Performance determinants of the Food, Beverages and Tobacco MNEs: A new
approach to testing firm level theory
Performance determinants of the Food, Beverages and Tobacco MNEs: A new
approach to testing firm level theory
Hesilda Discua Cruz 1474243
International Economics and Business
Faculty of Economics
University of Groningen, the Netherlands
ABSTRACT
What triggers firm performance of MNEs? Multinational enterprises are the key drivers and enablers of globalization. Previous studies suggest that firm specific factors are more important than corporate-parent effects in explaining performance of firms (Rumelt, Schendel & Teece, 1994a; McGahan & Porter, 1997b). From such standpoint, this study attempts to find the firm level factor(s) that have a significant weight in the influence of the financial performance of the multinational enterprises (MNEs) of the Food, beverages and tobacco industry.
Using a firm specific theory framework, firm size, operational knowledge, operating margin, capital intensity and board size are assessed to determine their weight on MNEs Return on Equity (ROE) by following the contrasting of different studies and their reviews made by earlier and recent accounts.
The contribution of this study falls on the testing of premises in mainstream theoretical considerations of firm specific factors to the application of real world scenarios by the measurement of ROE using the DuPont model of MNEs in the food, beverages and tobacco industries in 2004. There is a lack of studies tackling the approach of ROE as a performance proxy measured by the DuPont model, since the model is generally decomposed and tested against ROE by its three key financial ratios: Profit margin, total asset turnover and equity multiplier. Thus, this study contributes to the on-going discussion of existing research in academia on the factors that affect ROE as a financial performance measure by utilizing a more current firm sample than prior studies, and by covering a substantially different perspective by analyzing new variables such as operational knowledge and further examining the interactions of independent factors, how they are related to, and what implications do these interactions entail as determinants of ROE.
CONTENTS
I. INTRODUCTION ... 5
II THEORETICAL BACKGROUND ... 9
A. Consistent overview of Relevant and Related studies ... 9
B. The DuPont model ... 11
a. DuPont ratios... 12
III. THEORETICAL MODEL... 13
A. Dependent and Independent variables ... 13
a. Return on Equity ... 13 b. Firm size... 14 c. Operational knowledge ... 15 d. Operating margin ... 16 e. Capital intensity ... 17 f. Board size ... 18 B. Control variables ... 19 a. Country GDP... 19 b. Industry ... 20
IV. RESEARCH METHODS ... 23
A. Data collection and Sampling ... 23
B. Measures ... 25 C. Econometric techniques ... 27 V. EMPIRICAL RESULTS ... 28 A. Descriptive Statistics... 28 B. Correlations ... 29 C. Regression Results ... 31 VI. CONCLUSIONS ... 35
A. Aim and Questions... 35
B. Limitations and future research... 35
REFERENCES……….………..37
APPENDIX A ………...42
I. INTRODUCTION
In an increasing industrialized world, companies thrive for superior performance; we see multinationals encompass efficiency guidelines on which their goals and interests are set for the years to come. In the last two decades, there has been a higher importance in the performance of companies; such performance, both locally and abroad still drives researchers to further consider their inspection. Certainly, an inquiry ever present on the study of international businesses and economics is on the influences that a performance-driven strategy has on these companies. Illustration of the different aspects that need to be taken into consideration to understand the superior performance of multinationals can pose a puzzle to researchers.
Then, a pinpointing aspect in such endeavors would be to pose the following question: What triggers firm performance of MNEs? Multinational enterprises are the key drivers and enablers of globalization, in this manner this study attempts to find the firm level factor(s) that have a significant weight in the influence of the financial performance of the multinational enterprises (MNEs) of the Food, beverages and tobacco industry. Forces of global integration appear to be strengthening for these industries, driven by the increasing proliferation of regional and international brands (Ghoshal & Nohria, 1993). One of the key characteristics of MNEs is, as Dunning (1980a) claims, that multiple locations of value-added activities are perceived by management to yield gains. Thus, the proliferation of a company across many countries might be an indication of greater MNE performance.
For the most part, MNEs begin by operating in a single country, and evolve into multifaceted organizations. In general, multinationality entails a lower cost location for firm operations; where for instance the costs of labor, taxes, and/or transportation may be lower (Buckley, 1988). Geringer, Beamish & daCosta (1989) find that highly international companies perform better; thus MNEs, due to the geographic and product diversity, generally exhibit superior returns on equity. At the same time, Palich, Cardinal & Miller (2000) and Dosi, Teece & Winter (1992) have stated that have a firm performs better when it can minimize costs, spread the prestige of its brands into a variety of products, distribute the costs of already capitalized assets, and/or exploit interrelationships between different technologies and marketing channels.
Interestingly, there have been numerous studies on whether industry, firm or country level effects have a significant impact on the financial performance of MNEs (Hawawini, Subramanian & Verdin, 2002b). Accordingly, several studies contend that certain criteria such as the portfolio investment and the resource based perspectives could be used for a more detailed and meaningful analysis (Chin-Chun & Boggs, 2003). However, then it is important to pinpoint in advance that in this study certain measures would be more meaningful than return on investment (ROI), or Earnings before interest , taxes, depreciation and amortization (EBITDA), when researching financial performance of firms. In recent studies, one measurement stands out as conclusive and meaningful to properly analyze performance because of its indication of efficiency by a company in investing its net worth: Return on Equity (Lee & Blevins, 1990; Firer, 1999) especially when considering publicly-listed companies. ROE, a measure of financial efficiency, considers the operating, investing, and financial decisions of a business. Teitelbaum& McDonald (1996) suggests that in industrialized countries, ROE is commonly used as a comparative measure of profitability and financial performance of companies.
average they are older and internationally more competent (Stopford & Dunning, 1983). In addition to their significant historical and adamant consideration, the International Labour Organization (ILO) has stated that just in 2002 the world’s top companies in the food and beverage sector accounted for US$679.4 billion in revenue, in that same year the top ten companies accounted for 37 percent of the revenues earned by the world’s top food and beverage companies. Furthermore, in 2003 alone, US$333.9 billion in revenue accounted to the eight largest global food, beverage and tobacco companies. Thus, the food, beverages and tobacco MNEs studied account for about one-third of world production. Lastly, leading food, beverages and tobacco MNEs, due to their strategies of diversification into a variety of industrial and geographic markets, by spreading into new industries in the domestic market and investing abroad, have survived for more than a century (Chandler, 1990).
The purpose of this study concentrates on determining which factor(s) influence ROE employing a firm specific model; thus, it analyzes a number of firm-level factors that have an effect on the financial performance of MNEs. As an initial step, this study examines MNEs headquartered in the USA, The Netherlands, France, Germany, Mexico, South Africa, Australia, Japan, Ireland, Canada, Spain, Denmark, Switzerland, Sweden and the UK. In order to test to what extent firm size, operational knowledge, operating margin, capital intensity and board size have a critical weight in the financial performance of these multinational enterprises of the food, beverages and tobacco industry in 2004. The MNEs and countries stated are taken by the annual list provided by Forbes and Fortune magazine, which annually perform an audit and rankings in different aspects of the top companies in the world.
