1
The Influence of CEO Narcissism on
Accounting Conservatism
Bastiaan Janssen
1Master Accounting & Controlling, University of Groningen
January 2017
Abstract: This study investigates the relationship between narcissistic characteristics in CEO´s of Standard and Poor´s 500 companies and accounting conservatism measures of balance sheet conservatism and earnings conservatism. Using panel data with 1,280 firm-years from 276 firms from 2005 through 2014, I find that CEO narcissism is positively associated to earnings conservatism. The findings suggest that narcissistic characteristics of CEO’s affect accounting conservatism measures through the decisions and influence they have over a firm’s accounting policies.
Keywords: narcissism; accounting conservatism; balance sheet conservatism; earnings conservatism. Supervisor: Karaibrahimoglu, Y. Co-assessor: Hussein, N. Word count: 6.407 1 s2774534, email: B.Janssen.3@student.rug.nl
2
1. Introduction
Chief Executive Officers (CEO’s) often find themselves in the limelight for they are seen as the figureheads of their organization. CEO’s are a vital part of an organization, as they literally lead the company. The way in which a CEO leads the organization depends on his characteristics, which have an impact on the financial results. Famous and well-known CEO’s like Mark Zuckerberg (Facebook) and Elon Musk (Tesla) are visionaries whom, with a great sense of self-belief, founded world-changing companies. There is a thin line between self-belief and narcissism. Maccoby (2000) found that many narcissistic CEO’s lead highly successful companies. Is this because of the choices for bolder accounting policies made by a narcissistic CEO? Accounting policies like accounting conservatism are a matter of choice, made by an organization and the CEO. Conservatism is an important and much discussed accounting policy that leads companies to ´play safe´. These accounting policies are a way for CEO’s to impact the reported numbers. Why does one CEO choose these policies while others don’t?
´´Narcissistic CEO’s favor bold actions that attract attention, resulting in big wins or big losses´´ (Chatterjee & Hambrick, 2007, p.351). Previous research in management find that narcissism affects company strategy and performance, entrepreneurial orientation and corporate strategy (e.g. Chatterjee & Hambrick, 2007; Wales, Patel, & Lumpkin, 2013; Zhu & Chen, 2015). Recent literature examines the impact of narcissistic CEO’s on firm performance (Olsen, Dworkis, & Young, 2014), executive compensation (O'Reilly, Doerr, Caldwell, & Chatman, 2014) and tax sheltering (Olsen & Stekelberg, 2016). I extend this line of research by investigating the effects of CEO narcissism on accounting conservatism. I focus on accounting conservatism because it is one of the most influential principles of accounting and it reflects a manager’s attitude toward risk (Sterling, 1970; Francis, Hasan, Park, & Wu, 2015). I find consistent and robust evidence of a significant positive effect of CEO narcissism on earnings conservatism.
It is important to investigate the effects of narcissism on accounting policies, because CEO narcissism can influence the choice for certain accounting policies. For example, Ahmed & Duellman (2013) found that overconfidence (a trait of narcissism) is negatively related to both conditional (earnings) and unconditional (balance sheet) conservatism. Similarly, Hales, Hobson & Resutek (2012), find that participants with narcissistic characteristics have a bigger chance of exaggerating reported performance to receive positive feedback such as praise.
I hypothesize that if CEO’s show signs of narcissism they are likely to use more (less) accounting conservatism, under efficient contracting (rent extraction) theory. For example,
3
narcissistic CEO’s may feel the need to show off greater equity and earnings, because they have a strong need for praise (Resick, Whitman, Weingarden, & Hiller, 2009). Seeing that conservatism has an understating effect on both book value of equity and earnings, a narcissistic CEO might want to avoid this, under the rent extraction theory.
The aim of this study is to examine the effect of CEO narcissism on accounting conservatism using a sample of 3,966 firm-years over 2005 till 2014 from S&P 500 companies. I contribute to the literature by demonstrating that CEO narcissism significantly affects earnings conservatism. My findings suggest that CEO’s are more conservative towards earnings when they are considered narcissistic. As CEO’s manage the firm’s activities, these findings can aid firms when appointing a CEO. It also supports the efficient contracting theory as a mechanism to prevent managers from behaving opportunistic. Prior work by Ham, Lang, Seybert &, Wang (2015) show that CEO narcissism is not associated with accruals management and conditional conservatism, but does correlate with real earnings management. This paper complements their work.
