CEO Brand: A Paragon or Parasite for Organizational Performance?

(1)

FACULTY OF ECONOMICS AND BUSINESS

CEO Brand: A Paragon or Parasite for Organizational

Performance?

Miroslava Kopečná S2753766

MSc Business Administration – Organization & Management Control Supervisor: Dr. Yasemin Karaibrahimoglu

Second assessor: Dr. Hilco van Elten

(2)

ABSTRACT

CEO Branding is a subject that is attracting a lot of attention these days. As previous research shows, brands are important assets because of their value creation capabilities and the competitive advantage they can provide to the company. However not lot of research depicted the CEO as separate brand. There is little knowledge about the CEO brands and how it relates to company value. The goal of this research is to show this relationship and provide empirical evidence to explain it. The aim is to identify whether a strong CEO brand can affect company value. The data sample is composed of CEO’s of S&P 500 non-financial companies from 2005 to 2014. First, I used factor analysis to conduct a single measure of the CEO brand. Then, I used pool regression to analyze the data. The results were inconclusive. I did not find a positive relationship between strong CEO brand and company value, however I did find a negative relationship between CEO brand and company value with longer CEO tenure.

JEL classification: C23,C33, M19, M29, M39

Key words: branding, CEO brand, company value, value creation, factor analysis

(3)

ACKNOWLEDGMENTS

(4)

1. Introduction

A product is something that is made in a factory, a brand is something that is bought by customers. A product can be copied by a competitor, a brand is unique. A product can be quickly outdated, a successful brand is timeless.

Stephen King

Creating value is the main aim of any business. From the financial perspective value is created when return on capital exceeds the cost of capital, but recent research shows that there is a broader definition of value creation which goes beyond financial metrics (Hillstrom, 2016). According to the extensive literature, non-financial metrics such as management capabilities, marketing, innovations, and brands are the main drivers of value creation in today’s world (Luo et al. 2014). Nowadays, in this technological advanced era, being innovative and investing in marketing is not enough because innovations and marketing are easily replicated by competitors. What makes all the difference is a good reputation and a strong brand. Good corporate reputation and a strong brand can provide a competitive advantage because they are unique to each organization, thus hard to imitate, providing a superior position in the market (De Castro et al., 2006). That’s why it is prudent for a company to establish a strong reputation and brand.

(6)

6 From literature and empirical evidence, (Fetscherin, 2015; Bendisch et al., 2013; Bendisch, 2011; Rosenberger, 2015) we know that both the corporate brand and the CEO brand add value to the company but there is increasing public acknowledgement that the CEO brand can have even more power that the corporate brand in terms of value creation. A good example can be illustrated by the company Apple and its founder Steve Jobs. On the day he announced his resignation as CEO, Apple stocks dropped by 3%. This equates to $10 billion of the company’s value (Fetcherin, 2015). The study of Delgado-García et al. (2015) shows that up to 50% of corporate reputation can be attributed to the CEO’s reputation. Another study (Shandwick, 2016) shows that at least half of the executive board expects that CEO reputation will matter more to a company’s reputation in the next few years. The reason for this is that the CEO, as the public face of the company, sends a signal about the expected performance of the company to the public. This is like the effect that a famous movie star has after they’ve starred in a major film. They send a signal to the audience about the quality of the film and more people end up watching it in the end (Fetcherin, 2015).

(7)

7 Although the CEO brand is and increasingly important subject for academia and corporations themselves (because of its value creation capabilities), there is still relatively little research conducted in this field (Olsen et al.,2014; Bendisch et al., 2013; Fetscherin, 2015; Rosenberger, 2015). Therefore, the aim of this research paper is to contribute to existing literature by providing empirical evidence on the relationship between the CEO brand and the corporate brand in context of value creation and overemphasizing one over the other one.

Given the recent research on CEO branding (regarding how the CEO brand became more and more important for corporations), it’s important to understand if managerial efforts to build the CEO brand can somehow enhance or harm the company. We already have knowledge from previous studies (Ramaswamy and Ozcan, 2016; Gaines-Ross, L., 2000; Fetcherin, 2015) that the corporate brand has significant effect on company’s value, but we don’t know if the company value would be effected when the CEO brand is overemphasized beyond the company brand; thus it is important to identify this relationship in the context of company value. This research attempts to make several contributions to the CEO brand literature. It provides a deeper understanding on the relationship between the CEO brand and value creation. It identifies what kinds of value creating capabilities present the CEO brand and if it should be emphasized over the corporate brand. Researchers can use the empirical evidence for future research to find a deeper understanding of the CEO brand in the context of value creation. Managers, on the other hand, can use this evidence to decide which brand they prefer to build so they can implement it the overall firm strategy.

(8)

8

2. Literature review

2.1. Brand

Branding has been around for centuries. There is a lot of evidence from ancient history that names were put onto goods to identify their unique qualities. The power of the brand can be seen as early as the 16th century when whiskey producers burned their company names onto the wooden barrels and shipped them across the seas. This practice was established for decades to signal to the customers who the makers are and to prevent imitation. Therefore, the purpose of a brand was to distinguish the goods of one producer from another. Although brands have been around for centuries, it was not until the 20th century that researchers started to pay attention to them (Aaker, 1991). This is due to an up-rise in new technologies. Throughout the years, there have been many research studies on branding. One of the most important research studies was done by Aaker (1991) and Keller (1993). According to Keller (1993), a brand can be defined as “a name, term, sign, symbol or design, or a combination of them which intended to identify the goods and services of one seller or group of sellers and to differentiate them from those of competitors” (Keller, 1993). Good identification and good knowledge of the brand provides companies with a huge advantage. If we think about buying a soft-drink, for example cola, we immediately think about Coca-Cola or Pepsi, without even considering other brands. People gladly pay much more for these two brands compared to others because they acknowledge that these brands are the best in their industry, thus the taste of that cola would be good no matter what. Brand reputation is the overall goal of all companies, so the consumers don’t have to think twice about the brand they consume. Keller (1993) suggested that brand knowledge can be defined in terms of two components: brand awareness and brand image.

(9)

9 affect the purchasing decision, even when there are no other brands in consideration. For example, even when you are not thirsty, you may buy Coca-Cola anyways. Thirdly, brand awareness is associated with brand image. A necessary condition for promoting brand image is that brand awareness is already established in customer memory. (Keller, 1993; Baker et al., 2007).

Brand image is the overall impression in customers’ minds. It is a set of beliefs consumers have about specific brands. It is not just mental image, but also an emotional connection. It can be developed over time through extensive and consistent marketing campaigns. An image is formed about a particular brand based on subjective perceptions and associations that customers have about the brand. Brand association should be positive, unique, strong, and favorable. That’s why four unique dimensions of brand association can be defined: types of brand association, the favorability of brand association, the strength of brand association, and the uniqueness of brand association (Keller, 1993). The figure 3 shows all dimensions of brand knowledge according to Keller (2013).

(10)

10 Figure 1

Dimensions of Brand Knowledge

Source: Keller (1993)

2.2. Corporate reputation

(11)

11 customers and a positive reputation increases stakeholder willingness to invest in a company since it enables the company to attract higher quality employees and gain better returns (Dijkmans et al., 2015).

