The effect of CEO and CFO on Corporate Policies

Transcription

1 The effect of CEO and CFO on Corporate Policies Better policies if CEO and CFO link Master Thesis Name: Milan Paulissen Anr: Draft-version: Final version Supervisor: dr. O.G. Spalt

2 Table of Contents Abstract: : Introduction : Literature Review Literature on the role of the CEO and the CFO Age Education Inside company experience : Data Collection : An Initial Look at the Data Big and small companies Age groups : Regression Analysis Age Education Inside company experience : Limitations : Conclusion References: Appendix Abstract: This paper analyzes whether firm variables are influenced by personal characteristics of executives. The research is based on a large hand collected database on CEO and CFO characteristics of S&P 500 firms over the years 2001 until The results show that some variables are influenced by solely the characteristics of the CEO or the CFO, while other variables are influenced by the combination of two executives. 2

3 1: Introduction In this thesis I will take a look at the CEO s and CFO s of the companies that have been in the S&P 500 index in the period To be more precise I will be evaluating if and how the characteristics of the CEO and CFO influence firm variables. This research is relevant because of the fact that it can help companies with respect to how they form their executive team. In this case with respect to probably the two most important people namely the CEO and the CFO. I find that the CFO seems to be more responsible for the financial decision, and that when both the executives are old, firm variables differ from when both the executives are young. Specifically, the results show a negative effect of inside company experience on cash, while it shows a positive effect on the market value. Next to that I tested if there is any difference in the firm variables between big and small companies. The results are based on a novel hand-collected dataset containing information for both the CEO and the CFO. The variables that are used as firm characteristics in this paper are: leverage ( in the form of debt to asset ratio) investments, the amount of cash and the market value of the company. The reason I took the market value is because a book value can be influenced by the accountant, the market value of these company is much more accurate since the companies in the dataset are publically traded. A publicly traded company has reasonable value because the value will react to the behavior of the investor. If a certain investor is convinced that the company is overvalued he will sell his shares. The market will react to this by lowering the market value of the company. A question that needs to be answered first is do managers even matter. Especially on the role of the CEO and whether the CEO matters at all, has been a lot of research. A paper by Bertrand and Schoar from 2003 shows that the CEO matters and that he does influence 3

4 firm characteristics. A question answered in this thesis is what happens when we introduce the CFO to the research. This should give us a valuable insight in the role of the CEO and the CFO. I therefore analyze a unique hand collected dataset that consist of 6820 CEO and CFO combinations, divided over 709 different companies. Included in this dataset are the characteristics of both the CEO and CFO, next to that the dataset also contains a number of firm variables. First I will execute the research with the characteristics of the CEO and CFO separately, later on I will link this data. With this linking I construct pairs of the CEO and the CFO for a certain year in a specific company. Dummy variables are used to show whether they share a certain characteristic. It shows if it is really the CEO or the CFO that matters or whether it is their combination. This way of executing the research has one drawback namely that it is focused on hard numbers only, and thereby leave out the sociological criteria such as mood, character and whether the managers are friends. This research is not completely new. There is a paper by Frank and Goyal from 2007 which uses almost the same dataset. The difference lies in the fact that they only look at the CEO and therefore forget about the role of the CFO. The reason to introduce the CFO is that a paper from Simons, Hope Pelled and Smith in 1999 shows that it is not only the CEO who makes all the decisions, they show that it is the whole management team that makes the decisions which is confirmed by a paper from Finkelstein and Hambrick in I only add the CFO because I mainly look at financial firm variables, where the CFO is responsible for the financial policy of the firm. Another reason to look at the CFO is a paper by Favaro from 2001 in which he discusses the growing role of the CFO. 4

5 2: Literature Review In this part of the thesis I will indicate how the current literature stands with respect to the function of the CEO and the CFO. I will also indicate what is already written about the executive characteristics (age, education and inside company experience) that will be discussed in this paper. 2.1 Literature on the role of the CEO and the CFO Recently a lot of research on the effect of the CEO incentives on the firm s investment and leverage choice has been done ( Core and Guay (1999) and Coles et al. (2004)). In a paper by Frank and Goyal from 2007 it s concluded that the board of directors will try to assign the best person for the job, meaning that the CFO and CEO have to fit the profile that the board of directors has set up for the job. Apart from that most decisions are made in a team by the board of directors. Because of that people started to investigate the linkage between a non- CEO and his influence on the corporate policies (Bertrand and Schoar (2003) and Geczy, Minton and Schrand(2004)). In a paper by Bertrand and Schoar from 2003, they argue that an individual manager has no influence on the policies at all. They argue that the firm policies and decisions are always made by the entire board of directors. Although in the same paper they argue that a manager sometimes is hired because of his capabilities, therefore he should have a larger influence on the corporate policies. A paper in 2007 by Chava and Purnanandam, shows that the CEO is more responsible for the bigger decisions such as the holding of cash and the capital structure. While the CFO is responsible for the more specialized decisions, such as investments. Another aspect that needs to be considered in the policy making trajectory is the agency theory, which gives according to Chava and Purnanandam in 2007 certain incentives to the managers on which they base their decision. A Paper by Frank and Goyal investigates if certain characteristics of 5