The theoretical background and relation to financial performance of each factor is developed in the subsequent section.
Justification
Earlier studies have evaluated the impact of accounting and business practices of firm performance, where ROE has been emphasized (Hirschey & Koch, 1995). On the international level, the study of Lee & Blevins (1990) examined the ROE of 400 firms in Japan, the Republic of Korea, Taiwan, and the United States (US) suggesting that variation in the ROE in each country was due to firm-specific measures. Their study tested firm level determinants of performance for the 1980-1987 period including variables such as firm size, credit activity, capital intensity and debt ratio; the most important and consistent determinants of performance found were the debt equity ratio and capital intensity.
In order to consider ROE as a proxy for firm performance, the DuPont model of financial analysis developed in 1914 by F. Donaldson Brown, an executive in the treasury department at the E.I. du Pont de Nemours Company, was assessed (Goldwyn, 1998). Although financial performance has been researched using ROE, few studies have employed the DuPont model to examine a specific industry or to study an international variation of financial performance. This study utilizes the DuPont model to measure ROE of the MNEs as a related development on the research done by Achaempong & Epperson (1998), furthermore, this study considers the examination of the largest worldwide multinational enterprises in three specific sectors (food, beverages and tobacco) in the year of 2004 and test firm level factors affecting ROE.
The DuPont conceptual model is a financial analysis and planning tool designed to help to understand the factors that drive ROE of a firm following accounting relationships. In this model ROE is the product of three factors: the profit margin, the asset turnover, and the equity multiplier (EM) of the firm and can also be written as ROE = ROA*EM where ROA is defined as Return on Assets (Firer, 1999). Recent studies by Chin-Chun & Boggs (2003), Walters (1997) and Achaempong & Epperson (1998) applied this model and provided contributions to the study of performance by analyzing the forces that shape ROE within countries, across countries, and over time.
model, of the MNEs in the food, beverages and tobacco industries in 2004. There is a lack of studies tackling the approach of ROE as a performance proxy measured by the DuPont model, since the model is generally decomposed and tested against ROE by its three key financial ratios. Thus, this study adds to existing knowledge of the factors that affect ROE as a financial performance measure by utilizing a more current firm sample than prior studies, and by covering a substantially different perspective by analyzing new variables such as operational knowledge and further examining the interactions of independent factors, how they are related to, and what implications do these interactions entail as determinants of ROE.
The paper is organized as follows: The next section presents an overview of previous studies relevant to firm performance, and explains how the conceptual model of DuPont will measure the ROE. Subsequently in section III, I discuss the theoretical model, followed by the description of the data and measures for the empirical tests of the hypotheses in section IV. The afterward section V discusses the empirical results obtained. The final section VI provides conclusions, limitations and guidance for future research.
II. THEORETICAL BACKGROUND A. Consistent overview of relevant and related studies
Since the early 90s, the study of firm performance has given rise to ample literature in the field of financial performance of large MNEs, which is an area of persistent and enduring interest to researchers in the field of international business (Han & Lee, 1998). Table 1 depicts the range of findings by authors who have conducted empirical studies on the financial performance of firms, following an array of perspectives.
The supporters of the firm-specific resource view, which have considered the industry structure unimportant relative to a firm’s resources, have suggested that firm-level specific factors have a superior effect than industry factors on the variance of firm performance (Hawawini, Subramanian & Verdin (2003a) and McGahan & Porter (2002a). Prior research on industry effects relative to firm effects in explaining performance, by firm profitability, began with papers by Schmalensee (1985) making use of data from the US Federal Trade Commission (FTC).
relationship between learning orientation and firm performance, measured by either return on assets (ROA), return on investment (ROI) or return on sales (ROS). The empirical tests revealed a nil effect of an organizations’ age on the relationship between learning orientation and performance.
Table 1
Summary of selected items of previous research on determinants of performance
Studies Performance measure
Sample Results
Schmalensee (1985) ROA 1975 Federal Trade
Comission (FTC) firms Industry effects dominate firm and market share effects. Slater (1988) ROA 1987 financial results
for US grocery stores
An improvement in operating margin will yield an improvement in performance.
Wernerfelt & Montgomery (1988)
Tobin’s q 1976 firms sample in the FTC
Industry effects are the major determinants of firm success. Bhagat & Black (2002) Tobin’s q,
ROA 1985-1995 largest US firms There is no significant relationship between the board of directors and firm performance.
Calantone, Cavusgil & Zhao
(2002) ROI, ROA, and ROS CorpTech Directory of 400 sample of Technology Companies
in 2002.
There is no significant relationship of MNEs age on learning/knowledge and performance.
McGahan & Porter (2002a) Tobin’s q Firms in the American economy from
1981-1994
Firm-specific effects, which arise from competitive positioning and other factors, have a large influence on performance.
Lee & Bose (2002) ROE, ROA,
and ROS 438 firms from the International Data Group from 1989–1993
Capital and R&D intensity show strong relationships with most performance measures.
Hawawini, Subramanian &
Verdin (2003a) Tobin’s q companies ranging 6 countries, listed from 1989-1997
Firm-specific factors dominate performance both across and within countries.
More recent empirical work regarding firm performance and capital intensity by Lee & Bose (2002) in a study based on accounting-based and market-based performance, while controlling for some key variables, such as firm size, capital intensity, and research and development intensity in major US firms. The control variables of capital and R&D intensity showed strong relationships with most performance measures.
Despite the fact that prior research offers evidence regarding the importance of firm level factors in explaining the profitability of firms and industries at the national or international level, there is virtually no prior research examining interactions of such effects, broadly defined, in determining the performance of multinational firms.
The present work provides an initial examination at the global industry level for the food, beverages and tobacco sectors.
B. The DuPont model
In order to measure the financial performance, based on Return on Equity (ROE), of the MNEs of the industry under study recent studies make use of the DuPont model (Achaempong & Epperson, 1998). In this model, ROE is measured by three ratios: profit margin, total asset turnover and equity multiplier (Walters, 1997); thus granting a way to measure ROE by three key areas of MNE management. The ratios measuring ROE are discussed in the subsequent sub-section.
The ratios indicate that there are several paths that a MNE can use to gain a return for its owners: margin, volume and leverage (Chin-Chun & Boggs, 2003). This study will address the measurement of ROE by these three key areas, and thus attempt to analyze how MNEs maximize the value of its business.