The remainder of the paper proceeds as follows. Section 2 provides a theoretical background on narcissism and accounting conservatism. Section 3 provides the hypothesis development. The next section describes the methodology of this paper and section 5 shows the results of the analyses. The robustness analysis can be found in section 6. Finally, section 7 concludes the paper and section 8 describes the limitations of the research.
2. Theoretical background 2.1. CEO Narcissism
The word ‘narcissism’ finds its origin in Greek mythology, where a young man named Narcissus fell in love with his own reflection and committed suicide because his love could not be returned. In ‘On Narcissism’ (1914) Freud spoke of narcissism as emotional investment in the self. The American Psychiatric Association (APA) defines NPD (Narcissistic Personality Disorder) in the Diagnostic and Statistical Manual of Mental Disorders—5 (DSM–5; APA, 2013) as ´´a pervasive pattern of grandiosity (in fantasy and behavior), need for admiration, and lack of empathy, beginning by early adulthood and present in a variety of context´´ (p. 645). Narcissists are ´´masters at creating ways of getting what they do need to exist: positive feedback and stroking from other´´ (Pinsky and Young, 2009, p. 100).
Maccoby (2000) found that this is not a bad thing per se when it comes to CEO’s with this trait, for they are often found to be leading highly successful companies. While narcissism can
4
in fact have negative effects on companies, it also has facets such as pride and grandiosity which can have a positive effect (Rijsenbilt & Commandeur, 2013). Narcissists often make for great leaders stated by Maccoby (2002), as they have a very clear and captivating view of companies and through their charisma can attract followers. Narcissism does have its darker traits as ´´researchers have linked increases in narcissism to unsatisfactory task performance (Judge, Lepine & Rich, 2006), job dissatisfaction (Soyer, Rovenpor, Kopelman, Mullins, & Watson, 2001), toxic leadership (Schmidt, 2008), and a host of other negative work attitudes and outcomes.´´ (O’Boyle, Forsyth, Banks, & Daniel, 2012, p. 560). Narcissists pursue their own goals at the long-term cost of others because of their self-centered attitude (Campbell, Bush, Brunell, & Shelton, 2005).
As in any stock listed firm agency costs between principal and agent arise. Shareholders will want to protect themselves against any self-serving behavior from the CEO, by aligning incentives between agent and principal. In line with this contracting view, employing stock-based executive compensation can prevent undesirable behaviors by executives (Westphal & Zajac, 1994). Shareholders will want to align the interests of the CEO with those of the company trough to use of incentives. When this is done right, CEO’s (regardless of narcissistic characteristics) will not show less accounting conservatism, as they don’t get rewarded for this self-serving behavior. This is especially important for narcissistic CEO’s as they have shown to be self-centered (Resick et al., 2009). In contrast, the rent extraction theory states that CEO’s will want to optimize their compensation which depends on their ´´ability to extract private benefits in excess of the optimal compensation´´ (Chalmers, Koh, & Stapledon, 2006, p. 260). This is consistent with the agency-theory, where due to differences in interests between principal and agent problems may arise, because the agent has imperfect information about the contribution of the agent.
Recent work in accounting examines the effects of narcissism on fraud (Rijsenbilt & Commandeur, 2013), performance (Olsen et al., 2014) and managerial reporting (Hales et al., 2012). Most directly related to my study Ahmed & Duelmann (2013) find evidence of CEO overconfidence being negatively related to accounting conservatism, as well as the fact that this effect is not reduced by external monitoring. I add to the literature by examining overconfidence more broadly in the form of narcissism, and its effects on accounting conservatism. Previous literature have also examined the effects of efficient contracting on conditional conservatism (earnings conservatism). Brown, He & Teitel (2006) find evidence that support the notion that opportunistic behavior by managers, in the use of accruals, can be contained due to efficient contracting. Regarding the rent extraction theory, recent literature including Desai &
5
Dharmapala (2006) and Desai & Dharmapala (2009) state that rent extraction by managers can be expedited through aggressive tax avoidance policies. Following, Chalmers et al. (2006) find evidence showing that the CEO is able to extract rents through non-cash incentives over a one year period.