That’s why it is important to know how a positive reputation can be created. Castro et al. (2006) in their research find two components of corporate reputation: business reputation and social reputation. Social reputation is the result of the perception and opinions of stakeholders that are not so close to the business activities of the firm such as investors and the wider community. Business reputation relates to the agents and other stakeholders of the company that are closely associated to business activities, such as employees and the CEO. According to Fetscherin (2015), the CEO is very important for corporate reputation because all the firm activities are reflected and associated with them. The researchers found that CEOs can represent up to 50 percent of total corporate reputation (Delgado-García et al., 2015). A CEO can influence not just corporate reputation, but also financial aspects of a company such as stock returns (Johnson et al., 1993), company profits (Jian and Lee, 2011), and other non-financial aspects of the company, for example firm strategy and risk-taking (Nadkarni and Heemann, 2010) etc.

2.3. CEO Brand

(12)

12 such as Oprah play an important role in modern society serving as arbiters of taste, style and public opinion, thus influencing buying behavior. Their endorsement and creative input enables them to bring attention and credibility to a brand that no other way can. The celebrity that can play an important role in corporate brand building is the CEO of the company. Celebrity CEOs raise the company’s popularity, but their status is also an answer to the leadership challenges arising from cultural changes (Litter, 2007). CEOs are now required to appear more accessible, promoting bottom-up empowerment instead of centralized hierarchies while maintaining the celebrity/superhero status (Austman, 2014).

The CEO brand, as any other brand, can be created and managed. If managed correctly, it can create value for the company. A CEO brand shares certain characteristics with a product or corporate brand, but encompasses also certain characteristics that cannot be traced to non-humans. Fetcherin (2015) identified the CEO brand as a combination of CEO image and CEO reputation. He pointed out that it is important to distinguish between these two dimensions, because the image is an overall mental picture a person has about another person and reputation is the judgment of the individual action. Based on this, he decomposed the CEO brand to four main elements that can be controlled and managed. The 4 P’s of CEO branding are: CEO person, and CEO personality (related to CEO image), CEO prestige, and CEO performance (related to CEO reputation). These four elements make up the CEO branding mix which helps the CEO and the company measure and assesses the CEO brand and establish relationships between the CEO brand and corporate brand. It can also assess the gap between what the CEO thinks and feels about his/her image and reputation compared to what different stakeholders think or feel about the CEO (Fetscherin, 2015).

(13)

13 for regular people (Treadway et al., 2009). As famous people are instantly recognizable and attract interest, they can bring attention to a brand in a way that no other type of advertising can. Therefore, by including social networking sites in their corporate strategies, the companies can increase their performance (Dijkmans et al., 2015). Dijkmans et al. (2015) findings were that engagement in a company’s social media activity was positively correlated with corporate reputation and company performance, especially among non-customers. The study of Leung and Bai (2013) shows that the more people use social media, the more likely they are to become a “friend” or “follower” of the company, thus actively engaged in the company’s online activities. As argued by Rosenberger (2015), CEOs can control personal availability to media and stakeholders depending on his or her personality. One of the personality traits that is worth examining and which can affect the availability of CEOs on social media is overconfidence.

Overconfidence is defined in the literature (Brown and Sarma, 2007) as an overestimation of one’s own ability. Langer (1975) suggests that people are overconfident and overestimate their own abilities because they think they are better-than-average. These assumptions are based on extensive literature in psychology, which finds that people are generally overconfident (Weinstein, 1980).

(14)

14 Based on the empirical evidence we argue that overconfidence will have a positive effect on engagement in social media activities. If the CEOs are overconfident, they are more self-centered, and they will overestimate their abilities, thus they would want to share their achievements with the world –by social media these days.

Empirical findings from previous research (Dijkmans et al., 2015; Burson-Marsteller, 2012; Leung and Bai, 2013) of corporate reputation and their involvement in social media suggest that the more a company uses social media, the more beneficial it is for their reputation, increasing their revenue and improving market capitalization. As the CEOs are perceived as the face of the company and are an integral part of the company’s reputation, they usually receive the most publicity (Fetcherin, 2015). A CEO’s promotion on social media would promote their own reputation as well as the reputation of the company. Straughan and Bleske (1996) found that CEOs increase appeal of news stories since they can related more interesting information and they are more trustworthy and persuasive than any other source. Utilizing CEO reputation as a means of promoting the corporate reputation has been justified by the notion that a CEO, literally and symbolically, becomes the organization itself in the eyes of consumers and other stakeholders (Sohn and Lariscy, 2015). Considering the above discussion and empirical findings, I propose the following hypothesis:

H1: If the CEO brand is strong, it would have a positive impact on company value.

(15)

15 the same biological and environmental factors that shape appearance can also shape personality. Thus a CEO who looks competent can also behave like it, leading to higher financial performance. Psychological research shows that the people form judgments about others based not only on objective physical characteristics such as gender and age, but also more subjective characteristics such as education, social status, and appearance. Focusing on gender, research in psychology shows that because of the gender bias, people associate men with more authority and more capable than women, thus they would appear as better leaders. Research of Faccio et al. (2016) shows that firms run by female CEO have lower leverage, less volatile earnings, and engage in less risk-taking activities than otherwise similar firms with male CEOs in charge. As research of Hirshleifer et al. (2012) suggests, overconfident CEO engage with more risk-taking behavior because they overestimate their own abilities. Females are associated with less-risky behavior, therefore the empirical evidence would suggest that they are not as overconfident that men. This can be associated with the gender gap, which shows that only a small number of the top executive positions are occupied by women. Thus, we argue that female CEOs are less overconfident in their abilities because they have so much more to prove with no room for mistakes. Based on the discussion above and the empirical evidence, I propose the following hypothesis:

H2: If the CEO brand is strong it would have a positive impact on the company value and this effect will be stronger when the CEO is a male.

(16)

16 Meyers, 1987; DeAngelo, 1988; Pourciau, 1993) shows that new CEOs associated with non-routine executive changes overstate expenses/losses of their firms in their first year of services and attribute them to the previous CEOs to take credit for the resulting higher reported earnings in the subsequent years. The findings of Ali and Zhang (2013) suggest that new CEOs try to favorably influence the market perception of their abilities in their early years of service, when the market is uncertain of their abilities. Based on this empirical evidence, one can argue than the CEO at the beginning of their tenure would be more confident in their abilities to run a firm because they would feel privileged that they have been chosen for this position, with so much to prove to the uncertain market. I propose following hypothesis:

H3: If a CEO brand is strong it would have a positive impact on the company value and this effect would be stronger with shorter tenure

Figure 2 presents a conceptual model of the proposed relationship between CEO overemphasis and the impact on the company’s value.

Figure 2

(17)

17

3. Data and Methodology

3.1. Sample selection

The data set is composed of CEOs of all S&P 500 non-financial companies. First, I begin my selection by selecting all CEOs of S&P companies available on Execucomp database. The initial sample consists of 920 CEOs from 500 firms over the period from 2005 to 2014, which makes a 4755 CEO-year observation in total. Next step was to exclude all the data from financial firms and CEOs of subsidiaries. I excluded 867 CEO-year observation from financial companies using the SIC (standard industry classification) code. The SIC codes excluded where SIC 6000-6799. Table 1 summarizes all industries and number of the firms in each industry based on SIC codes. Thirdly, I obtained the list of CUSIPs (Committee on Uniform Security Identification Procedures) codes. CUSIP codes were used to download firm-specific financial information for each company from the COMPUSTAT database through the WARTHON website. The COMPUSTAT database provides a broad range of financial, statistical, and market information from both active and inactive global companies. During this process I eliminate 1556 observations with missing firm-specific variables. The final sample consists of 3199 firm-year observation from 749 CEOs and 407 firms over the period 2005 to 2014.