Conklin, Prabhakar & North (2002) suggest that analysts need to be able to compare the values between companies to understand the significance of the ratios. In this study these DuPont model ratios will generate ROE, depicted in the self-made figure 1. Therefore, ROE will be measured by the Dupont models ratios.
well as other possible factors are critical in the performance of multinationals in the industry under study.
Figure 1
Ratios utilized to measure ROE
Profit Margin
Total net income / Total sales
a. DuPont ratios
As stated in the previous section, a system of analysis has been developed which focuses on elements of a good financial condition. This analysis is called the DuPont Formula, which comprises:
ROE = PMargin * TAT * EquityM (1)
PMargin = Total net income / Total sales TAT = Total sales/ Total assets
EquityM = Total assets/ Total equity
ROE will be measured by the three ratios in equation (1), asset management, expense management, and capital structure management, providing a way to group three key areas of MNE management into ROE; including all three of these areas to determine the ROE in the model, will tend to maximize the value of the company under study.
- Net Profit Margin: A higher profit margin indicates a more profitable company that has better control over its costs compared to its competitors; hence it has a higher probability to
ROE
Total Asset Turnover
Total sales / Total assets
Equity Multiplier
Total assets / Total equity
*
increase its sells. Thus, when this ratio is positive, it is a good sign that operating management is going well. Profit margin is useful when comparing companies in similar industries (Firer, 1999). Profit margin is measured by the ratio of total net income to sales in US billion dollars.
- Total Asset Turnover: Measured by the ratio of sales per total assets in US billion dollars. Firer (1999) states that a rising ratio of Total Asset Turnover (TAT) means that the firm is able to produce more and more sales from its assets. In other words, the firm is becoming more efficient in using its assets.
- Equity Multiplier: Measured by the ratio of total assets per total equity in US billion dollars. It is possible for a company with terrible sales and margins to take on excessive debt and artificially increase its return on equity; the equity multiplier, a measure of financial leverage, allows the investor to see what portion of the return on equity is the result of debt (Kennon, 2006).
Return on Equity represents the profitability of funds invested by the owners of the firm. ROE is valuable for comparing the profitability of a company to that of other firms in the same industry measured by the return on equity (ROE), in this study it will equal the natural logarithm of the result of each MNE net profit margin multiplied by total asset turnover and by the equity multiplier.
III. THEORETICAL MODEL
In this section, ROE along with firm size, operational knowledge, operating margin, capital intensity, board size and the control variables of GDP and industry are described for their selection in the study. The measurement definition of these concepts is later explained in section IV with greater detail.
A. Dependent and Independent variables
a. Return on Equity (ROE)
exogenous conditions (Berger & di Patti, 2004). In that sense, profitability performance ratios (e.g., return on equity and/or return on assets) may offer more valuable information than raw earnings data; where “the financial statement analysis is embedded in terms of these performance ratios rather than in terms of earnings changes or price-deflated earnings” (Lev, 1989: 165).
In order to asses and analyze the performance of MNEs, particularly when these are headquartered in different countries, I make use of Return on Equity (discussed in section II) instead of profits. Profits are mainly the gain achieved in a business over and above the expenditures carried, whereas ROE is viewed as a broad description of a business’s profitability with regard to capital employed, and the effectiveness of management. The performance signals given by return on equity are plain and intuitive (Ciesielski, 2003).
The ratios that measure ROE, discussed in section II, reflect three major performance dimensions of interest to all analysts, namely: i) income statement management (or how much profit a company can generate per dollar sales) and two aspects of balance sheet management: ii) how well assets can generate total sales and iii) the amount of solvency risk (Chin-Chun & Boggs, 2003).
b. Firm Size (number of employees)
The relationship between performance, measured by ROE, and firm size is a subject that has been explored as one of the main factors that maximize firm profits (Ammar et al., 2003; Orlitzky, 2001).
A study made by Mitchell states that “larger businesses tend to have larger pools of financial and managerial resources that help overcome problems that threaten their survival” (1994:577). Furthermore, Brüderl & Schüssler study suggests that “large firms have advantages in raising capital, face better tax conditions and government regulations, and are in better position to compete for qualified labor” (1990: 535). In the global economy, governments must make adjustments to economic policies according to the domestic economic situation (either monetary and/or fiscal policy); Political institutions can enhance the stability of economic policy, directed toward multinationals (levels of taxation, for instance) which affect the domestic market. Hence, institutions affect policies, and policies affect multinational operations (Jensen, 2006).
Based on these previous researches, the relationship between firm size and profitability may be positive over some firm size ranges and negative for others. Due to economies of scale and market power, firm size has been assumed to have positive effect on financial performance (Winn, 1977). This is due to the fact that economies of scale enable companies with a larger size to operate with greater geographical reach; and thus, access to a larger market.
Such findings then suggest that the size of the firm, measured by number of employees, would have an effect on ROE. In this study, it is hypothesized that firm size has a positive effect on financial performance.
H1: Larger MNEs, in terms of employed individuals, will tend to increase
the ROE level of performance.
c. Operational Knowledge (years in operation)
Operational knowledge is assessed to understand the nature of financial performance in multinational enterprises (MNEs) with respect to companies’ age, in terms of years of operation, and its impact in the context of a model involving variables that sustain firm-level effects. In that sense, Porter suggests that knowledge and experience “can lower costs in marketing, distribution and other areas as well as production or operations within production” (1980a: 34).
knowledge obtained from international experience in foreign markets and strategic planning provides a particularly appealing variable to consider in this study.
As the firm increases its years of operation then it can distribute and employ, throughout the organization, the knowledge acquired (Darr & Argote, 1995). At the same time as a learning organization, the company constantly renews the knowledge assets acquired in order to promote superior guidelines of operation (Lundvall, Johnson, Andersen & Dalum, 2002). Thus, the ability to learn is frequently mentioned as a key factor of performance and long term competitive advantage (Teece, 1998). However, contrasting studies suggest that the older the organization the weaker the relationship between learning orientation and firm performance, where empirical tests do not reveal an effect of the organization age on the relationship between learning orientation and performance (Calantone, Cavusgil & Zhao, 2002).
Interestingly, various studies made into these distribution processes have also analyzed the steps of building, maintaining, and transferring knowledge across the organization (Argote, 1999). In this study, it is hypothesized that a MNE with longer or extended years in operation will tend to accumulate knowledge and therefore continually increase its performance across the organization; thus, having a significant and positive effect on ROE for the year in question.
H2: A higher ROE level is perceived as MNEs increase their knowledge by
years in operation.
d. Operating Margin (operating income/ total revenue)
previous research would suggest, ROE, by the ROA composition for it from the DuPont model, is highly related to operating margin.