2.2. Accounting Conservatism
Beaver and Ryan (2002) define bias (conservatism) as book value being persistently higher (lower) than market value, so that the book-to-market ratio is persistently above (below) one. The literature makes a distinction between two types of conservatism called balance sheet conservatism (unconditional conservatism) and earnings conservatism (conditional conservatism). Balance sheet conservatism focuses on the book value of equity and takes place with the costs of intangibles being instantly expensed. (Ryan, 2006). Balance sheet conservatism is defined by Feltham & Ohlson (1995) as the book value being persistently understated in relation to the market value. For example, if an item in the inventory has a lower market value than is recorded in the financial statements, according to the conservatism principle the item should be adjusted to the market value and the loss should be reported.
The second type, earnings conservatism or conditional conservatism, ´´involves firms writing down the book value of net assets in a timely fashion upon receiving sufficiently bad news, but not writing up net assets so immediately upon receiving corresponding good news´´ (Ryan, 2006, p. 512-513). Basu (1997) interprets it as resulting in earnings reflecting ‘bad news’ more quickly than ‘good news’. By recognizing these bad and good news periods differently it causes asymmetry in the timeliness and persistence of earnings (Basu, 1997).
In short, the difference between the two types of conservatism is that balance sheet conservatism has the effect of consistently understating the firm’s equity, while earnings conservatism affects the income of a firm.
3. Hypothesis Development
Conservatism is a property of accounting systems globally (Bushman and Piotroski, 2006) and managerial estimates play a critical role in applying conservative accounting (Ahmed & Duellman, 2013). Accounting conservatism has the effect of always reporting a lower net income and lower financial future benefits. Managers tend to delay the release of bad news to outside investors (Kothari, Shu & Wysocki, 2009), resulting in delayed loss recognition.
6
As said, narcissists are constantly looking for praise so a CEO with narcissistic tendencies will want to present a market value as high as possible. For example, a company finds herself in a lawsuit and anticipates winning a large settlement. A narcissistic CEO might record the expected gain on the financial statements as he/she is confident in winning, where as a non-narcissistic CEO might follow the conservatism principle, which means the expected gain will only be recorded on the financial statements when the case is officially settled. Besides feeling confident, a narcissistic CEO will want to report the highest possible net income to receive praise.
However, the efficient contracting theory predicts that CEO’s will want to report a higher discretionary quality, because they risk losing personal reputation and will therefore avoid actions that result in higher cost of capital for their firms (Francis, Huang, Rajgopal & Zang, 2008). Francis et al. (2008) hypothesize that firms managed by reputed CEO’s have better discretionary earnings quality. This suggests that under efficient contract theory, firms with narcissistic CEO’s will show more accounting conservatism.
An opposing view is the rent extraction theory, which states that reputed CEO’s could report a lower discretionary earnings quality, because they overvalue their personal reputation. This could result in a poorer accruals quality to meet certain incentives. Following, Francis et al. (2008) hypothesize that firms managed by reputed CEO’s have poorer discretionary earnings quality. This suggests that under rent extraction theory, narcissistic CEO’s will show less accounting conservatism. The discussed above suggests the following hypotheses:
H1: Under efficient contracting (rent extraction) theory CEO narcissism is positively (negatively) associated with balance sheet conservatism.
H2: Under efficient contracting (rent extraction) theory CEO narcissism is positively (negatively) associated with earnings conservatism.
4. Research Methodology 4.1. Sample
My sample consists of the top executives of all Standard & Poor’s (S&P) 500 companies, excluding financial firms. S&P 500 companies are the 500 largest public companies in the United States and lead by CEO’s with a large variety of characteristics. These firms are known world-wide and receive much respect, which makes for a perfect environment for individuals with narcissistic characteristics. Following Olsen et al. (2014), the CEO’s must have at least four or more years of tenure, which makes it possible to measure the narcissism score in years
7
two and three of the CEO´s tenure. I exclude the first year of tenure, because it could impact the reliability of the results due to CEO turnover and succession.
