CEO之生肖與星座對公司風險承擔的影響 - 政大學術集成

全文

(1)4. 治Bx 政 The impact of CEO’s constellation and 大 Chinese zodiac on firm 立 risk-taking ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 1. E. 0. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(2) . 政治大. 立. ‧ 國. 學. ‧. sit. n. al. er. io. y. Nat. Ch. engchi. i n U. v. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(3) Bx. r. 4. d. 2. Bx Bx. 2. 政 O治大 O o. r. x. O O. B. a. ‧ y. Nat. io. sit. Bx. n. al. er. Bx. 學. C. 立. B. ‧ 國. O. . Ch. engchi. i n U. v. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(4) The impact of CEO’s constellation and Chinese zodiac on firm risk-taking Abstract Compared with prior studies mostly focus on the influences of CEO demographic characteristics on firm risk-taking, this study uses Chinese zodiac traits and constellation traits as the proxy variables for CEO psychological traits and takes into consideration demographic characteristic variables to analyze the relationship between CEO traits and firm. 政治大. risk-taking. My sample consists of CEOs who are covered by the BoardEx database over the. 立. period 2000-2014. The results show that CEOs with more conservative Chinese zodiacs are. ‧ 國. 學. less risky and prefer more conservative firm policies, and CEOs with more conservative constellation signs prefer more conservative firm policies. Further analyses reveal that these. ‧. findings are more pronounced among firms in the high-tech sector. Overall, my finding. Nat. sit. n. al. er. io. policies.. y. suggests that a CEO’s personal trait is associated with the riskiness of the firm and firm. Ch. engchi. i n U. v. Keywords: CEO traits, firm risk-taking, Chinese zodiac, constellation. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(5) Contents 1. Introduction………………………………………………………………………………1 2. Literature Review………………………………………………………………………...5 2.1. CEO Traits………………………………………………………………….……….5 2.2. Constellation………………………………………………………………….……..6 2.3. Chinese Zodiac……………………………………………………………….……..8 2.4. Firm Risk-taking Behavior and Firm Policies……………………………….…….11 3. Hypotheses Development………………………………………………………….……13 4. Data and Research Design…………………………………………………….………..16. 政治大. 4.1. Sample Selection…………………………………………………….…………….16. 立. 4.2. Risk Proxies and Firm Policies…………………………………….………..…….18. ‧ 國. 學. 4.3. Control Variables………………………………………………………………….20 4.4. Empirical Model…………………………………………………………………...21. ‧. 5. Empirical Results…………………………………………………………………….....24. Nat. sit. y. 5.1. Descriptive Statistics……………………………………………………………....24. n. al. er. io. 5.2. Chinese Zodiac, Firm Risk-taking Behavior and Firm Policies……………...........28. i n U. v. 5.3. Constellation, Firm Risk-taking Behavior and Firm Policies……………..............32. Ch. engchi. 6. Additional Analysis………………………………………………………………..……37 7. Conclusion……………………………………………………………………………….49 References……………………………………………………………………………….......51 Appendix……………………………………………………………………………..….......54 Appendix 1. Variable Definitions……………………………………………………....54 Appendix 2. Behavioral Traits Associated with the 12 Chinese Zodiacs…………........56 Appendix 3. Behavioral Traits Associated with the 4 Elements of Constellation……...58. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(6) List of Tables Table 1. Sample Selection……………………………………………………………………16 Table 2. Sample Distribution based on Chinese Zodiac or Constellation……………………17 Table 3. Descriptive Statistics………………………………………………………………..25 Table 4. Correlation Analyses……………………………………………………………......27 Table 5. Regressions of Conservative Chinese Zodiac and Corporate Risk and Firm Policies………………………………………………………………………….......29 Table 6. Regressions of Conservative Constellation and Corporate Risk and Firm Policies…………………………………………………………………………...…34. 政治大. Table 7. Regressions of Conservative Chinese Zodiac and Corporate Risk and Firm Policies within High-tech Industry……...……………………………………………………39. 立. ‧ 國. 學. Table 8. Regressions of Conservative Chinese Zodiac and Corporate Risk and Firm Policies within Non-high-tech industry……………………………………………………....41 Table 9. Regressions of Conservative Constellation and Corporate Risk and Firm Policies within High-tech industry………………………………………………………..................43. ‧. n. al. er. io. sit. y. Nat. Table 10. Regressions of Conservative Constellation and Corporate Risk and Firm Policies within Non-high-tech industry……………………………………….......................46. Ch. engchi. i n U. v. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(7) 1. Introduction The Chief Executive Officer (CEO) is the main decision-maker in a firm and integrates views of the top management team, requiring high cognitive complexity(Calori, Johnson, & Sarnin, 1994 ). CEOs’ managerial characteristics affect not only firm’s strategies and investment decisions but also risk-taking behaviors and firm performance (Malmendier & Tate, 2005, Coel & Thakor, 2008, Colbert, Barrick & Bradley, 2013). In addition, CEOs play a distinct role in influencing organizational financial performance and corporate organizational. 治政 commitment (Colbert, Barrick& Bradley, 2014) . In addition 大 to demographic 立 characteristics, CEOs’ personal traits also have impact on corporate policies and ‧. ‧ 國. 學. performance.. From recent research and practices, CEO’s personal trait matters. For instance,. sit. y. Nat. CEO’s charisma is related to firm performance via transformational leadership. al. er. io. climate within an organization, and increases a firm’s organizational identity strength. v. n. (OIDS) and relate positively to firm performance (Boehm, Dwertmann, Bruch, &. Ch. engchi. i n U. Shamir, 2015). Many studies have found that specific manager traits are related to decision-maker behavior, such as gender, age, education background and career experience (Barber & Odean, 2001; Bertrand & Schoar, 2003; Cadenilas et al., 2004; Malmendier & Tate, 2005; Matta & Beamish, 2008; Antia, Pantalis, & Park, 2010; Lin, Lin, Song, & Li, 2011; Buyl, Boone, Hendriks, & Matthyssens, 2011). In a moral hazard model created by information asymmetry, managers who prefer short-term results tend to sacrifice the interest of company in long term for the purpose of fulfilling personal optimization from his/her personal perspective (Jenson & Meckling, 1976). Due to hubris hypothesis of Roll (1986), overconfident CEOs. 1. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(8) aggressively make acquisitions (Malmendier & Tate, 2008). Besides, Malmendier and Tate (2008) has indicated that overconfident CEOs tend to overestimate the return on investment plans and be overinvest in a optimism bias. The trade-off theory (Myers, 2011) has pointed out that the personal behavior of the manager is not reflected in corporate structure decisions. Contrast to trade-off theory, the agency theory proposed by Jenson and Meckling (1976) suggested that the self-interested behavior of CEOs may harm the interests of the company. Furthermore, demographic characteristics and psychological traits have been discussed to be influential in corporate financing. 政治大 and career experience (Hambrick & Mason, 1984; Berger, Ofek, & Yermack, 1997; 立. decisions. Demographic characteristics include age, tenure, educational background. Pegels & Yang, 2000; Graham & Harvey, 2001; Barker & Mueller, 2002; Bertrand &. ‧ 國. 學. Schoar, 2003; Chen, Hsu, & Huang, 2010). As for psychological traits, they are. ‧. composed of risk aversion, time preference and overconfidence (Heaton, 2002;. y. Nat. Malmandier & Tate, 2005, 2008; Graham, Harvey, & Puri, 2013; Landier & Thesmar,. er. io. sit. 2009; Malmendier, Tate, & Yan, 2011). According to prior empirical studies, there are significant correlations between corporate risk-taking behavior and traits of CEOs.. al. n. v i n According to the matchingC theory, CEOs who doU h e n g c h i not match their firms will. leave, while those who match their firms will stay (Allgood & Farrell, 2003). Graham et al. (2013) suggest that it is the CEO’s behavioral traits that match the firm, i.e., a CEO will choose a firm that matches his/her personal traits, or a firm will employ a CEO with specific traits it demands. The paper also indicates that CEO traits such as optimism and managerial risk-aversion are related to corporate financial policies. Firms which prefer risk-taking CEOs have a higher degree of risks, such as highgrowth firms. When firms hire employees, especially executives, they not only focus on the demographic characteristics, but also their psychological traits owing to matching theory. 2. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(9) When it comes to our inborn personality, several factors matter, such as birth date, blood type. Based on our birth date, we can find our corresponding Chinese zodiac and constellation. It is commonly believed that Chinese zodiacs and constellation are connected with our personality traits (See Appendix 2 and 3). According to prior studies, blood type has little or no relationship with personality (Rogers & Glendon, 2002). Contrary to blood type, constellation and Chinese zodiac have been commonly and usually connected with our personality although there are few empirical studies related . Since constellation and Chinese zodiac has existed. 政治大 our personality and Chinese and western people have always been commonly believed 立 actively in our culture, I believe that Chinese zodiac and constellation have impact on. that we can understand others through their constellation and Chinese zodiac.. ‧ 國. 學. The purpose of this study is to examine whether CEO’s constellation and. ‧. Chinese zodiac have an impact on firm’s risk-taking behavior. Prior studies indicate. sit. y. Nat. that CEO traits affect firm’s risk-taking policies; however, there is no direct. io. er. indication that CEO’s Chinese zodiac and constellation, from Western and Chinese points of view, have an impact on the firm risk-taking behaviors and I attempt to fill. n. al. this gap in the literature.. Ch. engchi. i n U. v. Following Bova, Kolev, Thomas, and Chang (2014), I use two measures of firm risk to reflect the extent to which CEO’s Chinese zodiac or constellation affecting corporate risk-taking behavior, usually defined as the volatility of stock returns in an efficient market. My first risk measure is the standard deviation of monthly stock returns over the 12 months following the disclosure of audited financial statements. To the extent that the market may not be fully efficient, stock volatility will be affected by other factors and represent a noisy measure of corporate risk. As an alternative proxy, I use an accounting measure: the standard deviation of seasonallydifferenced quarterly accounting return on assets over the next 12 quarters. 3. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(10) Furthermore, I also consider other indirect measures of corporate risk, which reflect a firm’s policy, such as the level of R&D expenditures and capital expenditures. Previous studies have pointed out that the demographic characteristics and psychological traits of CEOs influence their behaviors and firm performances. However, prior studies mostly focus on the influences of demographic characteristics, which is easier to observe and measure. To supplement prior research, this study focuses on psychological traits which are not commonly discussed. In addition, this study uses the personality traits respectively represented by Chinese zodiacs and. 政治大 characteristics of CEOs, decision-makers of companies, have influence on the firm’s 立. constellation as a starting point for research to observe whether the personal. risk-taking behavior. This is an interesting and worthy research question that I believe. ‧ 國. 學. shareholders, board of directors and potential investors always want to seek for. ‧. suitable CEOs who can both match with firms’ missions and objectives and create. sit. y. Nat. values for the firms. The findings will also contribute to the discussion to the. io. er. discussion on influences of CEO personality traits on corporate risk-taking behavior. This study is organized as follows. Section 2 reviews related literature. Section 3. al. n. v i n C4hdescribes the sample develops the hypothesis. Section selection and research engchi U design. Section 5 describes the descriptive statistics and presents the empirical results, Section 6 describes the additional analysis, and Section 7 concludes.. 4. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(11) 2. Literature Review 2.1.. CEO personal traits Prior studies have long presumed that CEOs have heterogeneous talents and. capabilities that map into firm performance. Empirical studies have discovered that managerial heterogeneity is important for corporate actions and performance. Many studies have discussed about the relation between some specific CEO demographic and psychological traits and firm performance. For instance, Bertrand and Scholar (2003) identifies executives from earlier birth cohorts appear on average to be more. 政治大. conservative; on the other hand, managers who hold on MBA degrees seem to follow,. 立. on average, more aggressive strategies. Adams, Almeida, and Ferreira (2005), and. ‧ 國. 學. Bennedsen, Pérez-González, and Wolfenzon (2008) find that CEOs’ characteristics affect firm performance.. ‧. However, only a few studies indicate explicitly specific managerial. y. Nat. sit. characteristics. Bolton, Brunnermeier, and Veldkamp (2013) develop a theory of. n. al. er. io. leadership that resoluteness is a valuable attribute of a leader because we all. i n U. v. understand that resolute leaders will stick to their guns. Resoluteness is a form of. Ch. engchi. overconfidence that occurs when CEOs believe their initial information is more precise than it truly is. Bolton et al. (2013) analyze the challenges and tradeoff between adapting to changing circumstances and coordinating employees. In their conclusion, more resolute and overconfident CEOs perform better than CEOs who are better listeners and communicators. Besides, firms can get resolute managers to take more risks because choosing resolute managers allows firms to reconcile differences between theirs and the managers’ incentives. In addition, prior studies have analyzed and classified multiple managerial characteristics relating to corporate performance. Kaplan (2012) finds there are two 5. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(12) dominant factors that have intuitive interpretations by using the traditional empirical approach in studies of personality traits. In general, success and better performance is positively related to CEOs’ general talent or ability, which is the first factor. The second factor comprises two categories: communication and interpersonal abilities, and execution and resoluteness skills. In execution and resoluteness skills, there are several components, including fast, efficiency, aggressive, persistence, and proactive. Gervais (2011) finds that overconfident managers are more attractive to firms than their rational counterparts because overconfidence commits them to learn more about. 政治大. projects. However, too much overconfidence is damaging to the manager because it leads to excessive risk.. ‧ 國. 學. 2.2.. 立. Constellation. ‧. In practice, when hiring CEOs, many firms not only evaluate the traditional. sit. y. Nat. demographic characteristics of recruits, but also deliberate upon their psychological. al. er. io. traits, in order to select the right person to match the firm. In general, it is believed. v. n. that the constellation, also called twelve zodiac signs, are associated with different. Ch. engchi. i n U. kinds of personality traits. Moreover, in Chinese culture, it is also believed that twelve Chinese zodiac signs are also associated with their corresponding personality traits. The constellation is based on a range of dates that lasts about a month. According to countless ancient and modern studies, we can gain a better understanding of our potential traits with the zodiac signs. There are four basic elements in nature—Earth, Water, Fire and Air. Each of the twelve signs of the zodiac belongs to one of the four elements, and each element represents one form of energy and exhibit profound influence on basic character traits, emotions, behavior and thinking. 6. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(13) Earth signs are mostly conservative and realistic, but they can also be very emotional. They are practical, focused, disciplined, and eager to get the job done. Still, they are loyal, stable and they stick by their people through hard times. The Earth signs, also called pragmatists of the zodiacs, are: Taurus, Virgo and Capricorn. Water signs are exceptionally emotional, sensitive and intuitive. Water signs love profound conversation and intimacy, and they are always there to support their loved ones. The Water signs, also called the intuitives of the zodiac, are: Cancer, Scorpio and Pisces. Fire signs tend to be passionate, dynamic, and temperamental. Ruled by. 政治大 forgive easily. They are adventurers with tremendous energy, shine bright light on life 立 the quick and active fire element, these three signs get angry quickly, but they also. and, command attention wherever they go. They are intelligent, self-aware, creative. ‧ 國. 