6 a CEO influences the leverage choice of the company. As a result they find that some of these characteristics have a significant influence on the firm s leverage choice, however the effect of these characteristics is quite small. Significant variance in the firm s leverage is explained by the compensation the CEO receives. (Frank and Goyal; 2007). Since this compensation influences the decisions they made, probably the agency theory will play a role as well, which can be linked to the paper of Chava and Purnanandam in which they state that the agency theory gives certain incentives to the top managers. According to Bertrand and Schoar (2003) it is important that managers can execute their own policy. Combining this with the agency theory leads to the fact that the impact of a manager becomes larger if they are controlled less. The agency theory plays a role here since executives try to maximize their own value, while it is their job to maximize the shareholders value. Another point in the agency theory is that sometimes the (largest) shareholders are also the top executives of the company as described by Bathala, Moon and Rao in Recent papers take a look solely at the role of the CFO in the corporate policies. A paper by Indjejikian and Matejka in 2009 examines if it is appropriate for a CFO to receive a bonus, since this bonus is related to the publications of which he is responsible, meaning that in a way he influences his own bonus. They also mention in their paper that with the late corporate failures the regulation for financial reporting are strengthened, this has a huge influence on especially the CEO and the CFO. As described earlier the CFO plays a role in this research as well. Although it is in less papers than the CEO, who is seen as the leader. Some papers (Favaro 2001, Zorn 2004) also take a look at the CFO and his role in the organization. Especially the paper by Zorn in 2004 in which he argues that the role of the CFO keeps growing the coming years. Especially when 6

7 we look at the following sentence in the Favaro paper: Today, C E O s are delegating highlevel management responsibilities at such a rapid rate that CFOs a r e being stretched ev e n further (Favaro, 2001). Another quote that strengthens my believe to add the CFO to my research is the following: the CFO seems to play at least as important a role as the CEO in determining corporate leverage (Frank and Goyal, 2007). Bertrand and Schoar in their 2003 paper find that managers with high investment related to Tobin s Q fixed effects, have lower investment to cash flow sensitivity. Which means that differences in the benchmark can lead to differences in their investment decisions (Bertrand and Schoar, 2003). As a conclusion Bertrand and Schoar argue that the investment, financing, and the organizational strategy strongly depend on the executive that is in charge of the company/ department that makes the decision. A question that can be raised in this situation, and is the main question in my research. Do these decisions change if the CEO and CFO are linked together, by education, age or another variable that can influence the decision making process. Aside from all financial variables that explain the firms policy I also need to take a closer look at the manager itself. By example the difference between men and women with respect to their risk-aversion level. In combination with the financial decisions, papers show that: Women tend to be more risk averse (Byrnes et al.; 1999). Women choose less aggressive options in the context of investment decisions (Riley and Chow; 1992/ Sunden and Surette; 1998). Women tend be more in line with regulations (Baldry; 1987/Lenney; 1977). A paper by Barua et al. in 2010 shows that female CFO s have better accruals quality. Another significant variable in decision making, is the age of the CEO/CFO, younger people tend to be more risk averse. Holmström says in his 1999 paper, that this is due to the fact that younger people, normally have the ambition to grow in the firm/ stay longer in their 7

8 job, therefore they tend to take less risk, which leads to a non-negative performance, but on the other hand it can offset large gains. Note that this is against the public opinion in which it s said that older people tend to take less risk. 2.2 Age There is already a lot of research on age. The main reason for that is that age is not only of importance in the field of economy but also in the field of psychology and sociology. Next to these three fields there is the public opinion, like old people are more risk averse, less flexible, think that in the past everything was better. In this paper the economic side is of most importance therefore I will mainly focus on the economic side. Although for the best understanding we need to take psychology and sociology into account as well. Especially the learning ability seems to play an important role at least for the economic decision making. A paper by Salthouse in 2000 shows that cognitive abilities are best, when executives are still young. Next to that there is a paper that argues that older people are less flexible and are skeptical/ not open to new technologies, leading to a slower process of adapting the new technologies (De Lange et al, 2006). The other side of the story seems to be true as well. A paper by Goetzmann & Kumar from 2008 argues that this process of cognitive learning leads to more accomplished knowledge when older and therefore older people can make better and more balanced financial decisions. Conclusion is that the most important literature on this field can t give us a clear view on how the age will affect our financial company benchmarks, it could go either way. For now on I assume that the general thought that older will be more risk averse is true. 8

