The executive compensation: pay-for-performance or innovation in high technology firms

UNIVERSITY OF PORTO SCHOOL OF ECONOMICS AND MANAGEMENT PhD in Business and Management The executive compensation: Pay-for-performance or innovation in high-technology firms BY Paula Isabel Santos Faria A DISSERTATION Presented to the School of Economics and Management at the University of Porto in the partial fulfilment of requirements for the degree of Doctor of Business Management. Supervisors: Elísio Brandão Francisco Vitorino Martins June 2014 ii BIOGRAFICAL NOTE Paula Isabel Santos Faria was born on May 1, 1967 in Germany. At the age of 7, she came to live in Portugal with her parents and younger brother. Currently, she is married and has one child. Academically, driven by her early interest in economic studies, she obtained a degree in Economics at the University of Porto. In 2010, she obtained a Master’s Degree in Finance and Taxation from the School of Economics and Management of the University of Porto. Professionally, she has been working since 1990 at INESC’s accounting and finance department. Over the years, she progressively was assigned tasks of department responsibility and she is responsible for INESC TEC’s Accounting and Finance area since 1998, a position she holds to this day. Moreover, she is the President of INESC P&D Brasil’s supervisory council. iii ACKNOWLEGMENTS Completing my PhD degree is probably the most challenging thing I have ever done. The best and worst moments of my doctoral journey have been shared with many people, which I would like thank: My first debt of gratitude must go to my advisors, Professor Elísio Brandão and Professor Francisco Vitorino Martins, both from the School of Economics and Management, University of Porto. They patiently provided the vision, strategy; encouragement and advice I needed to in order go through with the doctoral program and complete my dissertation. Without their care and belief in the ideas pursued here, this work would not have come to fruition. My second acknowledgment must go to Professor José Manuel Mendonça and Professor João Claro at INESC TEC, who believed in me from the very beginning, and gave me all the courage and confidence I needed, which was very important to me. I also thank my friends (too many to list here but you know who you are!) for providing the support and friendship that I needed. I would like to thank Rute, Rita and Paula for their support, help and encouragement. Last, but not least, I must thank my husband, who has always relied on me and has given me strength to continue this long and difficult journey. And for all the time that I did not spend with my daughter, I appreciate your understanding and patience. This work is for you – Ângelo and Mafalda, I hope that this will make you proud. iv ABSTRACT Chef Executive Officer (CEO) compensation has become a very interesting topic of debate in the finance literature. This work addresses several research topics, some of which are still unexplored, regarding the compensation related to performance and innovation, performance in high-tech firms, corporate innovation and risk taking in high-tech firms. In order to reduce this gap, this investigation analyzes different components of CEO compensation in high-technology firms, performance, innovation and risk taking, and different econometrics models are tested. Firstly, executive compensation in high-technology firms in the US is studied and the S&P index is chosen in the period between 2000 and 2010. A panel data methodology is used to analyze the relationship between corporate performance and CEO compensation in hightechnology firms. Total CEO compensation and shortand long-term compensations are tested according to corporate performance models. Total shortand long-term CEO compensation in high-technology firms is compared to other industrial sectors from standard classification codes and analyzed for each year. Then, the analysis focuses on the effects of introducing the Standard Financial Accounting, statement 123 (R), on corporate performance, and all its accounting rules and underlying obligations. The effects of the statement 123 (R) are tested for the periods before and after implementation, and the importance of stock options accounting in high-technology firms is assessed. This study leads to a better understanding of the relationship between CEO compensation through its components and performance in high-technology firms. Finally, this study analyzes the relationship between corporate performance and innovation in CEO compensation in high-technology firms in the same firm sample. Finance variables are used as return on assets to measure performance and R&D expenses, and the number of patents and brands is used to measure innovation in a system equation and test the econometric model. The overall results confirm the importance of executive compensation in firm performance and the influence that executive compensation has on innovation. Moreover, the results show that innovation is selected as opposed to performance by the CEOs when they are responsible for promoting the firms’ goals. v RESUMO A compensação dos executivos tornou-se uma questão muito interessante em debate na literatura financeira. Este trabalho analisa vários tópicos de investigação, alguns dos quais ainda pouco explorados, como a relação da performance e inovação da empresa com a compensação dos executivos, a performance nas empresas de alta tecnologia, a inovação empresarial e a tomada do risco nessas empresas. Ao longo desta investigação, são usadas as empresas do índice S&P nos Estados Unidos, são analisadas diferentes componentes da remuneração do executivo, a performance da empresa é avaliado através de várias variáveis financeiras, a inovação é medida pela investigação e desenvolvimento (I&D), número de patentes e marcas, e são testados por diferentes modelos econométricos. Em primeiro lugar, é estudada a compensação dos executivos em empresas de alta tecnologia nos Estados Unidos e escolhido o índice S&P no período entre 2000 e 2010, utilizando uma metodologia de dados em painel para analisar a relação entre a performance das empresas e a compensação do CEO em empresas de alta tecnologia. É testada a compensação total do executivo no curto e no longo prazo em relação à performance da empresa. A compensação total do executivo no curto e longo prazo nas empresas de alta tecnologia é comparada com outros setores industriais com o mesmo código de classificação e analisados para cada ano. Em seguida, a análise centra-se nos efeitos da introdução da norma contabilística 123 (R) na performance das empresas, assim como na regra de contabilização e obrigações subjacentes. São testados os efeitos da norma contabilística 123 (R) antes e após a implementação, e a importância da contabilização das opções de ações em empresas de alta tecnologia. Deste estudo resulta uma melhor compreensão da relação entre a remuneração do CEO e as suas componentes e a performance das empresas de alta tecnologia. Finalmente, é analisada a relação entre a performance das empresas e inovação na compensação do CEO em empresas de alta tecnologia para a mesma amostra. São usadas variáveis financeiras como a rentabilidade dos ativos para medir a performance e despesas de I&D, número de patentes e marcas para medir a inovação num sistema de equações para testar o modelo econométrico Em suma, os resultados evidenciam a importância da compensação dos executivos na performance das empresas, a influência da compensação dos executivos na inovação, e ainda a preferência da inovação à performance quando o CEO tem responsabilidade na prossecução dos objetivos da empresa. vi CONTENTS BIOGRAFICAL NOTE ................................................................................................................. ii ACKNOWLEGMENTS ............................................................................................................... iii ABSTRACT ................................................................................................................................... iv RESUMO ........................................................................................................................................ v LIST OF TABLES ........................................................................................................................ ix LIST OF APPENDIX ................................................................................................................... xi CHAPTER 1 ................................................................................................................................. 12 1. General introduction......................................................................................................... 13 1.1 Motivation and purpose of this research ...................................................................... 13 1.2 The role of CEO compensation in high-technology firms ........................................... 15 1.3 The methodology ......................................................................................................... 17 1.4 Plan of presentation ..................................................................................................... 18 CHAPTER 2 ................................................................................................................................. 20 Executive Compensation for the longand short-term in High-Technology Firms .............. 20 1. Introduction ....................................................................................................................... 21 2. Corporate Performance and CEO Compensation ......................................................... 23 2.1 Executive Compensations to Address Agency Problems ................................................. 24 2.2 Examining CEO Pay-for-Performance ............................................................................. 25 vii 2.3 Appropriate measurements of corporate financial performance in high-technology firms ................................................................................................................................................ 30 3. Research Hypotheses ........................................................................................................ 32 4. Methodology, Sample, and Data Collection.................................................................... 33 4.1 Total Compensation model ............................................................................................... 35 4.2 Total compensation and cash compensation models ........................................................ 39 5. Conclusion and Future Research..................................................................................... 45 6. References .......................................................................................................................... 47 CHAPTER 3 ................................................................................................................................. 51 CEO compensation in high-tech firms and changes in the Financial Accounting Standard SFAS No 123 (R) .......................................................................................................................... 51 1. Introduction ....................................................................................................................... 52 2. Literature review ............................................................................................................... 53 2.1 Accounting treatment of equity-based compensation before and after SFAS 123 (R) 54 2.2 Executive compensations to address agency problems and the income strategy impact ..................................................................................................................................... 55 2.3 Financial performance in high-technology firms ........................................................ 57 2.4 Developing a hypothesis .............................................................................................. 58 3. Empirical approach .......................................................................................................... 60 4. Data and summary statistics ............................................................................................ 63 5. Results ................................................................................................................................ 66 6. Conclusion and future research ....................................................................................... 72 7. References .......................................................................................................................... 73 viii CHAPTER 4 ................................................................................................................................. 77 CEO compensation in high-tech firms: The choice between performance and innovation .. 77 2. Literature review and hypothesis development ............................................................. 80 2.1 Executive compensations to address agency problems .................................................... 80 2.2 Corporate innovation ........................................................................................................ 82 2.3 Risk taking to performance and innovation ...................................................................... 84 3. Sample, Data and Method of Analysis ............................................................................ 87 3.1 Dependent variables .......................................................................................................... 88 3.2 Independent variables ....................................................................................................... 88 3.3 Method of analysis ............................................................................................................ 89 4. Results ................................................................................................................................ 91 5. Conclusion and future research ....................................................................................... 96 6. References .......................................................................................................................... 97 CHAPTER 5 ............................................................................................................................... 104 1. General conclusions ........................................................................................................ 105 2. Limitations and suggestions for future research .......................................................... 107 References ................................................................................................................................... 108 ix LIST OF TABLES CHAPTER 2 Table 1 - Executive Compensation Dependent and Independent Variables .................................. 34 Table 2 - Descriptive Statistics individual sample S&P500 .......................................................... 36 Table 3 - Total Executive Compensation Estimation .................................................................... 37 Table 4 - Total Executive Compensation Estimation in long-term ............................................... 38 Table 5 - Descriptive statistics individual sample S&P 1500 ........................................................ 40 Table 6 - Total compensation and cash compensation estimations using the SUR method for S&P500 .......................................................................................................................................... 40 Table 7 - Total compensation and cash compensation estimations using the SUR method for S&P1500 – SM, MD and SP .......................................................................................................... 43 CHAPTER 3 Table 1 – Executive compensation dependent and independent variables ................................... 60 Table 2 - Descriptive statistics and correlations for the sample of 1500 S&P over the period 20002010 ................................................................................................................................................ 65 Table 3 - Total compensation and cash compensation estimations using the SUR method ........................................................................................................................................................ 