Amsterdam Business School
Shareholder Influence on CEOs’ Short-termism: a Content
Analysis of Temporal Orientation in Annual Reports
By: Maarten Mees ten Oever Student Number: 10309632
Master Thesis Business Administration First supervisor: Robert Kleinknecht, MSc. Second supervisor: dr. Alan Muller
Abstract
This paper contributes to the academic literature on short-termism by analysing different
sorts of investors in combination with differences in the temporal orientation CEOs of Global Fortune 500 companies communicate. A new measurement is created to identify short-termism by performing a content analysis on the temporal orientation of CEOs in the letter to shareholders in annual reports. With a word count on terms that specifically relate to either the short-term or to a future outlook the temporal orientation in the letter to shareholders could be quantified. The influences of both shareholder types and the amount of ownership on temporal orientation are analysed. Contrary to the literature, which states that short-termism is largely influenced by institutional shareholders, this is not reflected in the communication. The results indicate that shareholders do not influence the temporal orientation CEOs communicate. No differences in the results are found when controlled for regions, industry types and publicly listed firms. This research helps to give managers a better awareness of the ownership structure and their orientation when addressing the shareholders and thereby could create a better fit between both.
Acknowledgements
I want to take the opportunity to thank my supervisor Robert Kleinknecht MSc. for the pleasant cooperation and supervision during the process of writing my thesis. I wish to thank dr. Alan Muller for reading and grading my thesis together with Robert Kleinknecht. Further, I would like to thank my family for their support throughout my studies and definitely during my thesis, which gave me the energy and perseverance to finish it.
Table of Contents page
---Front Page 1 Abstract 2 Table of Contents 4 1. Introduction 52. Theoretical Review and Hypotheses 7
2.1 Short-termism 7
2.2 Corporate Governance and Short-termism 8
2.3 Influences on short-termism 9
2.4 Hypotheses 12
2.4.1 Blockholdership 12
2.4.2 Different investors and investment horizons 17
2.4.2.1 Mutual and Pension Fund Shareholders 19
2.4.2.2 Insurance Fund Shareholders 20
2.4.2.3 Bank Shareholders 20
2.4.3 Non-institutional Investors 21
3. Dataset and Data Collection 23
3.1 Dataset 23
3.2 Database and data collection 25
4. Methodology 26
4.1 Content Analysis Method 26
4.2 Dependent, Independent and Control Variables 28
4.3 Regression Analysis 35 5. Results 36 5.1 Descriptive Analysis 36 5.2 Statistical Analysis 39 5.2.1 Correlation Analysis 39 5.2.2 Regression Analysis 40
5.2.3 Validity of the Measurement 42
6. Discussion 44
6.1 Discussion hypotheses 45
6.2 Potential Implications for Literature 46
6.3 Managerial Implications 49
6.4 Limitations and Future Research 51
1. Introduction
According to many the primary objective of companies is to maximize shareholder value or ownership wealth (Lazonick & O’Sullivan, 2000). This is always subject to some constraints, such as compliance with laws and adequate concern for employees, customers, and other stakeholders (Merchant & van der Stede, 2003). In the last decades, firms have increasingly become more focused on maximizing financial results in a shorter time period and some claim that over the past decades this has led to short-termism (Levitt, 2000; Barton, 2011) and even culminated in the recent economic crisis. Short-termism can be described as a systematic excessive focus on short-term earnings at the expense of long-term interests (Von Thadden, 1995). It has become a frequently debated topic in the Western World, which focuses on myopia (cognitive concept of recent past playing a unproportionally significant role in making judgments) or short-termism (Stein, 1988; Jacobs, 1991; Porter, 1992; Laverty, 1996, 2004; Marginson & McAulay, 2008; Demirag, 1998; Koller et al., 2010; Jackson & Petraki, 2011; Khanin & Turel, 2012).
Shortsightedness influences the business world in the US and Europe, where it is becoming the norm that CEOs are consuming much of their time by focusing on quarterly earnings, while some claim that a long-term perspective seems to have become the competitive advantage of other economies and businesses these days (Barton, 2011). Short-term focus can have negative consequences not only for a particular multinational enterprise (MNE) or industry, but also for the countries in which they are operating it can be an important problem (Jacobs, 1991; Porter, 1992). When there is general under- or overinvestment a countries economic and competitive position will deteriorate, because it not uses its full potential and quality of its labour pool to build a strong and competitive national economy (Porter, 1992). Barton (2011) comments on this as it can undermine the ability of companies to invest and grow, and those missed investments can have far-reaching
consequences, including slower GDP growth, higher unemployment, and lower return on investment for money-savers. Therefore it is important to intensify the duration and depth of the debate and expand academic research on short-termism.
Different causes for short-termism are frequently cited, like pressure from capital markets (Jacobs, 1991; Porter, 1992). Relieving this pressure, for example from institutional investors (banks, insurance companies, retirement or pension funds, hedge funds, investment advisors and mutual funds), would have a more long-term consequence. But the academic
literature is not univocal about this, as Hansen and Hill (1991) showed: more stock held by
institutional investors increases instead of decreases R&D spending, as a consequence of the input of institutional investors through corporate government engagement.
One of the challenges researches on short-termism face is the difficulty to measure the short-termism different companies exhibit and therefore this research presents a new
construct to capture short-termism in the communication of companies. This paper
contributes to the academic literature on short-termism by analysing different sorts of (institutional) investors in combination with differences (which might explain the contradictory results mentioned above) in the temporal orientation the CEOs of companies communicate. By performing a content analysis on the letter to shareholders in annual reports of the 2012 Global Fortune 500 firms this measure for temporal orientation is created. The added value of content analysis is that the method provides a unique look into the minds and cognition of CEOs, which is difficult to obtain through other methods (Yadav et al., 2007).
In the next section I will commence with a theoretical review of past research on the subject of short-termism, corporate governance and the root causes and consequences of short-termism. The literature review paves the way for the hypotheses, after which this paper continues with an explanation of the dataset and data collection, the methodology, the data
analysis and the results. Finally, this paper will close with practical and academic contributions in the discussion and finish with the conclusion.
2. Theoretical review and hypotheses
This section offers a literature review of the most important topics and the hypotheses covered in this research. Section one reviews the literature on short-termism, followed by a short review of corporate governance. In the third section the influences and causes of short-termism according to previous literature will be addressed. Section four provides literature on firm ownership and its influence on management orientation according to the literature; this is presented in combination with the hypotheses.