9 2.3 Education A paper by Goetzmann & Kumar from 2008 shows that low educated people tend to make better investment decisions. This contradicts with the public opinion that investing is almost rocket science. In order to make this part valuable for my research we need to translate education into knowledge. More knowledge of a certain subject should make a person more secure about the decision to be made regarding that subject. Although this is nowhere to be found in the literature. 2.4 Inside company experience A paper by Simsek from 2007 shows that there is a link between firm experience (tenure called in their paper) and the financial performance of a company. He even argues that when the complete management team has experience within the firm the financial results even get worse. A paper by Salancik and Pfeffer from 1980 shows that when the company is externally controlled the firm performs better when the executives have inside experience. Which contradicts partly with the conclusion that Simsek showed in his paper. Although the conclusion of externally controlled firms perform better needs to be seen in the market value which is tested later. There is a paper by Walsh from 1995 that shows that past success might influence future decisions. Next to that Miller shows in 1991 that the CEO with inside firm experience tends to be constrained in his/her decision making. This possibly submerges from the fact that he s comfortable with past decisions, and knows his employees to well to be critical. Next to that Miller shows that companies with tenure are less adoptive to problems raised by their direct environment, and therefore relates negatively to financial performance. Some contradicting results are found in the literature here. 9

10 3: Data Collection In order to execute a proper investigation I needed a specific dataset, in this chapter I will explain how the final dataset is formed, and some of the variables will be explained. Next to that I will present some summary statistics. The research in this paper is based on the CEO and the CFO of a company in a certain year, combined with the firm characteristics that year. In order to make this panel of data I used several sources of data. As starting point the Execucomp dataset, containing the 5 best paid executives annually, was used for the years 2001 until The firms that are mentioned in the dataset are the firms that are listed in the S&P In the Execucomp dataset are, amongst other the following data available: Age, Gender, Name of the executive, CEO or CFO, basic salary and bonus. In order to have the CEO and the CFO for all the years mentioned I searched for the annual reports of that company and manually filled in the missing variables. Since most of the personal information of the CEO and CFO needs to be collected manually, I decided to only use the companies that were listed in the S&P 500 for the years mentioned earlier. This hand collecting was done by using the Lexis-Nexis search engine. This engine allowed me to search in the various editions of Who is Who in finance (a yearly revised book that gives personal information about important people, under which are a lot of executives). Based on the information given by Lexis-Nexis data was filled with level of education, year of birth, living place, place of birth, previous jobs and some other. For some of the executives this was not enough to find all the data that is needed to do the research. These blanks where filled by simply searching the internet, especially resources such as NMBD, Wikipedia, BusinessWeek and the google search engine. Leading to a final dataset with 6820 CEO and CFO combinations, divided over 709 different companies. 10

11 For the research I also need several company data, this data was obtained from the COMPUSTAT dataset which includes all the company data as presented in the annual reports. The data that will be used in this thesis are financial variables such as cash holding, leverage ratio, investments and the firm size (logarithm of total assets). Table 1 presents the summary statistics for the personal CEO and CFO variables. Table 1: Summary Statistics Individual Characteristics Mean Std. Dev. Obs. Min. Max. Med. Panel A: CFO Personal Characteristics Age Gender Highest Degree Inside Firm Experience 0,708 0, Panel B: CEO Personal Characteristics Age Gender Highest Degree Inside Firm Experience 0,743 0, The variables that are shown in Table 1 present the statistics of the most important personal characteristics of both the CEO and the CFO for this thesis. Gender is presented by a dummy variable which is 1 if the executive is male, so the mean defines a percentage of the executives that are male. Following from this means is that there is a higher percentage male when it comes to CEO s in comparison to the CFO s. Table 1 shows that 92.8% of the CFO is male, while 98.5% of the CEO is men. It can be concluded that not a lot of women are represented in these two top function. Therefore it seems useless to add gender to the further research. A variable presented in table 1 that needs some more explanation is the variable Highest Degree, one of the variables that was manually collected to add to the dataset is the number 11