H3: A MNE increases its ROE performance level by its increasing
percentage in operating margin.
e. Capital Intensity (total assets / total receipts)
In earlier studies, such as the one conducted by Bettis (1981), performance differences (in terms of ROA) between related and unrelated diversified firms using a sample of 80 firms were investigated. The author concluded that performance differences are associated with capital intensity, for the reason that capital intensity of industries varies extensively, depending the nature of the technology, where some firms are more capital intensive than others. Thus, within an industry a firm has relatively a range of selection concerning the level of capital intensity
vis-à-vis competitors.
Harris (1994) suggests that capital intensity is related to asset specificity, an instrument of firm level profitability. The author tested the competing implication about the investment in, and financing of, firm specific assets in an empirical model for a sample of fortune 500 firms in a cross-sectional relationship between capital financing and asset specificity. His results showed that a higher ratio of capital intensity decreases the firm’s profitability.
as Research and Development (R&D); thus, financing by debt, asset specificity may result in a negative effect on firm performance.
Lee & Bose (2002) developed a study based on accounting-based and market-based performance, while controlling for some key variables, such as firm size, capital intensity, and research and development intensity in major US firms. The control variables of capital and R&D intensity showed strong relationships with most performance measures considered. Based on the importance of capital intensity highlighted by such studies, capital intensity is included to asses the expected negative relation to ROE.
H4: An increasing Capital intensive ratio will tend to negatively affect the
MNE performance in terms of ROE.
f. Board Size (members of the board)
A number of studies suggest that a larger board size would be less effective than a smaller board. In a larger board size members would find it difficult to effectively make use of their knowledge and skills due to the complexity of coordinating contributions; the board then becoming a more symbolic entity and a less part of the management process in the firm (Hermalin & Weisbach, 2001). Researches like Lipton & Lorsch (1992) believe that the effectiveness of the board declines as the size of the board of directors increases beyond a typical number of seven to nine members. Yermack (1996) reported a negative correlation between board size (boards may be too large) and several measures of performance. At the same time in a study by Bhagat & Black (2002) the authors found no significant relationship between the board and performance.
In contradictory perspective, a larger board size is recognized to be associated with a greater range of strategic decisions in the planning process; thus, smaller boards are assumed to be short of clear assessments of strategic decisions and alternatives, and lack confidence in proposing strategic change (Golden & Zajac, 2001). A study made by Anderson, Mansi & Reeb (2004) stated that the cost of debt is suggested to be lower for firms with larger boards, due to the fact that creditors observe these firms, as having more effective monitoring of their financial accounting processes.
a better quality of strategic decisions. In this study the factor of board size is considered because is regarded as the decision-making team of the company, where its size can affect the efficiency of the board, on firm performance, by its process of decision-making. Thus, it is hypothesized that a larger board size may have a significant effect on ROE.
H5: As a strategic decision group, a larger board size tends to be
significant and positively relate to higher MNE performance.
B. Control variables
Two control variables are included in this study in order for the regression estimation to determine what portion of the variation in the dependent variable ROE is explained by the control variables vs. the independent variables.
a. Country GDP
Different studies throughout the history of economical development strongly support the fact that there are many country-specific factors that may impact the performance of a firm in an economy, for instance: political stability, the institutional framework and/or the economic structure. These country-specific factors show the extent to which government policies can encourage the formation of a competitive business environment. However, firm-specific factors are relatively more significant than country factors when analyzing countries with higher levels of economic development and implicitly stronger political stability and/or institutional framework; the case countries in this study.
size, will outdo similar companies that operate in countries with smaller market sizes, being reflected by their ROE.
Recently, Hawanini et al., (2003a) analyzed the relevance of the size of the home market on firms; stating that the home market size is relevant because it indicates whether firms are able to achieve an industry minimum efficient scale (MES). The authors suggest that firms from a country where the size of the home market is less than the MES of the industry might suffer a competitive disadvantage, in terms of costs, when competing against rival firms from larger domestic markets that operate at minimum efficient scale. Therefore, in order to control for country differences, the size of the home market, measured by the Gross Domestic Product (GDP hereafter) of the country where the MNE is headquartered is used. In order to asses the country effect; in this study it is argued that the MNEs headquarter countries’ GDP will be positively related to financial performance.
b. Industry
Porter states that an industry can be viewed as “composed of clusters groups of firms, where each group consists of firms following similar strategies in terms of the key decision variables” (1979b: 215). The effects of such variable can be seen in numerous studies, for example, earlier studies such as Schmalensee’s (1985) study of the total variance of rates of returns on assets, reported that industry effects explained 20 percent of the variance in business returns. Schmalensee tested for evidence of business-specific differences through an exogenous measure of market share. As stated above the author found that industry effects accounted for about 20 percent of variance, at the same time market share effects were found to account for less than 1 percent of variance, and firm effects were not significant to variance. He concluded that managerial influences were not as significant against the differences in industry structure. Furthermore, recent studies suggest that country and industry effects account for about 10% of the total unexplained variations in industry growth, out of which country-specific factors account for more than 50% (Tadesse, 2004).
power of industry factors in explaining firm performance variance, and the contributions of specific industry factors. He found that industry factors explain about 20 percent of overall performance variance.
Based on the contradictory study by Rumelt (1991b), the inclusion of the industry variable is appealing to be assessed. The author found that industry membership is a much less important source of economic rents. He stated that although 20 percent of business-unit returns are explained by ‘industry effects’, its not known how much of this 20 percent is due to stable industry effects rather than to transient phenomena. Rumelt’s aim was to challenge the proposition made by Schmalensee (1985), showing that firm effects were in fact more significant than industry effects, by analyzing the business unit accounting profit of manufacturing firms in the period of 1974 to 1977.
In this study a dummy variable is construted with considerable arbitrariness, to control for industry association by a MNE in any of the three sectors of food, beverages and tobacco, in a way relevant to its influence on ROE.
Rationale of the study
This study focuses mainly on those firm-specific characteristics which previously have been found to be related to performance. To test the five hypotheses, a theoretical framework developed for the empirical analysis will identify categories of firm level factors that affect the financial performance of the MNE. The model will incorporate different attributes of firms in order to investigate their possible influence in the performance of MNEs in 2004 under the industry under study.
ln ROEi = β0+ β1ln (Firm size) + β2 Operational knowledge + β3 Operating margin
in the analysis, and summarizes the hypotheses arguments made in this study in the predicted signs.
Based on the findings of earlier studies, this study identifies firm level determinants of ROE in the food, beverages and tobacco industries and examines the direct and indirect effects of these practices on the ROE performance level, in particular: Firm size, is expected to have a positive effect on ROE because in accordance with Brüderl and Schüssler (1990) large firms are found to have advantages in raising capital, and are in better position to compete for qualified labor (1990). A larger number of employees might be needed to cope with demand as the company expands its global product distribution.