2.1 Short-termism
“Short-termism” or “myopia” are terms that describe decisions in which firms pursue short-term gains, quarterly profits for example, at the expense of long-short-term strategies, for example
R&D or innovation through basic research (Laverty, 1996). It involves situations where
corporate stakeholders – investors, managers, board members, and auditors – show a preference for strategies that add less value but have an earlier payoff relative to strategies that would add more value in the long run (Jackson & Petraki, 2011). When managerial myopia becomes systematically integrated in the way firms operate, it becomes short-termism (Laverty, 2004).
A definition of short-termism must not be viewed in terms of the prioritization of the short-term, but as actions taken in the short-term which damage the long-term effectiveness of the firm and hence its value by means of undervaluation of long-term consequences (e.g. Merchant, 1990; Laverty, 1996). This does not mean that short-term results are not important; the viability of firms depends often on these achievements (Merchant & Van der Stede, 2003).
The short term must be distinguished from short-termism because it is not necessarily disadvantageous to take actions that are beneficial within a one-year period. Short-termism describes systematic actions taken that are detrimental for the long-term and therefore create an inter-temporal trade-off.
It is important to note the distinction between myopic decisions and short-termism. The academic literature is not univocal about the difference between the two terms and both sides of the debate have been referring to the same fundamental arguments for two decades (Laverty, 2004). Miller (2002) explains “managerial myopia indicates cognitive limitations in relation to the temporal dimension of decision making.” Jackson and Petraki (2011) state that myopic decisions are the consequence of natural cognitive limits in the face of uncertainty that managers have when they have to make a decision, while short-termism occurs when
these decisions become institutionalized and are taken repeatedly. Short-termism reflects a
systematic character of decision making within an organization shaped by organizational
culture, processes, or routines and external influences (Laverty, 2004). Although the basic
definition is clear among researchers there is less consensus among them about how to
demonstrate the sub-optimal decisions that cause short-termism (Segelod, 2000; Marginson
& McAulay 2008). It is difficult to establish whether short-term decisions are actually
detrimental to long-term value creation and therefore managerial implications are few and widely scattered.
2.2 Corporate governance and short-termism
Short-termism is influenced in part by the rules, sets, and beliefs that come together in
corporate governance which can be different per country (Haxhi & van Ees, 2010). Porter
(1992) argued that the different corporate systems of America and Japan are responsible for
definitions for corporate governance (Gillan, 2006). The academic approach on this subject varies because of the different disciplinary perspectives like economics, management, law, political science, culture and sociology. Aguilera & Jackson (2010) define it broadly as the
study of power and influence over decision-making within the corporation. Shleifer & Vishny
(1997) focus on corporate governance in a narrower sense, namely that it deals with the actors as the ways in which suppliers of finance (shareholders) to corporations assure themselves of getting a return on their investment. The latter definition is used in this research, because this research looks at shareholder composition as an influence on short-termism, as it is the shareholders’ interest and their influence that the boards of MNEs will have to align their firms to.
2.3 Influences on short-termism
The debate on short-termism started in the 1980’s in the US with an article that blamed the management of American firms for their focus on short-term cost reduction instead of a focus on technological competitiveness on the long-term (Hayes & Abernathy, 1980). Jacobs (1991) showed the difference in the manner that capital investments were acknowledged and evaluated in Anglo-Saxon firms, by comparing these firms to Japanese and German investments in capital-intensive growth markets. His conclusion was that managers of MNEs have the responsibility to take actions that secure long-term value (through R&D, focus on new growth markets and competency development), but that they are influenced by capital markets and performance measurement systems that cause them to act in a more short-termist way (Jacobs, 1991; Porter, 1992). Porter (1992) stated that in general, the US system is geared to optimize short-term private returns, while the Japanese and German systems optimize long-term private and social returns. By focusing on long-term positioning of corporates and creating an ownership structure and governance process that incorporate the
interests of employees, suppliers, customers, and the local community (broader focus on stakeholder value creation instead of shareholder profit), the Japanese and German systems better capture the social benefits that private investment can create.
This started an academic gulf of research about what the different capital reasons for short-termism could be, like a higher cost of capital in the US than in Germany and Japan as well as higher rate on premium risk and higher inflation (Kester & Luehrman, 1992; Stulz, 1996). Performance measurement systems are the same in countries as the UK and Germany and make it hard to differentiate between as a cause for short-termism in the US compared to other countries (Coates et al., 1996). Miles (1993) further commented on the influence of the capital markets by showing that the stock market systematically underestimated the value of future cash flows. Those that would focus on growth six months ahead were undervalued by 5 percent and the percentages would even be higher on a longer time horizon. This was combined with the explanation that the use of short-term divisional performance measurements transmits the pressure from the stock market to divisional managers. Other research had already showed that the multidivisional structure of firms is also inherent to short-termism as managers feel pressure to take immediate steps to show more profit in the present (Loescher, 1984).
Testing market efficiency by looking at share prices would show possible market inefficiency when it comes to the reflection of short- and long-term focus (Marsh, 1990). When firms announced they had taken long-term investment decisions, a raise in the share price followed in the next days, thereby stating that capital markets are efficient in that sense
(Woolridge & Snow, 1990). But Woolridge & Snow (1990) also argued that it is difficult to
prove that the immediate rise has a long-term effect on share price and thus on the consequences of short-termism. The proposition that follows from these researches is that pressure of the stock market caused firms to focus on short run performance and thus invest
more than would be advisable for the firm in the long run, i.e. the underinvestment in R&D, competency development and the focus on potential new growing markets.
Laverty (1996) moved away from only looking at capital markets and performance measurement systems as influences on short-termism. He extended the conceptual knowledge on the subject by including individual- and organizational-level factors, attributing short-termism to individual psychological factors leading to irrational behaviour. Decisions may be within the norms of rational action, if they work effectively in particular types of situation (Todd & Gigerenzer, 2007). Giving attention to the organizational-level factors that influence how the time horizons and preferences of individuals are created and reinforced within particular settings furthered the focus on what causes short-termism. Vera and Crossan (2004) show that top management leadership has influence on a firms’ short-term and long-term performance. Marginson & McAulay (2008) provide empirical evidence in line with Laverty and further elaborated that there was no capital market effect on intra-firm short-termism, as measured through hierarchical influence and managerial level. They suggest that short-termism may be determined at the work group level, given the emerging significance in organizations of networks of autonomous teams.