12 of degrees and what different levels, if any, the executives obtained. Based on this information a dummy variable was created where 1 stands for a PHD-degree, 2 means a MBAdegree, 3 a master degree, 4 a bachelor degree, 5 means that the executive only obtained a high school diploma, and last 6 means any other diploma not classifiable in one of the categories mentioned before, most of the time this is a Juridicium degree. For completeness reasons the amounts of all the variables are presented in table 2. Table 2: Education Frequencies Number Of CEO s % Number of CFO s % Level Level Level Level Level Level Total As we can see in table 2 there are more CEO s who have obtained the highest degree possible namely the PHD-degree, on the other hand more CFO s have obtained a so called high education level, which is determined as a master degree or higher, or in this case level 3 or higher. In table 1 the inside firm experience variable is also presented, this is a dummy variable which takes a value of 1 if the executive worked in that particular company before. As the data shows more CEO s have worked in the company where they become executive before. In Table 3 the summary statistics of some of the firm variables are presented. The firms size is determined by the logarithm of the total assets, which is also presented as a number. The market value that is presented and used in the remainder of this paper is the logarithm 12

13 market value of equity (share price * number of shares). The number of cash and investment are presented as a ratio in which cash respectively investment are divided by the total assets. By calculating the numbers this way, it becomes a relative number in which the firm size is taken into account. Table 3: Summary Statistics Individual Characteristics Mean Std. Dev. Obs. Min. Max. Med. Tobins Q Cash Flow Market Value Cash Investment Total debt to Assets Total Assets* Size *Note: the number for total assets is scaled (*1000) 4: An Initial Look at the Data In this chapter I will present some initial analysis based on the method that is used in the 2004 paper by Moeller Schlingeman and Stulz. This simple analysis will be executed to see if there is a significant difference in some of the firm variables when they are divided into small and big companies, and on the other hand in young and old executives. The results of this analysis will be used to determine what is important to look at later on. 4.1 Big and small companies The first analysis will be based on the division between small and big companies. A big company is a company where the logarithm of the total assets is above average. The results of this very simple regression are presented in table 4. The most important column in this table is the last column in which the results of a t-test on mean difference is executed. In 13

14 other words if there is a negative significant result in column 4, the mean for a big company is different from the mean when it is a small company. In almost every case we see a negative significant result. Only for the Cash number a positive results is shown, meaning that there s on average more cash in a small company in comparison to a big company. For the other variables this means that the mean of a big company is higher for each of the four other firm variables. Table 4: Firm characteristics divides in small and big company Variable Small Company Big Company Difference (S-B) Market Value (413.67)*** (422.59)*** (-17.94)*** N Cash (62.88)*** (51.58)*** (11.80)*** N Investment (10.06)*** (24.08)*** (-0.26) N Total Debt to Asset (72.37)*** (83.20)*** (-3.48)*** N Size of the Board ( )*** ( )*** (-12.47)*** N The regression in column 2 and 3 are executed where I regress market value to market value, without any control variables, for respectively only small and big companies (as is done in the paper by Moeller Slingeman and Stulz). In the regression in column 3 are regressed on a dummy for small company, without any control variables. The standard errors for the regression in column three are clustered at firm level. The significance is denoted by the next symbols: * p<0.1; ** p<0.05; *** p<0.01 So my believe that the firm variables are higher for the so called big firms are enhanced by the results in table 4. A bigger company is usually better in acquiring external funds at lower rates, therefore it is normal that there level of leverage (total debt to asset) is higher than for the smaller companies. This is also a possible explanation for the lower number of cash in a big company, when it is easier to attract external money from the market, less cash needs 14

15 to be in the company, without the company getting into paying difficulties. I think that one remark should be in place here, namely that the market value of a bigger company is higher should be obvious since the market value of a company is for a part determined by his balance sheet, which includes the total assets. This little side part was only meant to show that almost every variable is significantly higher for big companies, and will not be used in the remainder of this thesis. 4.2 Age groups The second analysis will be based on the age of the executives. An old executive is determined as an executive with an age above average (the average age can be found in table 1). The result of this analysis based on age is presented in table 5. Also in this case the most important result can be found in column 4. Where by division in small and big company only significant t-test results show up. It is shown that this is not the case for a division in age groups. Looking for age groups I see different signs in both the CEO and CFO analysis meaning that for some variables the younger group has a higher average, and for other variables the old group has a higher average. Besides that fact there are only two variables significant at the 1% level, namely the market value with respect to the CFO and the cash number for the CEO, where the average of the old group is significantly higher than for the young group. Next to that we have one variable significant at the 5% level, namely the cash number for the CFO is positive significant. Although it is not significant it is good to notice that for the older groups the total debt to assets are lower than for the younger groups, which is in line with the general thought that older people take less risk. From this simple analysis there is no clear general view on how the age of the executive influences the firm variables, therefore further research will be executed on how the age influences the firm variables. Note that for the CEO and the CFO the same general results apply. 15