Based on Porter’s (1980a) study, knowledge and experience can lower costs distribution and of other areas of production, thus operational knowledge is expected to have a positive effect on MNEs performance. At the same time, a higher operating margin percentage is expected to have a positive effect on ROE as explained by Slater (1988) who concluded that profitability, based on ROA, is particularly sensitive to margin management; where an improvement in operating margin yields improvement in performance.
Table 2
Definitions and predicted signs of variables
Variable Definition Prediction
Independent
Firm size Natural logarithm of the total number of employees + Operational knowledge Difference from operation 1st year until 2004 +
Operating margin Ratio of operating income to total revenue + Capital intensity Natural logarithm of ratio of total assets to total receipts - Board size Number of members serving on a firm’s board +
Control
Country GDP Natural logarithm of the country GDP where MNE headquartered + Industry Dummy variable indicating the industry sector +
Controlling for home market size (GDP) and industry, I expect that these enclose a direct relation to ROE.
IV. RESEARCH METHODS A. Data collection and sampling
The unit of analysis was firm performance measured by ROE. My data set included information on MNEs in the food, beverages and tobacco industry for the fiscal year end of 2004. I make use of single year data because I consider that this will make the study more relevant and related to identify the relative importance of different categories of firm level factors that determine ROE. The data in this specific year contains the latest up-to-date information on the food, beverages and tobacco MNEs. Structural changes are prone to take place over long periods of time, in addition inefficiencies arise by the employment of an inadequate number of observations by time-series analysis (Lorek & McKeown, 1978); thus, single year data is utilized in an effort to lessen the possible bewildering that may result from such predicaments.
The goal of the data collection was to comprehensively cover the largest MNEs in the industry under study. The complete data set includes 70 MNEs headquartered in the US, the Netherlands, France, Germany, Mexico, South Africa, Australia, Japan, Ireland, Canada, Spain, Denmark, Switzerland, Sweden and the UK. The list of the 70 MNEs sample in consideration is given in Appendix A.1. The method of the 70 MNEs selection is as follows: the consideration that in the annual lists not every company was a MNE in the selected sector of food, beverages and tobacco was acknowledged; hence, only MNEs were taken from the 80 sample companies in the lists. A MNE was considered as having multiple foreign subsidiaries. The 2004 annual list by Fortune and Forbes, sources for business and financial news with company information and trends from international stock exchanges, of largest MNEs in the food, beverages and tobacco industry were used to extract the MNEs sample. Studies made by Orazem, Bouillon & Doran (2004) and Snider, Paul & Martin (2003), among others, have utilized such sources to address the largest companies worldwide.
firms; this resulted in the inclusion as follows 45 firms in food, 20 firms from beverages and 5 firms from the tobacco sector, resulting in a total sample of 70 firms. A list of the standard industrial codes of the sample in consideration is given in Appendix A.2.
I exercised caution in interpreting results obtained, by the data sample, from all three sectors, by including a dummy variable to differentiate between food, beverages and tobacco. As Koutsoyiannis states “Dummy variables are constructed to describe the development or variation of the variable under consideration” (1977: 218).
The methodology entails the evaluation, by an OLS estimation described in the following sub-section C, of the financial performance of a MNE by analyzing firm specific data for the year 2004 obtained from US securities and exchange commission ‘EDGAR’ and the Hoovers business search tool (www.hoovers.com). These include the income statement, balance sheet, and the annual cash flow. Measures of corporate governance, such as board size, are collected from the company official website annual reports. In addition, the macro economic data was collected from The World Bank annual country reports.
In order to estimate the logarithmic equations, statistical procedures are carried on using the latest statistical package up-to-date, SPSS 11.0; the software standard feature commands were used to transform the dependent and independent (ROE, firm size and capital intensity) variables into their logarithms. A point to take in consideration is when using natural logarithms, the coefficients are different in alternative (Semi-log and Double-log) functional forms; these are addressed in Appendix A.3.
made by Cortina states that “interaction terms X*Z involve multiplication and these examine nonadditive effects above and beyond linear and/or additive effects” (1994: 917).
Finally, model 3 entails the OLS analysis of the factors found in models 1 through 2 that were significant in determining firm performance.
B. Measures Definition Dependent variable
Firm performance: Return on Equity represents the profitability of funds invested by the
owners of the firm. ROE is valuable for comparing the profitability of a company to that of other firms in the same industry. In this study, described in section II, it will equal the natural logarithm of the result of each MNE net profit margin multiplied by total asset turnover and by the equity multiplier.
Independent variables
Firm Size: There has been particular attention on the relationship between firm size and
financial performance. Many authors have suggested that larger organizations generate stronger competition than their smaller competitors (Datta & Guthrie, 1994). Firm size has been viewed as an indicator of scale economies and market power, and empirical evidence exists linking firm size to profitability (Datta & Guthrie, 1994; Robins & Wiersema, 1995). In that sense, firm size can be measured in five basic ways: sales volume, net assets, capacity, number of clients served and number of employees (Kimberley, 1976).
Given the sample of the food, beverages and tobacco businesses under study, the natural logarithm of total number of employees will be utilized to measure firm size; this specification better captures the effect of size on firm performance. Kumar, Rajan & Zingales (2001) suggests that the number of employees as a measure of firm size is likely to be very similar to one based on value added. The author states that the number of employees is preferable to output (total assets), as a measure of firm size, because the complexity of the organization is related to the value of its contribution concerning rather than to the value of the output sold. This argues for a measure based on number of employees, which is used in the rest of the study.
Operational Knowledge: Knowledge intense firms reflect their dependency on knowledge
inherited in its activities and outputs as a source of competitive advantage (Autio et al., 2000). Thus I believe that operational knowledge should lead to better firm performance. In this study it will be measured by the difference of the first year of operations of the MNE and the year under study, 2004.
Operating Margin: A firm’s operating margin is likely to affect the earnings components used to
manage earnings. In this study it will be measured by the ratio of operating income to revenue, expressed as a percentage.
Firms with a high operating margin percentage are more likely than firms with a low operating margin percentage to use sales to manage earnings upward because each dollar of sales will have a greater effect on bottom-line earnings for high margin firms (Plummer & Mest, 2004).
Capital Intensity: In this study it will be measured by the natural logarithm of the ratio of total
assets to total receipts. Total receipts are measured by the cost of goods sold plus total income (The IRS, Internal Revenue Service). Ravenscraft (1983) stated that the capital intensity is important in explaining variations in profits.
It has been argued that in capital intensive industries, producers tend to compete away profit margins more vigorously, presumably in the hope of keeping their plants operating at high activity levels (Scherer, 1980); thus, capital intensity should be negatively related to ROE.
