Macro Economic Risk Factors

An executive summary for managers and executive readers can be found at the end of this article

Macro-economic risk factors in industrial markets: are eÂlite firms less susceptible? Richard A. Heiens

Assistant Professor of Marketing, School of Business Administration, University of South Carolina Aiken, Aiken, South Carolina, USA

Mark Kroll

Professor of Management, College of Administration and Business, Louisiana Tech University, Rusten, Louisiana, USA

Peter Wright

Professor of Management, Fogelman College of Business and Economics, The University of Memphis, Memphis, Tennessee, USA Keywords Macroeconomics, Risk, Arbitrage, Competitive advantage Abstract As far back as 1947, Alfred Marshall proposed that the disparity in income between those individuals with moderate ability and those with greater ability is larger than the disparity in talent. Building on Marshall's thesis, argues that marginal differences in firm capability may result not only in increased profitability, but also in lower susceptibility to macro-economic risk factors for basic manufacturing firms in industrial markets. The results seem to suggest that the firms with greater ability have in fact managed to combine resources in such a way as to create inimitable advantages. Specifically, through a commitment to product and process innovation and modern manufacturing facilities, the most successful firms in the study have been able to acquire key resources, and gain extensive control over the value creation process. The outcome is high relative product quality, relative pricing power, and lower susceptibility to macroeconomic risk.

Disparity in income

Introduction As far back as 1947, Alfred Marshall proposed that the disparity in income between those individuals with moderate ability and those with greater ability is larger than the disparity in talent. Building on Marshall's thesis, we argue that marginal differences in firm capability may result not only in increased profitability, but also in lower susceptibility to macro-economic risk factors for basic manufacturing firms in industrial markets. The basis for this conjecture is as follows. Initial marginal differences in firm capability may translate into marginally better products (Miles and Snow 1978; 1986). Over time, those firms with better quality outputs may build a reputation for superior quality, and might even further enhance their quality differential. Superior product quality may enable the firm not only to increase its market share (achieving scale economies and lower costs), but also to implement premium pricing policies (Buzzell and Gale, 1987). This may lead to increased profitability. In turn, higher profits may enable the firm to continue to invest in firm activities which might further lower costs and enhance differentiation All authors contributed equally to this article. The research register for this journal is available at http://www.mcbup.com/research_registers The current issue and full text archive of this journal is available at http://www.emerald-library.com/ft

246

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001, pp. 246-257, # MCB UNIVERSITY PRESS, 0885-8624

through progressively superior outputs. Because the most capable, or ``eÂlite'', firms can defend themselves better against general economic trends by forcing the burden of macroeconomic or industry decline on to their weaker rivals (Lubatkin and Rogers, 1989), they are likely to have not only higher profits, but also lower risk (Kroll et al., 1999). Risk Risk has been conceptualized in various ways. Most often, however, scholars tend to conceptualize risk either as the variance in accounting returns (Cool et al., 1989; Jemison, 1987) or as market related systematic and firm specific unsystematic risk, according to the Capital Asset Pricing Model (CAPM) (Kroll et al., 1999; Lubatkin and Rogers, 1989). Measuring risk

Many researchers, however, criticize the technique of measuring risk via CAPM (Fama and French, 1992; Jegadeesh, 1992). Specifically, because stock risks might have more than two dimensions associated with them, partitioning a firm's risk into systematic and unsystematic risk may be too narrow a view (Mei, 1993). In light of this limitation, an alternative measure of risk may be provided by the Arbitrage Pricing Theory (APT) (Ross, 1976). According to the APT, required rates of return are a function of a firm's sensitivity to several macro-economic factors. As further developed by Roll and Ross (1980; 1984), APT attributes a firm's required rate of return to as many as four economic forces: (1) the size of the spread between long-term and short-term interest rates; (2) the level of price inflation; (3) the level of industrial production; and (4) the yield spread between high risk and low risk bonds (i.e. the prevailing risk premium being demanded by investors).