16 O Table 5: Firm characteristics divided in young and old executives Young Old Difference Obs young Obs Old Panel A: CFO Market Value (359.71)*** (343.90)*** (-1.66)* Cash (53.65)*** (53.40)*** (2.39)** Investment (19.68)*** (11.39)*** (-1.70)* Total debt to asset 0,255 0, (77.17)*** (76.87)*** (0.66) Panel B: CEO Market Value (329.72)*** (372.35)*** (-1.13) Cash (53.81)*** (53.54)*** (3.56)*** Investment (10.33)*** (19.99)*** (0.640) Total debt to asset (71.22)**** (83.19)*** (0.180) The regression in column 2 and 3 are executed where I regress market value to market value, without any control variables, for respectively only young and old CEO/CFO (as is done in the paper by Moeller Slingeman and Stulz). In the regression in column 3 are regressed on a dummy for young CEO/CFO, without any control variables. The standard errors for the regression in column three are clustered at firm level. The significance is denoted by the next symbols: * p<0.1; ** p<0.05; *** p<0.01 5: Regression Analysis All the regressions that are executed in this chapter will be executed by the following formula = β Where i denotes the firm, j denotes the CEO of company i in year t and h denotes the CFO of company I in year t. is a set of firm control variables such as Tobin s Q, firm size and cash flow, note: not all control variables will be used in all regressions. The Tobin s Q represents a ratio between the market value and the book value of the assets a company possesses. Firm size is used in order to control for how big a company is, as is shown before the values for 16

17 the bigger companies are higher than for the smaller companies. and represents a vector with some executive characteristics for a certain year, like age, highest education for respectively the CEO and the CFO. Besides that all regression will be executed with industry fixed effects and year fixed effects. The industry fixed effects are based on a classification that divides the total market in eleven categories, the exact classification for each of the sic codes can be found in appendix B, I obtained this classification form the official sic-code website ( This is done in order to take away some effects that are caused by the industry that a company is in. There are industries where the company on average have higher leverage. Next to that some of the industries have more problems with the financial crisis that arose around So the combination of these two fixed effects variables should control for the possible variation as it is described above. The results of this regression will be presented in table 6. The results of this simple regression show that for every firm variable except for investment one of the executive characteristic is significant at the different levels. Because of this not significant variables for the investment category it is impossible to draw a conclusion about this variable. On the other hand there is one slight hint of what might be going on here. For the age of the CEO and the AGE of the CFO we see opposite signs. The age of the CEO shows a negative sign, while the age of the CFO gives a positive sign. When linked to what is written in the current literature, a possible explanation might be that the CFO has more cognitive learning about investment decision and has therefore more confidence in taking some more risky strategies. While the CEO who has less experience in investment decision and therefore less cognitive learning take the more risk averse strategy. For the other variables it is difficult to draw a conclusion. The market value is significant influenced by the highest education of the CEO, since the highest education has 17

18 O the lowest value this means that the lower the education level of the CEO is, the higher the market value is. Table 6: OLS Regression Results Investment Cash Leverage Market Value Age of the CEO (-0.52) (-2.23)** (-0.44) (1.44) Age of the CFO (1.22) (-1.20) (-1.87)* (0.74) Highest Education of the CEO (-0.77) (0.99) (-0.68) (2.08)** Highest Education of the CFO (0.71) (-1.31) (-0.62) (-0.24) Number of Observations R-Squared Year Fixed Effects Yes Yes Yes Yes Industry Fixed Effects Yes Yes Yes Yes Clustered by Firm Firm Firm Firm The regressions in this table are all executed with the first row as dependent variable and the other rows as independent variable. Next to that all the variables include control variable. For investment these control variables are cashflow lagged size of the firm and lagged Tobins Q. For cash these control variables are cash flow and the lagged size of the firm. For Leverage and the market value the control variables are lagged Tobin s Q and lagged size of the firm. The significance is denoted by the next symbols. * p<0.1; ** p<0.05; *** p<0.01 With respect to the firm variable cash we see that the age of the CEO has a significant effect and therefore has more influence than the age of the CFO. For the leverage we see a significant result for the AGE of the CFO, therefore it seems that the age of the CFO for the leverage has more influence. A conclusion with reservation is that it seems that the age of the CFO has more influence on the financial firm variables. Which is in line with a piece of the literature that claims that the role of the CFO has become much larger when looking at financial variables. Since all the conclusions above are based on well educated guesses I decided that further research is necessary. 18