Managers tend to think different in several countries and scholars point to different causes that can lead to short-termism. Segelod (2000) surveyed top-managers in Sweden who stated that the anticipated market growth and political risk in countries seem to be far more important for the willingness to make long-term investments than pressure from the stock market, the cost of capital, or the rate of inflation. Demirag et al. (1994) showed that top management is responsible for internal short-termism influenced by external short-termism by shareholders in the UK. This shows a difference between US and UK managers in comparison to Swedish managers. The exact reasons for this are unclear though, as the political risk in these countries tends to have the same insignificant influence. Another survey
showed that divisional managers tend to believe that corporate managers place more emphasis on short-term performance measures than they actually do. This is caused by the favouring of divisional projects that generate profits on the short-term (Barton et al., 1992).
2.4 Hypotheses
Multiple studies have researched different types of firm ownership as a possible cause for the short-term focus of companies (Chaganti & Damanpour, 1991; Kochhar & David, 1996; Bushee, 1998; Liljeblom & Vaihekoski, 2009). The relationship between ownership structure and firms’ focus is likely to vary across industries and countries, to the extent that there are differences in corporate governance, industry types and influence from the financial markets. For this research two meta-categories are created to analyse shareholder influence on CEO’s temporal orientation. The first is ownership composition (the amount of stock a shareholder owns) and the second is ownership type. The next section will first deal with ownership composition, which is divided in blockholders (i.e. at least one, one and multiple blockholders, majority ownership and full-ownership). Then ownership type will be addressed, this is divided in transient and dedicated shareholders, mutual/pension funds, insurance funds and banks (institutional shareholders) and non-institutional shareholders.
2.4.1 Blockholdership
Large shareholders are defined as blockholders. There is no unambiguous definition of a blockholder, but the empirical literature typically defines a blockholder as a shareholder that owns at least 5 percent of the shares (Edmans, 2013). Holderness (2009) finds that 96 percent of US firms contain at least one blockholder. Dlugosz et al. (2006) document that in 2001 the average S&P 1500 firm had 1.97 outside blockholders, who collectively owned 18 percent of the firm.
Shareholders can exert governance through direct intervention in a firm’s operations (del Guercio and Hawkins, 1999; Edmans, 2013). The well-known advantage of concentrated ownership is that a large or controlling shareholder has a strong incentive to monitor managerial misconduct and exert governance (Gelter, 2009). Blockholders are more capable of monitoring and aligning management to their objectives; therefore ownership concentration is expected to have a positive effect on firm performance (Jensen & Meckling, 1976; Shleifer & Vishny, 1986; Jiraporn & Gleason, 2007). However, others argue that in some cases no shareholder activism is observed (Parrino et al., 2003; Barber, 2007). An explanation that no measurable impact is observed is the ability of shareholders to “walk away” from the investment rather than spend costly effort to monitor management (Parrino et al., 2003). But, the possibility to walk away becomes considerably smaller as shareholder size becomes larger, as is the case with blockholders that own at least 5% of a company.
Although some prior research indicates that shareholders do not always monitor and influence firms through activism by governance, this does not refute that the concentration of shares with a blockholder creates an incentive to monitor the company and align management to its objectives and therewith create value for the firm as well as for the blockholder. Blockholders can hold up to 100% of the shares at a company. As ownership concentration gets higher it is easier for the management of a firm to align its goals to a long-term orientation. As the boards of these companies focus less on financial performance and the influence of financial markets, the communication towards its shareholders is expected to be orientated on the long-term rather than the short-term and therefore the first hypothesis is stated as follows:
Hypothesis 1: Firms with at least one blockholder are associated with a greater orientation on long-term performance than on the short-term.
To get better insight in the influence of blockholders a further separation is made between firms that have one blockholder and firms with multiple blockholders. Traditional academic theories argue that governance is strongest under a single (large) blockholder, as that shareholder has large incentives to undertake value-enhancing interventions (Edmans & Manso, 2011). Therefore the next hypothesis is:
Hypothesis 1a: Firms with one blockholder are associated with a greater orientation on long-term performance than on the short-term.
Although the value-enhancing interventions are largest at firms with one blockholder, most firms are held by multiple blockholders (Holderness, 2009). The fact that a firm has multiple owners can create a free-rider problem when it comes to the monitoring and control of managers. Blockholders know that the control should come from their side, they do not always feel the incentive to act upon it, as there are multiple blockholders and each blockholder individually has insufficient incentives to bear the cost of monitoring (Grossman & Hart, 1980). Literature shows that blockholders have more incentives to monitor and align management to their objective (Jiraporn & Gleason, 2007). But in spite of the incentives, research shows that active governance is not always identified (Barber, 2007) and this is related to the fact that there are multiple blockholders. As short-termism can be looked at as an agency problem, better control should lead to a greater long-term orientation. Therefore, firms with multiple blockholders have an ownership structure that can generate difficulties because of free-rider problems that hinder control. Other research shows that disciplining of managers through trading can weaken these difficulties (Edmans & Manso, 2011). Thus, the trading of blockholders strengthens the long-term decisions a manager made which is reflected in stock prices. Therefore the next hypothesis states:
Hypothesis 1b: Firms with multiple blockholders are associated with a greater orientation on long-term performance than on the short-term.
A further distinction is made in the size of shareholders, as some companies have large shareholders that are special cases of blockholdership. A shareholder that holds at least 50.01% of the shares in a company is called a majority shareholder and a shareholder that owns 100% of the shares in a company has full ownership or is full owner. The reason to separately test the influence of majority shareholders and full owners is that for blockholders in general shareholder control can be relatively smaller and therefore no effect may be found. This effect is expected to become larger when tested only for firms with majority shareholders and for firms with full-ownership shareholders.