Macro-environmental risk factors

Hypotheses Several studies have suggested that the most capable firms in a given industry have less susceptibility to macro-environmental risk factors. For example, many of the most capable firms in a given industry tend to create greater value through vertical integration. Because vertically integrated firms may have control over buying and selling costs, as well as distribution, production, and transaction costs, they might also experience lower levels of risk (Lubatkin and Chatterjee, 1994; Chatterjee et al., 1992). In addition, more profitable firms are likely to have the resources to make higher commitments to both process and product R&D, which tends to build entry barriers that might insulate such firms from competitive pressures and, concurrently, lower their risk levels (Amit and Livnat, 1989; Miller and Bromiley, 1990). Consequently, we offer the following hypotheses with respect to the most capable or ``eÂlite'' firms in industrial markets: H1: Elite firms' returns will be significantly less influenced by changes in the size of the spread between long-term and short-term interest rates than non-eÂlite firms. H2: Elite firms' returns will be significantly less influenced by changes in the level of price inflation than non-eÂlite firms. H3: Elite firms' returns will be significantly less influenced by changes in the level of industrial production than non-eÂlite firms.

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

247

H4: Elite firms' returns will be significantly less influenced by changes in the yield spread between high risk and low risk bonds than non-eÂlite firms. Methodology Sample At present, the two best available data sets which contain large samples and detailed business-level data are the Federal Trade Commission's Line of Business database and the Strategic Planning Institute's PIMS database. In the present study, we selected a cross-sectional sample of 365 manufacturing businesses from the PIMS database for the 1970 through 1983 time period. These businesses, with SIC codes from 3080 to 3864, include a broad crosssection of firms operating in industrial markets. Specifically, these firms include manufacturers in rubber and plastics, stone, glass and clay, primary metals, fabricated metals, industrial machinery, electrical equipment, transportation equipment, and instrumentation. Heterogeneous businesses

It has been suggested that pooled samples of very heterogeneous businesses may seriously distort the observed relationships between variables owing to the possibility of unique industry-specific forces (Bass et al., 1978). Because the business units chosen all manufacture component parts, the present study allows for an examination of multiple-year data for a large sample of business units subject to similar industry forces. Variables included in the study The businesses participating in the PIMS study use a standardized reporting system to describe their strategies and market environments. In order to identify the most capable or ``eÂlite'' firms among the sample of manufacturing businesses included in the study, several strategic variables were selected from the PIMS database. Specifically, the ``Relative Quality'' variable was used to represent the quality performance relative to the competition for each business unit in the sample. In order to measure the commitment of each business unit to product and process innovation, product R&D as a percentage of revenues and process R&D as a percentage of revenues were included in the analysis. As an additional measure of the business unit's commitment to innovation, the percentage of sales derived from new products was also included.

Value creation process

In order to measure the extensiveness of the business unit's control over the value creation process, the percentage of the business unit's products value added was included. The Strategic Planning Institute defines value added as net sales minus total purchases (Buzzell and Gale, 1987). The ``Plant and Equipment Newness'' variable was included in order to assess the business unit's commitment to modern manufacturing facilities. This variable is defined as the ratio of net book value of plant and equipment to gross book value (Buzzell and Gale, 1987). Finally, in order to assess the business unit's relative pricing power, we included the ``Relative Price'' variable, which simply assesses the business unit's product prices in relation to their leading competitors. Identification of strategic clusters In order to segment the SBUs in the sample into ``eÂlite'' and ``non-eÂlite'' groups, cluster analysis was employed. Specifically, all SBUs were evaluated in terms of their strategic orientation with regard to the PIMS variables described in the previous section. A non-hierarchical cluster analysis, utilizing

248

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

Anderberg's (1973) nearest centroid sorting method was employed. According to MoÈller et al. (1985), the non-hierarchical technique tends to simultaneously produce homogeneous clusters in terms of within-group differences, and heterogeneous clusters in terms of between-group differences. Because the variables included in the study were based on various scales, the data were normalized by first calculating the sample SBUs' standard deviations from the sample means for each variable. As per Romesburg (1984), the resulting standard deviation values were then used in the cluster analysis. The technique resulted in two distinct clusters of SBUs. One group appeared to be made up of 62 ``eÂlite'' firms, or those firms scoring highest on the strategic variables included in the analysis. The other cluster contained the remaining 303 firms, or the less capable, ``non-eÂlite'' firms in the sample. Macro-economic factors