I begin my sample selection process by selecting all CEO’s of S&P 500 companies available on ExecuComp database. My initial sample consists of 4,755 CEO-years observations from 920 CEO’s and 500 firms over the 2005-2014 period. I exclude 867 CEO-years observations from financial firms (SIC 6000-6799) and 72 CEO-years observations from firms with ambiguity regarding CEO information. I obtain financial data from Compustat. To measure the CEO narcissism, the annual reports of year two and three of the CEO´s tenure have to be available in digital form. I gathered annual reports from the company websites. The sample for evaluating CEO narcissism and accounting conservatism has 338 CEO´s and 276 firms with 1,280 firm-year observations from 2005 to 2014, after applying the just mentioned data requirements. 4.2. Narcissism Measure
The proxy I will use to measure narcissism will be the unobtrusive measures used by Olsen et al. (2014). They measured CEO’s narcissism using a composite measure based on the CEO’s relative cash pay to the second-highest paid executive, the CEO’s relative non-cash pay to the second-highest paid executive, and on the size and composition of CEO’s picture in the annual report.
Olsen et al. (2014) calculate the relative cash pay measure as the ratio of the CEO’s salary and bonus to that of the second highest paid executive. They calculate non-cash pay as the ratio of the CEO’s total compensation, excluding the cash compensation, to that of the second-highest paid executive. Both of the relative pay measures are averaged over the second and third year of the CEO’s tenure.
For the third measure, Olsen et al. (2014) created an additional category for CEO photographs that take up an entire page in the annual report, building on the classification used by Chatterjee and Hambrick (2007). They rate the presence of the CEO photograph in the annual report as follows:
1. No photograph of the CEO;
2. The CEO was photographed with other executives;
3. CEO’s photograph was him or her alone and occupied less than half of the page;
4. CEO’s photograph was him or her alone and occupied more than half of the page;
8
The photograph score measure is also averaged over the second and third year of the CEO’s tenure, as in Olsen et al. (2014). Panel C of Table 1 shows the breakdown of the CEO photographs scores.
Following Olsen et al. (2014), I conduct a factor analysis to make sure that the three elements have the same construct. I use an eigenvalue above 1.0 as my factor-loading
threshold. I find that the factor loadings are 0.813 for photograph size, 0.772 for relative cash pay and 0.310 for relative non-cash pay.
Panel A of Table 1 provides the descriptive statistics for my narcissism measure. Panel B of Table 1 shows the correlations among the three component measures of my narcissism score that are significantly and positively correlated at the 0,05 level and 0,01 level. Furthermore, I create a binary measure of the narcissism score to show that when the narcissism score is greater than the mean, it’s 1, otherwise 0.
TABLE 1
CEO Narcissism Measure Panel A: Descriptive Statistics
n Mean
Standard
Deviation Minimum Maximum
CEO Photo Sizea 338 2.40 1.06 1 5
Relative Cash Payb 338 1.58 0.68 0 8.22
Relative Non-Cash Payc 338 2.47 1.95 0 28.07
Narcissism Scored 338 -0.00 0.58 -1.37 4.79
Narcissism Binarye 338 0.47 0.50 0 1
a Photo size based on scaled of 1 to 5 as follows: (1) No photograph of the CEO; (2) The CEO was photographed with other executives; (3) CEO’s photograph was of him or her alone and occupied less than half the page; (4) CEO’s photograph was of him or her alone and occupied more than half of the page with text taking up some space on the page; and (5) CEO’s photograph was of him or her alone and occupied the whole page.
b CEO’s cash pay relative to the second-highest paid executive. c CEO’s non-cash pay relative to the second-highest paid executive. d
Computed as a summary measure of the CEO photo size, relative cash pay, and relative non-cash pay using the factor weightings from a factor analysis.
e
An indicator variable of 1 if Narcissism Score is greater than average, and 0 otherwise.