19 Therefore the next regression was executed by the next formula: (1) = β The only difference with formula 1 is the fact that this time denotes the CEO/CFO combination for firm i in year t. Also in this case the control variables as they were in formula 1 will be used. In this regression I will add a new variable, namely whether the CEO and the CFO both worked in the company before, this can be of importance because of the fact that when they both worked before in the company they probably know each other, and are more familiar with how the firm operates. The results of this regression are presented in table 7. The first major thing that can be noticed is that there are almost none significant values in the table. Based on these results there is no conclusion possible on how certain combinations of factors influence the company variables. By taking a closer look at the results there are some signs that show some influence by this combination of factors. Since all the variables in the table are dummy variables, the values given in the different regression indicate what happens with the firm variable when the requirement mentioned in column one is met. In the remainder of this part I will discuss panel by panel the main conclusions and how it links to the literature review. As a first result it can be concluded that by executing the regression all separately there can be no conclusion whether the effect on the firm variable really follows from the fact that the executives share this common characteristics or whether it s because one of the two has this characteristic. Therefore I made a new regression almost the same as the one of which the results are presented in table 7. This time I will also add a dummy variable weather the CEO and the CFO separately have this characteristic. The results of this regression are presented in table 8.1, table 8.2 and table

20 O Table 7: OLS Regression Results Investment Cash Leverage Market Value Panel A: Age Both Old (0.97) (-1.31) (-1.52) (1.44) Both Young (-0.48) (1.08) (0.23) (-0.96) Number of observations R-squared Panel B: Highest Education Both High Education (1.21) (0.08) (0.63) (-0.48) Both Low Education (-0.06) (-0.09) (-0.42) (0.47) Number of observations R-squared Panel C: Worked in the company before Both have inside company experience (-2.26)** (-2.15)** (-0.77) (3.69)*** Both have no inside company experience (-0.24) (2.34)** -(0.73) (-2.34)** Number of observations R-squared Year Fixed Effects Yes Yes Yes Yes Industry Fixed Effects Yes Yes Yes Yes Clustered by Firm Firm Firm Firm The regressions in this table are all executed with the first row as dependent variable and the other rows as independent variable. Next to that all the variables include control variable. For investment these control variables are cash flow lagged size of the firm and lagged Tobin s Q. For cash these control variables are cash flow and the lagged size of the firm. For Leverage and the market value the control variables are lagged Tobin s Q and lagged size of the firm. The significance is denoted by the next symbols. * p<0.1; ** p<0.05; *** p< Age As concluded earlier from table 7 there are no clear significant results. On the other hand we see that for all the four variables the signs of the different groups are opposite to each other. So from this fact we can t conclude that there s a difference between how the combination of two old executives run the firm in comparison to the combination of two young executives. 20

21 O Table 8.1: OLS Regression Results with respect to age Investment Cash Leverage Market Value Panel A: Age (Old) CEO (-0.29) (-0.54) (0.03) (0.75) CFO (0.56) (-0.84) (0.35) (0.26) Both (0.46) (-0.03) (-1.11) (0.28) Panel A: Age (Young) CEO (-033) (0.59) (1.30) (-1.04) CFO (-1.31) (0.91) (1.21) (-0.83) Both (0.46) (0.03) (-1.11) (0.28) Number of observations R-squared The regressions in this table are all executed with the first row as dependent variable and the other rows as independent variable. Next to that all the variables include control variable. For investment these control variables are cash flow lagged size of the firm and lagged Tobin s Q. For cash these control variables are cash flow and the lagged size of the firm. For Leverage and the market value the control variables are lagged Tobin s Q and lagged size of the firm. The significance is denoted by the next symbols. * p<0.1; ** p<0.05; *** p<0.01 In order to draw an even better conclusion I executed another regression in which I added a dummy for whether the CEO, CFO is old and the dummy when they are both old (all in one regression) Looking at investment in table 7 it seems that the combination of old executives leads to more investments then the combination of two young executives, which at first sight might not be in line with the general theory that how older people get, the more risk averse they behave. But it is related to the paper by Goetzmann & Kumar in 2008 who say that old executives have more investment knowledge because of their past experience, and therefore might become more secure about possible investment (the effect of cognitive learning). Table 8.1 shows that there is no significant effect, but it seems that especially the 21