Sensitivities to macro-economic factors As mentioned earlier, Roll and Ross (1980; 1984) identified four important macro-economic factors which systematically influence securities markets: unanticipated movements in the shape of the interest rate term structure, unanticipated changes in inflation, unanticipated changes in the level of industrial production, and unanticipated shifts in risk premiums. In order to assess the two groups of SBUs' sensitivities to the four macroeconomic risk factors, we regressed four surrogate measures of the risk factors against both ROI and Par ROI values for both the ``eÂlite'' and ``non-eÂlite'' sample groups. This resulted in two regression models for each cluster. The data used to represent each factor in the regression models were as follows: (1) The GNP deflator, labeled as GNPD, was used to test the firm's ROI and Par ROI sensitivity to changes in the rate of inflation. A statistically significant, positive or negative coefficient in the regression model would indicate earnings sensitivity to inflation, whereas a statistically insignificant coefficient would indicate no such sensitivity. (2) The index of industrial production, labeled here as IIP, was used to represent changes in industrial activity. The resulting regression coefficient represents the sample groups' sensitivities to unanticipated changes in the level of industrial production. (3) The variable labeled as YIELD is used to represent unanticipated changes in the yield curve. This variable was estimated by subtracting the quarterly rate on 90-day Treasury Bills from the rate on long-term (ten-year) Treasury Bonds. (4) Changes in the risk premium on private sector bonds, labeled here as RISK, was estimated by subtracting the rate on long-term Treasury Bonds from the average rate earned on Moody's AAA-rated corporate bonds. Because government bonds are free of the risk of default, the spread between the yields on the two instruments should provide a fairly accurate estimate of the risk premium on private sector bonds.

Treasury bond yields

Despite the contention of Roll and Ross (1984) that the four macro-economic risk factors are clearly not perfectly correlated, Treasury bond yields are nevertheless used to estimate both the yield curve and the risk premium. Consequently, due to the possibility of collinearity in the data, the various regression models were estimated using the AUTOREG procedure available in SAS. In order to test for differences between the results of each set of regression models specific to each of the four macro-economic risk factors,

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

249

Chow tests were employed (Chow, 1960). Significant Chow Test results would suggest that the models' betas represent truly unique relationships.

Cluster analysis

Results Cluster analysis and strategic groups As discussed earlier, the cluster analysis conducted on the firms included in the sample resulted in two distinct strategic groups. As gauged by differences in the values of the strategic variables included in the study, these distinct groups appear to pursue very distinct strategic agendas. As suggested by the results presented in Table I, the firms labeled as ``eÂlite'' clearly invest more heavily than their ``non-eÂlite'' counterparts in R&D, new product introduction, and plant and equipment. Additionally, the ``eÂlite'' firms contribute more to the total value added of the finished product than their ``non-eÂlite'' counterparts. The net result of these efforts is that the ``eÂlite'' firms successfully pursue higher quality levels than the ``non-eÂlite'' firms and are able to charge premium prices for their goods. Performance differences between groups As reported by Buzzell and Gale (1987), firms with superior product quality tend to enjoy growing market shares and superior returns on investment. Similarly, the present study verifies this finding. As presented in Table II, ``eÂlite'' firms tended to experience significantly better ROIs, Par ROIs, and market share growth figures than ``non-eÂlite'' firms.

Strategic variables 1. Product R&D 2. Process R&D 3. Proprietary products 4. New plant and equipment 5. Relative price 6. Value added 7. Product quality

Cluster 1 (N = 62) eÂlite firms

Cluster 2 (N = 303) non-eÂlite firms

1.142 (1.061) 0.306 (0.356) 0.044 (0.075) 7.500 (8.081) 3.650 (3.378) 41.835 (12.713) 35.261 (13.483)

0.515 (0.628) 0.134 (0.204) 0.0036 (0.025) 5.000 (3.877) 1.749 (2.108) 28.734 (11.478) 5.95 (6.313)

Probability value 0.004 0.026 0.038 0.011 0.001 0.001 0.001

Table I. Strategic variable profiles of the two clusters means, standard deviations (in parentheses), and probability values Cluster 1 (N = 62) Cluster 2 (N = 303) eÂlite firms non-eÂlite firms Probability value

Performance measures 1. ROI 2. Par ROI 3. Change in market share

19.756 (11.456) 8.856 (4.039) 26.337 (29.730)

13.500 (7.468) 6.415 (3.940) 11.883 (18.364)

0.001 0.009 0.002

Table II. Performance profiles of the two clusters means, standard deviations (in parentheses), and probability values 250