68 16 informed, timely actions by the CEO. This asymmetry in information allows CEOs to make adequate reductions in R&D spending, which Joseph P. O'Connor, Jr., Joseph E. Coombs, and Gilley (2006) believe is problematic because decisions that benefit short-term performance often do not lead to long-term benefits for shareholders. These consequences make reductions in R&D spending relevant because shareholders generally find R&D spending desirable due to their interests in greater risk-taking, as opposed to CEOs and their interests in long-term firm performance. As Makri et al. (2006) report, CEO total pay was associated with innovation behavior in hightechnology firms, while Balkin et al. (2000) suggest that compensation is more likely to align CEO pay with behaviors towards R&D in high rather than low R&D intensive firms. Therefore, reductions in R&D spending in high R&D intensive firms may not lead to short-term pay increases for underpaid CEOs. Managers in high-tech firms are faced with different sets of performance expectations such as innovation, new product development, integration of technology and research and development management (Shim et al., 2009). High-technology firms have their own special features which separate them from other firms. That poses unique corporate governance problems not only for managers, but also for claimants on R&D activity to create and use knowledge with the intention of improving a firm’s financial position (Belloc, F. 2013) In high-technology firms it is possible to find innovation, R&D investments and some assets with essential competitive advantage and there are, at the same time, some risks. Different R&D spending in firms is indication of a large variance in firm performance. The returns on high-tech investment are skewed and highly uncertain, in part because R&D projects have a low probability of succeeding financially and because there is asymmetric information shared between firms and potential investors (Percival, J., & Mcgrath, C., 2013). This happens because it is difficult to increase high-tech investments and often insiders will have much better information than outsiders about the prospects of the firm's investments. High-tech investments often have limited value R&D investment, which is predominantly salary payments (Carpenter & Petersen, 2002). For these reasons, it is pertinent and interesting to examine the role of the CEOs and their 17 compensation for managing high-tech firms, thus contributing to improving research in this area, as well as the understanding of CEO compensation. 1.3 The methodology This study uses a sample of US firms listed on the stock exchange, the S&P index. The classification for high-technology firms is the Standard Industrial Classification (SIC). The industry groups such as computer & office equipment, computer storage devices, terminals, services-computer programming services, services-prepackaged software and services-computer integrated systems design. The period under analysis is the period between 2000 and 2010. The data were collected from the same firm every year and the sample was organized as panel data. With panel data it is possible to combine time and cross-sectional data and make more credible statistical inferences. One of the advantages of panel data estimation is the fact that it point out individual heterogeneity. Thus, panel data suggest the existence of differentiating characteristics in the individuals studied and these features may or may not be constant over time. The second advantage of panel data is that it provides a larger amount of information, the data are more variable, the variables are not as collinear, the degrees of freedom are higher and the estimation is more efficient. Furthermore, it not only makes it possible to identify and measure effects that cannot be detected in pure cross-sectional or time studies, but also to build and test complex behavioral patterns, particularly using models with distributed lags with few restrictions. Different econometric estimation methods are used according to the sample and according to the specification of the performance, innovation and executive compensation equation that best explains the research question presented in each essay. 18 1.4 Plan of presentation This dissertation is the result of three essays developed over the last four years, and it is divided into five chapters. After a brief introduction addressing the main purposes of this work (provided in Chapter 1), Chapter 2 presents the first essay on CEO compensation and its relation to performance. The second essay is an extension of the first as it addresses the total compensation model for the shortand long-term. Panel data generalized least squares (GLS) and Seemingly Unrelated Regression (SUR) methods were estimated to create two other models, one for total compensation, and another for both total and cash compensation in the period between 2000 and 2010. Chapter 3 examines the effects of introducing the financial accounts standard, the statement 123 (R), and exploits the change in the accounting treatment of stock-option compensation, as well as the fair-value report which entered into force in December 2005, in the period of analysis. A Panel data SUR model is used to estimate total compensation and cash compensation as a proportion of total pay for the period between 2000 and 2010. It was found that there is an increase in CEO compensation after introducing the SFAS 123 (R). Although the change in the plan design was not analyzed, a new accommodation of CEO compensation was found as a result of the new rules of the SFAS 123 (R). To complete the main purpose of this dissertation, Chapter 4 investigates the relation between CEO compensation in high-tech firms and their choice between performance and innovation when managing high-tech firms. The literature on high-technology firms is examined, and the relationship between the CEO behaviors on corporate governance is discussed, along with the aggregate innovation activity of corporations. Innovation is measured as R&D expenses, number of patents and brands. After a brief review of the inter-relationship between CEO compensation, performance and innovation suggested that, from an econometric point of view, a system of three simultaneous equations should be formulated that specify the relationship between the abovementioned variables. Panel data Seemingly Unrelated Regression (SUR) and General Method of Moments (GMM) methods were used to estimate total compensation as function of performance and innovation. 19 Finally, Chapter 5 summarizes the results of the four essays, pointing out their limitations and presenting some topics for future research. 20 CHAPTER 21 Executive Compensation for the longand short-term in High-Technology Firms 1 Part of this chapter is published in China-USA Business Review, Volume 12, Number 11, 2013 (125). We are grateful to Professor Jerry Haar at Florida International University for discussing this paper at the International Academy of Management and Business, 15th Conference in Lisbon, in April 2013 and Contemporary Issues in Business, Management and Education ‘2013, in November. We are also grateful to anonymous attendees and professors present at this conference for their helpful comments. 21 1. Introduction While most management scholars would agree that technological innovation is a key source for competitive advantage in high-technology firms and that top executives in those firms should be rewarded accordingly, little is known about which executive pay policies are more appropriate for those organizations to promote such goals. The high-technology sector plays a pivotal role in the new economy and has become the major source of employment and productivity growth over the last years. Innovativeness is also one of the fundamental instruments for growth strategies to enter new markets and to provide the company with a competitive edge. The purpose of this study is to investigate the relation between the Chief Executive Officer (CEO) pay and the value, performance, and behavior of the firms in terms of innovation in hightechnologies. This work will contribute to this subject as it introduces a new measurement pertaining to the relationship between the CEO and the other members of the top executive team. Furthermore, this paper studies the relation between this measurement and the performance and behavior of firms in terms of innovation. For that, this paper will use a new data sample of hightech companies in the S&P for the period between 2000 and 2010. In their paper, entitled “The CEO pay slice”, Bebchuk, Cremers, and Peyer (2011) studied the relation between the CEO pay slice and the value, performance, and behavior of public firms, demonstrating a rich set of relations between these aspects. Furthermore, Makri, Lane, and Gomez-Mejia (2006) reported empirical evidence that high-technology firms that use outcomebased and behavior-based performance criteria to reward executives exhibit better market performance than those that do not. Their research on innovation CEO pay linkages in hightechnology firms has focused on aligning pay with the quantity of innovation inputs (R&D spending) and outputs (number of patents). In fact, authors show the importance of the quality of innovation outputs. They argue that for CEO pay-performance relations in high-technology firms these views are not incompatible, but represent two sides of the same coin (Makri, Lane, & Gomez-Mejia, 2006). Moreover, as pointed out by Makri, Lane and Gomez-Mejia to engage in innovative projects leading to innovations, the incentive schemes play a pivotal role in inducing 22 senior organizational managers. Furthermore, to secure the stream of innovations a firm needs to enhance its economic performance with a proper pay scheme to encourage executives (Makri et al., 2006). Appropriate incentives can be the tools in many cases, however, by basing compensation on changes in shareholder wealth. According to Graham (2012), managers often have better information than shareholders and boards in terms of identifying investment opportunities and assessing the profitability of potential projects. Furthermore, the fact that managers are expected to make higher investment decisions explains why shareholders relinquish decision rights over their assets by purchasing common stock (Graham et al., 2012). The theory summarizes that executive pay should be designed by the board to maximize shareholder value. The level and structure of executive pay have already been discussed in the literature, resulting in three dominant views. One strand of literature studies the pay-toperformance sensitivity. Jensen and Murphy (1990a) showed that CEO wealth is only weakly related to firm performance. Subsequently, another view provides abundant evidence of a significant increase in CEO pay in both absolute and relative terms since 1990, which is consistent with a better alignment of interests between managers and shareholders (Murphy, 1999; Bebchuk & Fried, 2004; Frydman, 2009). Another important strand of literature explains the level and the functional form of pay as skimming issues rather than optimal contracting outcomes. The inner workings of a top executive team and their importance for firm performance and innovation are hard to observe or quantify. As previously described, in order to promote firm growth, sustainable advantage, innovation and performance behavior, the role of the CEO is fundamental. Furthermore, over the last years, due to the effects of the global financial crisis, the role of the CEO has been called into question, as well as their behavior and their paycompensation as a result of their performance and objectives. Moreover, it is essential to maintain confidence in the executive for there to be a balance between the institutions that foster the best conditions for their employees and maximize the profits of their shareholders. For these reasons, and because this subject is pertinent, it is interesting to examine these issues and contribute to the enrichment of research in this area. 23 This study explores the performance determinants of the high-tech and all other CEO pays for long-term and short-term periods. This work also attempts to examine the systematic difference in CEO pays and the performance expectations of high-tech firms and other firms. Furthermore, this paper attempts to examine how in high-tech and others sectors, CEO pays are related to various performance measurements, such as assets and employment in their specificity in high-tech firms, sales growth, operating income before depreciation, net income before extraordinary items and discontinued operations, and earnings per share (Epstein & Roy, 2005).. This work is organized as follows: Section two contains a revision of the main theories in the literature, as well as an analysis of executive compensation in order to address agency problems. Furthermore, this section provides an analysis in order to examine CEO pays for performance and the appropriate measurements of corporate financial performance in hightechnology firms. Section three explains the research hypotheses and section four presents the methodology, sample, and data collection for the regression estimation, as well as the results of the econometric model in order to assess the influence that firm performance has on executive compensation. Lastly, the main conclusions are discussed, as well as some limitations and new perspectives for future research. 2. Corporate Performance and CEO Compensation In the period between 1970 and 2005, it was observed that executive compensations increased tremendously. The underlying reasons for these executive compensations need to be discussed and analyzed so as to provide a better understanding on this matter as we move into the future. Much literature on executive compensation has emerged since Jensen and Meckling (1976) published their work. According to Jensen and Murphy (1990b): There are serious problems with CEO compensation, but “excessive” pay is not the biggest issue. The relentless focus on how much CEOs are paid diverts public attention from the real problem—how CEOs are paid. In most publicly held companies, the compensation of top 24 executives is virtually independent of performance (pp 138). 2.1 Executive Compensations to Address Agency Problems The emergence and general acceptance of the agency theory and the parallel research on executive compensation began in the early 1980s. It was the evolution of the modern corporation with ownership separation and control that undermined the agency theory. Early studies in this area focused on documenting the relation between CEO pay and company performance. The problem of managerial power is analyzed in modern finance as an agency problem. The discussion of executive compensation must proceed with the fundamental agency problem afflicting management decision-making as background. According to Jensen and Murphy (1990a), there are two approaches to agency problems. The authors state that there is an optimal contracting approach, which is when boards use design compensation schemes to maximize shareholder value with efficient incentives. To connect the agency problem and the executive compensation, the authors use the managerial power approach, when this connection is seen as an integral part of the agency problems. It is important to remember that the principal-agent problems treat the difficulties that arise under conditions where information is incomplete and asymmetric whenever a principal hires an agent (Jensen & Murphy, 1990a). The agency theory is directed as an agency relationship between principal and agent in which one part—the principal—delegates work to another—the agent—, who performs that work. It is created at any company that is not owned by its manager. This theory may be summarized as having two problems: firstly, the agency problems arise when the desires or goals of the principal and agent are conflicting and when it is difficult or expensive for the principle to verify what the agent is doing; the second is the problem of risk sharing that arises when the principal and agent have different attitudes towards risk. Maybe the agent and the principal prefer different actions and different risk choices. Jensen and Meckling (1976) suggested that the agents of a company have the tendency to expropriate from the company because the benefits are higher than the cost as such costs are shared or undertaken by various shareholders. Therefore, there should be a balance, and both parties’ participation constraints should be satisfied. According to them, the 25 agency problem existed in all organizations and cooperatives, including universities (Jensen & Meckling 1976). The agency problem is a classic problem in corporate governance as a result to motivate executives to do what is best for their company when they themselves do not own the company. It is necessary to anticipate the agency problem as because of it company investors may try to specify how the manager should act. Furthermore, it is necessary to analyze this problem because the owner may not be able to predict the business and may not know the best action for their manager (Shleifer & Vishny, 1996). The contracts signed between shareholders and managers are usually general, specifying broad goals and the division of profits. These contracts do not specify how managers should behave in specific business situations. Some authors see the weakness of shareholder rights more generally and warn shareholders and their advisers to focus on the corporate governance provisions that really matter for the firm’s value (Bebchuk, Cohen, & Ferrell, 2009; Cremers & Nair, 2005). To help solve the apparent theoretical paradox in agency predictions on the normative consequences of performance-based pay, it is possible to create a common fate for the principal and the agent, or to make the agent overly conservative. The agency theory has been the foundation for both positive and negative answers to the key question: Does incentive compensation help hightechnology firms attain higher subsequent performance levels (Makri, Lane, & Gomez-Mejia, 2006)? Some authors assume CEOs to be more powerful when they serve as chair of the board, when they are the only member of the board, and when they have the status of a founder (Adams, Almeida, & Ferreira, 2005). 