Panel B: Correlations
CEO Photo Size
Relative Cash Pay
Relative Non-Cash Pay
CEO Photo Size 1.00
Relative Cash Pay 0.11* 1.00
Relative Non-Cash Pay 0.02 0.37** 1.00
*
Correlation is significant at the 0.01 level (two-tailed) **
9
Panel C: Distribution of CEO Photo Size
Average CEO
Photo Size Frequency
Percentage of CEO’s 1 89 26% 1,5 5 1% 2 62 18% 2,5 15 4% 3 120 36% 3,5 17 5% 4 16 5% 4,5 3 1% 5 11 3% Total CEO’s 338 4.3. Conservatism Measure
I first examine the effect of CEO narcissism on earnings conservatism, whereupon I examine the association between CEO narcissism and balance sheet conservatism. The year the CEO takes office is designated as time 𝑡. Following Olsen et al. (2014), the narcissism measure is calculated by taking the average narcissism score of year two and year three (time 𝑡 + 1 and 𝑡 + 2). CEO’s with a tenure of four years (𝑡 + 3) and more are used as the firm-years wherein narcissism is measured.
To make sure I incorporate other factors that can influence the results of my main model, I include multiple control variables. The characteristics of a CEO are unique and can influence the CEO´s decision making. To control for these characteristics I include control variables for the CEO´s age (𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡+𝑛), the CEO´s tenure (𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡+𝑛), the CEO´s gender (Gender,
women coded as 0) and whether the CEO is chairman of the board (𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡+𝑛). I also
include the following firm control variables; firm size (𝐿𝑛𝐴𝑇𝑖,𝑡+𝑛), measured as the natural log of total assets, leverage (𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡+𝑛), measured as the ratio of total liabilities to common
equity and current year performance (𝑅𝑂𝐴𝑖,𝑡+𝑛), measured as the return on assets. I use two-digit SIC codes to include industry dummies. I estimate the model using a random panel regression analysis, because it allows for individual effects.
I use the following model to test the association between CEO narcissism and conservatism: 𝐴𝑐𝑐𝑜𝑢𝑛𝑡𝑖𝑛𝑔 𝑐𝑜𝑛𝑠𝑒𝑟𝑣𝑎𝑡𝑖𝑠𝑚𝑖,𝑡+𝑛
= 𝛽0+ 𝛽1𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖+ 𝛽2𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡+𝑛
+ 𝛽3𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡+𝑛+ 𝛽4𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡+𝑛+ 𝛽5𝐺𝑒𝑛𝑑𝑒𝑟𝑖 + 𝛽6𝐿𝑛𝐴𝑇𝑖,𝑡+𝑛+ 𝛽7𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡+𝑛+ 𝛽8𝑅𝑂𝐴𝑖,𝑡+𝑛
10
4.3.1 Earnings conservatism
I draw on the unrestricted model of earnings conservatism from Basu (1997) using the following regression:
𝐸𝑃𝑆𝑖,𝑡 = 𝛼0+ 𝛼1𝐷𝑖,𝑡+ 𝛼2𝑅𝑖,𝑡+ 𝛼3𝐷 ∗ 𝑅𝑖,𝑡+ 𝜀𝑖,𝑡 (1)
Where;
𝐸𝑃𝑆𝑖,𝑡 = Earnings per share after extraordinary items scaled by the share price at the
beginning of the period.
𝑅𝑖,𝑡 = Yearly return measured as (𝑃𝑖,𝑡−𝑃𝑖,𝑡−1)/𝑃𝑖,𝑡−1
𝐷𝑖,𝑡 = Dummy variable with a value of ´´1´´ when annual return is negative (bad news) and ´´0´´ otherwise (good news).
In the model from Basu (1997), the coefficient (𝛼1) represents the impact of positive
returns (good news) and (𝛼2+ 𝛼3) represents the impact of negative returns (bad news) on earnings per share. The interactive coefficient 𝛼3measures the asymmetric timeliness.
I examine the relation between narcissism and earnings conservatism using the follow regression;
𝐸𝑃𝑆𝑖,𝑡 = 𝛼0+ 𝛼1𝐷𝑖,𝑡+ 𝛼2𝑅𝑖,𝑡+ 𝛼3𝐷 ∗ 𝑅𝑖,𝑡+ 𝛼4𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡+ (2)
𝛼5𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝐷𝑖,𝑡+ 𝛼6𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝑅𝑖,𝑡+ 𝛼7𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝐷 ∗ 𝑅𝑖,𝑡+ 𝜀𝑖,𝑡
Where;
𝐸𝑃𝑆𝑖,𝑡 = Earnings per share after extraordinary items scaled by the share
price at the beginning of the period.