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

Regression models

Differences in sensitivities to macro-economic risk factors As stated previously, two regression models were estimated for each cluster, one using ROI as the dependent variable, and the other using Par ROI as the dependent variable. The four resulting regression models are presented in Tables III and IV. For simplicity of presentation, the regression coefficients for the ``eÂlite'' and ``non-eÂlite'' firms have been paired. Based on the information presented in Table III, the results seem to support our a priori expectations of greater performance sensitivity to macro-economic changes on the part of the ``noneÂlite'' group. In fact, the ``non-eÂlite'' firms included in cluster 2 experience significantly negative reactions to unanticipated increases in inflation, unanticipated declines in industrial output, unexpected increases in longterm rates, and unexpected rises in risk premiums. In contrast, the ``eÂlite'' firms in cluster 1 demonstrate no significant sensitivities to changes in the four macro-economic risk factors.

Regression results

The regression results in which Par ROI is used as the dependent variable are presented in Table IV. Once again, the ``non-eÂlite'' firms tend to demonstrate greater sensitivity to the four macro-economic risk factors. However, as is apparent from the results presented in Table IV, ``eÂlite'' firms demonstrate some susceptibility to changes in industrial output and, to a lesser extent, changes in the level of inflation. The difference between these results and Independent variable GNPD IIP YIELD RISK

Cluster Elite firms Non-eÂlite firms Elite firms Non-eÂlite firms Elite firms Non-eÂlite firms Elite firms Non-eÂlite firms

Coefficienta

Probability value

Chow test

± 0.075 ± 0.156 0.025 0.053 ± 0.267 ± 0.799 ± 0.2777 ± 5.050

0.230 0.001 0.247 0.001 0.371 0.060 0.414 0.022

4.58b 5.94b 5.76b 5.81b

Notes: aDurbin-Watson results all approach 2.0, suggesting that auto-correlation was not a problem when using ROI as the dependent variable. bChow test significant at 0.05 level

Table III. Cluster sensitivities to macro-economic factors for return on investment Independent variable GNPD IIP YIELD RISK

Cluster Elite firms Non-eÂlite firms Elite firms Non-eÂlite firms Elite firms Non-eÂlite firms Elite firms Non-eÂlite firms

Coefficienta

Probability value

Chow test

± 0.051 ± 0.258 0.27 0.097 ± 0.475 ± 1.208 ± 4.505 ± 8.038

0.083 0.004 0.034 0.001 0.441 0.001 0.122 0.001

3.55b 2.15b 4.81b 3.55b

Notes: aDurbin-Watson results all approach 0.5, suggesting that auto-correlation in the models, and necessitating use of the AUTOREG procedure. bChow test significant at 0.05 level

Table IV. Cluster sensitivities to macro-economic factors for par return on investment JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

251

those reported for ROI probably stems from the structure of the Par ROI variable. Differences in profitability

The Strategic Planning Institute developed the Par ROI figures in order to control for a host of variables thought to cause differences in profitability across industries. In effect, Par ROI is intended to normalize performance data, permitting more accurate conclusions to be drawn when using pooled data from across the entire PIMS database. In an effort to correct for differences across industries, Par ROI includes adjustments made for vertical integration, pricing, and percent of sales from new products. Because these factors are included in our clustering procedure, the striking contrast between the two groups is not as obvious when Par ROI is used as the dependent measure. Nevertheless, as seen in Table IV, the ``non-eÂlite'' firms continue to emerge as the more economically sensitive and risky businesses. Managerial implications The results seem to suggest that the 62 ``eÂlite'' firms in cluster 1 have managed to combine resources in such a way as to create inimitable advantages. Barney (1991) argues that resources may be used to create sustainable competitive advantage, if those resources meet one of three criteria: (1) the firms' ability to acquire the resources is predicated upon ``unique historical conditions;'' in effect the firm is positioned to exploit key resources at an opportune time; (2) the ability of a firm to exploit a resource so as to achieve competitive advantage is not easily understood by those within or without the firm, and therefore not easily copied; or (3) the resource providing the competitive advantage is the product of a complex social process within the firm.