2.2 Examining CEO Pay-for-Performance For Murphy (1999), the components of CEO pay are substantially heterogeneous in pay practices across firms and industries. Most executive pay packages contain four basic components: a base salary, an annual bonus linked to accounting performance, stock options, and long-term incentive plans. Moreover, executives participate in employee benefit plans and also 32 income before extraordinary items and discontinued operation and earning per share suited the high-technology firms, as presented above. 3. Research Hypotheses As previously discussed, existing theories provide predictions on the outlined considerations related to firm value, allowing for two different selection hypotheses. The first research question will be: Hypothesis 1: The CEO compensation is positively correlated with firm performance for high-technology companies in the short-term. Rejection of the null hypothesis would mean that the relative weight in terms of total compensation of each compensation component (such as salary, bonus, stock options and other compensations) are different goals for executives, as opposed to performance in the short-term. It might be argued that powerful incentive models are especially valuable for high value firms with high opportunities for growth that need to be decisively and vigorously pursued. It is possible that high value firms have CEOs interested in long-term performance and in obtaining personal benefits in terms of total compensation. It might be argued that powerful incentive models are especially valuable for high value firms with high opportunities for growth that need to be decisively and vigorously pursued. It might also be that high value firms are especially likely to attract star CEOs and pay gold parachutes. The second research question will be: Hypothesis 2: The CEO compensation is positively correlated with firm performance in high-technology companies in the long-term. It is possible that high value firms have CEOs with an interest in long-term performance and with obtaining personal benefits in terms of total compensation. With less intensity and yet more persistent than long-term compensation, bonuses and salary are determinant and in the same effect related to accounting performance. 33 4. Methodology, Sample, and Data Collection The sample chosen is the ExecuComp database, which was used to find the variables and to create a sample of firms between 2000 and 2010. The ExecuComp database provides yearly data on salary, bonus, stock options and restricted stock grants, as well as managerial stock and option holdings for top executives in firms within the Standard & Poor’s Index (S&P 1500). Firstly, to test this hypothesis, the following specification is presented of the balanced panel of hightechnology firms, between 2004 and 2010. High-Technology firms are the firms that operate in an industry with a four-digit Standard Industrial Classification (SIC) code using the Fama and French classification of 48 industry groups (Fama & French, 1997). The ExecuComp database collects information about seven independent variables — total assets (ASSETS) and percentage change of assets (ASSETSCHG), employees (EMPL), total annual net sales (SALES) and yearly changes in sales (SALECHG), operating income before depreciation (OIBD), net income before extraordinary items and discontinued operation (NIBEX), earning per share (EPSEX) and return on assets (ROA) — and the independent, total compensation (TOTAL_COMP) cash compensation (CASH) variables are listed by each year and company. Several measurements were used in this study, such as control variables. Several measurements were used as control variables in this study. These include the number of employees, assets, increase in sales, net income, and the EPS, as a proxy of firm size, firm performance and wealth of the shareholder, which are the common predictors of executive pay. The High-Tech Dummy (DHTECH) is equal to one if the firm operates in an industry with a four-digit SIC code of 3570, 3571, 3572, 3576, 3577, 3661, 3674, 4812, 4813, 5045, 5961, 7370, 7371, 7372, or 7373, instead of four-digit SIC codes, in the following industry groups: computer & office equipment, computer storage devices, communication terminals, equipment, peripheral equipment, NEC, telephone & telegraph apparatus, radiotelephone communications, telephone communications, wholesale-computers & peripheral equipment & software, retail-catalog & mail-order houses, services-computer programming, data processing, services-computer programming services, services-prepackaged software and services-computer integrated systems 34 design. The main variable of the analysis is TOTAL_COMP and it is defined by the sum of the total compensations of the top executives in each company and it includes: salary, bonus, non-equity incentive plan compensation, grant-date fair value of option awards, grant-date fair value of stock awards, deferred compensation earnings reported as compensation, and other compensations. The table below identifies the updated variables that were used, including their definitions, measurement units and the expected signs, as reported by the theory. Table 1 - Executive Compensation Dependent and Independent Variables Name Expected variation Definition Units Ln (TOTAL_COMP) Total compensation Ln (the sum of the compensations of top executives includes: salary, bonus, non-equity incentive plan compensation, grant-date fair value of option awards, grant-date fair value of stock awards, deferred compensation earnings reported as compensation, and other compensations). Thousands Ln(CASH) Cash compensation Ln (SALARY + Bonus) The dollar value of the base salary plus bonus earned by the named executive officer during the fiscal year . Thousands Ln (ASSETS) (+) Ln (the total assets as reported by the company). Millions Ln (EMPL) (+) Ln (employees, the total employees as reported by the company (#)). Thousands EPSEX (-) EPS (Primary) excluding extraordinary items and discontinued operations. SALECHG (+) The year to year percentage change in Sales. Percentage Ln (OIBD) (+) Ln (the operating income before depreciation as reported by the company). Millions ROA Return on assets (+) The Net Income Before Extraordinary Items and Discontinued Operations divided by Total Assets. This quotient is then multiplied by 100. Percentage Ln(COMMEQ) (+) The sum of Common Stock, Capital Surplus, Retained Earnings, and Treasury Stock adjustments. Millions (OIBD/ASSETS)*100 (+) Ln (the Operating Income Before Depreciation as reported by the company/Assets).This quotient is then multiplied by 100. Percentage Ln(SALES) Ln(NIBEX) (+) (+) Ln (The Net Annual Sales as reported by the company). Ln (the Net Income Before Extraordinary Items and Discontinued Operations). Millions Millions 35 SIC Standard Industrial Classification Code. SPCODE (+) and (-) Current S&P Index membership "SP" = S&P 500 "MD" = S&P Midcap Index "SM" = S&P Small cap Index "EX" = not on a major S&P Index To test the research hypotheses, two models were used with a different approach to total compensation. The first is a total compensation model and the second is a total and cash compensation model. The first approach uses a dynamic model to explain total shortand longterm compensations. The second model divides the components of total compensation and analyzes total compensation and cash in two equations, which are used as a measurement of short-term compensation. 4.1 Total Compensation model The model presented below was used to test whether firm performance is relevant to explain executive compensation. Firstly, the model for the short-term: ln (TOTAL_COMP)it = a+ b1ln(ASSETS)it + b2ln(OIBD)it + b3ln(NIBEX)it + b4ln(EMPL)it + b5*ERPSEXit + b6*SALECHGit + uit (1) and the secondly, the model for the long-term: ln (TOTAL_COMP)it = a + b1ln(ASSETS)it + b2ln(OIBD)it + b3ln(NIBEX)it + b4ln(EMPL)it + b5*ERPSEXit + b6*SALECHGit + c*ln(TOTAL_COMP)it-1 +uit (2) where, i and t represent the year and the company, respectively. The coefficient a is a constant denoting the base level from which the sum of the compensations of top executives vary according to the changes in performance variables. The panel data model is used as it is the most suitable way of studying a large set of 36 repeated observations and because it assesses evolution over time. With panel data it is possible to simultaneously explore several variations over time and between different individuals. The use of such models has increased immensely and, in fact, combining time and cross-sectional data brings many advantages: it is possible to use a larger number of observations and the degree of freedom in estimates increases, thus making statistical inferences more credible. At the same time, the risk of multicollinearity is reduced since the data in companies present different structures. Moreover, this model provides access to further information and the efficiency and stability of the estimators increase, while enabling the introduction of dynamic adjustments (Gujarati, 2004, 2000; William, 2002, 2003). According to Bebchuk et al. (2011), in order to test the variables and to assess the abovementioned research hypotheses there are independent variables that will possibly be used by the regression model to perform the estimation. At an empirical level, this analysis focuses on a sample of 500 high-tech companies in the S&P index (S&P500), for the period between 2004 and 2010, which constitutes a sample of 3,356 observations. Table 2 - Descriptive Statistics individual sample S&P500 TOTAL_COMP ASSETS EPSEX OIBD NIBEX EMPL SALECHG DHTECH Mean 25,599.88 46,886.07 12.65983 3,249.064 1,215.441 46.30875 10.80346 0.121275 Median 19,561.74 10,698.19 2.090000 1,296.557 507.4820 17.59400 7.909000 0.000000 Maximum 264,964.7 2,264,909. 8,548.000 78,669.00 45,220.00 2,100.000 1,106.400 1.000000 Minimum 454.4000 182.7430 -37.84000 -76735.00 -99289.00 0.053000 -92.68800 0.000000 Std. Dev. 21,706.99 168,685.8 252.8151 6,878.305 3,586.554 110.5579 30.22776 0.326496 Skewness 3.365542 8.450905 26.43878 4.478257 -3.906328 11.92814 15.95401 2.320283 Observations 3,346 3,356 3,353 3,242 3,356 3,333 3,350 3,356 The descriptive statistics of the variables for total CEO compensation in high-tech firms are presented in Table 2. In the S&P500, in the period between 2004 and 2010, there are about 12% of high-technology firms, and it is possible to observe that the group of top executives in each company has a total average compensation around USD 25,600 million. Another interesting 37 finding is that in this period there was not always an increase in sales, but there was a 10.8% average growth in high-tech companies. The regressions presented below (see table 3, and table 4) was estimated using the Generalized Least Squares (GLS) with a fixed effect model for time. This means that the regression coefficients which were used with the fixed effect model for explanatory variables do not vary over time. The estimation was conducted assuming that the company’s heterogeneity is captured in the constant part and that it differs between companies. The fixed effect model is the most suitable when there is a correlation between errors and variables (Greene, William, 2003). For each shortand long-term model two scenarios were tested to confirm and show the internal stability between them concerning the influence that high-technology firms have on the set of variables for total compensation. Table 3 - Total Executive Compensation Estimation Ln (TOTAL_COMP) Coefficient Prob. Ln (TOTAL_COMP) Coefficient Prob. Constant 7,279,001 0.0000 Constant 7,302,197 0.0000 Ln(ASSETS) 0.054566 0.0002 Ln(ASSETS) 0.054500 0.0002 EPSEX -0.000620 0.0000 Ln(ASSETS)*DHTECH 0.014395 0.0000 Ln(OIBD) 0.204273 0.0000 EPSEX -0.000619 0.0000 Ln(NIBEX) 0.069080 0.0003 Ln(OIBD) 0.200525 0.0000 Ln(EMPL) 0.044388 0.0000 Ln(NIBEX) 0.070401 0.0002 SALECHG 0.002650 0.0000 Ln(EMPL) 0.043644 0.0000 DHTECH 0.140800 0.0000 SALECHG 0.002667 0.0000 Period fixed effects (dummy variables) Period fixed effects(dummy variables) Weighted Statistics Weighted Statistics R-squared 0.403318 R-squared 0.402732 Adjusted R-squared 0.400693 Adjusted R-squared 0.400104 S.E. of regression 0.553285 S.E. of regression 0.553552 F-statistic 1,536,448 F-statistic 1,532,713 Prob(F-statistic) 0.000000 Prob(F-statistic) 0.000000 Total panel (unbalanced) observations: 2969 Total panel (unbalanced) observations: 2969 In order to assess the abovementioned research hypotheses, the regression model was used and estimated with fixed effects. The first hypothesis for the positive influence of the CEO 38 compensation in firm performance is presented in Table 3. As it is possible to observe, the regressions are globally significant, with a 5% significance level. The following table presents the results of the estimation conducted by the generalized method using the fixed effect model for the studied data. The statistics are computed based on a panel data set of 484 firm-year observations, a total of about 2,969 companies that represent 14.08% of high-technology firms between 2004 and 2010. The total assets, the operating income before depreciation and the net income before extraordinary items and discontinued operations, the growth sales and employment, as reported by companies, are positive (see table 1 – expected variation) and significantly related to total executive compensations. Table 4 - Total Executive Compensation Estimation in long-term Ln (TOTAL_COMP) Coefficient Prob. Ln (TOTAL_COMP) Coefficient Prob. Constant 3,372,131 0.0000 Constant 3,379,249 0.0000 Ln(TOTAL_COMP(-1)) 0.545348 0.0000 Ln(TOTAL_COMP(-1)) 0.545797 0.0000 Ln(ASSETS) 0.026126 0.0388 Ln(ASSETS) 0.026164 0.0387 EPSEX -0.000275 0.0000 Ln(ASSETS)*DHTECH 0.006955 0.0177 Ln(OIBD) 0.085650 0.0006 EPSEX -0.000274 0.0000 Ln(NIBEX) 0.029623 0.0787 Ln(OIBD) 0.083758 0.0007 Ln(EMPL) 0.021974 0.0086 Ln(NIBEX) 0.030160 0.0733 SALECHG 0.003060 0.0000 Ln(EMPL) 0.021637 0.0096 DHTECH 0.066078 0.0138 SALECHG 0.003064 0.0000 Period fixed effects (dummy variables) Period fixed effects (dummy variables) Weighted Statistics Weighted Statistics R-squared 0.606381 R-squared 0.606318 Adjusted R-squared 0.604336 Adjusted R-squared 0.604273 S.E. of regression 0.440828 S.E. of regression 0.440872 F-statistic 2,966,106 F-statistic 2,965,326 Prob(F-statistic) 0.000000 Prob(F-statistic) 0.000000 Total panel (unbalanced) observations: 2517 Total panel (unbalanced) observations: 2517 As expected, the EPS are negative and significantly related to total compensation in hightech companies. This indicates that there are no explicit contractual arrangements linking compensations and EPS. The performance ratio of firms measured by return has a negative influence. Note that around 40.4% (R2 = 0.404) of the variance in degree of CEO compensation 39 can be explained by the group of variables for the short-term (see Table 3). However, it is important to highlight that around 60.6% (R2 = 0.606) of variance in the degree of CEO compensation for the long-term can be explained by the group of variables (see Table 4). These indicate that the variables addressed here play a significant role in explaining executive compensation for shortand long-term periods, as stated by the Chi-Square test (P-value = 0). 