Firms in which the ownership structure is concentrated tend to have shareholders that effectively monitor the company and when these shareholders have substantially larger equity holdings in a company the effectiveness increases (Sarkar & Sarkar, 2000; Short, 1994). Ownership concentration reduces agency costs and therefore the alignment with management suggests that increasing ownership concentration beyond the minimum level for effective control will reduce opportunistic behaviour and the incentive for managing earnings upward by controlling owners (Ding et al., 2007). On top of that, substantial mismanagement by majority shareholders and full-owners is ultimately reflected in declining firm size and financial difficulties and such firms would then be eliminated in the competition for survival. Therefore majority shareholders are also expected to have a long-term influence on a firm’s focus (Short, 1994).
Majority shareholders in general are also more active when it comes to corporate governance, the monitoring of the company and the board, and in influencing decision-making. One reason for this is that they are better informed and able to monitor at lower costs
in comparison with small shareholders. Although, Black (1998) showed that managers sometimes ignore a majority shareholder vote in favour of a proposal, controlling shareholders can always intervene with the way management runs a firm. Concentrated ownership implies strong shareholder influence on management decision-making (Gelter, 2009). As a majority shareholder owns at least 50.01% equity of a firm, it logically follows that they focus on long-term gains instead of short-term gains that are detrimental for the firm, particularly in comparison to smaller blockholders. Therefore, a positive relationship between long-term orientation and majority ownership is expected:
Hypothesis 1c: Firms that have a majority shareholder have a greater orientation on long-term performance than on the short-long-term than firms that have no majority shareholders.
The above explanations why firms with a majority shareholder have a long-term orientation definitely applies for firms with a full owner, as a full owner is the only shareholder that has a saying in the way a firm is run. As boards of these firms feel less pressure from the capital market than boards of firms that have to deal with the influence of multiple smaller (institutional) shareholders, the communication is expected to focus even more on the long-term. Therefore the following hypothesis is formulated:
Hypothesis 1d: Firms that have a shareholder with full ownership have a greater orientation on long-term performance than on the short-term than firms that have no shareholders with full ownership.
2.4.2 Different Investors and Investment Horizons
After analysing ownership composition, different shareholder types are analysed in this section. Firms are owned by different types of institutional and non-institutional investors. Institutional investors have become increasingly important equity holders in the financial markets. In the UK and US, domestic institutional investors hold 30 to 40 percent of equities, and in Germany, Japan and Canada institutional investors hold around 20 percent (Davis, 2002). Because they can coerce with management they can have a large influence on the way a company is run. Easing the pressure from institutional investors would have a more long-term consequence (Jacobs, 1990; Porter, 1991; Laverty, 2004; Bushee, 1998; Claessens et al., 2002). In contrast, some studies point out that institutional investors are well placed to minimize problems, because of their greater bargaining power over the firm compared to the relative influence individuals have (Davis, 2002). Others state that some even promote R&D, engage in monitoring, or improve the market value of firms with good future prospects but low current profitability through corporate government engagement (Hansen & Hill, 1991). But most of the academic literature argues that institutional investors are focused on maximization of equity in the short run (Segelod, 2000), because they rarely possess the industrial knowledge to engage in the long-term development of a company. Current shareholders may choose to incentivize the CEO for short-term stock performance, even if they understand that this also creates incentives for the CEO to manipulate earnings (Bolton et al., 2006). Lipton and Rosenblum, cited in Dent (2010), state that some of these shareholders are even focused on pushing corporations into steps that are designed to create a short-term rise in the company’s share price so they can turn a quick profit. As most of the literature shows, CEOs feel pressure from the capital market through institutional investors and it will therefore likely cause them to prioritize the short-term over the long-term.
Although some of the contradictory results from previous studies into shareholder influence on short-termism seems to stem from the different investment horizons different types of investors apply. Before analysing each type of institutional shareholder separately, a distinction is made between investors with a short investment horizon and investors with a long investment horizon. (Institutional) shareholders can be divided in transient (short-term) and dedicated (long-term) shareholders depending on their investment horizon (Chen et al., 2009). The transient shareholders in this research include mutual/pension fund shareholders and insurance fund shareholders as they have a turnover rate of at least 50% per year (Jackson & Petraki, 2011). The dedicated shareholders in this research include bank shareholders and non-institutional shareholders, as these types of shareholders are known to have a long-term focus. Porter (1992) showed that in contrast with American firms, capital in Japanese and German firms is more likely to be held by banks and other non-institutional shareholders and therefore these firms focus more on the long-term. Dedicated institutions do not affect institutions' investment decisions in a negative way (Bushee, 1998), while transient shareholders have a short-termist influence on investment decisions. Therefore the following hypotheses are stated:
Hypothesis 2a: Firms with a higher percentage of transient capital owners are associated with a greater orientation on the short-term than on long-term performance.
Hypothesis 2b: Firms with a higher percentage of dedicated capital owners are associated
2.4.2.1 Mutual and Pension Fund Shareholders
After dividing shareholders on the basis of the investment horizons they apply, this research
also makes a distinction between the most common types of institutional investors (pension
and mutual funds, insurance companies, and banks). The different institutional investors will
be considered monolithically, as different types of institutional investors are likely to have a different influence. Jackson and Petraki (2011) analysed the time horizons in which the different institutional investors trade their stocks and they showed that, paradoxically, pension funds have the highest level of turnover, at 70 to 90 percent per year. In general, pension funds have diversified portfolios and a high number of stocks in their portfolio (Jackson & Petraki, 2011). Pension funds focus on a maximization of the value of their assets (Romano, 1993). Mutual funds, with high liquidity requirements and strong focus on quarterly earnings, are also predominantly perceived to be short-term investors with 50 percent turnover per year. Findings from Bushee and Noe (2000) indicate that institutions that trade aggressively based on short-term trading strategies, invest more in firms with higher disclosure rankings, facilitating the realization of short-term trading gains. Pension funds, with their tendency to buy as much shares of a company in a short period and put pressure on the boards of MNEs, will likely have a negative effect on the communication of a long-term focus of firms. Mutual and pension funds tend to avoid shareholder activism (Ryan & Schneider, 2002). But when they influence decisions, they are more likely to oppose management in comparison with other investors, such as banks and insurance companies, who have current or potential business links with the firm (Sarkar & Sarkar, 2000). Although mutual funds do not have the same high turnover rate as pension funds, both types of shareholders have a short investment horizon and a short-termist influence on firms. Therefore, mutual and pension funds as shareholders are expected to cause CEOs to have a greater orientation towards the short-term in their communication:
Hypothesis 3a: Firms with a higher percentage of mutual and pension funds as shareholders are associated with a greater orientation on short-term performance than long-term performance.