Table 1

Descriptive statistics: Industry distribution S&P 500

SIC INDUSTRY No. of firms % of the firms in sample

10-14 Mining 27 6,65% 15-17 Construction 6 1,48% 20-39 Manufacturing 195 48,03% 40-49 Transportation 71 17,49% 50-51 Wholesale Trade 9 2,22% 52-59 Retail trade 39 9,61% 70-79 Services 59 14,53% Total* 407 100%

(18)

18

3.2. Model specification

In order to test the hypotheses and find answers to my research question, regression analysis is employed. The chosen data sample consist of observations of multiple phenomena obtained over multiple time periods for the same firms, therefore panel data models will be used. According to literature the examined dataset should be modeled as panel, when the dataset contains both time series and cross-sectional elements, because more precise estimators and test statistics with more power can be obtained (Wooldridge, 2015). The mathematical equation for testing the first hypothesis is described in the equation 1.

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t +βi,tMTBVi,t +

βi,tLEVERAGEi,t + βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

(1)

(19)

19 The equation 2 and 3 describe the econometric relationship between the independent and dependent variables in order to test the second and third hypothesis, respectively.

FIRMVALUEi,t = αi,t + βi,tCEOBRAND + βi,t(CEOBRAND)*CEOGENDERi,t + βi,tCEOGENDERi,t

+βi,tCEOAGEi,t + βi,t CEOCHAIRi,t + βi,t lnSIZEi,t +βi,tMTBVi,t + βi,tLEVERAGEi,t + βi,t DUMMYYi,t + βi,t

DUMMYINDi,t + εi,t

(2)

Where the CEOBRAND times the CEOGENDER is measurement of overemphasis of the CEO abilities, where the CEO abilities will be amplified with gender. CEOGENDER represents the gender of the CEO. All other variables are explained in equation 1.

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,t(CEOBRAND)*CEOTENUREi,t + βi,tCEOTENUREi,t

+βi,tCEOAGEi,t + βi,t CEOCHAIRi,t + βi,t lnSIZEi,t +βi,tMTBVi,t + βi,tLEVERAGEi,t + βi,t DUMMYYi,t + βi,t

DUMMYINDi,t + εi,t

(3)

Where CEOABILITY time CEOTENURE is measurement of overemphasis of the CEO abilities, where the CEO abilities will be amplified with length the CEO tenure. CEOTENURE represent the year the CEO had or has tenure. All other variables are explained in equation 1.

(20)

20 supposed to measure, and by which similar variables can be group into dimensions. By grouping the observations into dimensions it reduces the information in the model. This technique is widely used in empirical studies, especially the once that focus on the CEO brand (Olsen et al., 2014, Malmendier and Tate, 2008; Hribar and Yang, 2006).

After conduction of factor analysis, I can proceed with the regression analysis. Besides a pooled OLS regression, the simplest way how to deal with panel data is to use two classes of panel estimator approaches: fixed effects estimation and random effects estimation. Fixed effects estimators use a transformation to remove the unobserved effects prior to estimation. It allows for arbitrary correlation between the unobserved effect and the explanatory variables in any time period. Random effects model, on the other hand, is used when we assume that the unobservable effects are uncorrelated with each explanatory variables. In fact, the ideal random effects assumption include all of the fixed effects assumptions, plus the additional requirements that unobserved effects are independent of all explanatory variables, in all time periods (Wooldridge, 2015). To test the relevance of non-observable individual effects the LM test will be used. If the non-observable individual effects are not relevant we can use the pool OLS regression, otherwise fixed or random effect model have to be used. Furthermore, to determine if fixed effect or random effect model is more appropriate, I will use the Hausman test.

3.3. Dependent variable

Company value

(21)

21 of its assets, giving a Q value close to unity. The Tobin’s Q has been chosen among the other measurements due to its intrinsic linkage with the intangible assets of a firm. A systematic analysis of Tobin’s Q across firms provides concrete evidence on the highest impacting factors on the firms’ values, which can lead to useful managerial insights (Lin et al., 2014). Tobin’s Q is calculated as market value of the firm’s total assets plus the market value of equity, minus book value of equity, scaled by market value of total assets. To make sure the results are not robust, I will use alternative measure of firm’s value. For robustness check I will use total market value of a firm as measure. All of these measurements are retrieved using COMPUSTAT database and the CUSIP codes of each company. Detail description of the variables used and measurements are presented in appendix B.

3.4. Independent variables

CEO Brand

The CEO Brand is quiet a new research domain, thus there isn’t a lot of relevant literature that focuses on how to measure it. The Fetcherin model provides us with some theoretical guidance. He identified four main elements of the CEO brand that affects the company. These four elements – the 4 P’s - can be controlled and managed, therefore directly and indirectly measured. Those 4 P’s are: persona and personality (related to CEO image), and prestige and performance (related to CEO reputation).

(22)

22 CEO narcissism

As identified in prior literature (Olsen et al. 2014; Chatterjee and Hambrick 2007) narcissistic tendencies can be captured as the prominence of the CEO’s photograph in the company’s annual report. According to Olsen (2014) the prominence of a CEO’s photograph is an appropriate measure for narcissistic personality tendencies because CEOs with greater narcissistic tendencies seek out recognition and admiration to support their inflated self-concept, thus a more prominent photograph in the annual report would draw the attention of the financial statement user and promote the recognition that the narcissistic CEO desires. Following the approach of Olsen et al. (2014) to details the CEO' narcissism is measured as the size and composition of the CEO’s photograph on annual financial statements, the CEO’s relative non-cash pay to second highest paid executive, and the CEO’s relative cash pay to second highest paid executive. To collect the photograph, I examine annual reports from the S&P 500 companies and make five categories to distinguish the prominent of the CEO’s picture. The CEO’s pictures are rated as follows:

1) CEO photograph is in the annual report; 2) CEO was photographed with other executive;

3) CEO’s photograph was of him or her alone occupied less than half of the page; 4) CEO’s photograph was of him or her alone occupied with more than half of the

page with a text taking up some space on the page;

5) CEO’s photograph was of him or her alone and occupied the whole page.

(23)

23 Next a factor analysis is conducted to confirm that the components are capturing the same construct as intended. The starting point of factor analysis is to conduct a correlation matrix. The correlation matrix and the descriptive statistics of the variables are presented in appendix B. After conducting a correlation matrix, the second step is to choose number of factors to be retained. The importance of deciding on how many factors should be retained, when applying the exploratory factor analysis, has been addressed by many authors (Fields, 2000; Hayton et al., 2004; Ledesma and Valero-Mora, 2007). Hayton et al. (2004) pointed out that the number of factors retained can affect regression results if there are too much or too little factors. Given the importance of this decision, we selected the Guttman-Kaiser method. This method is most utilized in practice. According to this method only the factors that have eigenvalues greater than one are retained for interpretation. This method is also used in the study of Olsen et al. (2014). After extracting factor, it might be difficult to interpret the factors based on their factor loadings. Filed (2000) suggest as solution factor rotation. Factor rotation alters the pattern of the factor loading, thus can improve interpretation. According to Filed (2000) there are several methods that can be used for rotation as well. There are two main, the orthogonal (the factors are independent, correlation below 0.3) and oblique (the factors are related, correlations above 0.3). The correlations between the variables are low, therefore I used the orthogonal varimax rotation. The factor analysis loads to a single factor, using eigenvalue above 1.0 as the factor-loading threshold. By using the factor weightings I generated one measure of CEO narcissisms that summarize all variables above (Olsen et al., 2014).