4.2 Total compensation and cash compensation models The two primary measurements of CEO compensation were used. The short-term compensation consisted of annual salary and bonus, which represents the total cash compensation received during a specific year. Annual salary and bonus for 2000 and 2010 (in thousands of dollars) were taken from the ExecuComp data set. The long-term compensation represents the equity-based compensation of a CEO, as reported by Frydman, C (2008). As she reported in the case study of General Electric, salary and bonus are defined as the level of salaries and current bonuses, both awarded and paid out during the year. Long-term bonus measures the amount paid out during the year according to long-term bonuses awarded in prior years. Total compensation is the sum of salary, bonus, long-term bonus and the Black–Scholes value of stock options granted (Frydman, 2009). Other dummy variables are used, such as YEAR for the period between 2000 and 2010. The main variables of the analysis in the system equation are T_COMP (defined by the sum of Salary, Bonus, Non-Equity Incentive Plan Compensation, Grant-Date Fair Value of Option Awards, Grant-Date Fair Value of Stock Awards, Deferred Compensation Earnings Reported as Compensation and Other Compensations) and CASH (Salary plus bonus) of all top executives in each company. The models introduced by the system equation presented below were used to test whether firm performance is relevant to explain executive compensation for the long and shortterm. Firstly, the model for the long-term, 40 Ln (T_COMP)ij = b11+ b12*ln(ASSETS) ij + b13*ASSETCHG ij + b14*ROAij+b15*ln(OIBD/ASSETS*100)ij+ b16*ln(SALES) ij +b17*ln(NIBEX)ij + +b18*ERPSEX ij +b19*SPCODE ij + b10*DHTECH ij+ b31*ln(COMMEQ)ij + ∑∂j∗Yearj 2010 2001 + uij (1) and for the short-term Ln (CASH)ij = b21+ b22*ln(ASSETS) ij + b23*ASSETCHG ij + b24*ROA ij + b25*ln(OIBD/ASSETS*100)ij + b26*ln(SALES) ij +b27*ln(NIBEX)ij + b29*SPCODE ij + b20*DHTECH ij+ b32*ln(COMMEQ)ij + ∑∂j∗ 2010 2001 Yearj + vij (2) Where i and j represent the year and the company, respectively. The coefficients b11 and b21 are constants denoting the base level from which the sum of the compensations of top executive varies according to the changes in performance variables. Table 5 presents the descriptive statistics of the variables observed for the firms of the S&P1500 during 2000-2010, which constitutes a sample of 15,265 observations. Some interesting outcomes were found as a result of this study. Table 5 - Descriptive statistics individual sample S&P 1500 TOTAL_C OMP CASH ROA NIBEX EPSEX SALES ASSETCHG ASSETS COMMEQ OIBD DHTECH Mean 13727.19 4156.375 1.577071 288.9755 3.47042 5488.187 39.96929 15205.21 2714.774 1000.633 0.14415 Median 8089.794 2937.509 3.874000 58.40650 1.20000 1239.655 6.03400 1746.966 637.0890 176.3185 0.00000 Maximum 641446.2 199115.9 3551.351 45220.00 8548.00 425071.0 522050.0 3221972. 211686.0 124840.0 1.00000 Minimum 0.0000 0.0000 -10300.00 -99289.00 -231.670 -4,234.47 -99.4270 0.00000 -111403.0 -76735.00 0.00000 Std. Dev. 20478.12 5100.74 82.6438 2018.437 113.869 16956.91 3705.640 88055.59 9044.259 3833.163 0.35125 Skewness 8.7409 9.9984 -94.4612 -9.3868 55.3268 10.6209 140.802 15.88214 8.80257 9.601327 2.02623 Observ. 19678 19889 19869 19870 19842 19870 19855 19872 19872 19606 19889 The descriptive statistics of the variables for total CEO compensation in high-tech firms are presented in Table 5. In the S&P1500, in the period between 2000 and 2010, there are about 14.415% high-technology firms, and it is possible to observe that the group of top executives in 41 each company has an average total compensation around USD 13,727 million and earn around USD 4,156 million in cash. Another interesting finding is that, in this period and in this group of companies, there is an increase in assets and returns on assets around 39.96% and 1.57 %, respectively. To compare results of this model with the model presented in section 4.1, it was tested with the sample of the S&P 500 (see table 6) and then the sample was extended to the S&P1500 (see table 7) to understand the behavior in other dimensions of the firms. Table 6 - Total compensation and cash compensation estimations using the SUR method for S&P500 Ln (T_COMP) Coefficient Prob. L n (CASH) Coefficient Prob. Constant 6.2794 0.000 Constant 6.096 0.000 Ln(ASSETS) 0.1748 0.000 Ln(ASSETS) 0.198 0.000 ASSETCHG 0.0011 0.000 ASSETCHG 0.000 0.041 ROA 0.0118 0.007 ROA 0.007 0.066 OIBD/ASSETS*100 0.0086 0.000 OIBD/ASSETS*100 0.005 0.005 Ln(SALES) 0.0855 0.000 Ln(SALES) 0.054 0.000 Ln(NIBEX) 0.0563 0.010 Ln(NIBEX) 0.034 0.048 EPSEX -0.0004 0.000 - DHTECH 0.2927 0.000 DHTECH -4.890 0.335 Ln (COMMEQ) 0.0693 0.000 - 2001 -0.0126 0.784 2001 -1.110 0.267 2002 -0.0651 0.153 2002 0.501 0.616 2003 -0.0763 0.086 2003 3.239 0.001 2004 -0.0157 0.719 2004 4.617 0.000 2005 -0.0392 0.366 2005 3.329 0.001 2006 0.0240 0.574 2006 -12.016 0.000 2007 0.0531 0.214 2007 -13.805 0.000 2008 0.0161 0.714 2008 -13.973 0.000 2009 -0.0208 0.634 2009 -13.646 0.000 2010 0.0834 0.056 2010 -13.890 0.000 R-squared 0.3920 R-squared 0.4280 Adjusted R-squared 0.3894 Adjusted R-squared 0.4258 S.E. of regression 0.6069 S.E. of regression 0.4964 Durbin-Watson stat 0.8438 Durbin-Watson stat 0.5778 Included observations:4421 Included observations:4421 48 states and united kingdom, The Economic Journal 110, 640-671. Core, J. E., Guay, W. R., & Larcker, D. F. (2003). Executive Equity Compensation and Incentives: A Survey. Economic Policy Review, Vol. 9, No. 1, April 2003. Core, J. E., Guay, W. R., & Verrecchia, R. E. (2003). Price versus non-price performance measures in optimal CEO compensation contracts. The Accounting Review, 78(4), 957981. Cremers, K. J. M., & Nair, V. B. (2005). Governance Mechanisms and Equity Prices. The Journal of Finance, 60(6), 2859-2894. Eisenhardt, K. M. (1989). Agency Theory: An Assessment and Review. The Academy of Management Review, 14(1), 57-74. Epstein, M. J., & Roy, M. (2005). Evaluating and monitoring CEO performance: evidence from US compensation committee reports. Corporate Governance, 5(4), 75 - 87. Fama, E. F., & French, K. R. (1997). Industry costs of equity. Journal of Financial Economics, 43(2), 153-193. Fama, E. F., & French, K. R. (1997). Industry costs of equity. Journal of Financial Economics, Frydman, C. (2009). Learning from the past: Trends in executive compensation over the 20th century. CESifo Economic Studies, 55(3-4), 458-481. Frydman, C., & Jenter, D. (2010). CEO compensation. National Bureau of Economic Research Working Paper Series. Gabaix, X., & Landier, A. (2008). Why has CEO Pay Increased So Much? The Quarterly Journal of Economics, 123(1), 49-100. Gomez-Mejia, L. R, Gideon, D. M., & Balkin, D. B. (2000). Is CEO Pay in High-Technology Firms Related to Innovation? The Academy of Management Journal, 43(6), 1118-1129. 49 Graham, J. R., Li, S., & Qiu, J.. (2012). Managerial Attributes and Executive Compensation. Review of Financial Studies, 25(1), 144-186. Greene, William H., 2003. Econometric analysis. (Prentice hall) Grinstein, Y., Weinbaum, D., & Yehuda, N. (2011). The economic consequences of perk disclosure. SSRN eLibrary. Gujarati, Damodar N., 2004. Basic econometrics. Hall, B. J., & Liebman, J. B. (1998). Are CEOS really paid like bureaucrats?. Quarterly Journal of Economics, 113(3), 653-691. Hall, B. J., & Murphy, K. J. (2003). The Trouble with Stock Options. Journal of Economic Perspectives, 17(3), 49-70. Henry, T. F., Shon, J. J., & Weiss, R. E. (2011). Does executive compensation incentivize managers to create effective internal control systems? Research in Accounting Regulation, 23(1), 46-59. Jensen, M. C., & Murphy, K. J. (1990). Performance Pay and Top-Management Incentives. Journal of Political Economy, 98(2), 225-264. Jensen, M. C., & Murphy, K. J. (1990a). Performance pay and top-management incentives. Journal of Political Economy, 98(2), 225-264. Jensen, M. C., & Murphy, K. J. (1990b). CEO incentives: It’s not how much you pay, but how. Harvard Business Review, 3, 138-153. Jensen, M., & Meckling, W. (1976). Theory of the firm: Managerial behaviour, agency costs, and ownership structure. Journal of Financial Economics, 3(4), 308-360. Jensen, Michael C., (2004), Agency costs of overvalued equity, ECGI - Finance Working Paper No. 39/2004. Makri, M., Lane, P. J., & Gomez-Mejia, L. R. (2006). CEO incentives, innovation, and performance in technology-intensive firms: A reconciliation of outcome and behavior- 50 based incentive schemes. Strategic Management Journal, 27(11), 1057-1080. Malmendier, U., & Tate, G. (2009). Superstar CEOs. The Quarterly Journal of Economics, 124(4), 1593-1638. Murphy, K. J. (1999). Chapter 38 executive compensation. In C. A. Orley, & C. David (Eds.). Handbook of Labor Economics, 3, 2485-2563. Sesil, J., Lin, Y., & , Steven M. (2006). Stock Option Adoption and Irrational Exuberance: The Impact On Profitability. working papers series Shim, E. S., Lee, J., & Joo, I. K (2009). CEO compensation and US high-tech and low-tech firms’ corporate performance. Contemporary Management Research, 5, 93-106. Sundaram, R. K., & Yermack, D. (2007). Pay me later: Inside debt and its role in managerial compensation. Journal of Finance, 62, 1551-1588. Young, S., & Jing, Y. (2011). Stock repurchases and executive compensation contract design: The role of earnings per share performance conditions. Accounting Review, 86(2), 703733. 51 CHAPTER 3 CEO compensation in high-tech firms and changes in the Financial Accounting Standard SFAS No 123 (R) 52 1. Introduction Chief Executive Officer (CEO) compensation became common in the late 1970s and early 1980s and is often discussed in the literature ever since. Numerous stories have appeared recently in the financial press pointing out how many executives define contract remunerations. These news and striking reports have raised concerns on compensation. Nevertheless, no consensus view has emerged, and there is still much to learn about the determinants of CEO compensation. Appropriate incentives can be the tools in many cases, however, by basing compensation on changes in shareholder wealth. According to Graham (2012), managers often have better information than shareholders and boards in terms of identifying investment opportunities and assessing the profitability of potential projects. Furthermore, the fact that managers are expected to make higher investment decisions explains why shareholders relinquish decision rights over their assets by purchasing common stock (Graham, Li, & Qiu, 2012) . This study explores how the Financial Accounting Standards (FAS) Statement 123 (R) affects the performance determinants of CEO pays for long-term and short-term periods and points out the influence of high-tech firms. Furthermore, this paper examines how high-tech firms behave facing the cash based compensations and total CEO pays related to various performance measurements. The performance measurements pointed in this work are the usual accounting ratios of corporate finance. In 1990, Jensen and Murphy wrote that it is possible that CEO bonuses are strongly tied to an unexamined and/or unobservable performance measurement. When referring to the swings in CEO pay from year to year, the authors explain that the variations are consistent with the existence of an overlooked and yet important performance measurement, and that increase suggests that CEO pay is essentially unrelated to all relevant performance measurements (M. C. Jensen & Murphy, 1990, 1990b; K. J. Murphy, 1999). In many Standard & Poor´s (S&P) firms, employee stock option plans are an important component of employee remuneration. In 1999, 94% of companies in the S&P 500 offered stock options to their top employees (Brian J. Hall & Murphy, 2002; K. J. Murphy, 1999). In order to better understand this argument, this study 53 investigates the relation between the CEO pay and the performance against high-technology firms with the balance cash compensation and total compensation. The combination of salary, incentives and bonuses is often referred to as cash compensation for executives. The CEO behavior is different when we think in short-term and long-term periods. The main goal of this paper is to provide a broader perspective on the relationship between CEO pay and firm performance and how high-technology can improve that performance while analyzing the behavior with the implementation of the SFAS 123 (R) and all its accounting rules underlying obligations. The change in the accounting treatment of stock-option compensation is exploited as well as the fair-value report under the SFAS 123 (R), which was issued by the Financial Accounting Standard Board (FASB) and entered into force in December 2005. This paper contributes to the under-studied empirical literature on the accounting treatment of equity-based compensation, influenced by the change in accounting rules and its influence on executive pay of high-tech firms. This work is organized as follows: Section 2 contains a revision of the main theories in the literature, as well as an analysis of accounting treatment of equity-based compensation before and after the SFAS 123 (R), an analysis on executive compensation in order to address agency problems and the income strategy impact on CEO compensation. Furthermore, this section examines the appropriate measurements for corporate financial performance in high-technology firms. Section 3 explains the research hypotheses and section 4 presents the methodology, sample and data collection for the regression estimation, as well as the results of the econometric model to assess the influence that firm performance has on executive compensation before and after SFAS 123 (R). Lastly, the main conclusions are discussed, as well as some limitations and new perspectives for future research. 2. Literature review Based on the literature, a study was conducted in order to understand CEO compensation in high-technology firms. 54 2.1 Accounting treatment of equity-based compensation before and after SFAS 123 (R) In December 2004, FASB issued the FASB Statement No. 123 (revised 2004), Accounting for Stock-Based Compensation, to amend and replace the Financial Accounting Standards Statement No 123, which became mandatory for all firms toward the end of 2005 and supersedes the APB Opinion No. 25, Accounting for Stock Issued to Employees. Statement 123 as originally issued in 1995, which established that a fair-value-based method of accounting for share-based payment transactions with employees was preferable. The SFAS 123 (R) requires the use of a fair value accounting method to compute the value of option compensation. A similar approach is followed by international standards International Financial Reporting Standards (IFRS2) that states the same principle. Both standards require employee stock option to be recognized as an expense and measured at the fair value of the employee stock option determined at the time of grant. Prior to implementing the SFAS 123 (R), firms were required to report compensation expense due to stock options in an amount equal to the excess of the stock price at the grant date over the exercise price. This is allowed to as the intrinsic value method. Most options have an exercise price at least equal to the grant date stock price and so this method did not usually result in an expense reported on the income statement. In the originally issued SFAS 123, a company could choose to either report in its income statement the stock compensation expense calculated per the fair value method or the stock compensation expense calculated per the intrinsic value method and disclosing the impact in their footnotes. The SFAS 123 (R) covers a wide range of sharebased compensation arrangements including share options, restricted share plans, performance based awards, share appreciation rights, and employee share purchase plans. The SFAS 123 (R) leads to greater expenses as it increases the overall conservatism income. According to Heltzer (2010), different forms of conservatism have different implications on the quality of income. The author found that the SFAS 123 (R) causes an increased negative relation between economic gains and income, but it is mix on the quality of earnings in terms of conservatism (Heltzer, 2010). Since the publication of the SFAS 123(R) Share-Based Payment, which eliminates the alternative of using the intrinsic value based method, the IFRS and the US 55 GAAP have similar requirements for accounting for share-based payments. With this statement, the convergence between the IFRS in Europe and the GAAP in the US started. 