2.4.2.2 Insurance Fund Shareholders
Insurance companies are seen as less short-term investors, compared to other institutional investors like mutual funds, and known to have a relatively high engagement in corporate social responsibility (Brandes et al. 2006; Cox et al. 2004). Although, with an average turnover rate of 50 percent insurance companies are still identified as short-term investors (Jackson & Petraki, 2011). This stems from the fact that the core-business of insurance firms is the creation of profit through buying and selling of stock equity. Insurance firms tend not to coerce with the management of firms they own (Ryan & Schneider, 2002) and show a larger passivity with low concentrations of equity (Sarkar & Sarker, 2000). As insurance shareholders have the ability to sell their stock when they do not agree with the way a company is run, they do not have to engage in expensive shareholder activism and still influence the board with short-termist pressure (Parrino et al., 2003). Therefore the following hypothesis is stated:
Hypothesis 3b: Firms with a higher percentage of insurance funds as shareholders are associated with a greater orientation on short-term performance than on long-term performance.
2.4.2.3 Bank Shareholders
Banks do not affect the investment decisions of firms in a short-termist way (Bushee, 1998). In contrast with American firms, capital in Japanese and German firms is more likely to be
held by banks and therefore these firms focus more on the long-term (Porter, 1992). This provides European and Japanese firms with dedicated capital owners: shareholders who understand more about the firm's value and long-run prospects than investors in the United States. In many European countries bank shareholders are among the most stable type of investor, they invest capital in firms with the goal to enhance the long-term value of firms (Black, 1998). The continental European banks show higher levels of activism than UK banks, which have similarly low levels of activism as other institutional investors (Ryan & Schneider, 2002). Banks hold over 10 percent of shares in Germany, Japan and France, reflecting their strong influence on corporate governance (Davis, 2002). In Europe the long-term relationships with banks are thought to encourage bank financing (Maher & Andersson, 1999) and these long-term relationships create even more direct impact when banks acquire ownership stakes in industrial companies for various strategic purposes of securing influence (Vitols, 2005; Yanagawa, 2007). Although continental European banks have a greater long-term orientation through activism than their counterparts, both do not only focus on shareholder profit creation but also have a more long-term value enhancing function (Graves & Waddock, 1990). Next to that, governments in the US and Europe have been keen on banks when it comes to their responsibility since the recent financial crisis. Therefore, in contrast with the other institutional investors analysed in this research, the following hypothesis is stated:
Hypothesis 3c: Firms with a higher percentage of banks as shareholders are associated with a greater orientation on long-term performance than on short-term performance.
2.4.3 Non-institutional investors
In contrast to institutional investors, non-institutional shareholders are investors that not buy and sell debt, equity or other investments but only through any of the legal entities that are allowed to do so. In comparison with institutional investors, non-institutional investors are not focused on actively influencing stock market prices and are not dominant players on the financial markets. Research has shown that corporate short-termism is negatively associated with the extent to which long-term investors hold firms’ stock (Brochet et al., 2013) Long-term shareholders engage less in shareholder activism that is focused on short-Long-term gains (Gillan & Starks, 2003).
In this research, industrial shareholders, state or public authorities, and family owners are included as non-institutional shareholders. Industrial firms are non-institutional investors that are not focused on short-term gains by means of buying and selling of stock on the financial markets. When an industrial firm buys stock in other firms this is mainly for different reasons. Ellis and Fausten (2002) sum up the motivations for MNEs to buy stock in another MNE, which include a rational response to trade barriers, the desire to avoid transportation costs, opportunities to exploit multi-plant economies and political change making new low-cost locations available for production. These motivations do not include the active trading of firms on stock markets, which is the main influence on short-termism.
Research on state-owned enterprises in China showed that these firms are less sensitive to financial markets and that privately owned listed firms favour earnings boosting methods more than their state-owned counterparts (Ding et al., 2007). Evidence shows that CEO turnover rate is lower, as they are less likely to be dismissed, than at firms that are more dependent on the financial markets (Yuan, 2011), which increases a long-term orientation. Research on states as shareholders showed that on average a state holds shares for 3 to 4 years (Jackson & Petraki, 2011).
Family-owned firms apply longer time horizons in their decision-making than do their institutional owned counterparts (Zellweger, 2007). Family firms often focus on passing their ownership on to the next generation (Ward, 1997) and the rest of the ownership composition is mostly build out of shareholders that provide capital without the threat that it will be liquidated in the short-term (Zellweger, 2007). In most economies and especially Asian countries the bulk of companies are family-owned or most shares are owned by a small number of families, regardless of their size (Claessens et al., 2000) and CEOs stay seated as much as six times longer than at a typical nonfamily public company (Zellweger, 2007).
As non-institutional shareholders are less focused on the short-term for multiple reasons, there are fewer incentives for CEOs to be orientated on the short-term and their communication will thus be focused on the long-term. The hypothesis that follows this statement is:
Hypothesis 4: Firms with a higher percentage of non-institutional shareholders are associated with a greater orientation on long-term performance than on short-term performance.
3. Dataset and Data Collection
This section focuses on the dataset used for this research. Also the database from which the data is collected is described. Next to that, the process of the data collection for both the independent and the control variables, with which the analysis is performed to answer the hypotheses, is explained.
3.1 Dataset
Companies List 2013 are used as the primary data source. The Fortune 500 is compiled and published by the Fortune magazine and consists of the 500 largest companies in the world that are ranked from lowest to highest gross revenue. The 2013 Global Fortune 500 refers to fiscal year 2012 and is an appropriate sample for multiple reasons. First, the companies belong to all sorts of industries, from different regions and countries around the world, though predominantly from the US. Second, almost all of these companies are MNEs and most of them are active on different stock markets throughout the world. This gives them the obligation to report about and to their shareholders, which is a necessity regarding the data used for the independent variable. Third, the firms have to present an annual report each year, which consists of the company’s activities in the preceding year. Annual reports are written to give shareholders information about a company’s activities, the focus the management has for the company and the financial performance of the company. Further selection of the companies that are used in the creation of the dataset used for this research was made by the following criteria:
1. The annual report must be written with reference to the year 2012 and it must be available.
2. The annual report holds at least one letter to the shareholders written by the CEO of the company. When the chairman and CEO function are not embodied in the same person and the annual report holds a separate CEO letter, this letter is chosen. Due to the fact that the CEO is the person that is more involved with the daily business of the firm. If the annual report only includes a letter written by the chairman or president of the company this letter was analysed. Some annual reports do not hold a chairman or CEO letter but an interview. When this is the case, the interview is analysed. Because an interview addressed to the shareholders of a firm holds the same information in a different format; therefore an interview is a founded choice to analyse.