CEO overconfidence

(24)

24 Campbell et al (2011), Malmendier and Tate (2005) and Hirsheifer et al. (2012) are followed in calculating the average moneyness of the CEO’s option portfolio for each year. First step is to estimate the realizable value per option. This is calculated as the total realized value of option divided by the number of vested options. Next step is to estimate the average exercise price or the strike price, which is computed as year-end stock price minus the realized value per option. Finally, the average moneyness of the option is calculated as the realizable value per option divided by the average exercise price. I only include the vested options held by the CEO, because I am only interested in the options that the CEO can exercise following the approach of Hirsheifer et al. (2012). Hirsheifer et al. (2012) states that by using only exercisable options it allows for more firms can be included and also it allows covering a more recent period that includes the millennia technological boom. All the mention variables are retrieved using COMPUSTAT database.

CEO media coverage

(25)

25 To create a single measurement for the CEO BRAND I conducted another factor analysis. I followed the same steps as before in creating CEO Narcissism construct. The descriptive statistics and the correlation matrix are presented in table 2 and table 3, respectively. The factor analysis loads to a single factor1, therefore a one measure of CEO narcissisms that summarize all variables was generated.

Table 2

Descriptive statistics of CEO brand

Variable Mean Std. Dev Maximum Minimum Observation

CEO Narcissism 3.73e-10 0.492 6.417 -1.045 3809

CEO overconfid.. 0.451 0.498 1 0 3809

CEO media cover 637.311 1091.238 6000 0 3809

CEO twitter 0.125 0.331 1 0 3809

The table displays mean, max, min, standard deviation and number of observation of the CEO BRAND variables. CEO narcissism is a single factor of 3measurements the CEO’s relative non-cash pay to second highest paid executive, the CEO’s relative cash pay to second highest paid executive and CEO photograph measure. CEO overconfidence is measure of confidence, CEO is identify as overconfident if the CEO postpones exercise his or her vested option to the point when the options are at least 67% in the money. CEO media coverage in the measures how often the CEOs name is mentions in newspapers. CEO twitter represents if the CEO has twitter account.

Table 3

Correlation matrix CEO BRAND

CEO Narcissism CEO overconfid. CEO media cover. CEO twitter CEO Narcissism 1.0000

CEO overconfid. 0.045*** 1.0000

CEO media cover 0.085*** 0.058*** 1.0000

CEO twitter 0.083*** 0.0056** 0.112*** 1.0000

Notes: *** indicates significance level of 1%, ** indicated significance level of 5%

1

(26)

26

3.5. Control variables

Firm size

The size of a firm is used as a control variable because it enables us to control for firm-specific characteristics. According to the literature (Rego et al., 2009; Olsen et al., 2014), firm size can affect financial aspects of the firm; bigger firms are usually more established, more diversified, and have higher earnings attributed to less risk. In line with most of empirical studies (Rego et al., 2009; Kisgen, 2006), we will use the total assets as proxies for firm size. The data are obtained from the COMPUSTAT database (more details about all control variables can be found in appendix A). The natural logarithm of firm size is used in order to create normally distributed variables by which the results of regression are improved.

Leverage

Another variable that enables us to control for firm-specific characteristics is leverage. Leverage is known from empirical evidence (Kisgen, 2006; Rego et al., 2009; Olsen et al., 2014) to affect the composition of capital structure, equity risk, and the firm’s ability to repay debt. The proxy for leverage is measured as long-term debt plus short term debt divided by the market value of total assets. All variables are obtained from the COMPUSTAT database.

Market to book ratio

(27)

27 CEO chair

To control for specific CEO (personal) characteristic I included the CEO chair as control variable following the approach of the Olsen et al. (2014). According to (Fetcherin, 2015) CEO duality (when the CEO is also a chairman of the company can boost organizational performance because the CEO interests are in line with the board of director’s interest. The data are obtained manually using CEO full name from Execucomp.

CEO age

The CEO age is used in the model to control for specific personal characteristic of the CEO. Empirical evidence shows (Zebrowitz and Montepare, 2005) that the perception of maturity of faces can affect leadership selection, as people with younger faces look weaker. The data are collected using a CEO full name from Execucomp.

Other control variables

(28)

28

3.6. Descriptive statistic

To better understand the sample the descriptive statistics and the correlation matrix are analyzed. The descriptive statistics are presented in table 4 and the correlation matrix is presented in table 5. Looking at the dependent variable, Tobin’s Q mean is 2.13, which implies that most of the firm’s stocks in my sample are more expensive than the replacement cost of its assets, thus there are overvalued. This means that buyers are willing to pay more for the value. This is not surprising considering that we are examining the companies from S&P 500 index2. Leverage mean is 0.78, meaning that most of the firms are financed through debt. The high debt ratio can be preset some hazard to company, as to investors, although, if a company’s operations can generate a higher rate of return than the interest rate on the debt, then higher amounts of debt helps provide profits. Higher mean (3.5) for market to book ratio can be attributed to overvalued stocks. Looking at the CEO age the mean is 61 years, which is considered to be suitable for someone to be in the position of the CEO (experience).The mean of CEO chair measurement of 0.6 signals that more than half of the CEOs in this sample are part of the board of directors as well. The Mean of Gender is 0.9 (variable 1 was assigned for male and 0 for female) means that about 90% of the CEO’s in the sample are Male and only 10% are female (global gender gap3). Looking at CEO Tenure, on average, the CEOs have been in charge of their companies for about 10 years.

2

Very good investment portfolio made by standard & poor’s of 500 largest companies listed in NYSE.

(29)

29 Table 4

Descriptive statistics

Variable Mean Std. Dev Maximum Minimum Observation

CEOBRAND 0.021 0.338 2.999 -0.011 3199 TOBINS’Q* 2.13 1.317 15.506 0.323 3199 MV* 26609.35 45372.38 504239.6 0.00 3199 SIZE* 26084.78 54505.57 797769.0 163.89 3199 LEVERAGE* 0.787 2.892 46.407 -49.654 3199 MTVB* 3.596 5.932 76.441 -78.174 3199 CEOAGE* 61.256 6.693 97 31 3199 CEOCHAIR* 0.676 0.467 1 0 3199 CEOGENDER* 0.969 0.175 1 0 3199 CEOTENURE* 10.437 6.607 53 0 3199

The table displays mean, max, min, standard deviation and number of observation of dependent and independent variables Tobin’s Q if dependent variable (firm value), MV is market value, dependent variable for robust check, CEO brand is a single factor of 3measrures: CEO narcissism, CEO overconfidence and CEO media coverage, MTBV is market to book ratio, Leverage is total debt of the company, size is represented by total assets, CEOAGE is the age of the CEO, CEOCHAIR represents if the CEO is also on board of directors, CEOGENDER is gender of CEO, CEOTENURE is tenure of CEO in the company.