2.2 Executive compensations to address agency problems and the income strategy impact The general acceptance of the agency theory and the parallel research on executive compensation began in the early 1980s. It was the evolution of the modern corporation with ownership separation and control that undermined the agency theory. Early studies in this area focused on documenting the relation between CEO pay and firm performance. The discussion of executive compensation must proceed with the fundamental agency problem afflicting management decision-making as background. According to Jensen and Murphy (1990), there is an optimal contracting approach, which is when boards use design compensation schemes to maximize shareholder value with efficient incentives (M. C. Jensen & Murphy, 1990). To connect the agency problem and the executive compensation, the authors use the managerial power approach when this connection is seen as an integral part of the agency problems. It is important to remember that the principal-agent problems treat the difficulties that arise under conditions where information is incomplete and asymmetric whenever a principal hires an agent (Murphy, 1999, Eisenhardt (1989); Lucian Arye Bebchuk and Fried (2003)). Furthermore, the agency theory aims at solving two problems that can occur in agency relationships. The first is the desires or goals of the principal and agent conflict and it is difficult or expensive for the principal to verify what the agent is actually doing. The problem is that the principal is unable to check if the agent has behaved correctly. Secondly, it is the problem of risk sharing facing the different attitudes toward risk, because the principal and the agent have different actions according to different risk preferences (Eisenhardt, 1989). Hall and Liebman (1998) argue that the solution to the agency problem is aligning the incentives of executives with the interests of shareholders by granting (or selling) stock and stock options to the CEOs. The CEOs have the correct incentives on every margin, including effort, perquisites and project choice, and support that the optimal contract is a one-to-one 56 correspondence between firm value and CEO pay (Brian J. Hall & Liebman, 1998). In their work, Hall and Liebman (1998) conclude that the relationship between pay and performance is much larger than has previously been recognized, and that this includes both gains and losses in CEO wealth. The salary and bonus vary so little because corporate board members are often reluctant to reduce CEO pay, even in response to poor performance and that may attract unwanted media attention. Using salary and bonuses to reward and penalize CEOs may only be possible to create high-powered incentives that align CEO pay with shareholder objectives (Hall and Liebman, 1998). A large part of the executive pay literature argues that compensation and managerial interests should be aligned with shareholder interests in order to solve agency problems (see, for example, the surveys by Murphy and by Core et al. (2003a). Equity-based compensation is widely documented in the research examining pay versus performance. M. Jensen and Meckling (1976), Murphy (2003) and Jensen (2004) state that the increase in stock options pay is the result of the boards’ inability to evaluate the true cost of this form of compensation. The use of equity-based compensation is encouraged by all stakeholders, such as investors, regulators and academics. The controversy over CEO compensation reflects a perception that CEOs effectively set their own pay levels. In most companies, the last decisions over executive pay are made by members outside the board of directors who are keenly aware of the conflicts of interest between managers and shareholders over the level of pay. However, the CEOs and other top managers exert at least some influence on the level and on the structure of their pay (K. J. Murphy, 1999). In recent years, the use of restricted stocks in compensation executives has increased and has been widely criticized when these executives received dividend equivalents on restricted stocks before the vesting period. Agency cost benefits of dividend equivalent rights argue that this practice helps executives focus on the business, and rewards them for managing the business to produce cash. Therefore, this is encouraged because it is a way of distributing dividends by shareholders (Akpotaire, 2011). The SFAS 123 (R) is a change in accounting policy and represents an exogenous shock to the accounting benefits, and restricts the choice of accounting principles by managers (Zmijewski & 57 Hagerman, 1981). There are economic incentives to determine and motivate the managers’ concern with a set of accounting principal utilized to generate the firms’ financial statements. Under economic factors which influence the decision, managers will attempt to archive the optimal reported net income over time and will choose a set of income policies according to theirs goals. There are many variables that induce managers to use deflating policies while other variables encourage managers to choose income inflating solutions. That infers a conservative or liberal firm income strategy. This trade-off means that any combination of Generally Accepted Accounting Practice (GAAP) variables may be optimal for each firm. However, the SAF 123(R) prevents this income strategy by the imposing and restricting some variables as accounting treatment of stock-options compensation and the fair-value report. In their study, Zmijewski and Hagerman (1981) suggest that individual accounting choice decisions are part of an overall firm strategy and applicable in larger firms and in more concentrated industries. In this sense, Matsunaga (1995) suggests that some change in the financial reporting of treatment of stock options, as proposed by the FASB, is likely to reduce the use of the employees’ stock option for some firms (Matsunaga, 1995). 2.3 Financial performance in high-technology firms This chapter provides an analysis on the different forms of measuring performance in hightech firms and how these engage to the level of CEO pay. The behavior of high-tech firms and its contribution to CEO compensation for the shortand long-term are also analyzed. This is consistent with Shim (2009), who argues that it is possible to confirm that high-tech firms that depend more on managing assets are more successful. Some of these assets are technology innovation, continuous improvement, software development and knowledge-based management. High-tech firms must continuously innovate to survive and to sustain their growth (Shim et al., 2009). In high-technology firms it is possible to find innovation, R&D investments and some assets with an essential competitive advantage and there are, at the same time, some risks. Different R&D spending in the firms is indicative of a large variance in the firms’ performance. High-tech investment is particularly important because the returns on high-tech investment are 64 The ExecuComp database collects information on seven independent variables – Assets and year to year percentage change of assets (ΔASSETS), sales, operation income before depreciation (OIDB), net items and discontinued operation (NIBEX), earning per share (EPSEX), The sum of Common Stock Capital Surplus (COMMEQ), net annual sales (SALES) – and dependent total compensation (T_COMP) and cash compensation (CASH) variables are listed by each year and company. Several measurements were used in this study, such as control variables. These include assets, increase in sales, the net Income and the earning per share, as a proxy of firm size, firm performance and shareholder wealth, the common predictors of executive pay. The two primary measurements of CEO pay were used. The short-term compensation consisted of annual salary and bonus, which represents the total cash compensation received during a specific year. Annual salary and bonus for 2000 and 2010 (in thousands of dollars) were taken from the ExecuComp data set. The long-term compensation represents the equity-based compensation of a CEO, as reported by Frydman (2009). As she reported in the case study of General Electric, salary and bonus are defined as the level of salaries and current bonuses, both awarded and paid out throughout the year. Long-term bonus measures the amount paid out during the year according to long-term bonuses awarded in prior years. Total compensation is the sum of salary, bonus, longterm bonus and the Black–Scholes value of stock options granted (Frydman, 2009). The main variables of the analysis in the system equation are T_COMP (defined by the sum of salary, bonus, non-equity incentive plan compensation and other compensations) and CASH (Salary plus bonus) of all top executives in each company. Table 2 presents the descriptive statistics and correlations for variables used in the CEO compensation analysis. Some interesting outcomes were found as a result of this study. 65 Table 2 - Descriptive statistics and correlations for the sample of 1500 S&P over the period 20002010Table2 - Descriptive statistics and corr 1 Variable Mean Median Min Max Stdev N T_COMP 13,727.19 8,089.79 0641,446.20 20,478.12 19678 CASH 4,156.38 2,937.51 0199,115.90 5,100.74 19889 ASSETS 15,205.21 1,746.97 03,221,972.00 88,055.59 19872 Δ ASSETS 39.96 6.03 -99.43 522,050.00 3,705.64 19855 ROA 10.79 11.54 -1,000.00 138.87 76.41 19563 OIBD 1,000.63 176.32 -76,735.00 124,840.00 3,833.16 19606 SALES 5,488.19 1,239.66 -4,234.47 425,071.00 16,956.91 19870 NIBEX 2,889.75 58.41 -99,289.00 42,220.00 2,018.44 19870 EPSEX 3.47 1.2 -231.67 8,548.00 113.87 19842 COMMEQ 2,714.77 637.08 -111,403.00 211,686.00 9,044.26 19842 DHTECH 0.144 - - - - 18889 Panel A: Descriptive statistics - Full time period Variable Mean Median Stdev NMean Median Stdev NT-Test j Med Chi-Squa T_COMP 13,635.58 7,440.77 23,140.38 10687 13,762.41 8,973.70 16,254.54 7156 (-0,402) (100,22)*** CASH 4,751.23 3,352.65 5,594.83 10854 3,357.93 2,630.00 3,877.08 7159 (18,362)*** (502,02)*** ASSETS 13,008.18 1,491.48 64,758.39 10837 18,230.98 2,104.52 114,006.40 7159 (-3,909)*** (101,94)*** Δ ASSETS 65 7.04 5,019.63 10820 8.56 4.18 36.72 7159 (-0,951) (153,64)*** ROA 10.67 11.68 98.73 10571 10.61 11.26 39.15 7124 (-0,040) (7,75)*** OIBD 899.67 165.58 3,197.82 10571 11,101.00 190 4,445.22 7159 (-3,506)*** (15,35)*** SALES 4,822.67 1,127.75 14,207.38 10835 6,367.19 1,430.13 20,034.76 7159 (-6,047)*** (56,15)*** NIBEX 256.65 51.98 1,525.50 10835 287.62 63.7 2,612.11 7159 (-1,002) (25,26)*** EPSEX 3.11 1.17 94.12 10807 3.41 1.17 124.22 7159 (-0,186) (-0,001) COMMEQ 2,350.08 559.15 7,136.89 10837 3,182.20 753 1,153.48 7159 (-6,102)*** (100,43)*** DHTECH 0.148 - - 10854 0.141 - - 7159 - - Pre SFAS 123 (R) Post SFAS 123 (R) Difference Panel B: Descriptive statistics pre and post SFAS 123 (R) 1 2 3 4 5 6 7 8 9 10 1T_COMP 1 2CASH 0.559 1 3ASSETS 0.305 0.391 1 4Δ ASSETS -0.002 -0.004 -0.004 1 5ROA 0.013 0.013 -0.002 -0.004 1 6OIBD 0.398 0.432 0.003 -0.001 0.01 1 7SALES 0.354 0.333 0.476 -0.002 0.008 0.749 1 8NIBEX 0.195 0.176 0.205 -0.001 0.016 0.61 0.475 1 9EPSEX -0.009 -0.009 0.049 00.001 0.082 0.1 0.103 1 10 COMMEQ 0.409 0.387 0.631 -0.001 0.002 0.724 0.685 0.573 0.2 1 Pearson´s correlations in bold are significant at the 0,05 level Panel C : Correlation 66 The descriptive statistics of the variables for total compensation and cash compensation for high-tech firms are presented in Table 2. Panel A of table 2 provides descriptive statistics for the full sample. The firms in the sample are large with a mean of 15,201.20 million USD and median assets of 1,746.97 million USD. In the S&P1500, before and after SFAS (R), for the period between 2000 and 2010, there are about 14.415% high-technology firms, and it is possible to observe that the group of top executives in each company has an average total compensation around 13,727 million USD and earn in cash around 4,156 million USD. Other interesting finding is that, in this period and in this group of companies, and consistent with the overall economic growth, there is an increase in assets and returns on assets around 39.96% and 10.79 %, respectively. Panel B of Table 2 reports the same descriptive statistic partitioned by time period. Consistent with the overall economic growth, it is possible to observe that almost all firm performance measurements increase in the period after SFAS 123 (R). Total compensations maintain an average around 13,727 million USD (value for full period) for the periods before and after SFAS 123 (R). The same cannot be said about the compensation for the short-term as variable cash presents a significant decrease in value after SFAS 123 (R). Panel C of table 2 presents correlations between variables. The performance variables, such as assets, return on assets, operating income, sales and the common stocks, are positively correlated with total compensation and for the short-term they are positively correlated with operating income and sales with cash. 5. Results Using the sample presented above suggests that accounting played a significant role in the high-tech firms’ choice of equity compensation. The panel data model is used because it is the most suitable way of studying a large set of repeated observations and due to the fact that it assesses evolution over time. With panel data it is possible to simultaneously explore several variations over time and between different individuals. The use of such models has increased 67 immensely and, in fact, combining time and cross-sectional data brings many advantages: it is possible to use a larger number of observations and the degree of freedom in estimates increases, thus making statistical inferences more credible. At the same time, the risk of multicollinearity is reduced since the data in companies present different structures. Moreover, this model provides access to further information and the efficiency and stability of the estimators increase, while enabling the introduction of dynamic adjustments (Gujarati, 2004; William, 2003). The results are presented in table 3 and in Panel A the regression with full sample is analyzed in order to test the first hypothesis; the second hypothesis is tested in Panels B and C. As expected, there is a significant and positive correlation between performance variables and total compensation (T_COMP) and between them and the short-term compensation presented by cash. Therefore, it is possible to state that firm performance measured by assets, return on assets, sales and net income have a positive influence on the executive compensation for longand short–term periods. 68 Table 3 - Total compensation and cash compensation estimations using the SUR method Table 3 - Results oompensat 1 Dependent variable Prediction Sign Panel A - full sample Panel B - Pre SFAS 123(R) Panel C - Post SFAS 123(R) Ln_T_COMP Ln_CASH Ln_T_COMP Ln_CASH Ln_T_COMP Ln_CASH Ln_ASSETS + 0.069 0.1202 0.0616 0.108 0,071 0,121 (7,142)*** (18,82)*** (4,413)*** (12,36)*** (4,987)*** (12,42)*** Ln_Δ ASSETS + 0.001 0.0004 0.0001 0.0004 0,001 0,0003 (10,234)*** (4,090)*** (8,867)*** (3,46)*** (5,189)*** (2,37)** ROA + 0.007 0.004 0.0102 0.005 0,006 0,0002 (9,061)*** (6,643)*** (8,163)*** (6,592)*** (5,013)*** (2,93)*** Ln_SALES + 0.142 0.134 0.1238 0.168 0,157 0,090 (20,241)*** (26,83)*** (12,173)*** (11,53)*** (15,22)*** (11,54)*** Ln_NIBEX + 0.115 0.057 0.123 0.088 0,084 0,018 (14,897)*** (10,49)*** (10,81)*** (11,53)*** (7,726)*** (2,258) EPSEX - -0.0003 -0.0003 ~0,004 (-11,75)*** (-6,26)*** (-8,866)*** Ln_COMMEQ + 0.111 0.157 0,068 (12,62)*** (12,42)*** (5,326)*** DHTECH + 0.314 0.0119 0.449 0.053 0.1520 -0,063 (19,67)*** (1,038)*** (19,05)*** (3,345)*** (6,681)*** (-3,654)*** SP + / - 0.057 -0.047 0.0198 -0.033 0.1401 -0,015 (3,908)*** (0-4,444)*** (0,967) (-2,395)** (6,113)*** (-0,912) SM + / - -0.167 -0.016 -0.1531 -0.046 -0.207 -0,054 (-12,30)*** (-1,633) (-7,536)*** (-3,321)*** (.10,44)*** (-3,596)*** Year = 2001 + / - 0.0108 -0.029 0,020 -0,020 (0,443) (-1,618) (0,787) (-1,182) Year = 2002 + / - -0.045 0,052 -0,035 0,061 (-1,865)** (2,946)** (-1,376) (3,478)*** Year = 2003 + / - -0.0504 0,108 -0,047 0,115 (-2,148)** (6,342)*** (-1,884)* (6,714)*** Year = 2004 + / - 0.0214 0,152 0,019 0,150 (0,921) (8,995)*** (0,770) (8,841)*** Year = 2005 + / - 0.0183 0,134 0,012 0,120 (0,776) (7,790)*** (0,470) (6,960)*** Year = 2006 + / - -0.0072 -0,243 (-0314) -14,559)*** Year = 2007 + / - .00454 -0,306 (1,969)* (-18,217)*** Year = 2008 + / - 0.0521 -0,301 0,001 0,007 (2,140)* (-16,968)*** (0,068) (0,439) Year = 2009 + / - 0.0445 -0,208 -0,006 0,019 (1,824)* (-15,962)*** (-0,288) (1,203) Year = 2010 + / - 0.1514 -0,305 0,111 0,010 (6,229)*** (.17,208)*** (5,374)*** (0,686) Adj R2 0.565 0.546 0.538 0.603 0.611 0.443 N 15109 15265 8103 8223 5433 5435 T-Statistics are reported in parenthesis below the coefficient and White´s corrected for heteroskedasticity *, **, *** indicate significance at the 0,10, 0,05 and 0,01 level, respectly 69 A system equation was used for the dependent variables natural logarithm of total compensation (Ln_T_COMP) and natural logarithm of cash compensation (Ln_CASH), which are explained by performance measurement for longand short-term periods. The system equation presented was estimated using the Seemingly Unrelated Regression (SUR) method. The SUR is a generalization of a linear regression model that consists of several regression equations, each having its own dependent variable and potentially different sets of exogenous explanatory variables. The main motivations for using the SUR are: improving estimation efficiency by combining information on different equations; and imposing and testing restrictions that involve parameters in different equations. The model can be estimated for each equation considering the interdependence of distribution. The SUR model can be further generalized into the multiple regressions, where the variables on the right-hand side can also function as endogenous variables. The multiple-equation model is a system of equations where the assumptions made for the singleequation model apply to each equation. The regression coefficient, year, does not vary over time because the estimation was conducted using dummy variables for year, and assuming