學. and idealistic, and always ready for something new. The Fire signs, also called the. ‧. trailblazers of the zodiac, are Aries, Leo and Sagittarius. Air signs are rational, social,. sit. y. Nat. and love communication and relationships with other people. They love philosophical. io. er. discussions, social gatherings and good books. They rely greatly on their mind and like to give advice; however, they may be superficial and too quick to judge. The Air. al. n. v i n signs, also called the intellects of Libra, and Aquarius. Ctheh zodiac, are Gemini, engchi U. 7. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(14) 2.3.. Chinese zodiacs1 The Chinese zodiac is a classification scheme that assigns an animal and its. reputed features to each year in a 12-year cycle. The Chinese zodiac signs are determined by the lunar year in which you were born. The Chinese believed that the animal which rules one’s birth year has a profound influence on one’s personal traits. Unlike western astrology dealing with constellation, Chinese astrology copes with wuxing (metal, wood, water, fire, earth), Yin and Yang, Chi, and the cycles of time. Still, Chinese people are not used to categorize Chinese zodiac signs based on wuxing. 政治大 zodiac signs can offer a better insight into one’s ego and personality. Chinese zodiac 立 or yin and yang. They believe that each zodiac sign has its unique traits. Chinese. traditional begins with the sign of Rat and the following are Ox, Tiger, Rabbit,. ‧ 國. 學. Dragon, Snake, Horse, Goat, Monkey, Rooster, Dog, and Pig. Besides, the 12 Chinese. ‧. zodiac animals are in a fixed order according to Chinese Yin and Yang Theory and. sit. y. Nat. perceived attributes.. n. al. er. io. Rat people are quick-witted, resourceful and versatile. With strong intuition and. v. quick response, they easily adapt themselves to a new environment. Rats are cautious. Ch. engchi. i n U. and meticulous, so they tend to show a serious attitude toward their work. However, since Rats are lack of courage and good command skills, they are not capable as leaders. Ox people are known for diligence, dependability, strength and determination; besides, they regard discipline as their second nature. With eager to advance and great patience, Oxen can achieve their goals by consistent efforts. Oxen. 1. ZODIAC SIGNS AND ASTROLOGY SIGNS MEANINGS AND CHARACTERISTICS, Retrieved October 20 2017, from: http://www.astrology-zodiac-signs.com iZodiac Signs, Retrieved October 20 2017, from: https://www.izodiacsigns.com; Learn about the zodiac signs (Tarot. com), Retrieved October 28 2017, from: https://www.tarot.com/astrology/zodiac; Chinese zodiac wiki, Retrieved November 10 2017, from: https://en.wikipedia.org/wiki/Chinese_zodiac; Chinese zodiac signs and meanings, Retrieved November 10 2017, from: http://www.whats-your-sign.com/Chinese-zodiacsigns.html; The Chinese zodiac, Retrieved June10 2018, from: https://www.chinahighlights.com/travelguide/chinese-zodiac/ 8. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(15) tend to choose to work at where they can apply themselves for a long period of time. Short-term positions and project will make them uncomfortable. However, Oxen has poor communication skills, since they tend to be stubborn and stick to their ways. So while Oxen do not end up in managerial positions, they will eventually do well in their field of choice. Tiger people are brave, cautious, competitive, and self-confident. But they are likely to be irritable, and overindulged. They emphasize their appearance more than others, carefully crafting their looks to match their situation. With great confidence and indomitable fortitude, they can be competent leaders. Tigers prefer to. 政治大 Rabbit people are quiet, 立gentle, responsible but stubborn. Rabbits are faithful to. believe in themselves rather than others.. ‧ 國. 學. those around them, but reluctant to reveal their minds to others. Rabbits can hold steady as long as they have a plan. Therefore, they often occupy positions in life. ‧. where they are in charge of day to day operations, ensuring that projects go smoothly. sit. y. Nat. and plans go off without a hitch. With courage, tenacity and intelligence, Dragon. al. er. io. people are enthusiastic and confident. They are not afraid of challenges, and willing to. v. n. take risks. Dragons are the most honored animal in Chinese culture; hence Chinese. Ch. engchi. i n U. people believe that Dragon people are gifted and powerful. Nevertheless, Dragons are sometimes regarded as aggressive and not open to criticism. Instead of following tradition, they strive for a smooth future. In Chinese culture, the Snake is the most enigmatic animal among the twelve zodiac animals. People born in a year of the Snake are supposed to be the most intuitive. Snake people represent the symbol of wisdom. They are intelligent and wise. They are good at communication but say little. No matter what others say, Snakes will stick to their ways with their unique methods. For people like this, management and leadership positions are the best. When the team is confused, they’ll be able to guide their team members onto the right direction.. 9. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(16) Horse people are extremely active and energetic. With a deft sense of humor, Horses are masters of repartee. They love to take center stage and delight audiences everywhere. They stand for leadership, management and decision-making. Hence, Horses dislike taking orders. Rather than a simple and routine job, Horses are more suitable for technical and effective work since these jobs require quick reflexes and improvisation. Goat people are gentle, mild-mannered, stable and with a strong sense of kindheartedness and justice. Unlike Horses, Goats prefer not to be the center of attention, but to be a part of a group. The most notable trait of Goats is that they are. 政治大 therefore, they shouldn’t try competitive careers such as investment since high 立. able to persevere through any difficulty. Goats love helping and taking care of others;. intensity will throw them into confusion and anxiety. Monkey people are passionate,. ‧ 國. 學. confident and responsible. Monkeys are very intelligent, and they usually become. ‧. great leaders as adults. Due to this, they are sometimes arrogant, competitive and. y. Nat. irritable. Monkeys work very hard in their careers and they can adapt well to different. er. io. sit. working environments. Besides, Monkeys are never satisfied and complacent, and they always want to try something new. This makes them suitable for challenging and. n. al. risky careers.. Ch. engchi. i n U. v. Rooster people are very observant, confident and resourceful. Roosters are always active and popular within a crowd; they tend to be the center of attention and appear attractive. Roosters are very earnest with their work. They have professional skills and courage. They make smart choices, but instead of being more careful with important decisions, they often botch them. Dog people are loyal, honest, but stubborn because Dogs sometimes find it difficult to convey their thoughts to others. Dogs are always ready to help others and they tend to choose a job based on the principle of serving others. Pig people are diligent, compassionate and generous. Pigs have great. 10. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(17) concentration: once they set a goal, they will devote all their energy to achieving it. Pigs are relatively calm when facing trouble. No matter how difficult the problems are Pigs encounter, they can handle things properly and carefully. They have a great sense of responsibility to finish what they are engaged in.. 2.4.. Firm risk-taking behavior and firm policies Prior research studies discuss and analyze numerous indicators and measures of. firm risk-taking behavior. Claessens et al. (2000) use twelve indicators to capture corporate risk-taking, including measures of cash-flow risk, financial leverage, and. 政治大. liquidity. Li et al. (2013) find that culture can affect firms’ appetite for risk and. 立. therefore the riskiness of their corporate decisions. It also indicates that individualism. ‧ 國. 學. is positively and significantly associated with corporate risk-taking. In addition, many studies have found that CEO characteristics are highly correlated with firm. ‧. performance. Mohamed, Jarboui, Baccar, and Bouri (2015) provide evidence that. y. Nat. sit. CEOs’ personal characteristics can greatly affect firm value and policy makers are. al. n. firm value.. er. io. invited to provide the optimal CEOs’ characteristics that can avoid such distortions on. Ch. engchi. i n U. v. Besides, CEOs’ age has an impact on firm risk-taking behavior. Serfling (2014) indicates that risk-taking behavior and stock return volatility decreases as CEOs become older. Younger CEOs prefer high-risk strategies, while older CEOs are more conservative and prefer simpler solutions (Hambrick & Mason, 1984; Pegels & Yang, 2000; Barker & Mueller, 2002). Dechow & Sloan (1991), Barker & Muller (2002), and Matta & Beamish (2008) all find that CEOs close to retirement have higher degree of risk aversion and will reduce R&D investment. By contrast, younger CEOs, without consideration for position and financial gains, have broader long-term horizon and higher risk tolerance, and therefore tend to increase R&D investment. Ryan & 11. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(18) Wiggins (2002) indicated that corporate R&D investment and CEO age present a nonlinear inverted U-shaped correlation. Still, Goel & Thakor (2008), Kini & Williams (2012) point out there is a significantly positive relation between firm risk and tournament incentives. Other relevant studies have found that tenure also has effect on CEO’s willingness to take risks (Barker & Mueller, 2002). CEOs with shorter tenure will focus on short-term performance and be unwilling to invest in research and development costs with higher risks (Jensen & Meckling, 1979; Hirshleifer, 1993) On. 政治大 refuse changes and risky investments because they prefer to maintain status quo; thus 立 the contrary, CEOs with long tenures tend to neglect environmental changes and. causing an obstacle for company growth (Miller, 1991; Hambrick &Fukutomi, 1991;. ‧ 國. 學. Thomas, Litschert & Ramaswamy, 1991; Hambrick, Geletkanycz & Fredirckson,. ‧. 1993; Barker & Mueller, 2002).. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 12. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(19) 3. Hypotheses Development CEO, as the main decision-maker in a firm, integrates views of the top management team, requiring high cognitive complexity(Calori, Johnson, & Sarnin, 1994 ). CEO’ s managerial characteristics affect not only firm’s strategies and investment decisions but also risk-taking behaviors and firm performance (Malmendier & Tate, 2005, Coel & Thakor, 2008, Colbert, Barrick & Bradley, 2013). Prior literature has provided considerable evidence that CEO demographic characteristics may influence a firm’s risk-taking behavior, such as gender, age and. 政治大 take risks than the female. 立 Brookman and Thistle (2009) find that CEO tenure is. tenure. Byrnes, Miller and Schafer (1999) find out that the male are more likely to. ‧ 國. 學. related to firm value. Serfling (2014) points out that older CEOs decrease corporate risk with low-risk investment portfolios; however, the younger CEOs are on the. ‧. contrary. Dechow & Sloan (1991), Barker & Muller (2002), and Matta & Beamish. sit. y. Nat. (2008) all find that CEOs close to retirement have higher degree of risk aversion and. al. er. io. will reduce R&D investment. Besides, CEO psychological traits are also influential to. v. n. firm’s risk-taking behavior. For instance, overconfidence is a strong factor from. Ch. engchi. i n U. previous studies. Overconfident CEOs are likely to overestimate the return on investment plans and be overinvest in a optimism bias (Malmendier & Tate, 2008). In summary, CEOs’ personal traits have a great impact on a firm’s risk-taking behavior. While studying the relation between CEO characteristics and firm risk-taking behavior, few studies examine whether CEO’s constellation or Chinese zodiac has an impact on risk taking. According to Chinese and western astrology studies, we can predict one’s inborn personal characteristics through his/her Chinese zodiac and constellation and therefore arrive at his/her risk-taking behavior and decision. On the basis of prior studies, overconfident CEOs are likely to be overinvest in a optimism. 13. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(20) bias and vice versa. CEOs with conservative personality tend to have higher degree of risk aversion compared to others and they are more unwillingly to make risky strategies . In light of Chinese zodiacs, rabbits and dogs are more conservative and stubborn. Snakes, tigers and pigs prefer believing in themselves. Rather than taking others’ advice, they have their own and unique strategies. In summary, above five Chinese zodiacs (Tiger, Rabbit, Snake, Dog and Pig) are more conservative compared to others and according to their inborn traits, it is likely to believe they tend to dislike corporate risk and tend to have more conservative firm policies. When it comes to. 政治大 constellation is more conservative based on the traits of each element. Earth-sign 立 constellation, I use the same method as Chinese zodiac to determine which. people tend to be more practical and stable. On the contrary, fire-sign people are. ‧ 國. 學. confident, uncontrollable and active, and they may also be bold and brave. Water-sign. ‧. people are sensitive, intuitive and emotional. They tend to stick to their own ways and. y. Nat. thoughts. As for air-sign people, compared to water-sign people, they are more. er. io. sit. rational and social, and they tend to act with their intellect. Compared to fire-sign people and air-sign people, earth-sign people and water-sign people like to believe in. al. n. v i n CIhbelieve they are more themselves and act carefully, and conservative compared to engchi U the others. On the basis of their inborn traits, it is likely to believe earth-sign CEOs and water-sign CEOs tend to dislike corporate risk and tend to have more conservative firm policies.. On the basis of prior research, I examine risk-taking behavior and firm policies by several elements. Kini and Williams (2012) rely on return volatility, R&D intensity, capital expenditure intensity, firm focus and book leverage to measure firm risk-taking behavior and firm policies. Based on prior literature, it is still unclear whether CEO’s Chinese zodiac and constellation positively affect firm risk-taking behavior. If the empirical result of this paper is positive, it would benefit the firm 14. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(21) while hiring CEO or high-level executives since they could recruit their suitable and ideal CEO on basis of CEO’s constellation and Chinese zodiac and the firm’s risk appetite. In accordance to the matching theory, CEOs who do not match their firms will leave, while those who match their firms will stay (Allgood & Farrell, 2003) Therefore, I expect that CEO’s constellation and Chinese zodiac have an impact on firm risk-taking behavior. Moreover, I expect that CEOs with more conservative Chinese zodiacs tend to have higher degree of risk aversion; likewise, CEOs with more conservative constellation have higher degree of risk aversion. This leads to the. 政治大 H1: CEOs with conservative Chinese zodiacs are negatively associated with 立. first and second hypotheses.. corporate risk.. ‧ 國. 學. H2: CEOs with conservative constellation are negatively associated with. ‧. corporate risk.. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 15. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(22) 4. Data and Research Design 4.1.. Sample selection I use BoardEx database to obtain CEO data and corporate governance data of. companies in the US during the period 2000 to 2014. The CEO data I collect from BoardEx includes those who are both CEO and director in the firm. I exclude financial institutions due to their uniqueness in firm operation and the regulation in law. BoardEx is a business intelligence database and consolidates public domain information concerning the board of directors and senior management of publicly. 政治大 initial sample consists of BoardEx 立 firms having zodiac sign data or constellation data. quoted and large private companies. Table 1 illustrates the selection of final data. The. ‧ 國. 學. of CEO between 2000 and 2014. Financial firms (SIC 6000-6999) are excluded from the sample. I obtain financial statement data from the Compustat files and stock return. ‧. data from the Center for Research in Security Prices (CRSP) files. The initial sample. sit. y. Nat. of Chinese zodiac comprises 10,799 observations and the initial sample of. al. er. io. constellation comprises 2,076 observations. After merging the databases and. v. n. removing observations with missing values, the final sample of Chinese zodiac. Ch. engchi. i n U. comprises 6,768 observations from 1,302 firms and the final sample of constellation comprises 1,452 observations from 341 unique firms.. Table 1 Sample Selection Panel A: Sample of Chinese zodiac Description. Observations. BoardEx with Zodiac Sign Deduct: Missing Accounting/Financial Data Missing Corporate Governance Data Final Chinese zodiac data. 10,799 (4,021) (10) 6,768. 16. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(23) Panel B: Sample of Constellation Description. Observation. BoardEx with Constellation Deduct: Missing Accounting/Financial Data Missing Corporate Governance Data Final Constellation Data. 2,076 (624) (0) 1,452. The sample distribution based on Chinese zodiac from 2000 to 2014 is reported in Panel A of Table 2 and the sample distribution based on constellation from 2000 to. 