Board Size: Previously there has been a trend in assessing firm performance and board size by
the board composition, segregating the board of directors by the independent and inside directors; it has been stated that there is no relationship between the proportion of outsider to insider directors and various performance measures such as return equity and sales among others Fosberg (1989). Thus, in this study the total number of board members composing (without discriminating against inside or outside directors) a MNE is selected to measure board size. Control variables
Country, GDP: Country and industry effects account for about 10% of the total unexplained
positive and statistically large impact on growth in economic efficiency. The author suggests that “industries that account for a larger portion of the country's manufacturing have higher growth rates, which may reject the effects of other sources of comparative advantage” (2004: 716-717).
The home market size of each country is measured by natural logarithm of the GDP. The World Bank country reports (2004) provide GDP as the nominal value in billions of U.S. dollars. The World Bank study already provides the Dollar figures for GDP converted from domestic currencies using single year official exchange rates, of the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products of the country’s economy produced in a year. With an increase in economic activity, large industries attracting specialized resources lead to an increased firm profitability. Thus, GDP is expected to be directly related to ROE.
Industry: A dummy variable will measure the industry that the firm under study belongs to. 1
will equal the food industry, whereas 0 includes the beverages and tobacco manufacturing MNEs. Previous studies as the one of Rumelt (1991b), described in section III, find that industry membership is a much less important source of economic rents. He stated that although 20 percent of business-unit returns are explained by ‘industry effects’, its not known how much of this 20 percent is due to stable industry effects rather than to transient phenomena, thus showing that firm effects were in fact more significant than industry effects.
B. Econometric techniques
In order to test the three models, described in the previous sub section, OLS estimation will be used. Ordinary least squares (OLS) regression is arguably the most widely used method for fitting linear statistical models. Fox (1984) states that the ordinary least squares (OLS) estimation applies to the linear multiple regression model (3) as:
Yi = a0 + a1 X1i + a2 X2i + ... + ak Xki + ei , (3)
where Yi is observation i of dependent variable Y, X1i is observation i of independent variable X1, a0 is the constant term and equals the coefficient of an implicit explanatory variable with value 1, a1 etc. are the coefficients for the independent variables, ei is the residual.
The assumptions that underline OLS estimation are:
• Residuals are homoskedastic (no heteroskedasticity), E(e i2) = σ2= constant. • Residuals are independently distributed (no serial correlation), E(ei ej) = 0. • Explanatory variables are independent (no multicollinearity), cov(Xi,Xj) = 0. • Residuals are normally distributed, e ~ N(0, σ2).
Koutsoyiannis (1977) states that the ordinary least squares estimation method minimizes the sum of the squared residuals ∑e2i = ∑ (Yi - b0 - b1Xi)2 in which each residual is assigned
equal weight.
In applying this econometric method for the estimation of the model, an ‘experimental approach’ was used, where one starts with a simple model containing small number of equations and variables; these are modified gradually, on the basis of the statistical evidence accruing from the computations. An experimental approach is basically one that “combines the theoretical considerations with empirical observations; designed to extract the maximum of information from the available data” (Koutsoyiannis, 1977:23). As calculations are carried out, Koutsoyiannis states that by adding other explanatory variables, equations or alternative econometric models, the researcher is able to observe the effects of such changes in an attempt to achieve the best model, the best explanation of the phenomenon being analyzed (1977).
V. EMPIRICAL RESULTS
In order to validate the assumptions of the OLS estimation described in the previous section IV, a number of diagnostic checks were performed. These are explained in Appendix B.
A. Descriptive Statistics
Table 3 presents the mean, standard deviation and minimum and maximum values for the dependent, independent and control variables.
Table 3
Descriptive Statistics for MNEs ROE
Variable Mean s.d. Minimum Maximum
Constant 144.03 2.61 4.85 1719.86 Firm size 24343.01 2.83 1465.57 247706.53 Operational knowledge 92.97 45.49 4 256 Operating margin 11.46 6.50 2.10 32.50 Capital intensity 1.51 1.85 .38 9.78 Board size 11.99 3.74 6 24 GDP 8690.62 14.44 2.29 117008.28 Industry 0.64 0.48 0 1
Capital intensity x firm size 18163.22 21076.85 321.86 98800 Capital intensity x operational
knowledge 45.07 41.22 0.66 204.8
Capital intensity x operating margin 1.31 1.03 -1.09 3.92 Board size x operating margin 134.82 81.56 18.90 362.60 Board size x firm size 121.68 42.97 55.44 285.66 Board size x operational knowledge 1098.30 674.12 56.00 3335.00
B. Correlations
Table 4 presents the correlations of all variables tested by model 1. The data presented in this table suggest that operating margin is moderately related to ROE at the (p<0.01) level.
Table 4
Correlations of independent variables by models 1 and 2
Hence, a higher ratio of operating margin leads to superior ROE. Furthermore, operating margin and capital intensity are respectively negatively and positively correlated at the (p<0.01) to industry, suggesting that greater MNE association in various sectors of an industry will lead to a higher ratio of capital intensity while the percentage ratio of operating margin decreases.
Table 5 reports the correlations by model 2, interactions constructed from the firm-level factors found significant in the determination of ROE in model 1, where the interaction term of board size*operating margin at the (p<0.01) is moderately correlated to ROE. Thus, the effect of a larger board size, on ROE, is different for higher and lower operating margin percentages.
Table 5
Correlations of interactions by model 2
Variable 1 2 3 4 5 6 7
1. Constant 1
2. Capital intensity x firm size .029 1
3. Capital intensity x operational knowledge -.086 .354** 1
4. Capital intensity x operating margin .170 .441** .528** 1
5. Board size x operating margin .544** .273* .163 .664** 1
6. Board size x firm size .176 .352** -.105 -.083 .363** 1
7. Board size x operational knowledge .154 .145 .563** .011 .241* .497** 1 ** Correlation is significant at the 0.01 level (1-tailed).
* Correlation is significant at the 0.05 level (1-tailed).
The correlations results of the interaction effects seem to indicate that MNEs will perform better with lower capital intensity and a larger board size while facing either a higher or lower operating margin percentages. These results might be the reaction of investors who might perceive capital intensity, for instance, as a potential problem for assuring a superior firm performance. The findings might also reflect the fact that higher capital intensity and a lower board size might result in less company returns.
intensity*operational knowledge, capital intensity*operating margin, board size* operating margin, and board size*firm size. In the subsequent section, the regression results of significant interaction effects are addressed.
Furthermore, capital intensity*operating margin and board size*operating margin are highly correlated due to their common usage, in their measurement, of operating margin. Cortina states that “there will always be considerable overlap between an interaction term, X*Z, and its constituent parts, X and Z” (1993: 920).
C. Regression Results
Table 6 reports in each column the regression results for the three models, which are presented in an OLS estimation, and including the respective F and R2 values. The results indicate that the signs of the individual parameters are accurate (i.e. capital intensity should diminish ROE, thus having a negative sign), as indicated in table 2 in section III.