Sharing characteristics

The cluster analysis performed in the present study seems to indicate that firms which have been able to gain competitive advantage and establish themselves as eÂlite organizations share several characteristics. One important characteristic of the eÂlite firms in our sample was a commitment to innovation. Specifically, our eÂlite firms tended to spend more heavily than the non-eÂlite firms on both product R&D and process R&D as a percentage of revenues. In addition, eÂlite firms tended to derive a greater proportion of revenue from the sale of unique proprietary products and maintained higher relative product quality than competing firms. These firms also procured and maintained modern manufacturing facilities. Furthermore, the most successful firms in our study have been able to acquire key resources, and gain extensive control over the value creation process. As suggested by both conventional wisdom and empirical findings, the outcome is higher market share and higher return on investment. Although not included in the current study, numerous conceptual and empirical studies appear to suggest that pioneering or first-mover firms are often especially successful in achieving long-term competitive advantages. As a consequence, first-movers have frequently been shown to have higher market shares than market followers (Lambkin, 1988; Parry and Bass, 1990; Robinson, 1988). According to Lieberman and Montgomery (1988), firstmovers may gain competitive advantage through the preemption of various resources, such as technology, location and personnel, and through the

252

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

development of organizational capabilities that are crucial to the success of their products or services. Firms may strive to gain the competitive advantage often accruing to the first-mover in several ways. For example, the first firm to: (1) produce a new product; (2) use a new process; or (3) enter a new market can claim the distinction and potential benefits of being a first-mover (Lieberman and Montgomery, 1988). First-mover firms may also include those organizations that are the first to pursue opportunities deriving from the initiating of pricing changes or the adoption of new distribution ideas (Smith et al., 1989; McDaniel and Kolari, 1987). Competitive advantage

Whether predicated on managerial skill, order of entry, or perhaps merely unique historical conditions, some degree of competitive advantage is usually required in order to achieve superior market share and superior returns. The unique contribution of our study, however, is to demonstrate that competitive advantage also appears to insulate these firms from several important macro-economic risk factors. Specifically, firms with superior commitment to innovation, product quality, and the value creation process may minimize their susceptibility to unanticipated movements in the shape of the interest rate term structure, unanticipated changes in inflation, unanticipated changes in the level of industrial production, and unanticipated shifts in risk premiums. In summary, eÂlite firms are better able to protect their market positions when faced with economic downturns. One possible explanation is that customers prefer to make their purchases from larger firms and are predisposed to more frequently form loyalties towards the outputs of such firms (Donthu, 1994). Consequently, extending Alfred Marshall's Theorem, it may be accurate to say that the disparity in macro-economic risk susceptibility between those manufacturing firms with moderate ability and those with greater ability may be even larger than the disparity in income.

Lower susceptibility

One important implication that lower susceptibility to risk has for managers is that lower risk levels tend to lower the discount rate that the market applies in valuing the firm's earnings stream. As a result, such firms can further enhance their value and create additional wealth for their shareholders through an expansion of possible investment options, as these firms lower the internal rate of return which investments must generate in order to be profitable (Kroll et al., 1999). In addition, other stakeholders, particularly the firm's managers, are likely to favor lower levels of macro-economic risk. Managers, in contrast with investors, can only invest their human capital in one firm at a time. Because high susceptibility to macro-economic risk factors increases the likelihood of firm failure and consequently job loss, firms with low susceptibility to macro-economic risk will be more appealing to talented senior executives (Martin and McConnell, 1991). References Amit, R. and Livnat, J. (1989), ``Efficient corporate diversification: methods and implications'', Management Science, Vol. 35, pp. 879-97.