𝑅𝑖,𝑡 = Yearly return measured as (𝑃𝑖,𝑡−𝑃𝑖,𝑡−1)/𝑃𝑖,𝑡−1
𝐷𝑖,𝑡 = Dummy variable with a value of ´´1´´ when annual return is
negative (bad news) and ´´0´´ otherwise (good news).
𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡 = Narcissism score
I use the following models in my regressions analysis presented in Table 6: Model 1: 𝐸𝑃𝑆𝑖,𝑡 = 𝛼1𝐷𝑖,𝑡+ 𝛼2𝑅𝑖,𝑡+ 𝛼3𝐷 ∗ 𝑅𝑖,𝑡
Model 2: 𝐸𝑃𝑆𝑖,𝑡= 𝛼1𝐷𝑖,𝑡+ 𝛼2𝑅𝑖,𝑡+ 𝛼3𝐷 ∗ 𝑅𝑖,𝑡+ 𝛼4𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡+ 𝛼5𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗
𝐷𝑖,𝑡+ 𝛼6𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝑅𝑖,𝑡+ 𝛼7𝐶𝐸𝑂𝑛𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝐷 ∗ 𝑅𝑖,𝑡
I expect a positive relation between CEO narcissism and earnings conservatism under the efficient contracting theory, due to the mitigating effect on opportunistic behavior. When the rent extraction theory is applied, I expect a negative relation between CEO narcissism and
11
earnings conservatism, because of the opportunities for narcissistic CEO’s to draw attention by using accounting reporting mechanisms.
4.3.2 Balance sheet conservatism
As stated by Beaver and Ryan (2002), balance sheet conservatism is the book value being persistently higher (lower) than market value, so that the book-to-market ratio is persistently above (below) one. This means that when the market to book ratio, (𝑀/𝐵𝑖,𝑡), is higher than ‘1’, the existence of balance sheet conservatism is confirmed.
I expect the same relations between CEO narcissism and balance sheet conservatism and CEO narcissism on earnings conservatism, as discussed above. Narcissistic CEO’s will want to show a high market value instead of the lower book value to receive praise, but this effect could be mitigated under the efficient contracting theory.
4.3.3 Descriptive Summary
Table 2 presents the list of the industries, as well as the number of firms classified in each category. Panel A of Table 3 presents a list of variable descriptions.
TABLE 2
Two-Digit SIC Code Industry Breakdown
Code
Number SIC Code Description Firms
Percentage of Sample
01 Agricultural Production Crops 1 0.30%
10 Metal Mining 2 0.59%
12 Coal Mining 1 0.30%
13 Oil and Gas Extraction 20 5.92%
14 Mining and Quarrying of Nonmetallic Minerals 2 0.59% 15 Building Construction General Contractors An 2 0.59%
16 Heavy Construction, Except Building 3 0.89%
17 Construction Special Trade Contractors 2 0.59%
20 Food and Kindred Products 15 4.44%
21 Tobacco Products 2 0.59%
22 Textile Mill Products 2 0.59%
23 Apparel and Other Textile Products 4 1.18%
24 Lumber and Wood Products, Except Furniture 2 0.59%
25 Furniture and Fixtures 1 0.30%
26 Paper and Allied Products 3 0.89%
28 Chemicals and Allied Products 34 10.06%
29 Petroleum and Coal Products 4 1.18%
30 Rubber and Misc. Plastic Products 3 0.89%
31 Leather and Leather Products 1 0.30%
32 Stone, Clay, Glass and Concrete Products 1 0.30%
33 Primary Metal Industries 1 0.30%
12
35 Industrial Machinery and Equipment 23 6.80%
36 Electronic and Other Electronic Equipment 20 5.92%
37 Transportation Equipment 10 2.96%
38 Instruments and Related Products 26 7.69%
39 Misc. Manufacturing Industries 5 1.48%
40 Railroad Transportation 4 1.18%
42 Trucking and Warehousing 3 0.89%
44 Water Transportation 1 0.30%
45 Transportation by Air 2 0.59%
47 Transportation Services 3 0.89%
48 Communication 14 4.14%
49 Electric, Gas and Sanitary Services 29 8.58%
50 Wholesale Trade – Durable Goods 4 1.18%
51 Wholesale Trade – Nondurable Goods 6 1.78%
52 Building Materials, Hardware, Garden Supply 4 1.18%
53 General Merchandise Stores 6 1.78%
55 Automotive Dealers and Service Stations 5 1.48%
56 Apparel and Accessory Stores 4 1.18%
57 Furniture and Home Furnishings Stores 3 0.89%
58 Eating and Drinking Places 3 0.89%
59 Miscellaneous Retail 5 1.48%
70 Hotels, Rooming Houses, Camps and Other Lod 1 0.30%
72 Personal Services 1 0.30%
73 Business Services 34 10.06%
78 Motion Pictures 1 0.30%
79 Amusement and Recreation Services 1 0.30%
80 Health Services 5 1.48%
87 Engineering, Accounting, Research, Management 2 0.59%
99 Non-classifiable Establishments 2 0.59% Total Firms 338 TABLE 3 Variable Descriptions Variable Description
𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖 A summary measure for CEO photo size, relative cash pay, and
relative non-cash pay based on factor loadings.
𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚 𝐵𝑖𝑛𝑎𝑟𝑦𝑖 Indicator variable of 1 if Narcissism Score is greater than average,
and 0 otherwise. 𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡 The age of the CEO.
𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡 An indicator variable of whether the CEO is also the chairman of the
board.
𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 Length of the CEO’s tenure.
𝐺𝑒𝑛𝑑𝑒𝑟𝑖 An indicator variable of 1 for male, and 0 for female.
𝐸𝑃𝑆𝑖,𝑡 Earnings-per-share.
𝐿𝑛𝐴𝑇𝑖,𝑡 Natural logarithm of assets.
𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 Ratio of total liabilities to common equity.
𝑅𝑂𝐴𝑖,𝑡 Return-on-assets in the year prior to the CEO taking office.
𝑅𝑖,𝑡 Yearly return.
𝐷𝑖,𝑡 Dummy variable.
13
5. Results
Panel A of Table 4 presents descriptive statistics containing number of observations, mean, standard deviation, minimum and maximum of all variables. 𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 has a mean value of 8.13, indicating that the average tenure for CEO’s is more than 8 years. 𝐸𝑃𝑆𝑖,𝑡 has a mean value of 3.00 which indicates that the average return for the S&P companies is positive. 𝑅𝑖,𝑡 has a mean value of 0.14 which means that the average stock return of the sample is 14% and 𝐷𝑖,𝑡 shows that 68% has a negative return. The descriptive statistics on 𝑀/𝐵𝑖,𝑡 indicates,
with a mean value of 3.97, the existence of balance sheet conservatism for the S&P500 firms in the sample. In panel B of Table 4 shows the mean comparison between high- and low narcissistic CEO’s. 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 has a negative mean value with CEO’s deemed as
highly-narcissistic and a positive mean value of 1.357 with CEO’s with low narcissism. This indicates that for the S&P companies used in the sample, a CEO who shows low narcissism has a positive impact on the leverage (the mean difference is non-significant however). In Table 5 the variable correlations are presented. For example, results show that an increase in the age of the CEO is met with an increase in CEO narcissism. It shows that the age of the CEO has a significant impact at the 1% level on CEO narcissism.