3. To be able to analyse the shareholder letter or interview the annual report has to be in PDF format and the letter must be written in English.
4. For each company of which the chairman letter is analysed, shareholder data must be available.
This selection criteria creates a sample that consists of 397 companies from the 2013 Global Fortune 500 that have an annual report which contains a letter to shareholders (the list of the 397 firms can be found in Appendix A). 103 firms either do not have an annual report, the letter is not written in English or the annual report is submitted in a 10-K form; this obligatory form in the US does not hold a letter to shareholders most of the times.
3.2 Database and Data Collection
To collect all the necessary data for the independent variable and two of the three control variables, the Orbis database is used. Among many other things, the Orbis database holds extensive information on shareholder composition for the 397 companies that the dataset consists of. This is important for this research, because not all firms in the Global Fortune 500 are publicly listed. One of the advantages of Orbis is that it is possible to search for specific criteria like shareholder information on a specific date, for a specific firm. As every firm has a specific code it is possible to enter a code-list of the 397 companies and therefore the data can be collected at once.
As the annual reports that are analysed must refer to fiscal year 2012, the shareholder data is also collected for the year 2012. The information includes firms’ shareholder composition to a certain percentage (>0.10%) for which shareholders own a company. Further, the data states the name of the shareholder and denotes what type of shareholder it is. Next to that, the database shows information about the type of industry each of the 397 companies belongs to according to internationally standardized codes, which will be
explained further in the control variable section. The data from the Orbis database also indicates the home country, thus in which region of the world each of the 397 companies is located. Last, for the third control variable a different database was used. The information for the third control variable, namely whether a company in the dataset is publicly listed or not, can be found in the Bloomberg database.
4. Methodology
In this section the methods applied to test the hypotheses are described and discussed. The dependent variable is derived from a measurement that is created by means of a content analysis. Therefore, the next section will look at content analysis as a method, the appliance of content analysis in previous studies and how to overcome the pitfalls that exist when performing one. Content analysis is addressed more in-depth because it is not yet widely known and applied as a methodology. In the second section the different variables will be described and discussed. In the last section the regression analysis as a method to perform a quantitative analysis will be shortly looked at.
4.1 Content Analysis Method
Krippendorff (1980) was one of the first researchers to write a complete work on the methodology of content analysis. He stated that content analysis is indigenous to communication research and is potentially one of the most important research techniques in the social sciences. In 1988, Weber (cited in Holder-Webb et al., 2008, p.547) more specifically explained “content analysis is a way of codifying text and content of written narratives into groups or categories based on selected criteria, with the end goal of transforming the material into quantitative scales that permit further analysis.” Content
interest to management researchers (Duriau et al., 2007). Foremost to management research, content analysis provides a replicable methodology to access deep individual or collective structures such as values, intentions, attitudes, and cognitions (Kabanoff, 1997) Thus, content analysis can be used to identify groups of words revealing underlying themes and with that co-occurrences of keywords can be interpreted as reflecting an association between the underlying concepts. With the possibility to quantify these qualitative concepts, content analysis is a useful and valuable methodology.
The usage of content analysis in management studies started with the quantification of large volumes of disclosures, previous studies applied the method in different ways to quantify the disclosures. Bowman and Haire (1975), the first to use content analysis, examined the food processing industry and developed their own index based on the number of lines devoted to CSR in annual reports. Other studies applied content analysis by performing word counts, page counts or evaluated word constructs and created scales within which each sphere of information could be evaluated as a single unit and compared with other units on managerial importance (Holder-Webb et al., 2008; Campbell, 2000; Milne & Adler, 1999). Content analysis has the advantage that once the subjective part of choosing the particular variables has been done, the procedure is reasonably objective; the results are independent of the particular research. Although, objectivity depends on what type of content analysis is performed: word counts are more objective and larger sample sizes are possible than with the interpretation of texts (Cochran & Wood, 1984).
Content analysis also has its flaws, in many published social and environmental accounting literature exists a considerable difference with regards to dealing with matters of reliability and replicability (Milne & Adler, 1999). Krippendorff (1980) came up with three types of reliability for content analysis to reduce the possibility of errors that can question the reliability of a measurement, namely stability, reproducibility and accuracy. Stability is about
the ability of a content analyst to code data the same way over time. By performing test-retests reliability issues for this can be solved. The aim of reproducibility is to measure that when multiple analysts code a text they come up with the same results (Weber, 1990). The measurement of this type of reliability involves assessing the proportion of coding errors between the various coders. Reproducibility refers to the inter-rater reliability of multiple codes of the same text. The last one, the accuracy measure of reliability, involves assessing the coding performance of coders against a set or predetermined standard. When these three reliability checks are performed well, the content analysis method provides a reliable measurement and offers significant advantages for management research that outweigh potential limitations by carefully implementing a study (Brochet et al., 2013).
For this research the chance for errors in the reliability of content analysis has been reduced in two ways. First, by only counting words instead of interpreting texts subjectivity is decreased. Second, the surrounding context of the words is analysed; thereby checking whether a word explicitly refers to the category it is counted for. This will be explained in the next section.
4.3 Dependent, Independent and Control Variables
Dependent variable. A new measurement has been created for CEO orientation by
performing a content analysis on letter to shareholders in the annual reports of Global Fortune 500 firms. The dependent variable, CEOs temporal orientation, used for analysis in this research is derived from the new measurement. By analysing the letter to shareholders a unique glimpse into the minds of CEOs can be obtained and by performing a content analysis this is done in a way that is difficult to obtain through other means. Previous research shows the letters to shareholders in annual reports reflect CEOs’ orientation at that time and that this orientation can be judged in a meaningful way (Kabanoff, 1997; Yadav, et al., 2007).