In model 1, it is observable that the goodness of the fit indicator R2 has a value .46 which indicates that only 46.0% of variance in the ROE of the global food, beverage and tobacco MNEs can be explained by the independent and control variables; Even though, the result of the F statistics (7.45) suggests that overall the model entails enough significance in explaining the dependent variable, ROE. F statistics results for model 2 suggest it has enough significance (7.35) in explaining ROE, when including only interaction terms.
I have argued that the financial performance of multinational enterprises is influenced by five relevant variables, which would have significant effects on ROE. In model 1, I found the Return on Equity (ROE) to be highly determined by three primary factors, operating margin (p < .05), capital intensity (p < .05) and board size (p < .10).
Table 6
Results of Regression Analysis of MNEs ROE a
a t values are in parentheses. n = 70.
Variable Model 1 Model 2 Model 3
Constant 2.12* (2.09) 5.29* (5.51) 4.38* (8.44)
Firm size 8.75E-02
(0.98) Operational knowledge 2.09E-03
(1.03) Operating margin .102* (6.39) 1.93E-02 (4.60) Capital intensity -.64* (-3.79) -0.70* (-4.51)
Board size 4.52E-02**
(1.81) -7.49E-02 (-.93) GDP 2.14E-02
(0.61)
Industry 0.21 (0.98)
Capital intensity x firm size 0.12 (1.10) Capital intensity x operational knowledge -0.33
(-1.38) Capital intensity x operating margin -.27
(-1.23) Board size x operating margin 9.80E-03*
(4.27) 7.190E-03* (0.04) Board size x firm size -9.04E-03*
(-2.25)
4.790E-03 (0.47) Board size x operational knowledge 4.72E-04
(1.43) F 7.45 7.35 11.35 R2 .46 .41 .47 Adjusted R2 .40 .35 .43 * p < .05 ** p < .10
results (Model 1: β= -.64, t = -3.79) provided in table 6 confirms the assumption that as the ratio of capital intensity increases, ROE will decrease.
Thirdly, the results support hypothesis 5, where at the 10% significance level, the MNE board size is significant and positively related to ROE. This indicates that an increase in the MNE board size would result in a 4.5 percent increase change in ROE. Therefore the results contradict the study made by Hermalin & Weisbach (2001), where they concluded that members in larger boards would find it difficult to effectively make use of their knowledge and skills due to complexity of coordinating contributions; and therefore the board becoming a more symbolic entity and less part of the management process. The board size significant (Model 1: β= 0.04, t = 1.81) results follow the conclusions made by Golden & Zajac (2001) where a larger board size is recognized to be associated with a greater range of strategic decisions in the planning process, as having better assessments of strategic decisions and alternatives, and possessing confidence in suggesting strategic change.
Taken together, these three results suggest that the Return on Equity per MNE in the food, beverages and tobacco industries worldwide, is highly determined by an effective operating margin percentage across the organization and the carrying out of small capital intensity ratios while maintaining a relatively large board size.
As for the variables of firm size and operational knowledge, the results obtained suggest that they are not significant in the determination of performance. Concerning the variable of firm size, while there is a tremendous amount of prior studies on the relationship between firm size and firm performance (Hirschey & Koch, 1995; McGahan & Porter, 2002a), I am investigating the relationship between firm size and firm financial performance, ROE. Following Marcus (1969) result of non existent relationship between firm size and firm profitability, the statistical results do not support hypothesis 1. It is, possible that this assumption leads to underestimation of the size of the MNE resulting from the use of number of employees as a proxy; thus, the reason of this seemingly and slightly contradictory result may originate from the measurement of firm size in order to asses firm financial performance, ROE. Moreover, as the firm size grows, it becomes more difficult to sustain impressive performance (Banz, 1981).
in explaining firm performance. One explanation that the authors’ state for such result is that may be in terms of organizational resources, skills and knowledge, there is equal importance for young and old organizations.
As for the control variables of MNE headquarter countries’ GDP and industry, the results suggest that these do not have a significant effect in determining ROE. This is due to the fact that the companies under study do not differ significantly in their home market sizes, the financial performance of the MNEs in the food, beverages and tobacco industry is then not related to the macro economic conditions of the headquarter country. Concerning industry, the results obtained in table 6 ( β= 0.21) suggest that a MNE belonging to the food manufacturing sector causes its ROE to increase by 0.21 percent compared to being in the beverages and tobacco sector.
Is there any interaction effects between operating margin, capital intensity and board size variables that may help to predict firm performance? Operating margin, capital intensity and board size have significant beta weights in multiple regression predicting firm performance, as does the interaction term of board size*operating margin and board size*firm size. Thus, addressing model 2, the results indicate that the interaction terms of board size*operating margin and board size*firm size are respectively positively and negatively related to ROE. This indicates that the effect of a larger board size on ROE is different for higher and lower ratios of operating margin and that the effect of a smaller board size on ROE is different for smaller and larger firm sizes.
Lastly, in accordance with their results in models 1 through 2 the coefficients of capital intensity and the interaction term of board size*operating margin are respectively, negatively and positively related to ROE in model 3. The results of model 3 confirm that higher capital intensity will decrease the food, beverages and tobacco MNE returns on equity.
VI. CONCLUSIONS A. Aim and Questions
In the present research, I set out to study which and how firm level factors affect and determine the performance of MNEs, measured by ROE, in the industry sectors under scrutiny by OLS estimation. My study extends prior research of MNEs performance levels of the world’s largest food, tobacco and beverages companies, by assessing the variables developed in the DuPont model of profit margin, total asset turnover, and equity multiplier to measure ROE.
What triggers firm performance of the MNEs in the Food, beverages and tobacco industry in 2004? In this study the significant coefficients of operating margin and board size in model 1 provide evidence to support the existence of superior performance among the 70 global firms in terms of profit margin ratios and decision making groups. At the same time, the negative and significant coefficient of capital intensity throughout models 1 to 3 confirms the predicted assumption that as the ratio of capital intensity increases, ROE will decrease. Furthermore, the significant coefficients of the board size*operating margin and board size* firm size interaction effects suggest that the effect of a larger board size on ROE is different for higher and lower ratios of operating margin and that the effect of a smaller board size on ROE is different for smaller and larger firm sizes.
The results obtained in this study offer evidence about firm level effects on ROE by using a model that simplifies certain criteria, and a powerful explanation that allows the connection to managerial schemes inquired in the profitability of the MNEs under study.
B. Limitations and future research
This study aims at searching for most relevant factors influencing the performance of food, beverages and tobacco multinational enterprises. Due to restrictions in data access, I cannot carry out a study incorporating observations from many countries, which limits the application of this research to MNEs based in the countries stated.