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

253

Anderberg, M.R. (1973), Cluster Analysis for Applications, Academic Press, New York, NY. Barney, K. (1991), ``Firm resources and sustained competitive advantage'', Journal of Management, Vol. 17, pp. 99-120. Bass, F.M., Cattin, P. and Wittink, D.R. (1978), ``Firm effects and industry effects in the analysis of market structure and profitability'', Journal of Marketing Research, Vol. 15, pp. 3-10. Buzzell, R.D. and Gale, B.T. (1987), The PIMS Principles: Linking Strategy to Performance, Free Press, New York, NY. Chatterjee, S., Lubatkin, M. and Schoenecker, T. (1992), ``Vertical strategies and market structures: a systematic risk analysis'', Organization Science, Vol. 3, pp. 138-56. Chow, G.C. (1960), ``Tests of equality between sets of coefficients in two linear regressions'', Econometrica, Vol. 28, pp. 591-605. Cool, K., Dierickx, I. and Jemison, D.B. (1989), ``Business strategy, market structure, and riskreturn relationships: a structural approach'', Strategic Management Journal, Vol. 10, pp. 507-22. Donthu, N. (1994), ``Double jeopardy in television program choice'', Journal of the Academy of Marketing Science, Vol. 22 No. 2, pp. 180-85. Fama, E.F. and French, K.R. (1992), ``The cross-section of expected stock returns'', The Journal of Finance, Vol. 47, pp. 427-65. Jegadeesh, N. (1992), ``Does market risk really explain the size effect?'', Journal of Financial and Quantitative Analysis, Vol. 27 No. 3, pp. 337-51. Jemison, D.B. (1987), ``Risk and the relationship among strategy, organizational processes, and performance'', Management Science, Vol. 26, pp. 1238-49. Kroll, M., Wright, P. and Heiens, R.A. (1999), ``The contribution of product quality to competitive advantage: impacts on systematic variance and unexplained variance in returns'', Strategic Management Journal, Vol. 20 No. 4, pp. 375-84. Lambkin, M. (1988), ``Order of entry and performance in new markets'', Strategic Management Journal, Vol. 9 No. 4, pp. 127-40. Lieberman, M.B. and Montgomery, D.B. (1988), ``First-mover advantages'', Strategic Management Journal, Vol. 9 No. 4, pp. 41-58. Lubatkin, M. and Chatterjee, S. (1994), ``Extending modern diversification: does it apply?'', Academy of Management Journal, Vol. 33, pp. 756-79. Lubatkin, M. and Rogers, R.C. (1989), ``Diversification, systematic risk and shareholder return: a capital market extension of Rumelt's 1974 study'', Academy of Management Journal, Vol. 32, pp. 454-65. McDaniel, S.W. and Kolari, J.W. (1987), ``Marketing strategy implications of the Miles and Snow typology'', Journal of Marketing, Vol. 51, pp. 19-30. Marshall, A. (1947), Principles of Economics, Macmillan, New York, NY. Martin, K.J. and McConnell, J.J. (1991), ``Corporate performance, corporate takeovers, and management turnover'', Journal of Finance, Vol. 46 No. 2, pp. 671-87. Mei, J. (1993), ``A semi-autoregression approach to the arbitrage pricing theory'', The Journal of Finance, Vol. 48, pp. 599-619. Miles, R.E. and Snow, C.C. (1978), Organizational Strategy, Structure, and Process, McGrawHill, New York, NY. Miles, R.E. and Snow, C.C. (1986), ``Organizations: new concepts for new forms'', California Management Review, Vol. 28, pp. 66-73. Miller, K.D. and Bromiley, P. (1990), ``Strategic risk and corporate performance: an analysis of alternative risk measures'', Academy of Management Journal, Vol. 33, pp. 756-79. MoÈller, K.E., Lehtinen, J.R., Rosenquist, G. and Storbacka, K. (1985), ``Segmenting hotel business customers: a benefit clustering approach'', in Bloch, T.M., Upah, G.D. and Zeithaml, V.A. (Eds), Services Marketing in a Changing Environment, Chicago, IL, pp. 72-6. Parry, M. and Bass, F.M. (1990), ``When to lead and when to follow? It depends'', Marketing Letters, Vol. 1, pp. 187-98. Robinson, W.T. (1988), ``Sources of market pioneer advantages: the case of industrial goods industries'', Journal of Marketing Research, Vol. 25, pp. 87-94. 254

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

Roll, R. and Ross, S.A. (1980), ``An empirical investigation of the arbitrage pricing theory'', The Journal of Finance, Vol. 35, pp. 1073-103. Roll, R. and Ross, S.A. (1984), ``The arbitrage pricing theory approach to portfolio planning'', Financial Analyst's Journal, Vol. 41 No. 3, pp. 14-27. Romesburg, H.C. (1984), Cluster Analysis for Researchers, Lifetime Learning Publications, Belmont, CA. Ross, S.A. (1976), ``The arbitrage pricing theory of capital asset pricing'', Journal of Economic Theory, pp. 335-44. Smith, K., Guthrie, J. and Chen, M. (1989), ``Strategy, size, and performance'', Organizational Studies, Vol. 10, pp. 63-81.

&

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

255

256

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

JOURNAL OF BUSINESS & INDUSTRIAL MARKETING, VOL. 16 NO. 4 2001

257