(30)

30 Table 5 Correlation matrix FIRMS VALUE CEO BRAND CEO Gender CEO Tenure

CEO age CEO chair lnSIZE MTBV Lever-age FIRMs VALUE 1.00 CEO BRAND -0.058*** 1.00 CEO gender 0.032** 0.054*** 1.00 CEO Tenure 0.077*** 0.198*** 0.107*** 1.00 CEO age -0.078*** 0.141*** 0.07***6 0.297*** 1.00 CEO chair -0.001 0.105*** 0.0*** 0.143*** 0.241*** 1.00 lnSIZE -0.424*** 0.126*** -0.115*** -0.148*** 0.053*** 0.091** 1.00 MTBV 0.361*** -0.009 -0.019 -0.006 -0.057*** 0.004 -0.119*** 1.00 Leverage -0.085*** -0.047*** -0.033 -0.035 -0.041** -0.029* 0.068*** 0.472** 1.00

The table shows the correlation coefficient between the dependent and independent variables. Variables are negatively correlated when the sign upfront the coefficient is minus, positively correlated when it is plus. Firms Value is Tobin’s Q, CEO brand is a single factor of 3measrures: CEO narcissism, CEO overconfidence and CEO media coverage, lnSize is the natural logarithm of companies assets, MTBV is market to book ratio, Leverage is total debt of the company.

(31)

31

4. Conclusion

4.1. Results of regression

The results of the empirical analysis are presented and analyzed in this section. Before I conducted an regression analysis I did make sure that the model presented in equation 1,2 and 3 has number of desirable properties4, so that hypotheses tests regarding the coefficient estimates could be constructed with validity. If the model does not have these desirable properties it can cause that the standard errors estimates are wrong, which can underestimate their true variability leading to biased probabilities. Table 6 shows the regression results based on equation 1.

Table6

Pooled OLS model CEO Brand

Variable COEFICIENT t-Stat

CEOBRAND -0.077 -0.93 CEOAGE -0.006 -1.61 CEOCHAIR 0.123 5.61*** lnSIZE -0.357 -9.82*** MTBV 0.096 7.18*** LEVERAGE -0.125 -6.03*** R-SQUARED 0.397 Root MSE 1.025

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t

+βi,tMTBVi,t + βi,tLEVERAGEi,t + βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

The sample consist of 3199 observation from period 2005-2014, Method pooled OLS, regression with Driscoll-Kraay standard errors.

Notes: *** indicates significance level of 1%,

4

(32)

32 The coefficient of the CEO BRAND is negative and insignificant therefore, we fail to reject the first hypothesis of no relationship. I assumed that if the CEO brand is strong, it will have a positive impact on the company value, but the result shows that even when the CEO brand is strong it does not have any effect on company value. To make sure the results of the analysis are correct I performed other regression analysis using different models (random and fixed effects model). The results are displayed in appendix C and are the same (significance levels) as the results in table 6. The R-squared value from the pool OLS model is 39%, which means that our model can explain 39% of variances, the other models, fixed and random, have R-square value of 31% and 33% respectively. Compare to the empirical evidence (Olsen et al. 2014, Hribar and Yang, 2016; Malmendier and Tate, 2005a; 2008) the R-squared is about the same level thus I assume the regression is performed correctly.

Table7

Pooled OLS model CEO brand and gender

CEOBRAND -0.101 -0.51 CEOBRAND*gender 0.021 0.11 GENDER 0.058 3.23*** CEOAGE -0.006 -1.62 CEOCHAIR 0.122 5.58*** lnSIZE -0.356 -9.82*** MTBV 0.096 7.19*** LEVERAGE -0.125 -6.03*** R-SQUARED 0.397 ROOT MSE 1.025

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,t(CEOBRAND)*CEOGENDERi,t +

βi,tCEOGENDERi,t βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t +βi,tMTBVi,t +

βi,tLEVERAGEi,t + βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

(33)

33 Table 7 summarizes the result of the regression for the second hypothesis, using equation 2. The results are insignificant. I fail to find a positive relationship between strong CEO brand and company values, thus it’s logical that I could not find a relationship here as well. I assumed that if the CEO brand is strong it would have a positive effect on company value and it will be amplified when the CEO is a male because of gender biases issues. However, the same is not for the result of regression from the third hypothesis. The results are displayed in the table 8. I failed to reject the null hypothesis that stronger CEO brand has no effect on firm’s value. Although the coefficient of CEO brand multiplied by the tenure is negative and significant at the 1% level. This means that the strong CEO brand would have an negative effect the company value with longer tenure of the CEO. The reason why I found a negative relationship may be because since the CEOs make all the important decisions they can decide to pursue their own interest by focusing on building their own brand over the company brand. With a longer tenure of the CEO he or she has longer time to establish his/her own personal brand.

Table8

Pooled OLS model CEO brand and tenure

T h e s a m p l e c o n s i s t T h

e sample consist 3199 observation from period 2005-2014, Method pooled OLS, regression with Driscoll-Kraay standard errors.

CEOBRAND -0.121 0.88 CEOBRAND*tenure -0.021 -2.34*** TENURE 0.009 1.39 CEOAGE -0.007 -2.62 CEOCHAIR 0.109 4.42*** lnSIZE -0.348 -11.20*** MTBV 0.096 7.11*** LEVERAGE -0.125 -6.05*** R-SQUARED 0.397 ROOT MSE 1.025

(34)

34

4.2. Robustness test result

A common exercise in any empirical study is to check if certain regression coefficient and the results will be the same when the regression specification is modified. If the coefficients differ from the initial model, it is possible that there is evidence of structural validity (White and Lu, 2010). The robustness test was performed on the dependent variable. I used the same models (OLS, Fixed, Random effect models) to analyze the relationship between independent variables and dependent variables. As a proxy for dependent variable Instead of Tobin’s Q ratio I used the market value of the company, as it best describes the firm value. The result can be found in appendix C. The coefficients of the CEO brand are also insignificant for all three models, thus I again fail to reject the null hypothesis.

A second robust test was performed by modifying the CEO brand measure. As describe in the section above one single factor was computed from four constructs, namely CEO narcissism, CEO overconfidence CEO media coverage and CEO twitter. I performed a regression analysis using the pool OLS model as in the initial regression, but altered the equation 1 by changing the single variable CEO Brand to four separate variables. The results are presented in table 9.

(35)

35 Table 9

Pooled OLS model CEO narcissism, CEO overconfidence, CEO media coverage, CEO twitter

CEO narcissism -0.253 -5.43***

CEO overconfidence 0.058 0.88

CEO media coverage 0.000 6.49***

CEO twitter -0.0322 -0.64 CEOAGE -0.007 -1.83** CEOCHAIR 0.123 5.63*** lnSIZE -0.374 -12.48*** MTBV 0.095 7.12*** LEVERAGE -0.125 -6.38*** R-SQUARED 0.412

FIRMVALUEi,t = αi,t + βi,tCEONarcisism+ βi,tCEOoverconfidence + βi,tCEOmeadiacoverage +

βi,tCEOtwitter + βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t +βi,tMTBVi,t + βi,tLEVERAGEi,t +

βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

The coefficient of media coverage is positive and significant at 1%, which would suggest that present of the CEO is the media have a positive effect on company value. However, the coefficient is very small, suggesting that these results are significant only statistically not economically. In order for the company value to increase 1%, the CEO presence in the media would have to increase by 11723% (1/0.00008). The coefficient of the CEO twitter is insignificant even at 10% significance level, suggesting that there isn’t any relationship between the CEO presence on Twitter and the company value. In summary, we founded that only the CEO narcissism tendencies will affect the company values. However, if the CEO characteristic into are combined, the results show, that it does not affect the value of the firm.