To create the dependent variable for the temporal orientation of CEOs from the Fortune 500 firms, the content of 397 letters was analysed on 33 terms, which were divided in two categories. The two categories the dependent variable consists of are (1) terms that directly relate to a short-term focus and, (2) terms referring directly to a future outlook or a long-term orientation. To decide which terms to analyse the shareholder letters on, two things were done. First, a selection was made according to research done by Brochet et al. (2013) in which they analysed conference calls on temporal orientation. Second, three analysts randomly picked and read ten shareholder letters to add frequently occurring terms that were not used in the research by Brochet et al. (2013). Appendix B shows a complete overview of all the terms each category holds and the 397 letters to shareholders were searched for.
The shareholder letters were analysed on the terms using the program Acrobat Reader XI Pro. 26 of the 397 letters to shareholders could not be analysed, but by using ABBYY Finereader 12 Pro the 26 letters were converted into analysable letters. By using the advanced search option in Acrobat Reader XI Pro to search for occurring terms all the shareholder letters could be searched for at once. The counts of the 33 terms were manually entered in Excel per firm. Next to that, a word count of each letter was performed. The letters were converted to Word documents so the total words of each letter could be counted. The highest count was 13993 and the lowest count was 269 (M = 1677.37). By dividing the number of the found terms per letter by the total words of each letter the ratio of used terms in each letter compared to the other letters was corrected for. In other words, the measure was made relative with regards to the frequency of occurrence of a specific term.
The resulting spreadsheet was exported to SPSS Statistics. In SPSS the singular and plural terms referring to the same concept were transformed into one variable. A correlation matrix showed all significant correlations between the variables with r < -.10 and r > .10. Terms that can be considered as synonyms often showed negative correlations, this
confirmed the suspicion that often the same word is consistently used to refer to the same thing. When the qualitative definitions of variables were the same, the variables were computed together into a new variable. Here the fact that the correlations with all other variables should be similar was important. By analysing each variable and comparing it to all the other variables, quantitative and qualitative deduction led to a reduction of the total number of terms in the dataset. Dimensions holding similar words were created. For example, the ‘future outlook’ dimension held the terms ‘lie ahead’, ‘looking ahead’, ‘looking forward’ etc.
By calculating a new correlation matrix the dimensions were put together further, but only when there was a qualitative similarity and quantitatively the correlation with all the other variables was not contradictory. By repeating the process multiple times, two mutually exclusive dimensions were created and the initial 33 words were brought back to 28 words in the two dimensions. The resulting correlation matrix (Appendix C) shows the significant positive and negative correlations between the dimensions. The correlation matrix shows that the content analysis on shareholder letters is a valid measurement. The short-term dimension correlates positively with the financial dimensions and negatively correlates with the long-term dimension and CSR dimension. The index is a measurement for the relative weight of firms’ short-term vs. long-term orientation.
The dependent variable that is used in this research is derived from the measurement on CEO’s orientation. The topic of interest in this research is short-termism, therefore only the future-outlook and short-term categories were used. By dividing the ‘future outlook’ dimension by the sum of the ‘future outlook’ and ‘short-term’ dimension the relative weight of the long-term orientation in the letter to shareholders is measured and the dependent variable created.
Independent variables of interest. For the independent variables in this research
shareholder structure data of the 397 companies at precisely December 31st 2012 is used. One
company was traded at December 31st 2012 and therefore the latest public shareholder data
on that company is used. For this research and the timeframe in which the research must be
performed it is chosen to only collect data about shareholders holding ≥1% of the shares at
one of the 397 companies. These shareholders are ordered by the percentage of shares they hold at a company and the type of shareholder they are. Table 1 offers an overview and explanation of each of the independent variables of interest.
Table 1: explanation of independent variables
Variable Explanation Type of variable
Shareholder composition:
At least one blockholder Firms with at least one blockholder that owns a minimum of 5% of the shares. Dummy variable One blockholder Firms with one shareholder that owns at least 5% of the shares. Dummy variable Multiple blockholders Firms with multiple shareholders that each own at least 5% of the shares. Dummy variable Majority shareholder Firms with a shareholder that owns at least 50.01% of the shares. Dummy variable Full owner Firms with one shareholder, this shareholder owns 100% of the shares. Dummy variable
Shareholder type:
Transient shareholders Firms with shareholders (mutual/pension and insurance funds) that apply a short
investment horizon Continuous variable
Dedicated shareholders Firms with shareholders (banks and non-institutional shareholders) that apply a long
investment horizon. Continuous variable
Mutual/Pension shareholders Firms that have mutual/pension fund shareholders that own ≥1% of the shares.
Variable includes the total percentage of mutual/pension shareholders of a firm. Continuous variable Insurance shareholders Firms that have insurance fund shareholders that own at least ≥1% of the shares.
Variable includes the total percentage of insurance shareholders of a firm. Continuous variable Bank shareholders Firms that have bank shareholders that own at least ≥1% of the shares.
Variable includes the total percentage of bank shareholders of a firm. Continuous variable Non-institutional shareholders Firms that have non-institutional shareholders. Combined category that consists of
industrial firm, public authority, state or government, and individual or family-owned shareholders that own ≥1% of the shares.
Control variables. First, type of industry is used as a control variable. Research
conducted by economists has long suggested the impact of certain industry structural characteristics (e.g., production differentiation and rate of industry growth) on both individual firm and total industry performance levels (Hitt et al., 1982). Different types of industry bring with them unique external as well as internal conditions for companies that operate in a particular industry. For example, mining and quarrying companies can be more long-term because contracts for multiple mines can provide resources for over twenty years. Another example is US manufacturers of consumer durable goods (e.g. automobiles) and producer goods (e.g. retail products) that face severe economic restraints, whereas manufacturers in certain capital goods industries face completely different economic conditions (Hitt et al., 1982). This may influence the focus the management of a firm holds and therewith communicates, because the management of those companies must emphasize different conditions per industry to achieve organizational success. Whereas some industries impose and influence companies to have a long-term orientation, other types of industry can influence companies to have a short-term orientation. This can have an influence on the type of shareholders a company attracts and with that the influence of a shareholder on the communication of the chairman or CEO of a company. For example, previous research has shown that industry type explained 10% of the average variance of stock returns (Hitt et al., 1982). Thus, type of industry could impose a difference in the managerial short- and long-term orientation that comes forth out of the letter to shareholders and the influence of different types of shareholders on this orientation.