22 CFO has an influence on the number of investment. As shown in the literature part earlier theories point out in both ways. The most important theory in that line is a combination of two: a paper by Salthouse from 2000 shows that cognitive learning is best developed in the earlier years, and a paper by Korniotis & Kumar from 2009 shows that older people make more save choices. So there s in the current literature no clear view on how age influences the investment strategy of the company, but a majority of the papers point out to a negative effect between investment and age. Still my results seem to show a positive effect which I think can be caused by several things. First of all most of the papers looking at the financial perspective of age mentioned the effect of cognitive learning (which effect is greater at younger ages). Which can lead to a situation at a later age of I have seen this before, I learned from that situation and therefore I m more secure about how to act now. On the other hand there are the papers who look at investors, and conclude the negative effect between aging and investing. First of all they look at individual investors, what might give different results, then when the decision are made by a team. Besides that they invest their own money, and which should serve their old days. This could be a reason why they act more conservative. Looking at the amount of cash in the company it seems that the combination of young executives hold more cash than their old colleagues (table 7). The problem here is that there is no literature linking personal characteristics to the cash in the company. There are some papers who discuss to holding of cash in the company. But again the opposite conclusion can be found on the relation between age and cash (holding). On one hand there is the argument that holding cash is a sign of conservative thinking. In that way cash is thought of as a security to liquidity shortage (Acharrya, Davydenko & Strebulaev, 2012). On the other hand a paper by Opler, Pinkowitz, Stulz and Williamson from 1999 shows that the 22

23 holding of cash can be seen as a signal that company wants to grow, which can be a more risky strategy. Since for both sides there is a reasonable outcome all results could be linked to a part that is in the current literature. In table 8.1 the same opposite signs shows as discussed earlier, therefore it seems save to conclude that age has some influence on the amount of cash. With respect to the leverage of the company the conclusion is much clearer. Since in all the table that are presented earlier with respect to age no significant result is shown when the dependent variable is leverage. Therefore no conclusion for the influence of age on investment is possible. With respect to the market value we find a positive effect on the market value for the combination of two old executives. For the executives separately opposite signs for old and young executives are shown. Therefore it seems that for the market value the executives separately are important, especially the CFO. The market value that I used is based on the number of shares and the value per share. Which is normally not really a value that is influenced by the characteristics of an executive, the reason I include these characteristics in the research is that it has the highest r-squared out of the four variables. Since the market value normally gives an insight in the past performance and how investors thinks that the company will perform in the future. 5.2 Education Looking at investment we see a negative significant result for the low educated CEO, a combination of the CEO and the CFO sharing the same high/low education seems to have a positive effect on the investment number. Because of this significant result it seems that especially the education of the CEO influence the number of investments in the company. 23

24 O Table 8.2: OLS Regression Results with respect to education Investment Cash Leverage Market Value Panel A: Education (High) CEO (0.28) (-0.58) (0.25) (-1.59) CFO (-1.07) (0.59) (0.14) (0.87) Both (0.92) (0.08) (0.15) (0.19) Panel A: Education (Low) CEO (-2.19)** (0.53) (-0.48) (1.45) CFO (-0.29) (-0.81) (-0.39) (-1.04) Both (0.92) (0.08) (0.15) (0.19) Number of observations R-squared The regressions in this table are all executed with the first row as dependent variable and the other rows as independent variable. Next to that all the variables include control variable. For investment these control variables are cash flow lagged size of the firm and lagged Tobin s Q. For cash these control variables are cash flow and the lagged size of the firm. For Leverage and the market value the control variables are lagged Tobin s Q and lagged size of the firm. The significance is denoted by the next symbols. * p<0.1; ** p<0.05; *** p<0.01 With respect to cash it seems that the level of cash for the executives who have both a high education degree (explained in the data section) is negative while it is the other way around for the lower educated executives (table 7). I believe that in this case the holding of cash is linked to the more aggressive strategy, namely if both the executives have a higher degree, they normally have more knowledge about how the financial system works, and are therefore better able to manage the risk of expanding the company, which leads to less cash in the company. (Opler, Pinkowitz, Stulz and Williamson, 1999). On the other hand lower educated managers lack this knowledge and might be more afraid for the future and therefore keep a buffer in there cash holding just in case that something happens. When 24

25 adding the results of table 8.2 it seems that a high educated CFO and a low educated CEO have the highest number of cash. With respect to the leverage table 7 it seems that the leverage is positive for the higher educated executives, while it is negative for the lower educated executives. I think that for leverage the same reasoning as for the cash applies, namely that the high educated have more knowledge and therefore more confidence in holding the higher leverage ratio. This conclusion is consistent with the numbers present in table 8.2. With respect to the market value I find a surprising result in table 7 namely that it seems to be positive for the low educated executives while it is negative for the high educated executives. From table 8.2 it seems to follow that the high educated CEO has a negative effect on the firm s market value. Following from this can be concluded that investors have more confidence in the lower educated executives, but is this really the way investors look at the firm. Probably not, the only thing investors care about is the performance of the company and there prognosis for the future, and as mentioned earlier lower educated executives take less risk, therefore the prognosis for the future is more certain and less risky. 5.3 Inside company experience For the variable inside company experience the regression results show significant results for every firm variable except for leverage. For investments a negative effect with respect to the CFO is shown. For cash negative significant effects for both the CEO and the CFO are shown. For investment a negative significant result with respect to the CFO is shown. For the variable inside company experience the regression results show significant results for every firm variable except for leverage. 25