that the company’s heterogeneity is captured in the constant part (William, 2003). The results in table 3 reflect the estimation of equations (1) and (2). Panel A is consistent with hypothesis 1, the estimated coefficient for total compensation for long–term periods and cash compensation for short-term periods. As it is possible to observe the regressions are globally significant, with a 5% significance level. The following table presents the results of the estimation for the studied data. The sample includes 15109 observations for full time, the period before SFAS 123 (R) represented by 8103 observations and period after SFAS 123 (R) represented by 5433 observations. The result of the SUR model is depicted for total compensation and cash compensation in the period between 2000 and 2010. Adjusted R2 is 0.565, which means that the dependent variables total compensation is explained by this set of regressors present in the model. For the period after SFAS 123 (R), the adjusted R2 is 0.611, meaning that the model can be explained by the group of variables and is higher than that the adjustment or the model is better for this sample. These indicate that the variables addressed here play a significant role in 70 explaining executive compensation for shortand long-term periods, as stated by Chi-Square test (Probability=0). It is possible to note that in full period in high-tech firms CEO compensation is higher than in other firms of the S&P 1500 at about 31.4%, but in the period before SFAS 123(R) it was about 44.9% higher and dropped to 15.2% in the period after SFAS 123(R). CEO compensation in high-technology firms is higher than in the other firms but with a smaller difference than previously. It is important to highlight that the implementation of the SFAS 123 (R) has an influence on awards in the long-term, but not for short-term periods. Stock options represent awards for the long–term, and the negative influence on CEO compensation in high-tech firms after the SFAS 123 (R) is confirmed. However, for long-term S&P 500, for the biggest S&P firms, CEO compensations are higher than in S&P small firms. When the annual effects are analyzed, it is possible to find a decrease in CEO compensation for the long-term in 2002 and 2003 over 2000 and an increase in the period between 2007 and 2010. In table 3, the coefficient signs are similar in both specifications. However, the magnitudes of the coefficients are sensitive to the specification. As expected, earnings per share are negative and significantly related to total compensation for the long-term. This indicates that there are no explicit contractual arrangements linking compensations and earnings per share. The performance ratio of firms measured by return has a negative influence on CEO Compensation (Core, Guay, & Verrecchia, 2003a; Young & Jing, 2011). According to Aboody, Barth, and Kasznik (2004), there is a significant negative relation between share price and the SFAS 123 expense when it is relevant to investors and well measured (Aboody et al., 2004). A positive and statistically significant relationship was found between sales, asset growth and return on assets, and for adding the same level of total CEO compensation and cash compensation Gabaix and Landier (2008) also empirically test the relation between the level of pay and firm size. Ln (assets), a variable proxy for firm size is positively related to pay with a coefficient total compensation and cash compensation in the regression.When the adjustment is performed for the long-term compensation, it is possible to understand that when firm sizes are compared using the current S&P index membership, S&P500 firms have an increase around 5.7% and for the S&P 71 Small caps 600 there is a decrease around 16.7%, comparatively to the S&P mid firms. In terms of cash compensation, the S&P 500 firms are 4.7% below mid cap, and the S&P small firms are 1.6% below, comparatively to the same group of S&P mid firms. Another finding is that the influence on CEO pay for the short-term between the year 2006 and 2008 does not have the same meaning in long-term compensations. As expected, there is an increase around 5% for each year between 2007 and 2010 as a result of the introduction of the SFAS 123(R). There is a positive relation between CEO compensation and firm performance in high-tech firms after the implementation of the SFAS 123 (R), but with less intensity than before. For the S&P 500 firms, the implementation of the SFAS 123 (R) is profitable to CEOs because it increases their compensations, while for high-tech CEOs it increases the value, although not as strongly, and it is possible to verify the normalization for all S&P 500 firms. Some authors, such as Hall and Murphy (2002), advocate that this adjustment of stock options is necessary to restrict options and to consequently increase CEO compensation. That suggests that firms find it difficult to downsize the executive pay packages and shift toward restricted options to provide more incentives for long-term CEO compensation (Carter, Lynch, & Tuna, 2007). Restricted stock awards are profitable for executives because the income tax consequences can be more favorable to employees than stock options. The special case of the USA and the consequences of a restricted stock mean that in some cases the award can be structured to allow for the deferral of all tax until the time of stock sale, and for all appreciation to be taxed at capital gain rates even if the stock is appreciated prior to vesting. In contrast, stock options can result in ordinary income to the recipient the stock has appreciated prior to vesting, with only the post-exercise appreciation being deferred to the time of sale at capital gain rates. Furthermore, the preferred stock usually carries no voting rights but may carry a dividend and may have priority over common stock in the payment of dividends and upon liquidation. The preferred share investor is entitled to a preset rate of dividend that must be paid out of earnings before any dividends are distributed to common shareholders. 72 6. Conclusion and future research This paper will contribute to a better understanding of the relationship between compensation and performance in high-technology firms, and of the behavior caused by the new role of expensing stock options with the SFAS 123 (R).The main purpose of this study was to examine whether the total compensation paid to CEOs in high-technology firms in the S&P 1500 is related in corporate finance and how it is influenced by the introduction of the SFAS 123 (R). The results presented are consistent with those achieved by Carter et al. (2007), who stated that the favorable accounting treatment for stock options possibly lead to overall higher CEO compensations. There is no evidence of a decrease in total compensation combined with the positive association between financial reporting. They find that after controlling standard economic determinants of compensation, expensing options in firms decrease compensation from options and increase compensation from restricted stock. These results suggest that accounting plays an important role in executives plan design (Carter et al., 2007). There was an increase in CEO compensation after the introduction of the SFAS 123 (R). Although the change in the plan design was not analyzed, a new accommodation of CEO compensation was found as a result of the new rules of the SFAS 123 (R). The influence of firm performance on the CEO compensation is positive and consistent in this group of hightechnology firms in the period between 2000 and 2010. As concluded by Graham et al. (2012), there are differences in corporate culture and in the managers’ latent traits, which are difficult to observe or measure. These latent traits could be an innate ability, personality, risk aversion or, in this case, propensity to innovation, managing uncertain times in order to boost (enhance) returns to the firm and reaction to stakeholders. The CEOs of high-tech firms had to maximize returns, facing a big competition with new technological solutions, thereby warranting a higher compensation than others. However, this work is not without limitations. This study focuses only on high-technology firms in the S&P 1500 in the period between 2000 and 2010, and the results of this study may not be 73 generalized to include other sectors due the specificity of high-tech firms. Another limitation is the definition of high-technology used in this study that can be extended, as performed by Shim, Lee, and Joo (2009), to include other important item measurements, such as value of R&D expenditures, number of patents by firm and citation of patents (Gomez-Mejia et al., 2000; Shim et al., 2009). The level of R&D expenditures and new product introductions are viewed as proxies for innovation, risk-taking and long-term decision-making, which are crucial to characterize high-technology firms. Furthermore, innovation constitutes an indispensable component of corporate strategies. In the future, it will be important to analyze other developments, such as the effect of managerial attributes for the shortand long-term in executive compensation (Graham et al., 2012). Furthermore, it will also be important to broaden the period of analysis in order to investigate the effect of the financial crisis in the USA, which started in 2007. 7. References Aboody, D., M. E. Barth, and R. Kasznik. 2004. SFAS No. 123 Stock‐Based Compensation Expense and Equity Market Values. The Accounting Review 79 (2):251-275. Akpotaire, B. U. 2011. Agency Cost Problems in Executive Compensation: An Evaluation of Dividend Equivalent Rights on Restricted Stocks. Social Science Research Network. Bebchuk, L. A., M. Cremers, and U. Peyer. 2011. The CEO pay slice. Journal of Financial Economics In Press, Accepted Manuscript. Bebchuk, L. A., and J. M. Fried. 2003. Executive Compensation as an Agency Problem. National Bureau of Economic Research Working Paper Series No. 9813 (published as Bebchuk, Lucian Arye and Jesse M. Fried. "Executive Compensation As An Agency Problem," Journal of Economic Perspectives, 2003, v17(3,Summer), 71-92.). 80 testable predictions. The first prediction is about how the CEO is encouraged to promote the goals of maximizing the shareholder’s wealth to increase firm performance and to improve innovation policy. The second prediction is about how R&D as measure of innovation is related to CEO compensation and performance. The third prediction is about how the CEO manages risk taking to promote innovation. Surprisingly, unlike the other strands of study on the economics of innovation, this field of research has not benefited so far from a systematic discussion and review of its major contributions. This paper aims to fill this gap. This work is organized as follows: Section 2 contains a revision of the main theories in the literature, as well as an analysis on executive compensation in order to address agency problems. Furthermore, this section provides an analysis which examines corporate innovation and risk taking in terms of performance and innovation. Section 3 explains the research hypotheses and section 4 presents the methodology, sample and data collection to estimate regression, as well as the results of the econometric models to assess the influence that firm performance and innovation has on executive compensation. Lastly, the main conclusions are discussed, as well as some limitations and new perspectives for future research. 2. Literature review and hypothesis development 2.1 Executive compensations to address agency problems The general acceptance of the agency theory and the parallel research on executive compensation began in the early 1980s. It was the evolution of the modern corporation with ownership separation and control that undermined the agency theory. Early studies in this area focused on documenting the relation between CEO pay and firm performance. The discussion of executive compensation must proceed with the fundamental agency problem afflicting management decision-making as background. The shareholders’ primacy view of the firm, built on the principal–agent paradigm, states that shareholders (the principals) engage managers (the agents) to run the firm on the shareholders’ 81 behalf (M. Jensen & Meckling, 1976). According to Jensen and Murphy (1990), there are two approaches to agency problems. The authors state that there is an optimal contracting approach, which is when boards use design compensation schemes to maximize shareholder value with efficient incentives (Jensen and Murphy, 1990). To connect the agency problem and the executive compensation, the authors use the managerial power approach when this connection is seen as an integral part of the agency problems. It is important to remember that the principalagent problems treat the difficulties that arise under conditions where information is incomplete and asymmetric whenever a principal hires an agent (Eisenhardt (1989); Lucian Arye Bebchuk and Fried (2003); K. J. Murphy (1999)). The first is the desire or goal of the principal and agent conflict and it is difficult or expensive for the principal to verify what the agent is actually doing. The problem is that the principal is unable to check if the agent has behaved correctly. Secondly, there is the problem of risk sharing facing the different attitudes toward risk, because the principal and the agent have different actions according to different risk preferences (Eisenhardt, 1989). Brian J. Hall and Liebman (1998) argue that the solution to the agency problem is aligning the incentives of executives with the interests of shareholders by granting (or selling) stock and stock options to the CEOs. The CEOs have the correct incentives on every margin, including effort, perquisites and project choice, and support that the optimal contract is a one-to-one correspondence between firm value and CEO pay (Brian J. Hall & Liebman, 1998). It is reasonable for small firms but it is not appropriate for large firms because optimal contracts represent a trade-off between incentives and risk-sharing (Eisenhardt, 1989). In their work, (Brian J. Hall & Liebman, 1998) conclude that the relationship between pay and performance is much larger than has previously been recognized, and that this includes both gains and losses in CEO wealth. The salary and bonus vary so little because corporate board members are often reluctant to reduce CEO pay, even in response to poor performance and that may attract unwanted media attention. Using salary and bonuses to reward and penalize CEOs may only be possible to create high-powered incentives that align CEO pay with shareholder objectives (Brian J. Hall and Liebman (1998)). A large part of the executive pay literature argues that compensation and managerial interests should be aligned with shareholder interests in order to 82 solve agency problems (see, for example, the surveys by ((Core et al., 2003) and by (K. J. Murphy, 1999)). Managers in high-technology firms have different goals, such as managing intangible assets, continuous improvement, and software and product development. As a result, they must continuously innovate and sustain growth in an increasingly competitive and global market (Shim et al., 2009). Studies that link corporate governance to innovation form a corpus of research that is difficult to disentangle for two interrelated reasons. Firstly, a well-received theory of the innovative enterprise is still missing, which implies the absence of a single coherent conceptual framework for understanding the phenomenon of corporate technological innovation at firm level. Secondly, because such a theory is lacking, contributions to this issue have remained separate and relate to various and different aspects of corporate governance. As previously discussed, existing theories to provide predictions on the outlined considerations related to CEO compensation, allowing for the hypotheses formulated, support and enhance this evidence: H1: CEO compensation is determinant and negatively correlated with firm performance in hightechnology firms. 