria. While we provide some evidence on the relative worth of an incremental increase in corporate reputation, we do not provide a method for evaluating and measuring, in dollar terms, an individual rm's reputation. However, a rm could choose to disclose non- nancial factors related to its internally generated intangible assets in the Management Discussion and Analysis section of the annual report, or in other ways that are appropriate. Avenues for future research include Page 40

11 Black, Carnes and Richardson isolation and valuation of speci c factors a ecting corporate reputation in order to determine which non nancial information is more value-relevant to the markets and investigation of the market e ects of shifts in reputation. The longevity, or permanence, of these assets is also an area of interest to researchers, managers, and investors. ACKNOWLEDGMENT The authors would like to thank Mark Hirschey, Jim Guthrie and Kelly Welch at the University of Kansas, participants at the Central States Accounting Workshop, and workshop participants at the University of Arkansas for their helpful comments on an earlier version of this paper. ENDNOTES 1 We also ranked the residuals and used the ranking rather than the residual, with no qualitative e ect upon the results. 2 It is possible that corporate reputation varies by industry. As a test of robustness, we include industry dummy variables in the estimation of models 3 and 5 and nd similar results. As another test of robustness, we control for the possible endogeneity between market value and the estimation of the Fortune reputation score by estimating the incremental market valuation of the nancial (FREP t±1 ) and non- nancial components of reputation (NonFrep t±1 ) to still be positive and signi cant. 3 As an additional control for size, we converted all our data to z-scores and performed the analysis using these scores. The tenor of the reported results was not changed. REFERENCES Antunovich, P. and Laster, D. S. (1999) `Do Investors Mistake a Good Company for a Good Investment?', Federal Reserve Bank of New York sta reports, No. 60. Barth, M. E., Beaver, W. H. and Landsman, W. R. (1998) `Relative Valuation Roles of Equity Book Value and Net Income as a Function of Financial Health', Journal of Accounting and Economics, 25: 1± 34. Barth, M. E. and Landsman, W. R. (1995) `Fundamental Issues Related to Using Fair Value Accounting for Financial Reporting', Accounting Horizons, 9: 97±107. Beatty, R. P. and Ritter, J. R. (1986) `Investment banking, reputation, and underpricing of initial public o erings', Journal of Financial Economics, 15: 213±232. Bernard, V. (1994) `Accounting-based valuation methods, determinants of market-to-book ratios, and implications for nancial statement analysis', Working paper, University of Michigan. Brown, B. and Perry, S. (1994) `Removing the nancial performance halo from Fortune's ``Most Admired Companies'',' Academy of Management Journal, 37: 1347±1359. Burgstahler, D. C. and Dichev, I. D. (1997) `Earnings, adaptation, and equity value', The Accounting Review, 72: 187±215. Caves, R. E. and Porter, M. E. (1977) `From entry barriers to mobility barriers', Quarterly Journal of Economics, 91: 421±434. Drumwright, M. E. (1996) `Company advertising with a social dimension: The role of noneconomic criteria', Journal of Marketing, 60: 71±87. Edvinsson, L. and Malone, M. S. (1997) `Intellectual capital: Realizing your company's true value by nding its hidden brainpower', HarperBusiness, New York. Fama, E. F. and French, K. R. (1992) `The crosssection of expected stock returns', Journal of Finance, 47: 427±465. Fombrun, C. and Shanley, M. (1990) `What's in a name? Reputation building and corporate strategy', Academy of Management Journal, 33: 233±258. Fombrun, C. and van Riel, C. (1997) `The Reputational Landscape', Corporate Reputation Review, 1: 5±13. Ittner, C. D. and Larcker, D. F. (1998) `Are non nancial measures leading indicators of nancial performance? An analysis of customer satisfaction', Journal of Accounting Research, Supplement: 1±35. McGuire, J., Schneeweis, T. and Branch, B. (1990) `Perceptions of rm quality: A cause or result of rm performance?' Journal of Management, 16: 167±180. McGuire, J., Sundgren, A. and Schneeweis, T. (1988) `Corporate social responsibility and rm nancial performance', Academy of Management Journal, 31: 854±872. Milgrom, P. and Roberts, J. (1986) `Relying on the information of interested parties', Rand Journal of Economics, 17: 18±32. Ohlson, J. A. (1995) `Earnings, Book Values and Dividends in Security Valuation', Contemporary Accounting Research, 11: 661±687. Perrow, C. (1961) `Organizational prestige: Some Page 41

12 The Market Valuation of Corporate Reputation functions and dysfunctions', American Journal of Sociology, 66: 335±341. Shefrin, H. and Statman, M. (1995) `Making sense of Beta, Size, and Book-to-Market', Journal of Portfolio Management, 21: 26±34. Shefrin, H. and Statman, M. (1998) `Comparing Expectations about Stock Returns to Realized Returns', working paper, Santa Clara University. Shenkar, O. and Yuchtman-Yaar, E. (1997) `Reputation, image, prestige and goodwill: An interdisciplinary approach to organizational standing', Human Relations, 50: 1361±1381. Wilson, R. (1985) `Reputations in games and markets', In `Game-theoretic models of bargaining', edited by A. E. Roth, Cambridge University Press, New York. Page 42