(36)

36 CEO has narcissistic tendencies and is a male and has a longer tenure it would have a stronger negative effect on company value. Surprising results are on the CEO overconfidence. By multiplying the CEO overconfidence by gender I founded a negative, but significant relationship, meaning that if the CEO is male and he is overconfident it can impact the company’s value in a negative way. The opposite goes for tenure. We wounded a positive and significant relationship, therefore with longer tenure overconfident CEO impacts the company’s value in a positive way. These results are surprising because in testing the first hypothesis I failed to find significant relationship between CEO overconfidence and company value. CEO media coverage is significant statistically, but insignificant economically for both equations 2 and 3. CEO twitter has insignificant coefficients also for both equations 2 and 3.

4.3. Conclusion and Discussion

(37)

37 startups, a strong CEO brand can have a big impact. However, I founded a negative relationship between CEO brand and company value when this relationship is amplified by tenure. This signals that with longer tenure of the CEO, a strong CEO brand would have a negative effect on the company value. I argue that CEOs with longer tenure, have more time to build their personal brand that goes beyond organizational boundaries. Building personal brand can sometimes not be aligned with corporate brand which can affect the company negatively. The probability that the CEO leaves is higher with longer tenure, therefore the CEO will selfishly focus on building his or her personal brand. The CEOs with longer tenure have also stronger brands, and when they leave they take their personal brands with them which can invoke uncertainty on the market. Different results may be presented by using sample of start-up companies. As I argued before, the new CEOs try to favorably influence the market perception of their ability in their early years of service, when the market is uncertain of their abilities, therefore they do not focus on building their personal brand but the company brand. However, in the early of tenure, the CEO brand may not be as strong as the corporate brand of the well-established company.

(38)

38

4.4. Limitation

There are several particular limitations to this study that have to be considered. The first limitation is the CEO brand variable computation. Two factor analyses where used to compute a single factor for the CEO brand measure. If a mistake was made in the process the variable it-self can be omitted and the results may be robust. The second limitation is presented by modelling issues. If we used the wrong model or the model does not have number of desirable properties, the coefficient estimates may be wrong, the associated standard error may be wrong, thus the distribution that was assumed for the test statistics are inappropriate and we could reject/not reject a hypothesis that should not be rejected/should be rejected. Another limitation is the collection of the data. Although the process of the collection is described in details, we cannot ensure that the data would also be available later. The last limitation is presented by the measurement of the variables. The measurements are based on prior literature, but different empirical studies use different measurement approach and it was beyond of this paper’s scope to approach all of the different measurements.

4.5. Future research

(39)

39

References

Aaker, D. (1991). Managing brand equity. New York: Free Press.

Ali, A. and Zhang, W. (2015). CEO tenure and earnings management. Journal of Accounting and Economics, 59(1), pp. 60-79.

Austman, J. (2014). From fat cats to cool cats? CEOs and micro-celebrity practices on Twitter. Master of science. London School of Economics and Political Science.

Bacharach, S. (1989). Organizational Theories: Some Criteria for Evaluation. The Academy of Management Review, 14(4): 496.

Banducci, S.A., Karp, J.A., Thrasher, M. and Rallings, C. (2008), Ballot Photographs as Cues in Low Information Elections. Political Psychology, 29 (6), pp. 903–917.

Bendisch, F., Larsen, G. and Trueman, M. (2013). Fame and fortune: a conceptual model of CEO brands. European Journal of Marketing, 47(3/4), pp.596-614.

Bendisch, F. (2011). BRANDING CEOS: HOW RELATIONSHIPS BETWEEN CHIEF EXECUTIVE OFFICERS, CORPORATE BRANDS AND STAKEHOLDERS. IMAGE CAN INFLUENCE PERCEIVED BRAND VALUE.

Brockman, P., Lee H.S., Salas, J.M. (2015). CEO turnover and firm performance. CEO Branding: Theory and Practice. Fetscherin, M. (ed.). Routledge.

Brown, R. and Sarma, N. (2007). CEO overconfidence, CEO dominance and corporate acquisitions. Journal of Economics and Business, 59(5), pp. 358-379.

Budesheim, T.L. and Depaola, S.J. (1994), Beauty or the Beast – The Effects of Appearance, Personality, and Issue Information on Evaluations of Political Candidates. Personality and Social Psychology Bulletin, 20 (4), pp. 339–348.

Burson-Marsteller. (2012). Global social media check-up 2012 New York: Burson-Marsteller. Retrieved from http://www.burson-marsteller.com/social/PressRelease.aspx

Campbell, T. Colin, Michael Gallmeyer, Shane Johnson, Jessica Rutherford, and Brooke Stanley (2011). CEO optimism and forced turnover, Journal of Financial Economics 101, pp. 695– 712.

(40)

40 Close, A., Moulard, J. and Monroe, K. (2011). Establishing human brands: determinants of placement success for first faculty positions in marketing. Journal of the Academy of Marketing Science, 39(6), pp. 922-941.

Davis, S. (1995) A Vision for the Year 2000: Brand Asset Management. Journal of Consumer Marketing, 12(4), pp. 65-82.

DeAngelo, L. (1988). Managerial competition, information costs and corporate governance: the use of accounting performance measures in proxy contests. Journal of Accounting and Economics 10, pp. 3–36.

De Castro, Gregorio Martín, José Emilio Navas López, and Pedro López Sáez. (2006). Business And Social Reputation: Exploring The Concept And Main Dimensions Of Corporate Reputation. Journal of Business Ethics 63(4). pp.361-370.

Delgado-García et al. (2015). How does CEO Reputation matter? Impact of CEO reputation on corporate reputation and performance . CEO Branding: Theory and Practice. Fetscherin, M. (ed.). Routledge.

Dijkmans, C., Kerkhof, P. and Beukeboom, C. (2015). A stage to engage: Social media use and corporate reputation. Tourism Management, 47, pp.58-67.

Fang, L. and Peress, J. (2009). Media Coverage and the Cross-section of Stock Returns. The Journal of Finance, 64(5): 2023-2052.

Faccio, M., Marchica, M. and Mura, R. (2016). CEO gender, corporate risk-taking, and the efficiency of capital allocation. Journal of Corporate Finance, 39, pp.193-209.

Fetscherin, M. (2015). The CEO branding mix. Journal of Business Strategy, 36(6), pp.22-28. Fetscherin, M. (2015). The 4Ps of CEO branding. CEO Branding: Theory and Practice. Fetscherin,

M. (ed.). Routledge.

Field, A. (2000). Discovering Statistics using SPSS for Windows London – Thousand Oaks – New Delhi: Sage publications.