政治大 Chinese zodiac samples is 立 even; however, the distribution of constellation samples is 2014 is reported in Panel B of Table 2. According to Table 2, the distribution of. ‧ 國. 學. not as even as the former. Based on Panel B of Table 2, the percentage of Leo and Libra is higher than others, which means that there are more Leo CEOs and Libra. ‧. CEOs compared to the others in my sample.. Nat. sit. y. Table 2 Sample Distribution based on Chinese zodiac or Constellation. n. al. er. io. Panel A Sample Distribution based on Chinese zodiac Rat Ox Tiger Rabbit Dragon Snake Horse Goat Monkey Rooster Dog Pig. Ch. i n 0.085 U. v. Percentage. e n g c h i 0.083 0.081 0.082 0.074 0.080 0.095 0.074 0.095 0.093 0.081 0.077. (Continued on next page). 17. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(24) Panel B Sample Distribution based on Constellation Percentage Taurus Virgo Capricorn Cancer Scorpio Pisces Aries Leo Sagittarius Gemini Libra Aquarius. 立. 0.085 0.092 0.091 0.079 0.084 0.095 0.060 0.144 0.071 0.041 0.100 0.058. 政治大. ‧ 國. 學. The descriptive statistics of Chinese zodiac variables, constellation variables, and industry characteristics are provided in Section 5, Table 3. All variables are. ‧. winsorized at their 1% and 99% values, except for dummy indicators.. n. al. er. io. Risk Proxies and Firm Policies. sit. y. Nat. 4.2.. i n U. v. I use one-year return volatility and seasonally differenced quarterly accounting. Ch. engchi. return on assets (ROA volatility) as proxies for firm risk (Risk). Return volatility (SD_RET) is defined as the standard deviation of monthly CRSP stock returns over the 12 months starting from the fifth month after fiscal year-end. Because return volatility is affected by factors other than CEO’s personal traits to influence stock market volatility, such as unexpected revisions in discount rates and forecasts of future cash flows, we consider the volatility of accounting rates of return (e.g., Beaver, Kettler, and Scholes 1970 ) as an alternative measure. The ROA volatility is the standard deviation of seasonally differenced quarterly accounting return on assets over the subsequent three years for the period 2000-2016, where 18. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(25) return on assets is income before extraordinary items, scaled by average total assets for the quarter (!"_∆%&'). !"_∆%&' is the degree of risk-taking in firms’ operations measured by volatility of corporate earnings. Riskier corporate operations lead to more volatile earnings (John et al., 2008; Zhang, 2009). I also examine whether firm policies (Policy), specifically R&D intensity, capital expenditures intensity, are influenced by CEOs’ Chinese Zodiac or Constellation. Prior studies usually used R&D intensity as a measure of risky corporate policies (Bhagat and Welch, 1995; Coles et al.,2006; and Bargeron et al.,. 政治大 positively related to R&D levels. R&D intensity is constructed as annual research and 立 2010). Many studies confirm that the incentive to increase risk is expected to be. ‧ 國. 學. development expense scaled by total assets in year t+1 (%"*+, ). Unlike R&D studies’ results, arguments about the relation between capital expenditure (CAPEX) and the. ‧. incentive to increase risk have been made for both sides, positive and negative. Both. y. Nat. positive (Guay, 1999; Bargeron, Lehn, and Zutter, 2010; Cohen, Dey, and Lys, 2013). er. io. sit. and negative (Coles et al., 2006; Hayes et al., 2012) links have been observed for CAPEX. Capital intensity is constructed as annual capital expenditures scaled by total. n. al. assets in year t+1 (-'./0*+, ).. Ch. engchi. i n U. v. 19. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(26) 4.3.. Control Variables The control variables that we use as determinants of the firm characteristics and. corporate governance characteristics are all based on existing literature. Size (Size) of the firm is proxied by natural log of the book value of assets. Historical growth of the firm size is proxied by sales growth (Growth). Profitability is measured by ROA, which is computed as income before extraordinary items scaled by average total assets. I consider leverage (LEV) and free cash flows (CF) as measures of financial distress and capital availability, respectively. I define LEV as book value. 政治大 debt + long term debt)/total assets), and CF as cash flow from operations minus cash 立 based leverage ratio ((short term interest bearing debt + current portion of long term. flow from investing and cash dividends scaled by total assets (Core and Guay 2001).. ‧ 國. 學. Moreover, I control for the company’s stock return (RET) over year t and set loss. ‧. dummy (LOSS) as an indicator variable equal to one if the firm has loss.. y. Nat. As for corporate governance characteristics, I use the number of directors of the. er. io. sit. board as the measure of board size (BOD). I consider the percentage of independent directors of the board (INDEP%) and set a dummy if has the same person who is CEO. al. n. v i n and has the chairman position ofCthe board to control the h e n g c h i U corporate governance. (SAME). Last, as mentioned in previous section, CEO age can also affect CEO’s risk preference. CEO age (AGE) is measured by taking natural log of age. In all of the regressions, I include industry (two-digit SIC code) fixed effects, year dummies due to the concern that unobserved heterogeneity attributable to time invariant characteristics of the industry, and time specific shocks may drive the results. To control for firms-specific factors, I also further include firm fixed effects.. 20. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(27) 4.4.. Empirical Model My empirical models of firm risk and firm policies are as follows: %1234*+, = 67 + 6, 9:;_-:<2=>?4* + 6@ !1A=4* + 6B %&'4* + 6C D/E4* + 6F -G4* + 6H I>:JKℎ4* + 6M D:224* + 6N %/O4* + 6P QR"/.%4* + 6,7 !'T/4* + 6,, U&"4* + 6,@ 'I/4* + Q<;V2K>W X1Y=; =XX=ZK + [=\> X1Y=; =XX=ZK + ]4* %1234*+, = 67 + 6, 9:;_-:<2=>?4* + 6@ !1A=4* + 6B %&'4* + 6C D/E4* + 6F -G4* + 6H I>:JKℎ4* + 6M D:224* + 6N %/O4* + 6P QR"/.%4* + 6,7 !'T/4* + 6,, U&"4* + 6,@ 'I/4* + Q<;V2K>W X1Y=; =XX=ZK + [=\> X1Y=; =XX=ZK + G1>^ X1Y=; =XX=ZK + ]4*. 立. (1). (2). 政治大. ‧. ‧ 國. 學. .:_1ZW4*+, = 67 + 6, 9:;_-:<2=>?4* + 6@ !1A=4* + 6B %&'4* + 6C D/E4* + 6F -G4* + 6H I>:JKℎ4* + 6M D:224* + 6N %/O4* + 6P QR"/.%4* + 6,7 !'T/4* (3) + 6,, U&"4* + 6,@ 'I/4* + Q<;V2K>W X1Y=; =XX=ZK + [=\> X1Y=; =XX=ZK + ]4*. y. Nat. sit. n. al. er. io. .:_1ZW4*+, = 67 + 6, 9:;_-:<2=>?4* + 6@ !1A=4* + 6B %&'4* + 6C D/E4* + 6F -G4* + 6H I>:JKℎ4* + 6M D:224* + 6N %/O4* + 6P QR"/.%4* + 6,7 !'T/4* (4) + 6,, U&"4* + 6,@ 'I/4* + Q<;V2K>W X1Y=; =XX=ZK + [=\> X1Y=; =XX=ZK + G1>^ X1Y=; =XX=ZK + ]4*. Ch. engchi. i n U. v. The first four equations are for the test of H1, where the dependent variables are risk taking behavior and firm policies. Risk represents the two primary measures of corporate risk, SD_∆ROA and SD_RET. Policy represents the two primary measures of corporate policies, %"*+, and -'./0*+, . 9:;_-:<2=>?4* is the dummy variable measured as an indicator variable equal to one if CEO’s Chinese zodiac is “Tiger”, “Rabbit”, “Snake”, “Dog”, or “Pig” of the firm i in fiscal year t . According to previous section and Appendix 2, these five zodiacs are rather stubborn, cautious and conservative. Compared to other Chinese zodiacs, people who are born in years. 21. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(28) belonging to those five animals ought to be more conservative. In order to control for industry and macro-economic factors, I include industry fixed effects (two-digit SIC) and year fixed effects. Equation 1 and 3 include industry fixed effect and year fixed effect and firm fixed effect. Besides, to control firm factors, equation 2 and 4 include firm fixed effect in addition to industry fixed effects (two-digit SIC) and year fixed effects. %1234*+, = 67 + 6, -_-:<2=>?4* + 6@ %&'4* + 6B D/E4* + 6C -G4* + 6F I>:JKℎ4* (5) + 6H D:224* + 6M %/O4* + 6N QR"/.%4* + 6P !'T/4* + 6,7 U&"4* + 6,, 'I/4* + Q<;V2K>W X1Y=; =XX=ZK + [=\> X1Y=; =XX=ZK + ]4* %1234*+,. 政治大 = 6 + 6 -_-:<2=>? + 6 %&' + 6 D/E + 6 -G + 6 I>:JKℎ + 6 D:22立 + 6 %/O + 6 QR"/.% + 6 !'T/ + 6 U&" 7. ,. 4*. H. 4*. M. @. 4*. 4*. N. B. 4*. C. 4*. 4*. P. F. 4*. ,7. 4*. 4*. (6). ‧. ‧ 國. 學. + 6,, 'I/4* + [=\> X1Y=; =XX=ZK + G1>^ X1Y=; =XX=ZK + G1>^ X1Y=; =XX=ZK + ]4*. er. io. sit. y. Nat. .:_1ZW4*+, = 67 + 6, -_-:<2=>?4* + 6@ %&'4* + 6B D/E4* + 6C -G4* + 6F I>:JKℎ4* + 6H D:224* + 6M %/O4* + 6N QR"/.