REFERENCES
Acheampong, Y.J., & Epperson J.E. 1998. International variation of return on equity in the
food and beverages industries. Working paper no 98-07.University of Georgia, Department of
Agricultural and Applied Economics, Athens,GA.
Agmon, T., & Lessard, D. R. 1977. Investor recognition of corporate international diversification. Journal of Finance, 32(4):1049-1056.
Ammar, A., Hanna, A.S., Nordheim, E.V., & Russell, J.S. 2003. Indicator Variables Model of Firm’s Size-Profitability Relationship of Electrical Contractors Using Financial and Economic Data. Journal of Construction Engineering & Management, 129(2).
Anderson, R. C., Mansi, S.A., & Reeb, D. M. 2004. Board characteristics, accounting report integrity, and the cost of debt. Journal of Accounting & Economics, 37(3):315-342.
Argote, L. 1999. Organizational Learning: Creating, Retaining and Transferring Knowledge. Kluwer, Boston, MA.
Autio, E., Sapienza, H.J., & Almeida, J.G. 2000. Effects of age at entry, knowledge intensity, and imitability on international growth. Academy of Management Journal, 43(5):909-924. Banz, R. W. 1981. The Relationship between Return and Market Values of Common Stocks.
Journal of Financial Economics, 9: 3-18.
Berger, A. N, & Bonaccorsi di Patti, E. 2006. Capital structure and firm performance: A new approach to testing agency theory and an application to the banking industry. Journal of
Banking & Finance, 30(4):1065-1102.
Bettis, R. A. 1981. Performance Differences in Related and Unrelated Diversified Firms. Strategic Management Journal, 2 (4):379-393.
Bhagat, S., & Black, B. 2002. The Non-Correlation Between Board Indpendence and Long-Term Firm Performance. Journal of Corporation Law, 27(2).
Blaine, M.1993. Profitability and competitiveness: Lessons from Japanese and American firms in the 1980s. California Management Review, 36(1).
Bland, J.M., & Altman, G.D. 1996. Education and Debate: Statistics Notes on Logarithms. BMJ, 312: 700
Brüderl, J., & Schüssler, R. 1990. Organizational mortality: The liabilities of newness and adolescence. Administrative science,35: 530-547.
Brush, T.H., Bromiley, P., & Hendrickx, M. 2000. The free cash flow hypothesis for sales growth and firm performance. Strategic Management Journal, 21(4).
Buckley, P. J. 1988. The Limits of Explanation: Testing the Internalization Theory of the Multinational Enterprise. Journal of International Business Studies.
Calantone, R.J., Cavusgil, S.T., & Zhao, Y. 2002. Learning orientation, firm innovation capability, and firm performance. Industrial Marketing Management, 31(6):515-524.
Ciesielski, J.T. 2003. Return on equity remixed: The S&P 500’s real numbers. The analyst’s accounting observer,12(14).
Chandler A. D. J. 1990. Scale and scope: The dynamics of industrial capitalism. Cambridge, MA: Harvard University Press.
Chin-Chun, H., & Boggs, D. J. 2003. Internationalization and Performance: Traditional Measures and Their Decomposition. Multinational Business Review, 11(3):23-49.
Conklin, N., Prabhakar, S., & North, C. 2002. Multiple Foci Drill-Down through Tuple and
Attribute Aggregation Polyarchies in Tabular Data. Paper presented at the IEEE Symposium
Cortina, J.M. 1993. Interaction, Nonlinearity, and Multicollinearity: Implications for Multiple-Regression. Journal of Management,19(4): 915-923.
Darr, E. D., & Argote, L. 1995. The acquisition, transfer and depreciation of knowledge in service organizations. Management Science, 41(11):1750–1762.
Datta, D., & Guthrie, J.1994. Executive Succession: Organizational Antecedents of CEO Characteristics. Strategic Management Journal, 15:569-577.
Dosi G., Teece D., & Winter S. G. 1992. Toward a theory of corporate coherence: Preliminary remarks. In Dosi, Giannetti, R., & Toninelli, P.A. (Eds.), Technology and enterprise in a
historical perspective,Oxford: Clarendon Press.
Dunning, J.H. 1980a. Toward and eclectic theory of international production: some empirical tests. Journal of International Business Studies, 11(1): 9-31.
Dunning, J.H. 1973b. The Determinants of International Production. Oxford Economic Papers, 25(3):289-336.
Errunza, V. R., & Senbet, L. W. 1984. International Corporate Diversification, Market Valuation, and Size-Adjusted Evidence. Journal of Finance, 39(3):727-746.
Firer, C. 1999. Driving Financial Performance Through the duPont Identity: A Strategic Use of Financial Analysis and Planning. Financial Practice & Education, 9(1):34-45.
Fosberg, R. H. 1996. Agency problems and capital expenditure financing. Arkansas Business &
Economic Review,29 (4)15-24.
Fox, J. 1984. Linear Statistical models and related methods: with application to social
research. New York: Wiley Series.
Geringer J. M., Beamish P., & daCosta R. 1989. Diversification strategy and internationalization: Implications for MNE performance. Strategic Management Journal, 10: 109-119.
Ghoshal, S., & Nohria, N. 1993. Horses for Courses: Organizational forms for Multinational corporations. Sloan Management Review, 34(2):23-35.
Golden, B. R., & Zajac, E. J. 2001. When will boards influence strategy? Inclination Power = Strategic change. Strategic Management Journal, 22(12):1087-1117.
Goldwyn, R. 1998. ‘Tis the gift to be simple: why the 80-year-old DuPont model still has fans. http://www.cfo.com/article.cfm/2990236. CFO Magazine, Accessed Feb 25, 2006.
Hall, M., & Weiss, L. 1967. Firm size and profitability. Review of Economics & Statistics, 49 (3).
Han, K.C., & Lee, S.H. 1998. Multinationality and Firm Performance. Multinational Business
Review, 6(2).
Harris, F.H. 1994. Asset specificity, capital intensity and capital structure: An empirical test.
Managerial and Decision Economics, 15(6): 563-576.
Hayes. A.F., & Cai, L. 2000. Using heteroscedasticity-consistent standard error estimators in
OLS regression: An introduction and software implementation. Ohio state University
dissertation paper.
Hawawini G., Subramanian V., & Verdin P.J. 2003a. The relative importance of country,
industry and firm effects on firm performance. DTEW Research Report, K.U.Leuven.
Hawawini, G., Subramanian, V., & Verdin, P. 2002b. Is performance driven by industry – or firm- specific factors? A new look at the evidence. Strategic Management Journal, 24 (1):1-16. Hermalin, B., & Weisbach, M. 1991. The effects of board composition and direct incentives on firm performance; Finance Management, 20(4):101-112.
Hirschey, M., & Koch P.K 1995. The Global Variation In Financial Ratios. Advances in