Gaines-Ross, L. (2013). Get Social: A mandate for New CEOs. MIT SLOAN MANAGEMENT, 54(3), 1.

Gaines-Ross, L. (2000). CEO reputation: A key factor in Shareholder Value. Corporate Reputation Review, 3(4), pp.366-370.

Gaines-Ross, L. (2015). The emergence of the social CEO. CEO Branding: Theory and Practice. Fetscherin, M. (ed.). Routledge.

(41)

41 Hayton, J.C., Allen, D.G. & Scarpello,V. (2004) Factor Retention Decisions in Exploratory Factor Analysis: A Tutorial on Parallel Analysis. Organizational Research Methods, 7, pp. 191-205 Hayward M, Hambrick D. (1997). Explaining the premiums paid for large acquisitions: Evidence

of CEO hubris. Administrative Science Quarterly 42, pp. 103-127.

Hirshleifer, D., Low, A. and Teoh, S. (2012). Are Overconfident CEOs Better Innovators? The Journal of Finance, 67(4): 1457-1498.

Hribar, Paul and Holly Yang (2015) CEO Overconfidence and Management Forecasting. Contemporary Accounting Research 33(1), pp. 204-227.

Jian, M. and Lee, K.W. (2011). Does CEO Reputation Matter for Capital Investments? Journal of Corporate Finance, 17(4), pp. 929–946.

Johnson, B., Young, M. and Welker M. (1993). Managerial Reputation and the Informativeness of Accounting, Contemporary Accounting Research, 10(1), pp. 305–332.

Keller, K.L. (1993). Conceptualizing, measuring, and managing customer-based brand equity. Journal of Marketing, 59, pp. 1-22.

Keller, K. (2013). Strategic brand management. Harlow: Pearson.

Kisgen, Darren J. (2006). Credit Ratings and Capital Structure. Journal of Finance, 61 (3), pp. 1035–1072.

Langer, E. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311– 328.

Ledesma, Rubén Daniel and Pedro Valero-Mora (2007). Determining The Number Of Factors To Retain In EFA: An Easy-To-Use Computer Program For Carrying Out Parallel Analysis. Practical Assessment, Research & Evaluation 12(2).

Leung, X. Y., & Bai, B. (2013). How motivation, opportunity, and ability impact travelers' social media involvement and revisit intention. Journal of Travel & Tourism Marketing, 30(1-2). Lin, Y., Wang, Y., Chiou, J. and Huang, H. (2013). CEO Characteristics and Internal Control

Quality. Corporate Governance: An International Review, 22(1), pp.24-42.

Loroz, P. and Braig, B. (2015). Consumer Attachments to Human Brands: The “Oprah Effect”. Psychology & Marketing, 32(7), pp.751-763.

Luo, X., Kanuri, V. and Andrews, M. (2014). How does CEO Tenure Matter? The Mediating Role of Firm Employee and Firm Customer Relationships. Strategic Management Journal, 35 (4), pp. 492–495.

(42)

42 Malmendier, U. and Tate, G. (2005a). CEO overconfidence and corporate investment. Journal of

Finance, 60(6).

Malmendier, U. and Tate, G. (2005b). Who makes acquisitions? CEO overconfidence and the market’s reaction, Working Paper. Stanford University.

Malmendier, U. and Tate, G. (2009), Superstar CEOs. Quarterly Journal of Economics, 124 (4), pp. 1593–1638.

Nadkarni, S. and Herrmann, P. (2010). CEO Personality, Strategic Flexibility, and Firm Performance: The Case of the Indian Business Process Outsourcing Industry. Academy of Management Journal, 53 (5), pp. 1050–1073.

Nielsen. (2012). Global trust in advertising and brand messages. Retrieved from

http://www.nielsen.com/us/en/insights/reports-downloads/2012/globaltrust-in-advertising-and-brand-messages.html.

Olsen, Kari Joseph, Kelsey Kay Dworkis, and S. Mark Young (2014). CEO Narcissism And Accounting: A Picture Of Profits. Journal of Management Accounting Research, 26(2), pp. 243-267.

Ramaswamy, Venkat and Kerimcan Ozcan (2016). Brand Value Co-Creation In A Digitalized World: An Integrative Framework And Research Implications. International Journal of Research in Marketing, 33(1), pp. 93-106.

Re, Daniel and Rule, Nicolas (2015) CEO facial appearance and firm performance. CEO Branding: Theory and Practice. Fetscherin, M. (ed.). Routledge.

Rego, Lopo L, Matthew T Billett, and Neil A Morgan (2009). Consumer-Based Brand Equity And Firm Risk. Journal of Marketing 73(6), pp. 47-60.

Rijsenbilt, A. and Commandeur, H. (2013). Narcissus Enters the Courtroom: CEO Narcissism and Fraud. Journal of Business Ethics, 117(2), pp.413-429.

Rosenberger, B. (2015). The 4Ps of CEO branding. CEO Branding: Theory and Practice. Fetscherin, M. (ed.). Routledge.

SHANDWICK, W. (2016). The CEO Reputation Premium: A New Era of Engagement - Weber Shandwick. [online] Webershandwick.com. Available at: https://www.webershandwick.com/news/article/the-ceo-reputation-premium-a-new-era-of-engagement [Accessed 23 Nov. 2016].

(43)

43 Strong,J.,Meyer,J. (1987). Asset writedowns: managerial incentives and security returns .Journal

of Finance 42, pp. 643–661.

Sohn, YJ. And Lariscy R.W. (2015). The 4Ps of CEO branding. CEO Branding: Theory and Practice. Fetscherin, M. (ed.). Routledge.

Thomson, M. (2006). Human brands: investigating antecedents to consumers’ strong attachments to celebrities. Journal of Marketing, 70, pp. 104–119.

Treadway, D.C., Adams, G.L., Ranft, A.L. and Ferris, G.R. (2009). A Meso-Level Conceptualization of CEO Celebrity Effectiveness. The Leadership Quarterly, 20(4), pp. 554–570.

Pourciau, S. (1993). Earnings management and non-routine executive changes. Journal of Accounting and Economics 16, pp. 317–336.

Walsh, G., Mitchell, V.-W., Jackson, P. R., & Beatty, S. E. (2009). Examining the antecedents and consequences of corporate reputation: a customer perspective. British Journal of Management, 20(2), pp. 187-203.

Weinstein, N. (1980). Unrealistic optimism about future life events. Journal of Personality and Social Psychology, 39, pp. 806–820.

Wooldridge, J. (2013). Introductory econometrics. 1st ed. Mason, OH: South-Western Cengage Learning

Zebrowitz, L.A. and Collins, M.A. (1997). Accurate Social Perception at Zero Acquaintance: The Affordances of a Gibsonian Approach. Personality and Social Psychology Review, 1(3), pp. 204–223.