%4* + 6P !'T/4* + 6,7 U&"4* (7) + 6,, 'I/4* + Q<;V2K>W X1Y=; =XX=ZK + [=\> X1Y=; =XX=ZK + ]4* .:_1ZW4*+, = 67 + 6, -_-:<2=>?4* + 6@ %&'4* + 6B D/E4* + 6C -G4* + 6F I>:JKℎ4* + 6H D:224* + 6M %/O4* + 6N QR"/.%4* + 6P !'T/4* + 6,7 U&"4* + 6,, 'I/4* + [=\> X1Y=; =XX=ZK + G1>^ X1Y=; =XX=ZK + G1>^ X1Y=; =XX=ZK + ]4*. n. al. Ch. engchi. i n U. v. (8). Equations (5) to (8) are for the test of H2. As aforementioned, Risk represents the two primary measures of corporate risk, SD_∆ROA and SD_RET. Policy represents the two primary measures of corporate policies, %"*+, and -'./0*+, . -_-:<2=>?4* is the dummy variable measured as an indicator variable equal to one if CEO’s constellation belongs to earth element and water element of the firm i in fiscal year t. Based on section 3 and Appendix 3, people whose constellation signs belonging to earth element and water element are more conservative and careful, and 22. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(29) rather believe in themselves compared to others. In order to control for industry and macro-economic factors, I include industry fixed effects (two-digit SIC) and year fixed effects. Equation 5 and 7 include industry fixed effect and year fixed effect and firm fixed effect. Besides, to control firm factors, equation 6 and 8 include firm fixed effect in addition to industry fixed effects (two-digit SIC) and year fixed effects.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 23. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(30) 5. Empirical Results 5.1.. Descriptive statistics The descriptive statistics of the main variables are reported in Table 3. Panel A. reports that the mean value of the primary Chinese Zodiac variable (Zod_Conserv) is 0.400 and the total observation is 6,768, which means that the number of CEOs whose Chinese Zodiac belongs to the conservative Chinese Zodiacs are 2,707. Panel B of Table 3 reports that the mean value of primary constellation variable (C_Conserv) is 0.524 and the total observation is 1,452, which means that the number of CEOs whose. 政治大. constellation signs belong to the conservative constellation signs are 761.. 立. Turning to the primary risk measures, the mean (median) value for the standard. ‧ 國. 學. deviation of stock returns from Panel A of Table 3 is 0.068 (0.090) and from Panel B of Table 3 is 0.101 (0.082). The mean (median) value for the standard deviation of. ‧. seasonally-differenced quarterly ROA from Panel A of Table 3 is 0.024 (0.010) and. y. Nat. sit. from Panel B of Table 3 is 0.023 (0.009). Regarding the firm policies, the mean. n. al. er. io. (median) value for the R&D in year t+1 from Panel A of Table 3 is 0.030 (0.000) and. i n U. v. from Panel B of Table 3 is 0.029 (0.001). The mean (median) value for the CAPEX in. Ch. engchi. year t+1 from Panel A of Table 3 is 0.049 (0.036) and Panel B of Table 3 is 0.051 (0.039).. 24. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(31) Table 3 Descriptive Statistics Panel A: Sample of Chinese zodiac (Obs.= 6,768) Variable. Mean. Stdev. Min. Q1. Median. Q3. Max. SD_RET SD_∆ROA R&D CAPEX Zod_Conserv SIZE ROA LEV CF Growth LOSS RET INDEP% SAME BOD. 0.107 0.024 0.030 0.049 0.400 7.086 0.033 0.525 0.139 0.042 0.193 0.129 0.813 0.008 9.214. 0.068 0.044 0.064 0.046 0.49 2.023 0.113 0.223 0.157 0.221 0.395 0.514 0.106 0.086 2.519. 0.02 0.000 0.000 0.000 0.000 1.975 -0.694 0.047 -0.633 -2.233 0 -0.903 0.4 0 2. 0.063 0.005 0.000 0.019 0.000 5.706 0.015 0.371 0.074 -0.011 0 -0.151 0.75 0 7. 0.09 0.01 0.000 0.036 0.000 7.207 0.046 0.533 0.137 0.055 0 0.071 0.857 0 9. 0.129 0.023 0.028 0.064 1 8.437 0.082 0.676 0.218 0.134 0 0.298 0.889 0 11. 0.744 0.711 0.489 0.316 1. 0.135. 3.584. 3.951. 4.043. 4.127. 4.401. Panel B: Sample of Constellation. 0.674 1.223 0.608 0.754 1 6.234 0.941 1 22. Stdev. Min. Q1. Median. Q3. Max. 0.101 0.023 0.029 0.051 0.524 8.303 0.043 0.564 0.154 0.004 0.165 0.116 0.83 0.008. 0.067 0.053 0.058 0.042 0.5 1.951 0.107 0.22 0.158 1.102 0.371 0.454 0.097 0.091 2.731 0.125. 0.018 0.000 0.000 0.000 0.000 2.116 -1.613 0.074 -0.94 -40.6 0 -0.881 0.25 0 3 3.611. 0.061 0.005 0.000 0.023 0.000 7.22 0.02 0.426 0.086 -0.006 0 -0.137 0.786 0 9 3.97. 0.082 0.009 0.001 0.039 1 8.322 0.052 0.579 0.15 0.046 0 0.072 0.857 0 10 4.06. 0.12 0.02 0.035 0.068 1 9.753 0.085 0.701 0.232 0.122 0 0.29 0.9 0 12 4.127. 0.869 0.985 0.9 0.365 1. 10.478. 4.048. Ch. engchi. er. n. al. sit. Mean. io. SD_RET SD_∆ROA R&D CAPEX C_Conserv SIZE ROA LEV CF Growth LOSS RET INDEP% SAME BOD AGE. Nat. Variable. (Obs.= 1,452). y. 4.033. ‧. ‧ 國. 立. 學. AGE. 政治大. 12.509. i n U. v. 12.627. 0.468 1.605 0.912 0.952 1 5.813 0.938 1 22 4.394. 25. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(32) Notes: Financial data from 2000 to 2016 are from Compustat and stock return data from 2000 to 2016 are from CRSP. CEO data and corporate governance data from 2000 to 2014 are from BoardEx. SD_RET is the standard deviation of monthly CRSP stock returns over the 12 months starting from the fifth month after fiscal yearend. SD_∆ROA is the standard deviation of seasonally differenced quarterly accounting return on assets over the subsequent three years for the period 2000-2016. R&D is research and development expense scaled by total assets in year t+1. CAPEX is capital expenditures scaled by total assets in year t+1. Zod_Conserv is a dummy variable measured as an indicator variable equal to one if CEO’s Chinese zodiac is “Tiger”, “Rabbit”, “Snake”, “Dog”, or “Pig”. C_Conserv is the dummy variable measured as an indicator variable equal to one if CEO’s constellation belongs to earth element and water element.. 政治大. Table 4 reports the Pearson correlations between pairs of the main variables of. 立. interest. There are negative correlations both between conservative Chinese zodiacs. ‧ 國. 學. and my two measures of risk-taking behavior, and between conservative Chinese zodiacs and firm policies. This suggests that people whose zodiac signs belong to. ‧. conservative Chinese zodiacs choose to reduce firm risk and make less risky firm. y. Nat. sit. policies. In addition, there are negative correlations between conservative. n. al. er. io. constellation and both measures of firm risk, and between conservative constellation. i n U. v. and R&D intensity. This suggests that people whose constellation signs belong to the. Ch. engchi. conservative constellation choose to reduce firm risk and make less investment in R&D expenditure. As for the correlation between conservative Chinese zodiacs and control variables, and between conservative constellation and control variables, it is as follows. There is a negative correlation between conservative Chinese zodiac and corporate governance variables (INDEP%, SAME, and BOD), and a positive correlation between conservative Chinese zodiac and CEO age. In summary, this suggests that CEOs with conservative Chinese zodiacs care less about corporate governance and their age tends to be above average. There is a negative correlation 26. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(33) between conservative constellation and two of the corporate governance variables (SAME and BOD), but a positive correlation between conservative constellation and the percentage of independent directors of the board (INDEP%). Besides, the correlation between conservative constellation and CEO age is negative. In summary, this suggests that CEOs with conservative constellation care about the independence of the board but don’t put much attention on other corporate governance issues, and their age tends to be under average. Table 4 Correlation Analyses. 政治 C_Conserv 大 -0.004 0.004. Zod_Conserv. n. engchi. y. sit. io. Ch. -0.008 -0.005 0.002 0.02 0.072*** -0.046* 0.01 0.041 -0.064** -0.018 0.056** -0.004 -0.035 -0.079***. er. Nat. al. -0.017 -0.029** -0.008 -0.046*** 0.024** -0.025** -0.013 -0.001 -0.030** -0.011 -0.066*** -0.005 -0.026** 0.008. ‧. ‧ 國. 立. 學. SD_RET SD_∆ROA R&D CAPEX SIZE ROA LEV CF Growth LOSS RET INDEP% SAME BOD AGE. i n U. v. Notes: Financial data from 2000 to 2016 are from Compustat and stock return data from 2000 to 2016 are from CRSP. CEO data and corporate governance data from 2000 to 2014 are from BoardEx. SD_RET is the standard deviation of monthly CRSP stock returns over the 12 months starting from the fifth month after fiscal yearend. SD_∆ROA is the standard deviation of seasonally differenced quarterly accounting return on assets over the subsequent three years for the period 2000-2016. R&D is research and development expense scaled by total assets in year t+1. CAPEX is capital expenditures scaled by total assets in year t+1. Zod_Conserv is is the dummy variable measured as an indicator variable equal to one if CEO’s Chinese zodiac is “Tiger”, “Rabbit”, “Snake”, “Dog”, or “Pig”. C_Conserv is the dummy 27. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(34) variable measured as an indicator variable equal to one if CEO’s constellation belongs to earth element and water element. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively, using a two-tailed test.. 5.2.. Chinese zodiac, firm risk-taking behavior and firm policies In this section, I examine the extent to which conservative Chinese zodiac. influences risk-taking behavior and firm policies. As stated earlier, I expect that those CEOs with conservative Chinese zodiac will prefer less firm risk, where the coefficients of SD_RET and SD_∆ROA are both negative, and more conservative firm. 政治大 Table 5 reports both the relationship between conservative Chinese zodiac and 立. policies, which are lower R&D and higher CAPEX.. corporate risk, and between conservative Chinese zodiac and firm policies. Panel A of. ‧ 國. 學. Table 5 reports the relationship between conservative Chinese zodiac and corporate. ‧. risk. Panel B of Table 5 reports the relationship between conservative Chinese zodiac. y. Nat. and firm policies. Panel A of Table 5 indicates that the coefficient of Zod_Conserv is. er. io. sit. -0.002 when the dependent variable is ROA volatility (SD_∆ROA) and the result is significant at 10% level under both conditions of controlling only industry fixed effect. al. n. v i n C hand firm fixed effect. and controlling industry fixed effect Zod_Conserv is -0.002 engchi U. when the dependent variable is R&D in year t+1 and the result is significant at 1% level under the condition of controlling industry fixed effect and firm fixed effect. However, Zod_Conserv does not present significant influence on capital expenditure. (CAPEX) and return volatility (SD_RET). Consistent to my expectation, the result confirms that CEO who belongs to more conservative Chinese zodiac, which is “Tiger”, “Rabbit”, “Snake”, “Dog”, or “Pig”, prefers less risk and more conservative firm policies.. 28. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(35) Table 5. Regressions of Conservative Chinese Zodiac and Corporate risk and Firm Policies Panel A: The relationship between conservative Chinese zodiac and corporate risk Dependent Variable: Corporate risk measures SD_RET. LEV. (4). -0.002 (0.154) -0.007*** (0.000) -0.112*** (0.000) 0.024*** (0.000) 0.021*** (0.000) 0.004 (0.251) 0.028*** (0.000) 0.015*** (0.000) -0.031*** (0.000) -0.013* (0.079) -0.002*** (0.000) -0.017*** (0.000) 0.295*** (0.000) YES YES. -0.003 (0.261) 0.004** (0.032) -0.094*** (0.000) 0.032*** (0.000) 0.012** (0.039) -0.003 (0.338) 0.009*** (0.000) 0.018*** (0.000) -0.011 (0.374) -0.009 (0.171) -0.001 (0.150) -0.008 (0.276) 0.335*** (0.000) YES YES YES. -0.002* (0.062) -0.003*** (0.000) -0.092*** (0.000) -0.009*** (0.001) -0.015*** (0.000) 0.008*** (0.001) 0.009*** (0.000) -0.007*** (0.000) 0.004 (0.426) -0.006 (0.238) -0.000* (0.068) -0.009** (0.011) 0.105*** (0.000) YES YES. -0.002* (0.064) 0.001 (0.224) -0.028*** (0.000) -0.010*** (0.008) -0.004 (0.199) -0.002 (0.251) 0.000 (0.832) -0.002*** (0.007) -0.006 (0.388) -0.009** (0.012) 0.000 (0.625) -0.003 (0.410) 0.178*** (0.000) YES YES YES. 6768 0.568. 6768 0.239. 立. Growth. n. al. SAME BOD AGE constant Industry FE Time FE Firm FE N Adj. R2. Ch. 6768 0.417. engchi. ‧. io. INDEP%. Nat. RET. 政治大. 學. ‧ 國. CF. LOSS. (3). y. ROA. (2). sit. SIZE. (1). er. Zod_Conserv. SD_∆ROA. i n U. v. 6768 0.715. (Continued on next page) 29. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(36) Panel B: The relationship between conservative Chinese zodiac and firm policies Dependent Variable: Firm policy measures R&D. LEV CF. -0.002*** (0.007) -0.006*** (0.000) -0.043*** (0.000) -0.004 (0.124) -0.006*** (0.007) 0.004*** (0.002) 0.000 (0.704) -0.003*** (0.000) -0.003 (0.486) 0.000 (0.955) 0.000 (0.603) -0.004 (0.179) 0.067*** (0.003) YES YES YES. 0.000 (0.686) -0.002*** (0.000) -0.012** (0.036) 0.004* (0.059) 0.070*** (0.000) 0.005** (0.017) -0.003* (0.051) -0.002** (0.050) -0.007 (0.142) -0.006 (0.242) 0.001*** (0.003) -0.007** (0.023) 0.116*** (0.000) YES YES. 0.000 (0.837) -0.007*** (0.000) 0.016*** (0.007) -0.017*** (0.000) 0.018*** (0.000) 0.005*** (0.002) -0.002 (0.120) -0.003*** (0.000) 0.001 (0.881) -0.002 (0.524) 0.000 (0.280) -0.005 (0.239) 0.340*** (0.000) YES YES YES. 6768 0.923. 6768 0.441. io. SAME. n. al. BOD AGE constant Industry FE Time FE Firm FE N Adj. R2. Ch. 6768 0.483. engchi. ‧. ‧ 國. 政治大. Nat. INDEP%. -0.002 (0.199) -0.002*** (0.000) -0.221*** (0.000) -0.035*** (0.000) -0.019*** (0.000) 0.019*** (0.000) 0.001 (0.717) -0.008*** (0.000) 0.020*** (0.001) -0.001 (0.844) -0.001* (0.061) -0.038*** (0.000) 0.178*** (0.000) YES YES. 學. RET. (4). 立. Growth LOSS. (3). y. ROA. (2). sit. SIZE. (1). er. Zod_Conserv. CAPEX. i n U. v. 6768 0.719. (Continued on next page). 30. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(37) Notes: SD_RET is the standard deviation of monthly CRSP stock returns over the 12 months starting from the fifth month after fiscal year-end. SD_∆ROA is the standard deviation of seasonally differenced quarterly accounting return on assets over the subsequent three years for the period 2000-2016. R&D is research and development expense scaled by total assets in year t+1. CAPEX is capital expenditures scaled by total assets in year t+1. Zod_Conserv is the dummy variable measured as an indicator variable equal to one if CEO’s Chinese zodiac is “Tiger”, “Rabbit”, “Snake”, “Dog”, or “Pig”. Numbers in parentheses are p-values. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively, using a two-tailed test.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 31. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(38) 5.3.. Constellation, firm risk-taking behavior and firm policies In this section, similar to the previous section, I examine the extent to which. conservative Constellation influences firm risk-taking behavior and firm policies. As stated earlier, I expect that those CEOs with conservative constellation signs will prefer less firm risk, where the coefficients of SD_RET and SD_∆ROA are both negative, and more conservative firm policies, which are lower R&D and higher CAPEX. Table 6 reports both the relationship between conservative constellation and. 政治大 Table 6 reports the relationship 立 between conservative constellation and corporate risk. corporate risk, and between conservative constellation and firm policies. Panel A of. ‧ 國. 學. Panel B of Table 6 reports the relationship between conservative constellation and firm policies. Panel A of Table 6 indicates that C_Conserv does not present. ‧. significant influence on SD_RET and SD_∆ROA under the condition of controlling. sit. y. Nat. industry fixed effect. However, C_Conserv has positive influence on SD_RET. al. er. io. (return volatility) under the condition of controlling industry fixed effect and firm. v. n. fixed effect. The possible and reasonable explanation of this result may be as follows:. Ch. engchi. i n U. Controlling firm fixed effect means that the firm makes changes to its CEO and the change of CEO may increase risk under the perception of the market. C_Conserv is 0.006 when the dependent variable is R&D in year t+1 and the result is significant at 5% level under the condition of controlling industry fixed effect. C_Conserv does not present significant influence on capital expenditure (CAPEX). There is no evidence showing that CEOs with conservative constellation signs, which belong to earth element and water element, prefers less risk. However, there evidence indicating that CEOs with conservative constellation signs, which belong to. 32. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(39) earth element and water element, prefer more conservative firm policies, which is consistent with my expectation. The empirical evidence presented in Table 5 and 6 is partially consistent with the predictions of H1 and H2. CEOs with more conservative Chinese zodiacs are negatively associated with corporate risk. In addition, CEOs with more conservative constellation signs only prefer more conservative firm policies.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 33. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(40) Table 6. Regressions of Conservative Constellation and Corporate Risk and Firm Policies Panel A: The relationship between conservative constellation and corporate risk Dependent Variable: Corporate risk measures SD_RET. LEV. (4). 