Industry will be controlled for according to the separation into different types of industry by the NACE rev. 2 code. The NACE-code is a statistical classification of economic activities in the European community. In other words, it is the European industry standard classification system consisting of a 6-digit code. The first 4 digits of the classification
system are the same in all European countries and the last 2 digits are country-specific. In this research a categorization is made according to the first 2 digits. NACE is similar in function to the Standard Industrial Classification (SIC) and North American Industry Classification System (NAICS). With the NACE rev. 2 code every company can be subdivided into a specific industrial category. By categorizing the sample per industry, the orientation per industry is controlled for. It is important to note that many companies employ different activities belonging to different types of industry, but they can only receive one NACE-code. The NACE-code is appointed to the ‘primary activity’ of the company, this is the activity that contributes the most to the total added value of that economic entity. As already explained above, it is expected that financial and insurance firms have more focus on the short-term than do companies that are active in the mining and quarrying industry.
The companies in the dataset originally belong to 12 categories in total. Because agriculture, forestry and fishing (N = 1) would leverage a too small number of firms this type industry was added to mining and quarrying. The reason for this is that both types of industry have more in common with each other than other types of industry. Both types of industry focus on the land as a supplier of economic value. Therefore the firms belong to 11 different categories (1 = mining and quarrying, agriculture, forestry and fishing, 2 = manufacturing, 3 = electricity, gas, steam and air conditioning supply, 4 = wholesale and retail trade; repair of motor vehicles and motorcycles, 5 = transportation and storage, 6 = construction, 7 = accommodation and food services, 8 = information and communication, 9 = financial and insurance activities, 10 = professional, scientific and technical activities, 11 = administrative and support service activities). To keep the variables large enough for proper analysis, the six smallest industrial categories (N < 19) are put together in one variable. The categories that are put together are electricity, gas, steam and air conditioning supply; transportation and storage; construction; professional, scientific and technical activities; accommodation and
food services; administrative and support service activities. The number of companies in this compiled sample is N = 50 (12.6%) and it is used as a reference category. The control variable contains six different categories and type of industry is used as a dummy variable in this research.
Second, home region of each company is used as a control variable. Research shows that in developing economies, ownership is also heavily concentrated (La Porta et al., 1998). Looking outside the United States, particularly at countries with poor shareholder protection, even the largest firms tend to have controlling shareholders. Sometimes that shareholder is the state; but more often it is a family, usually the founder of the firm or his descendants (La Porta et al., 1999). Because these firms have large controlling shareholders, they feel less pressure to focus on the short-term. Next to that, the regions differ when it comes to corporate governance and dedicated institutions as shareholders. These types of shareholders characterize European and Asian firms (Bushee, 1998). This provides European and Asian firms with shareholders who understand more about the firm's value and long-run prospects than investors in the United States. Therefore it is realistic to state there is a difference between shareholder influences on the CEO orientation that comes forth out of the letter to shareholders per region. For this research was chosen to divide the companies in this dataset into three different regions. These regions are North America together with Middle and South America, Europe, and Asia. Russian companies were included in the Asian region, as well as one Middle-Eastern firm. The American region is used as the reference category. The home region of firms is used as a dummy variable in this research.
The third control variable used in this research is whether a company is publicly listed. Bhide (1993) showed that non-listed firms are more often owner-managed than publicly listed firms and even when they are not owner-managed they are less liquid and have a higher concentration of ownership than their public counter-parts. Le Breton-Miller and
Miller (2006) showed that family firms exhibit a long-term orientation and a better performance on environmental and social metrics than their counterparts. The long-term orientation influences the investment decisions firms make, compared to publicly listed firms, non-listed firms invest less and are even less responsive to changes in investment opportunities. Publicly listed firms invest even more in industries where firms’ earnings news has a high influence on the sensitive stock prices and short-termist pressures negatively influence investment decisions (Asker et al., 2013). There is a difference in the concentration of ownership and the allocation of resources for investment between both types of companies. Privately held companies experience less pressure from capital markets and less need to focus on short-term gains. When the letter to shareholders is a good reflection of a firm’s orientation, being publicly listed as a firm will influence the orientation CEOs display. Whether a firm is publicly listed is used as a dummy variable in this research.
4.3 Regression analysis
In this research a regression analysis will be performed. The regression analysis is a commonly known statistical technique that shows the best fitting straight line though a set of points. It estimates the relationship among variables in a research model. This method explains how the value of a dependent variable will change, when one or more variables in the model change. The equation of a multiple regression is:
Y = a + β1 * X1 + β2 * X2 + βi * Xi + ε
In this equation Y, the dependent variable, is the degree of CEO orientation on either the short- or long-term in the letter to shareholders. Each X stands for one of the independent variables, which are the size of the shareholder (i.e. at least one blockholder, one blockholder, multiple blockholders, majority shareholders and full-ownership) and the type of shareholder
(i.e. transient, dedicated, mutual, bank, insurance and non-institutional shareholders). After that come the control variables (type of industry, home region and publicly listed firms).
To answer the hypotheses stated in this research, 6 regression models are constructed. The first model includes the control variables. In the second model the influence of firms with at least one blockholder is analysed. In the third model one and multiple blockholders are compared. In the fourth model the focus is on firms with majority and full owners. The fifth model analyses the influence of shareholders with different investment horizons. The sixth and last model analyses the influence of different types of shareholders i.e. transient, dedicated, mutual and pension fund shareholders, insurance shareholders, bank shareholders and non-institutional shareholders.
5. Results
This section describes the results obtained from the 6 different regression analyses that were performed. The first section provides a descriptive analysis of the different variables. The second part analyses the correlations between the variables accompanied with a correlation matrix and in the third section the regression analyses are shown and explained. Only one significant correlation between the dependent and independent variables showed up in the performed analyses. Therefore validity checks were performed on the measurement, which have been added to the results section.
5.1 Descriptive Analysis
In this section the descriptive statistics of the shareholder data are analysed. Table 2, on page 37 and 38, provides the descriptive statistics of all the variables that are taken into account in the different models of this research. With most of the independent variables the standard
Table 2. Correlations and descriptive statistics (continues on next page)
Table 2. Correlations and descriptive statistics (continuation of previous page)