26 O Table 8.3: OLS Regression Results with respect to inside firm experience Investment Cash Leverage Market Value Panel A: Inside firm experience (Yes) CEO (-1.14) (-2.07)** (-1.13) (2.68)*** CFO (-2.34)** (-2.93)** (0.83) (2.35)*** Both (0.34) (1.13) (-0.39) (0.43) Number of observations R-squared The regressions in this table are all executed with the first row as dependent variable and the other rows as independent variable. Next to that all the variables include control variable. For investment these control variables are cash flow lagged size of the firm and lagged Tobin s Q. For cash these control variables are cash flow and the lagged size of the firm. For Leverage and the market value the control variables are lagged Tobin s Q and lagged size of the firm. The significance is denoted by the next symbols. * p<0.1; ** p<0.05; *** p<0.01 For the variable inside company experience the regression shows significant results for every firm variable except for leverage. For investments a negative effect with respect to the CFO is shown. For cash negative significant effects for both the CEO and the CFO are shown. For investment a negative significant result with respect to the CFO is shown. Contrary to this, positive effects for the CEO and CFO with respect to market value are shown. For these 3 firm variables only significant variables for the CEO and CFO separately are shown. The general conclusion therefore is that the personal inside the company experience is of much more importance than the combination of the executives. The positive effect for market value can be explained that investors have more confidence in executives with inside firm experience. A remark here is that industry experience might play a role as well. An executive with inside firm experience automatically has industry experience, while the other way around is not necessarily needed. 26

27 6: Limitations The first limitation to this thesis comes from the fact that I only looked at companies who are listed on the S&P 500, so they are listed on the stock exchange. From this follows that there are shareholders who all are allowed to vote for certain decision that the management team wants to take, taking away some of the decision power the executives have The second limitation is related to the first one, namely I only looked at the CEO and the CFO and not the entire management team. A third limitation is the fact that I haven t looked at the CEO and CFO turnover. It is possible that for a certain company the same executive is there in all the 11 years that are in the dataset, therefore their personal influence cannot be tested. A fourth limitations is that I divided all the different sic codes into eleven industry categories as is shown in appendix B, with as consequence that one can argue if a chemical production company is in the same category as a cookie production company ( both manufacturers), because of this the industry fixed effect might not give a clear picture of that specific industry. Last there is the financial crisis that raised in 2007/2008 which might have changed the firm variables over the past years. This problem was taken away for a part by the year fixed effects. Although not for all the firms the impact of the crisis is as large as it is for other companies. 7: Conclusion This thesis shows that certain combinations of CEO and CFO characteristics do influence some firm variables. Although most of these results are not significant there are still some conclusions that can be drawn from this thesis. Therefore most of the results are presented 27

28 with precaution. Further research could lead to further enhancing of the conclusion. One of the things that needs to be considered here is that in the literature until now, there is also no clear view in which direction this effect should point. First I will present all the significant results that are shown in this thesis. First of all it is shown that inside company experience influences the firm variables the most. For cash a negative effect for both the CEO and the CFO separately is shown. For market value the exact opposite signs are shown, while investment is only significant influenced by the CFO. For education only a negative significant effect for the CEO, when the depend variable is investment is shown. For age no significant variable is shown. From this it can be concluded that the CEO and CFO separately do influence the firm variable, for their combination no significant variable can be found. The following conclusion are with precaution. It seems that age has influence on all the four firm variables that are tested in this thesis. For investments it seems that when both executives share a characteristic it has a positive influence on the investments. This goes for all three characteristics tested. For age it seems that an old CFO has a positive effect, while both the young and old CEO have a negative effect. For education a negative significant result for the low educated CEO can be concluded. While the inside company experience seems to draw a negative number on investment when it is solely for the CEO or CFO, the combination once more shows a positive results. Although the numbers are not significant it seems that the combination of characteristics has a positive influence on the investments. With respect to cash and age a simple non-significant conclusion can be made namely the number for old CEO and old CFO and the combination have a negative number, while they re opposite have a positive number all the time. Education seems to have no clear impact on the number of cash, although another table shows that higher educated people have a lower level of cash in the company..looking at education, low educated CEO and CFO seem to have lower 28