2.2 Corporate innovation This section briefly discusses how the most influential corporate innovation deals with technological innovation, in order to outline the theoretical ground on which the debate on this issue develops. The traditional economics of innovation treats firms as if they were similar and considers innovation as a direct consequence of profit-maximizing behavior (Nelson, 1991). Conversely, the literature on corporate governance and innovation recognizes that firms differ in their internal governments’ structure and organization, and recognizes that differences are very important for a firm’s economic performance. By definition ‘innovation’ means technologies or practices that are new to a given society and they are not necessarily new in absolute terms. These technologies or practices are being disseminated in that economy or society. Innovation has always played a decisive role in the economic and social development of countries: it is responsible for economic growth, for improving productivity and welfare, and it is the foundation 83 of competitiveness. As Belloc (2012) reports, technological innovation is the development of an original product or process, through the utilization of productive resources and the embodiment, combination or synthesis of knowledge in a new object or method. The author argues that innovation is generated through a collective and cumulative process of learning as R&D programs, which requires the commitment of resources for a prolonged period of time. He defines that technological innovation involves three elements: specificity of the investments, uncertainty about the result and impossibility of anticipating future returns. In most firms, if somewhere deep in the corporate hierarchy an innovator has a very daring and promising innovation idea, the traditional advice is to first obtain top management commitment in order to overcome the resistance that is to be expected later on during the innovation project. This is this the traditional approach; however, there are several issues associated: often there is only one chance to pitch an idea at the top management. If the idea is not convincing enough, not only are the executives going to say "no", they are probably going to remain negative indefinitely, due to anchoring bias and confirmation bias. Another issue associated is the political power played, when people are forced to follow extensive corporate procedures, distortion by certain stakeholders with differing procedures and pressure for short-term results. A major difficulty in observing the effect of innovation on growth is that a firm may require a long period of time to convert economically valuable knowledge increases into economic performance (Coad, 2007). Leveraging innovation is particularly important today, in what is the most severe global economic crisis. History has shown that times of crisis are also times of innovation, when institutional, mental, and other obstacles are more easily removed. The time is thus ripe for mobilizing creativity and entrepreneurship to meet the challenges ahead. Government and other leaders play a key role in promoting innovation in public goods and in finding ways to conduct business more effectively. Most importantly, the government should help provide the right environment for innovation. Dealing with innovation variables in econometric estimation can be problematic. The first problem involves measuring innovation. Generally, two indicators are used to measure innovative activity: an input measurement, R&D spending (Lanjouw & Schankerman, 2004), and an output measurement, the number of patents and brands. According to Belloc (2013), both of 84 these indicators have some disadvantages. R&D spending is an imperfect proxy for innovative activity, because not all innovations are generated within formal R&D programs and the net book value of brands, patents and trademarks does not capture all innovations. In our estimation, the aggregated number of net book value of brands, patents and trademarks awarded by the firms and R&D spending in high-tech are used as indexes of innovative performance. The second problem is the time around innovation programs. While R&D spending should be immediately affected by investment decisions, innovation programs take time to get to a patent, or the brand is awarded within one year after an investment decision and, on average, the duration of innovation projects is between five and 10 years. The finance for innovation usually comes from internal sources as cash flow, but when substantial investment is required, external investment may be necessary. Reductions in R&D spending merits considerable attention because R&D spending is a primary input into innovation (Heeley, Matusik, & Jain, 2007) and thus a firm’s competitive advantage (Makri et al., 2006). As discussed in previous theories, providing predictions on the outlined considerations related to corporate innovation makes it possible to select a second hypothesis. H2: R&D is determinant and positively correlated with firm performance and CEO compensation in high-technology firms. 2.3 Risk taking to performance and innovation Agency research recognizes that the interests of CEOs and shareholders diverge with regard to firm risk, with CEOs preferring less firm risk and shareholders preferring more firm risk (Tosi & Gomez-Mejia, 1989). These differences occur because CEOs are less diversified than shareholders and both the CEO’s pay and employment are tied to the firm. Therefore, CEOs prefer short-term outcomes that have inherently less risk than long-term outcomes. Furthermore, Hill and Snell (1988) argue that institutional investors are risk-averse and so when they are major stockholders they also wield pressure on management to obtain good short-term performance to the detriment of long-term projects and innovation. Only one party has both the right to make residual management decisions and the right to claim the residual profits of the production, and 85 the remaining parties lose the ability to make opportunistic threats. The firm as a centralized structure of governance is only a second-best solution to the extent that under a one-party-owner regime, the non-owner firm members lose the ability to hold up, as well as the incentive to invest. This deeply affects innovation activities because innovation is a process of collective and specific investment (Belloc, 2012). The innovation literature reports high failure rates for innovation, ranging from 50% to 90% as reported by Spieth and Hedenreich (2013). They show that individuals will be less likely to adopt new technological products if they perceive a significant risk associated with such exploration. Offering warranties may also be effective in reducing perceived risk associated with innovations independent from the possibility of a new product trial or demonstration. Moreover, the problem is innovations that fail cannot generate future revenues and, therefore, they can hinder the competitiveness of companies in the long run. Although innovation, and thus R&D expenses, is important for creating firm value and a sustainable competitive advantage, research shows that CEOs will opportunistically target R&D spending because R&D projects are associated with information asymmetry and risk (Aboody & Lev, 2000). Innovation is an inherent risk and reductions in R&D can quickly increase a firm’s short-term market performance at the expense of innovation and long-term returns. R&D investments also inherently involve information asymmetry between principals and agents, even in high technology firms. Aboody and Lev (2000) and M. Jensen and Meckling (1976) suggest that CEOs should possess the highest knowledge about the firm, which implies that CEOs should have a better understanding than owners about the optimal level of R&D spending, thus allowing for more informed, opportunistic action by the CEO. This information asymmetry allows CEOs to make opportunistic reductions in R&D spending, which Joseph P. O'Connor, Jr., Joseph E. Coombs, and Gilley (2006) believe is problematic because decisions that benefit short-term performance often do not lead to long-term benefits for shareholders. These consequences make reductions in R&D spending relevant because shareholders generally find R&D spending desirable due to their interests in greater risk-taking than CEOs and their interests in long-term firm performance. 86 Another important argument is the US generally-accepted accounting practice (GAAP) that requires the immediate expensing of R&D spending, and thus a reduction in R&D spending increases short-term performance through increases in current market and accounting performance. Sometimes R&D spending has been found to be negatively associated with CEO pay. In fact, research on R&D spending shows that when a CEO approaches retirement or when the firm faces a potential reduction in firm performance CEOs tend to reduce R&D spending and suggest that compensation committees respond to, and effectively mitigate, potential opportunistic reductions in R&D spending (Cheng, 2004). CEOs approaching retirement no longer financially gain from the long-term benefits associated with current R&D spending; instead, they financially gain when R&D is reduced due to the immediate expensing of R&D based on generally accepted accounting principles (GAAP) requirements. Moreover, if firm performance decreases, the CEO could face termination, which again reduces the CEO’s opportunity to gain from current R&D spending. Given the strategic importance of R&D spending on long-term firm performance, Cheng (2004) shows that compensation committees will adjust CEO compensation contracts to reduce opportunistic manipulations of R&D spending when the firm encounters the horizon problem or the myopia problem, conditions under which R&D manipulation is likely to occur. However, the arguments provided by the agency theory suggest that CEOs may be encouraged to behave opportunistically towards R&D spending outside the specific conditions presented by the horizon and myopia problems, which may be when the CEO pay deviates from the labor market rate. As previously discussed, existing theories provide predictions on the outlined considerations related to risk taking, allowing for the hypotheses presented below. H3: Risk taking positively influences firm performance and innovation in high-technology firms. In summary, it was found that firms are subjected to the agency problem in which the CEO may not work in favor of the shareholders to maximize their wealth by improving firm performance and innovation. The risk taking to make decisions and theirs goals are important factors. Furthermore, the decisions related to CEO compensation are based on the firms’ accounting performance and innovation in high-technologies firms. 87 3. Sample, Data and Method of Analysis The ExecuComp and the DataStream databases were used to find finance performance and innovation variables and to create a sample of firms between 2000 and 2010. The ExecuComp database covers current, historic and total compensation data such as salary, bonus, stock options and restricted stock grants, as well as managerial stock and option holdings on top five executives of more than 2600 firms in the Standard & Poor’s Index in the United States. The total sample collected was 5,500 observations for the five hundred companies of the S&P 500 for the eleven years. The DataStream provides access to a large variety of financial and economic data, such as financial statements, accounting ratios, price information of listed companies in the world; exchange rates, interest rates and thousands of economic time series from various countries in the world. The match between these two databases was made by comparing the same variables for each firm and year in order to confirm the values to the same firm. If the firm short name, number of employees and the asset value was equal in the two database, it is guaranteed that the firm is the same. To test this hypothesis, the following specification is run on the balanced panel of hightechnology firms. High-Technology firms are the firms that operate in an industry with a fourdigit SIC code of 3570, 3571, 3572, 3576, 3577, 3661, 3674, 4812, 4813, 5045, 5961, 7370, 7371, 7372, or 7373, DHTECH variables using the Fama and French classification of 48 industry groups, instead of the four-digit Standard Industrial Classification (SIC) code (Fama & French, 1997). A new DHTEC economy dummy is also included to control for the possibility that opportunistic timing was more prevalent among new economy (hi-tech) firms (definition of the new economy follows Murphy (2003)). Another classification about high-technology firms is the OECD's classification (stable since 1973), of R&D intensity, which is presented next. The definition of R&D intensity is the ratio of expenditures by a firm on research and development to the firm's sales. This classification defines high-technology firms when the value of R&D intensity is greater than 7%. The absolute levels of R&D expenditures indicate the level of effort dedicated to producing future products and process improvements while maintaining current market share and increasing operating 88 efficiency. By extension, such expenditures may reflect the firms' perceptions on market demand for new and improved technology. However, R&D intensity is the most frequently used measurement "to gauge the relative importance of R&D across industries and among firms in the same industry”. William N. Leonard, Economics professor since 1971, found that research intensity is measured usually by ratios of scientific personnel to total employment or by R&D expenditures/sales, and gains in variables such as productivity, profits, sales, assets, and other variables (Leonard, 1971). 3.1 Dependent variables The dependent variables used for the tests are CEO compensation as the total year pay for the top executives, return on asset to measure performance (G. B. Murphy, Trailer, & Hill, 1996) and the R&D expenses to measure innovation (Lanjouw & Schankerman, 2004) for each firm in the S&P 500 for the period between 2000 and 2010, as defined in table 4. Total compensation the sum of the compensations of top executives includes: salary, bonus, non-equity incentive plan compensation, grant-date fair value of option awards, grant-date fair value of stock awards, deferred compensation earnings reported as compensation, and other compensations. The return on assets in calculated for each firm and per year as the net income before extraordinary items and discontinued operations divided by total assets. Research and development expenses represent all direct and indirect costs related to the creation and development of new processes, techniques, applications and products with commercial possibilities. 3.2 Independent variables The key independent variables are firm level factors reports by each firm, presented in table 4 in the appendix. A set of variables explain corporate finance as common equity that represents common shareholders' investment in a company, the number of employees give a perception of company sizes and the A return index shows a theoretical growth in value of a shareholding over a specified period, assuming that dividends are re-invested to purchase additional units of an equity or unit trust at the closing price applicable on the ex-dividend date. Market capitalization represents the total market value of the company based on year end price and number of shares 89 converted to U.S. dollars using the year end exchange rate, and for companies with more than one type of common/ordinary share, market capitalization represents the total market value of the company. Another set of variables to explain corporate innovation are patents/sales as total amount of patents owned in the current year for each firm divided by net sales or revenue in US dollars. The R&S/sales five-year average is the arithmetic average of the last five years of research and development by sales. Brands and patent net represent the net book value of brands, patents and trademarks. Cash-flow represents the net cash receipts and disbursements resulting from the operations of the company, fundamental to make investments. The tool to measure risky corporate policies is R&D expenses over total assets for each year, which is commonly employed (Bargeron, Lehn, and Zutter (2010) and Coles, Daniel, and Naveen (2006)). 