(44)

I

Appendixes

Appendix A Variable descriptions and measurement

Table B1

Description of the variables used from COMPUSTAT

Variable Definition COMPUSTAT code

1 Total assets Book value of total assets* AT

2 Size Natural logarithms of AT See item 1

3 Debt in current liabilities Book value of current liabilities* DLC 4 Long-term debt Book value of long term-debt* DLTT 5 Stockholders’ Equity Book value of total equity * SEQ

6 Leverage (DLTT+DLC)/SEQ See items 3,4 and 5

7 Number of shares Outstanding shares* CSHO

8 Closing price Annul, fiscal, market value* PRCC_F

9 Market value CSHO*PRCC_F See items 7 and 8

10 Common Equity Book value of equity* CEQ

(45)

II Table A2

Description of other variables

Variable Definition SOURCE

CEO Chair 1 if the CEO is also a chairman of the board , 0 otherwise Execucomp CEO Tenure Length of CEO’s tenure in years (based on the year

he/she took office)

Execucomp

CEO Age Age of the CEO (based on year of the birth) Execucomp

CEO gender 1 if the CEO is male, 0 women Execucomp

(46)

III

Appendix B Factor analysis

Table C1

Correlation matrix CEO Narcissism

CEO Photo size Relative Cash pay Relative Non/cash pay CEO Photo size 1.00

Relative Cash pay 0.098*** 1.00

Relative Non/cash pay 0.049*** 0.022*** 1.00

NOTES: ***correlation is significant at the 1% significance level

Table C2

CEO Narcissism measure descriptive statistics

Variable n Mean St.deviation Min Max

CEO Photo size 3814 0.997 0.967 0 2

Relative Cash pay 3814 1.538 0.759 0 15.32

Relative Non/cash pay 3814 2.398 2.225 0 34.68

(47)

IV

Appendix C Robust check of methodological approach

Table C1

Fixed and random effects model CEO Brand Variable

Fixed Random

COEFICIENT t-Stat COEFICIENT t-Stat

CEOBRAND -0.018 0.34 0.011 0.21 CEOAGE -0.010 2.96 0.005 1.41 CEOCHAIR 0.072 1.37*** 0.103 2.15** lnSIZE -0.547 -12.03*** -0.484 -17.16*** MTBV 0.043 15.37*** 0.051 18.01*** LEVERAGE -0.048 -8.07*** -0.598 -10.64*** R-SQUARED 0.307 0.330

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t +βi,tMTBVi,t + βi,tLEVERAGEi,t +

βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

The sample consist of 3199 observation from period 2005-2014, Method fixed effects and random-effect GLS regression

(48)

V Table C2

Fixed and random effects model CEO Brand and gender Variable

CEOBRAND 0.128 0.42 0.128 0.43 CEOBRAND*gender -0.121 -0.36 -0.121 -0.40 CEOGENDER 0.023 0.19 -0.009 -0.08 CEOAGE 0.0102 2.96*** 0.005 1.40 CEOCHAIR 0.071 1.35 0.103 2.16** lnSIZE -0.54 -11.97*** -0.484 -17.16*** MTBV 0.044 15.35*** 0.051 18.01*** LEVERAGE -0.048 -8.68*** -0.059 -10.63*** R-SQUARED 0.307 0.331

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,t(CEOBRAND)*CEOGENDERi,t + βi,tCEOGENDERi,t + βi,tCEOAGEi,t +

βi,tCEOCHAIRi,t + βi,tlnSIZEi,t +βi,tMTBVi,t + βi,tLEVERAGEi,t + βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

(49)

VI Table C3

Fixed and random effects model CEO Brand and tenure Variable

CEOBRAND -0.016 -0.19 0.013 0.17 CEOBRAND*tenure 0.005 0.70 -0.001 -0.03 CEOtenure -0.006 -1.21 -0.001 -0.23 CEOAGE -0.012 3.29*** -0.005 1.46 CEOCHAIR 0.085 1.59 0.107 2.21** lnSIZE -0.549 -11.97*** -0.486 -17.13*** MTBV 0.043 15.36*** 0.051 18.00*** LEVERAGE -0.048 -8.63*** -0.059 -10.61*** R-SQUARED 0.303 0.329

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,t(CEOBRAND)*CEOTENUREi,t + βi,tCEOTENUREi,t + βi,tCEOAGEi,t +

βi,tCEOCHAIRi,t + βi,tlnSIZEi,t +βi,tMTBVi,t + βi,tLEVERAGEi,t + βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

(50)

VII Table C4

Pooled OLS model CEO narcissism, CEO overconfidence, CEO media coverage, CEO twitter multiplied by gender

CEO narcissism 0.0877 0.62*

CEO narcissism*gender -0.352 -2.32**

CEO overconfidence 0.463 2.41

CEO overconfidence*gender -0.404 -2.65***

CEO media coverage -0.000 -3.32***

CEO media coverage*gender 0.000 6.75***

CEO twitter 0.065 0.54 CEO twitter*gender -0.094 -0.74 CEO gender 0.175 0.29 CEOAGE -0.007 -2.03** CEOCHAIR 0.141 5.80*** lnSIZE -0.373 -12.49*** MTBV 0.095 6.99*** LEVERAGE -0.125 -6.41*** R-SQUARED 0.412

FIRMVALUEi,t = αi,t + βi,tCEONarcisism+ + βi,t(CEONarcisism)*CEOgender + βi,tCEOoverconfidence + βi,t(CEOoverconfidence)*CEOgender + βi,tCEOmeadiacoverage βi,t(CEOmeadiacoverage)*CEOgender + βi,tCEOtwitter + βi,t(CEOtwitter)*CEOgender + βi,tCEOgender + βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t +βi,tMTBVi,t + βi,tLEVERAGEi,t + βi,tDUMMYYi,t +βi,t DUMMYINDi,t + εi,t

(51)

VIII Table C5

Pooled OLS model CEO narcissism, CEO overconfidence, CEO media coverage, CEO twitter multiplied by tenure

CEO narcissism -0.111 -1.21

CEO narcissism* tenure -0.011 -1.72**

CEO overconfidence -0.119 -3.34

CEO overconfidence* tenure 0.016 3.85***

CEO media coverage 0.000 13.50***

CEO media coverage* tenure 0.000 -0.92

CEO twitter -0.127 -1.65

CEO twitter* tenure 0.008 1.06

CEO Tenure -0.004 -0.78 CEOAGE -0.008 -2.64*** CEOCHAIR 0.134 4.82*** lnSIZE -0.365 -15.61*** MTBV 0.095 7.02*** LEVERAGE -0.124 -6.34*** R-SQUARED 0.412

FIRMVALUEi,t = αi,t + βi,tCEONarcisism + βi,t(CEONarcisism)*CEOtenure + βi,tCEOoverconfidence +

βi,tCEOoverconfidence*CEOtenure + βi,tCEOmeadiacoverage + βi,tCEOmeadiacoverage*CEOtenure +

βi,tCEOtwitter + βi,tCEOtwitter* CEOtenure + βi,tCEOtenure + βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t

(52)

IX

Appendix D Robust check of dependent variable

TableD1

Pooled OLS model CEO Brand

CEOBRAND 0.023 -0.49 CEOAGE -0.004 -1.43 CEOCHAIR 0.072 2.69*** lnSIZE 0.77 64.56*** MTBV 0.051 8.34*** LEVERAGE -0.078 -5.23*** R-SQUARED 0.699 Root MSE 1.025

FIRMVALUEi,t = αi,t + βi,tCEOBRAND+ βi,tCEOAGEi,t + βi,tCEOCHAIRi,t + βi,tlnSIZEi,t

CEO Brand: A Paragon or Parasite for Organizational Performance?

FACULTY OF ECONOMICS AND BUSINESS