0.002 (0.496) -0.008*** (0.000) -0.114*** (0.000) 0.014* (0.06) 0.025** (0.013) 0.001 (0.474) 0.030*** (0.000) -0.004 (0.267) -0.024 (0.142) -0.019 (0.206) -0.003*** (0.000) -0.003 (0.797) 0.254*** (0.000) YES YES. 0.016** (0.01) 0.006 (0.224) -0.063** (0.018) 0.022 (0.157) 0.022* (0.065) 0.002 (0.735) 0.018*** (0.001) 0.003 (0.451) 0.012 (0.690) -0.018 (0.222) -0.004*** (0.000) 0.104*** (0.000) -0.330*** (0.005) YES YES YES. 0.001 (0.672) -0.003*** (0.000) -0.115*** (0.000) -0.005 (0.494) 0.016* (0.096) -0.002* (0.096) 0.007 (0.102) -0.006* (0.076) 0.017 (0.265) -0.001 (0.927) -0.001* (0.054) -0.001 (0.936) 0.082* (0.100) YES YES. 0.002 (0.587) -0.001 (0.576) 0.002 (0.853) -0.013* (0.062) 0.003 (0.530) -0.006** (0.013) 0.004 (0.149) -0.003 (0.103) -0.012 (0.380) 0.000 (0.950) -0.002*** (0.002) 0.020* (0.084) 0.031 (0.559) YES YES YES. 1452 0.556. 1452 0.224. 立. Growth. io. INDEP%. n. al. SAME BOD AGE constant Industry FE Time FE Firm FE N Adj. R2. Ch. 1452 0.453. engchi. ‧. Nat. RET. 政治大. 學. ‧ 國. CF. LOSS. (3). y. ROA. (2). sit. SIZE. (1). er. C_Conserv. SD_∆ROA. i n U. v. 1452 0.851. (Continued on next page) 34. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(41) Panel B: The relationship between conservative constellation and firm policies Dependent Variable: Firm policy measures R&D. LEV CF. -0.006** (0.019) 0.000 (0.776) -0.073*** (0.000) -0.007 (0.301) 0.016* (0.051) -0.017*** (0.000) 0.015*** (0.000) -0.006* (0.051) 0.047*** (0.001) 0.013 (0.295) -0.002*** (0.000) -0.047*** (0.000) 0.177*** (0.000) YES YES. -0.001 (0.596) -0.012*** (0.000) 0.024** (0.010) 0.002 (0.772) 0.004 (0.299) 0.001 (0.593) 0.007*** (0.000) 0.001 (0.367) -0.009 (0.365) 0.005 (0.363) 0.000 (0.760) -0.003 (0.751) 0.090** (0.026) YES YES YES. -0.003 (0.165) -0.002*** (0.001) 0.000 (0.981) -0.004 (0.445) 0.063*** (0.000) 0.000 (0.768) -0.005 (0.124) -0.002 (0.335) -0.017 (0.117) 0.003 (0.732) 0.000 (0.914) -0.014* (0.076) 0.183*** 0.000 YES YES. 0.002 (0.480) -0.006** (0.013) 0.035** (0.011) -0.009 (0.230) 0.020*** (0.001) 0.003 (0.224) -0.003 (0.227) -0.002 (0.260) 0.014 (0.371) -0.006 (0.392) -0.001* (0.081) 0.02 (0.122) -0.017 (0.776) YES YES YES. 1452 0.929. 1452 0.399. ‧ 國. io. SAME. n. al. BOD AGE constant Industry FE Time FE Firm FE N Adj. R2. Ch. 1452 0.481. engchi. ‧. Nat. INDEP%. 政治大. 學. RET. (4). 立. Growth LOSS. (3). y. ROA. (2). sit. SIZE. (1). er. C_Conserv. CAPEX. i n U. v. 1452 0.704. (Continued on next page). 35. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(42) Notes: SD_RET is the standard deviation of monthly CRSP stock returns over the 12 months starting from the fifth month after fiscal year-end. SD_∆ROA is the standard deviation of seasonally differenced quarterly accounting return on assets over the subsequent three years for the period 2000-2016. R&D is research and development expense scaled by total assets in year t+1. CAPEX is capital expenditures scaled by total assets in year t+1. C_Conserv is the dummy variable measured as an indicator variable equal to one if CEO’s constellation belongs to earth element and water element. Numbers in parentheses are p-values. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively, using a two-tailed test.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 36. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(43) 6. Additional Analysis Based on my main result, it is reasonable to believe that CEOs with more conservative Chinese zodiacs are less risky and CEOs with more conservative constellation signs prefer more conservative firm policies. In this section, I would like to further analyze whether my main results are driven by firms in the high-tech industry. Research has shown that CEO characteristics can systematically vary by industry (Rajagopolan and Datta, 1996) and that firms in high-technology industries certainly have different CEO profiles than firms in low-technology industries. 政治大. (Hambrick et al., 1992). According to Kile and Phillips (2009), I use three-digit SIC. 立. code to distinguish high-tech firms from non-high-tech firms.2. ‧ 國. 學. The relationship between conservative Chinese zodiac and corporate risk and. ‧. between conservative Chinese zodiac and firm policies within high-tech industry are. y. Nat. reported in Table 7. The relationship between conservative Chinese zodiac and. er. io. sit. corporate risk and between conservative Chinese zodiac and firm policies within nonhigh-tech industry are reported in Table 8. The relationship between conservative. al. n. v i n constellation and corporate risk C and between conservative h e n g c h i U constellation and firm. policies within high-tech industry are reported in Table 9. The relationship between conservative constellation and corporate risk and between conservative constellation and firm policies within non-high-tech industry are reported in Table 10. On the basis of Table 7 and Table 8, I find out that the result is consistent with my main Chinese zodiac result, and that CEOs with more conservative Chinese zodiacs prefer less risk and more conservative firm policies. The result indicates that 2. Three-digit SIC codes of high-tech firms include 261, 272, 282 , 283, 284, 286, 287, 289, 351, 352, 353, 354, 355, 356, 357 , 358, 359, 360, 361, 362, 363, 364, 365, 366, 367, 369, 371, 372, 373, 374, 375, 376, 377, 378, 379, 381, 382, 384, 387, 481, 482, 489, 491, 492, 493, 573, 737, 762, 781, 783, 791, 871 and 873. 37. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(44) my main result from Table 5 is mainly driven by firms in the high-tech industry. Furthermore, based on Table 9 and Table 10, the result is consistent with my main result showing that CEOs with more conservative constellation signs prefer more conservative firm policies. Similar to Chinese zodiac result, the result from Table 9 and Table 10 indicates that my main result is mostly composed of firms in the hightech industry. Prior research has shown that CEOs from high-tech industry have much influence on firm policies and corporate risk and this result also confirms the idea that CEO’s Chinese zodiac and constellation has influence on corporate risk and firm policies within high-tech industry.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 38. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(45) Table 7. Regressions of conservative Chinese zodiac and corporate risk and firm policies within high-tech industry Panel A: Conservative Chinese zodiac and corporate risk in the high-tech industry Dependent Variable: Corporate risk measures SD_RET. LEV. (4). -0.004** (0.032) -0.006*** (0.000) -0.108*** (0.000) 0.009* (0.052) 0.017*** (0.008) 0.000 (0.902) 0.026*** (0.000) 0.016*** (0.000) -0.031*** (0.003) -0.021** (0.041) -0.002*** (0.000) -0.021*** (0.003) 0.379*** (0.000) YES YES. -0.005* (0.093) 0.006** (0.035) -0.070*** (0.000) 0.026*** (0.004) 0.009 (0.203) -0.003 (0.379) 0.008** (0.023) 0.023*** (0.000) -0.018 (0.338) -0.015 (0.114) 0.000 (0.907) -0.032*** (0.004) 0.204*** (0.003) YES YES YES. -0.004** (0.012) -0.004*** (0.000) -0.097*** (0.000) -0.011*** (0.007) -0.020*** (0.001) 0.010*** (0.001) 0.009*** (0.001) -0.008*** (0.000) 0.013 (0.174) -0.011 (0.227) -0.001* (0.093) -0.008 (0.194) 0.099*** (0.001) YES YES. -0.002 (0.411) -0.002 (0.152) -0.032*** (0.000) -0.005 (0.366) -0.009** (0.045) 0.002 (0.486) 0.000 (0.833) -0.002 (0.116) -0.01 (0.415) -0.004 (0.502) 0.000 (0.446) 0.005 (0.464) 0.045 (0.285) YES YES YES. 立. Growth. io. INDEP%. n. al. SAME BOD AGE constant Industry FE Time FE Firm FE N Adj. R2. Ch. engchi. 3150 0.469. 3150 0.619. ‧. Nat. RET. 政治大. 學. ‧ 國. CF. LOSS. (3). y. ROA. (2). sit. SIZE. (1). er. Zod_Conserv. SD_∆ROA. i n U. v. 3150 0.286. 3150 0.742. (Continued on next page) 39. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.

(46) Panel B: Conservative Chinese zodiac and firm policies in the high-tech industry Dependent Variable: Firm policy measures R&D. LEV CF. -0.004** (0.029) -0.011*** (0.000) -0.060*** (0.000) -0.008 (0.150) -0.007 (0.121) 0.004** (0.047) 0.001 (0.565) -0.006*** (0.000) -0.013 (0.267) 0.000 (0.995) 0.000 (0.918) -0.009 (0.160) 0.067*** (0.003) YES YES YES. 0.000 (0.914) 0.000 (0.406) -0.003 (0.639) 0.002 (0.431) 0.044*** (0.000) 0.005** (0.031) -0.002 (0.278) -0.001 (0.533) -0.015** (0.020) -0.005 (0.361) 0.001 (0.102) -0.017*** (0.000) 0.213*** (0.000) YES YES. -0.002 (0.207) -0.003** (0.027) 0.000 (0.999) -0.017*** (0.000) 0.009*** (0.008) 0.006*** (0.002) -0.002 (0.209) -0.002** (0.016) -0.007 (0.462) -0.004 (0.424) 0.000 (0.667) -0.011** (0.049) 0.340*** (0.000) YES YES YES. io. SAME. n. al. BOD AGE constant Industry FE Time FE Firm FE N Adj. R2. Ch. engchi. 3150 0.472. 3150 0.904. ‧. ‧ 國. 政治大. Nat. INDEP%. -0.003 (0.153) -0.002*** (0.010) -0.254*** (0.000) -0.040*** (0.000) -0.009 (0.254) 0.017*** (0.000) 0.008** (0.042) -0.013*** (0.000) 0.030** (0.018) -0.005 (0.664) -0.002*** (0.001) -0.064*** (0.000) 0.142*** (0.000) YES YES. 學. RET. (4). 立. Growth LOSS. (3). y. ROA. (2). sit. SIZE. (1). er. Zod_Conserv. CAPEX. i n U. v. 3150 0.224. 3150 0.624. (Continued on next page). 40. DOI:10.6814/THE.NCCU.ACCT.047.2018.F07.