3.3 Method of analysis The archived sample has led to do the analysis of the panel partly because panel data provide such a rich environment for the development of estimation techniques and theoretical results. In more practical terms, however, researchers have been able to use time-series cross-sectional data to examine issues that could not be studied in either cross-sectional or time-series settings alone. The fundamental advantage of a panel data set over a cross section is that it will allow the researcher great flexibility in modeling differences in behavior across individuals. The panel data model is used because it is the most suitable way of studying a large set of repeated observations, and due to the fact that it assesses evolution over time, consisting of cross-section observations from different points in time. With panel data it is possible to simultaneously explore several variations over time and between different individuals. The use of such models has increased immensely and, in fact, combining time and cross-sectional data brings many advantages: it is possible to use a larger number of observations and the degree of freedom in estimates increases, thus making statistical inferences more credible. At the same time, the risk of multicollinearity is reduced since the data in companies present different structures. Moreover, this model provides access to further information and the efficiency and stability of the estimators increase, while enabling the introduction of dynamic adjustments (Gujarati (2004) and William (2003)). This study starts by using the Generalized Linear Square (GLS) to estimate the model, and then the 96 Table 5 (appendix) presents Pearson´s correlations (all coefficients in bold are statistically significant at 0.05 level). Employees and net Income before extraordinary items and discontinued operation are statistically significant and the positive correlation with CEO compensation, return of assets and R&D variables is high, and statistically significant but with relatively low correlations with dummy variables of high-technologies. Sales growth, brand, patent net have statistically significant low correlations with CEO compensation, return on assets and R&D. 5. Conclusion and future research The main purpose of this study was to examine whether the total compensation paid to CEOs in high-technology firms in the S&P 500 is related to corporate innovation and performance. This work aims at contributing to explain the influence that innovation technologies and performance has on CEO compensation in these companies. In conclusion, according to the results that were obtained there is empirical evidence to state that in high-technology firms in the S&P500, during the period between 2000 and 2010, innovation determined total CEO compensation. Results suggest that in high-tech firms innovation determines CEO compensation and CEOs tend to use more sophisticated performance measurements. The findings indicate that CEO compensation in high-tech firms chooses innovation over performance, and that there is a strong and positive relation between CEO compensation and innovation. This econometric study provides a better understanding of the CEO risk taking and the relationship between CEO compensation, innovation and performance in high-technology firms. In the future, to enhance this study it is necessary to analyze the shortand long-term period of CEO compensation to understand their goal throughout the duration of R&D projects, brands and patent processes. This study only focuses on high-technology firms in the S&P 500 in the period between 2000 and 2010, but an extended set of observations with different sectors or industries was important to confirm our conclusion. Another aspect would be understanding the influence that the 2007 crisis had in this period. Organizing such a body of literature was difficult because studies on this issue form a heterogeneous puzzle that covers interrelated aspects of corporate organization. This study started 97 by briefly discussing how different approaches to the analysis of the firm deal with technological innovation, in order to outline the theoretical ground on which the various studies linking innovation to corporate governance are developed. This paper describes the main corporate risk taking through which a system of corporate governance shapes innovation activities, classifying them in the three dimensions of the agency problem, corporate innovation and risk taking. Finally, the literature on high-technology firms has been examined, and the relationship between the CEO behavior of corporate governance and aggregate innovation activity of corporations has been discussed. As reported by Belloc (2012), the more recent and rather heterogeneous literature recognizes the importance of corporate governance for a firm’s performance. The author states that differences in the various dimensions of a corporation’s governance are important for its innovation activity. Contrasting with this opinion, it was found that in fact innovation in hightechnology firms appears as a result of technological determinism in a context of profitmaximizing firms, and it can emerge because individuals decide to invest in innovative projects, where these investment decisions are shaped by the corporate governance system and by CEO decision. 6. References Aboody, D., & Lev, B. (2000). Information Asymmetry, R&D, and Insider Gains. The Journal of Finance, 55(6), 2747-2766. doi: 10.1111/0022-1082.00305 Balkin, D. B., Markman, G. D., & Gomez-Mejia, L. R. (2000). Is CEO Pay in High-Technology Firms Related to Innovation? The Academy of Management Journal, Vol. 43 (No. 6), 1118-1129. Bargeron, L. L., Lehn, K. M., & Zutter, C. J. (2010). Sarbanes-Oxley and corporate risk-taking. Journal of Accounting and Economics, 49(1–2), 34-52. doi: http://dx.doi.org/10.1016/j.jacceco.2009.05.001 98 Bebchuk, L. A., & Fried, J. M. (2003). Executive Compensation as an Agency Problem. Journal of Economic Perspectives, 17((3,Summer), 71-92.). Belloc, F. (2012). Corporate Governance and Innovation: A Survey. Journal of Economic Surveys, 26(5), 835-864. doi: 10.1111/j.1467-6419.2011.00681.x Belloc, F. (2013). Law, finance and innovation: the dark side of shareholder protection. Cambridge Journal of Economics, 37(4), 863-888. doi: 10.1093/cje/bes068 Carpenter, R. E., & Petersen, B. C. (2002). Capital Market Imperfections, high-tech investiment, and new equity financing. The Economic Journal, 112(477), F54-F72. doi: 10.1111/14680297.00683 Cheng, S. (2004). R&D Expenditures and CEO Compensation. The Accounting Review, 79(2), 305-328. doi: 10.2308/accr.2004.79.2.305 Coad, A. (2007). Firm Growth: A Survey. Papers on Economics and Evolution. Coles, J. L., Daniel, N. D., & Naveen, L. (2006). Managerial incentives and risk-taking. Journal of Financial Economics, 79(2), 431-468. doi: http://dx.doi.org/10.1016/j.jfineco.2004.09.004 Core, J. E., Guay, W. R., & Larcker, D. F. (2003). Executive Equity Compensation and Incentives: A Survey. Economic Policy Review, Vol. 9, No. 1, April 2003. Eisenhardt, K. M. (1989). Agency Theory: An Assessment and Review. The Academy of Management Review, 14(1), 57-74. Fama, E. F., & French, K. R. (1997). Industry costs of equity. Journal of Financial Economics, 43(2), 153-193. doi: 10.1016/s0304-405x(96)00896-3 Fong, E. A. (2010). Relative CEO Underpayment and CEO Behaviour Towards R&D Spending. Journal of Management Studies, 47(6), 1095-1122. doi: 10.1111/j.14676486.2009.00861.x 99 Gabaix, X., & Landier, A. (2008). Why has CEO Pay Increased So Much? The Quarterly Journal of Economics, 123(1), 49-100. doi: 10.1162/qjec.2008.123.1.49 Gomez-Mejia, L. R., Gideon, D. M., & Balkin, D. B. (2000). Is CEO Pay in High-Technology Firms Related to Innovation? The Academy of Management Journal, 43(6), 1118-1129. Gujarati, D. N. (2004). Basic Econometrics (Vol. Fourth Edition). Hall, B. J., & Liebman, J. B. (1998). Are CEOS Really Paid Like Bureaucrats?*. Quarterly Journal of Economics, 113(3), 653-691. doi: doi:10.1162/003355398555702 Hansen, L., Peter. (1982). Large sample properties of generalized method of moments estimators. Econometrica, 50, 1029--1054. Heeley, M. B., Matusik, S. F., & Jain, N. (2007). Innovation, Appropriability, And The Underpricing Of Initial Public Offerings. Academy of Management Journal, 50(1), 209225. doi: 10.5465/amj.2007.24162388 Hill, C. W. L., & Snell, S. A. (1988). External control, corporate strategy, and firm performance in research-intensive industries. Strategic Management Journal, 9(6), 577-590. doi: 10.1002/smj.4250090605 Jensen, M., & Meckling, W. (1976). Theory of the firm: managerial behaviour, agency costs, and ownership structure. Journal of Financial Economics(3), 308-360. John, K., Litov, L., & Yeung, B. (2008). Corporate Governance and Risk-Taking. The Journal of Finance, 63(4), 1679-1728. doi: 10.1111/j.1540-6261.2008.01372.x Joseph P. O'Connor, Jr., R. L. P., Joseph E. Coombs, & Gilley, K. M. (2006). Do CEO Stock Options Prevent or Promote Fraudulent Financial Reporting? ACAD MANAGE J, 49(3), 483-500. 100 Lanjouw, J. O., & Schankerman, M. (2004). Patent Quality and Research Productivity: Measuring Innovation with Multiple Indicators*. The Economic Journal, 114(495), 441465. doi: 10.1111/j.1468-0297.2004.00216.x Leonard, W. N. (1971). Research and development in industrial growth. Journal of political economy., 79(4). Makri, M., Lane, P. J., & Gomez-Mejia, L. R. (2006). CEO incentives, innovation, and performance in technology-intensive firms: a reconciliation of outcome and behaviorbased incentive schemes. Strategic Management Journal, 27(11), 1057-1080. doi: 10.1002/smj.560 Murphy, G. B., Trailer, J. W., & Hill, R. C. (1996). Measuring performance in entrepreneurship research. Journal of Business Research, 36(1), 15-23. doi: http://dx.doi.org/10.1016/01482963(95)00159-X Murphy, K. J. (1999). Chapter 38 Executive compensation. In C. A. Orley & C. David (Eds.), Handbook of Labor Economics (Vol. Volume 3, Part 2, pp. 2485-2563): Elsevier. Nelson, R. R. (1991). Why do firms differ, and how does it matter? Strategic Management Journal, 12(S2), 61-74. doi: 10.1002/smj.4250121006 Percival, J., & Mcgrath, C. (2013). Short-term and long-term returns to innovation from the aplication of technology and trading practices. International Journal of Innovation Management, 17(05), 1350023. doi: doi:10.1142/S1363919613500230 Shim, D., Lee, J., & Joo, n. K. (2009). CEO Compensation and US High-tech and Low-tech Firms’ Corporate Performance. Contemporary Management Research, Vol. 5, 93-106. Solow, R. M. (1988). Growth Theory and After. The American Economic Review, 78(3), 307317. 101 Spieth, P., & Hedenreich, S. (2013). Why innovations fail - The case of passive and active innovation resistance. International Journal of Innovation Management, 17(05), 1350021. doi: doi:10.1142/S1363919613500217 Tosi, H. J., & Gomez-Mejia, L. R. (1989). The Decoupling of CEO Pay and Performance: An Agency Theory Perspective. Administrative Science Quarterly, 34(2), 169-189. William, G. (2003). Econometric analysis. New Jersey: Prentice-Hall. Young, S., & Jing, Y. (2011). Stock Repurchases and Executive Compensation Contract Design: The Role of Earnings per Share Performance Conditions. Accounting Review, 86(2), 703733. doi: 10.2308/accr.00000024 102 Table 4 – Dependent and independent executive compensation Name Definition Units Assets The Total Assets as reported by the company. Millions Brand and patent net Represents the net book value of brands, patents and trademarks. This item is available in the annual original and annual restated time series for detailed companies, full update. Millions Cash flow Represent the net cash receipts and disbursements resulting from the operations of the company. It is the sum of Funds from Operations, Funds From/Used for Other Operating Activities and Extraordinary Items Millions CEO_Comp Total Compensation (Salary + Bonus + Other Annual + Restricted Stock Grants + LTIP Payouts + All Other + Value of Option Grants) Thousands Common equity Represents common shareholders' investment in a company. Millions Employees Number of employees as reported by companies Millions Market Capitalization Represents the total market value of the company based on year end price and number of shares outstanding converted to U.S. dollars using the year end exchange rate. For companies with more than one type of common/ordinary share, market capitalization represents the total market value of the company. Percentage NIBEX The Net Income Before Extraordinary Items and Discontinued Operations. Millions OIBD The Operating Income Before Depreciation as reported by the company. Millions Patents/Sales Total U.S. patents owned in the current year divided by net sales or revenue in US dollars Percentage R&D Research and development expense, represents all direct and indirect costs related to the creation and development of new processes, techniques, applications and products with commercial possibilities. Millions R&D/Assets R&D expenditures over total assets as a proxi for risk-taking in long-term corporate investments Percentage R&D/Sales Research and Development Expense / Net Sales or Revenues * 100 Percentage R&D/Sales - 5 yr avg Arithmetic average of the last five years of Research and Development SALES Percentage Return Index A return index (RI) is available for individual equities and unit trusts. This shows a theoretical growth in value of a shareholding over a specified period, assuming that dividends are re-invested to purchase additional units of an equity or unit trust at the closing price applicable on the ex-dividend date. Percentage ROA The Net Income Before Extraordinary Items and Discontinued Operations divided by Total Assets. This quotient is then multiplied by 100. Percentage SALES The Net Annual Sales as reported by the company. Millions SALES GROWTH Annual growth of Sales as reported by the company. Percentage 103 Table 5 - Correlations between variables Table 5 - Corre lation s betwe en varia bles 1 Correlation t-Statistic Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (14) (15) (16) (17) (1) CEO_Comp 1 (2) ROA 0.573 1 (11.717) ----- (3) R&D 0.656 0.568 1 (14.568) (11.577) ----- (4) Common Equity 0.672 0.430 0.753 1 (15.219) (7.980) (19.193) ----- (5) Cash-Flow 0.646 0.456 0.757 0.939 1 (14.177) (8.581) (19.409) (45.685) ----- (6) Employees 0.720 0.451 0.688 0.775 0.838 1 (17.402) (8.469) (15.875) (20.562) (25.777) ----- (7) Return Index 0.540 0.558 0.546 0.385 0.413 0.378 1 (10.756) (11.284) (10.913) (7.002) (7.611) (6.837) ----- (8) Market Capitali 0.696 0.559 0.847 0.894 0.903 0.765 0.513 1 (16.226) (11.312) (26.734) (33.414) (35.314) (19.916) (10.022) ----- (9) R&D/Sales -5yr 0.064 0.052 0.007 0.035 0.021 0.004 0.012 0.029 1 (1.081) (0.870) (0.110) (0.585) (0.359) (0.073) (0.207) (0.477) ----- (11) NIBEX 0.6739 0.5861 0.8326 0.8963 0.9423 0.8147 0.4832 0.9462 0.0239 1 (15.288) (12.127) (25.203) (33.872) (47.198) (23.547) (9.252) (48.994) (0.400) ----- (12) Sales growth 0.415 0.256 0.253 0.257 0.223 0.232 0.327 0.273 0.018 0.216 1 (7.648) (4.430) (4.385) (4.460) (3.829) )4.003) (5.799) (4.758) (0.297) (3.712) ----- (13) R&D/Sales 0.584 0.509 0.591 0.401 0.359 0.327 0.528 0.470 0.054 0.400 0.348 1 (12.045) (9.908) (12.274) (7.332) (6.444) (5.798) (10.415) (8.928) (0.913) (7.315) (6.218) ----- (14) R&D/Assets 0.549 0.559 0.564 0.342 0.339 0.349 0.527 0.441 0.010 0.387 0.327 0.947 1 (11.017) (11.305) (11.452) (6.095) (6.044) (6.238) (10.399) (8.246) (0.174) (7.037) (5.791) (49.223) ----- (15) Assets 0.514 0.244 0.533 0.856 0.864 0.729 0.240 0.724 0.013 0.756 0.135 0.217 0.184 1 (10.038) (4.213) (10.557) (27.737) (28.747) (17.855) (4.1458) (17.573) (0.218) (19.375) (2.278) (3.723) (3.141) ----- (16) Patents/Sales 0.472 0.366 0.549 0.451 0.554 0.644 0.342 0.536 0.007 0.561 0.268 0.380 0.372 0.396 1 (8.983) (6.587) (11.020) (8.460) (11.143) (14.113) (6.091) (10.653) (0.117) (11.363) (4.669) (6.893) (6.711) -7.23 ----- (17) Brands patent n 0.381 0.210 0.422 0.553 0.415 0.360 0.120 0.485 0.012 0.479 0.187 0.204 0.148 0.341 0.119 1 -6.916 -3.606 -7.803 -11.123 -7.646 -6.465 -2.025 -9.293 -0.199 9.136878 -3.199 -3.495 -2.499 -6.072 -2.008 ----- Pearson´s correlations in bold are significant at the 0,05 level 104 CHAPTER 5 105 1. General conclusions The main purpose of this study was to examine whether the total remuneration paid to CEOs in high-technology firms in the S&P 1500 is related to corporate finance. This work tries to understand how CEOs are compensated when managing high-tech firms and how their performance and innovation are driven to improve shareholders’ wealth and to promote their own goals. The first essay aims at explaining the influence that performance has on CEO compensation for shortand long-term periods in these groups of companies. It was found that there is a strong and positive relation between CEO compensation and firm performance. According to the results obtained, there is empirical evidence to state that in high-technology firms in the S&P, during the period between 2000 and 2010, performance determined total CEO compensation in shortand long-term periods, together with accruals of financial performance measurements. Results suggest that high-tech firms tend to use more sophisticated performance measurements than other firms to determine CEO compensation. The method used has potential implications in finance and accounting, for instance, where it is preferable to separately capture the specific effects of firm and performance. The main purpose of the second essay, included in chapter 3, is to examine whether the total compensation paid to CEOs in high-technology firms is related to corporate finance, and tries to understand how compensations are influenced by the introduction of the Standard Financial Accounting, SFAS 123 (R). This study will contribute to a better understanding of the relationship between compensation and performance in high-technology firms caused by the new role of expensing stock options with the SFAS 123 (R). This is consistent with the results achieved by Carter et al. (2007), who stated that the favorable accounting treatment for stock options possibly lead to overall higher CEO compensations. They state that there is no evidence of a decrease in total compensation combined with the positive association between financial reporting. The finding that after controlling standard economic determinants of compensation, expensing options in firms decrease compensation from options and increase compensation from