TABLE 4 Summary Statistics Panel A: Descriptive Statistics
Variable n Mean
Standard
Deviation Minimum Maximum
𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡 1280 60.69 6.51 31.00 84.00 𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡 1280 0.68 0.47 0.00 1.00 𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 1280 8.13 1.88 4.00 10.00 𝐺𝑒𝑛𝑑𝑒𝑟𝑖 1280 0.96 0.18 0.00 1.00 𝐸𝑃𝑆𝑖,𝑡 1280 3.00 3.34 -29.72 32.16 𝐿𝑛𝐴𝑇𝑖,𝑡 1280 9.33 1.17 6.30 13.59 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 1280 -9.10 429.92 -15091.00 2014.42 𝑅𝑂𝐴𝑖,𝑡 1280 0.76 0.78 -0.85 0.36 𝑅𝑖,𝑡 1280 0.14 0.38 -0.81 2.98 𝐷𝑖,𝑡 1280 0.68 0.47 0 1 𝑀/𝐵𝑖,𝑡 1280 3.97 31.43 -527.02 710.01
14
Panel B: Interpretations
Mean Full sample
High narcissism value
Low narcissism
value Mean Difference
Significance (2-tailed) 𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡 60.69 61.89 59.62 -2.272 0.00* 𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡 0.68 0.77 0.61 -0.164 0.00* 𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 8.13 8.11 8.15 0.018 0.078 𝐺𝑒𝑛𝑑𝑒𝑟𝑖 0.96 0.96 0.97 0.045 0.673 𝐸𝑃𝑆𝑖,𝑡 3.00 3.213 2.820 -0.393 0.035* 𝐿𝑛𝐴𝑇𝑖,𝑡 9.33 9.333 9.323 -0.010 0.877 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 -9.10 -20.831 1.357 22.188 0.357 𝑅𝑂𝐴𝑖,𝑡 0.76 0.0788 0.729 -0.006 0.175 𝑅𝑖,𝑡 0.14 0.111 0.169 0.058 0.006* 𝐷𝑖,𝑡 0.68 0.65 0.70 0.050 0.056 𝑀/𝐵𝑖,𝑡 3.97 4.372 3.618 -0.744 0.669
I created a binary for narcissism score to divide the sample in high and low narcissism CEO’s. If the narcissism score is above the mean of the narcissism score, which is -.001, than ‘1’, otherwise ‘0’. For high narcissism CEO’s n=603 firm-year observations between the years 2008-2014. For low narcissism CEO’s n=677 firm-year observations between the years 2008-2014. Variable descriptions can be found in Table 3.
15
TABLE 5 Variable Correlations
Panel A: Correlations Variables 𝑵𝒂𝒓𝒄𝒊𝒔𝒔𝒊𝒔𝒎 to 𝑳𝒏𝑨𝑻𝒊,𝒕
Variable 𝑵𝒂𝒓𝒄𝒊𝒔𝒔𝒊𝒔𝒎 𝑪𝑬𝑶𝒂𝒈𝒆𝒊,𝒕 𝑪𝑬𝑶𝒄𝒉𝒂𝒊𝒓𝒊,𝒕 𝑪𝑬𝑶𝒕𝒆𝒏𝒖𝒓𝒆𝒊,𝒕 𝑮𝒆𝒏𝒅𝒆𝒓𝒊 𝑬𝑷𝑺𝒊,𝒕 𝑳𝒏𝑨𝑻𝒊,𝒕 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚 1 𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡 0.188** 1 𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡 0.152** 0.226** 1 𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 -0.004 -0.014 0.117** 1 𝐺𝑒𝑛𝑑𝑒𝑟𝑖 0.002 0.028 -0.047 0.045 1 𝐸𝑃𝑆𝑖,𝑡 0.108** -0.016 0.128** -0.002 0.018 1 𝐿𝑛𝐴𝑇𝑖,𝑡 0.007 0.050 0.145** -0.142** -0.152** -0.070* 1 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 -0.016 -0.008 -0.020 -0.006 -0.006 -0.071 0.019 𝑅𝑂𝐴𝑖,𝑡 0.005 -0.013 -0.013 0.118** 0.013 0.512** -0.221** 𝑅𝑖,𝑡 -0.060* -0.082** 0.022 0.086** 0.037 0.112** -0.051 𝐷𝑖,𝑡 -0.024 -0.092** 0.026 0.089** 0.013 0.147** 0.010 𝑀/𝐵𝑖,𝑡 0.003 -0.045 -0.029 -0.030 0.001 -0.012 -0.016 TABLE 5 (continued) Panel B: Correlations Variables 𝑹𝑶𝑨𝒊,𝒕 to 𝑴/𝑩𝒊,𝒕
Variable 𝑳𝒆𝒗𝒆𝒓𝒂𝒈𝒆𝒊,𝒕 𝑹𝑶𝑨𝒊,𝒕 𝑹𝒊,𝒕 𝑫𝒊,𝒕 𝑴/𝑩𝒊,𝒕 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 1 𝑅𝑂𝐴𝑖,𝑡 -0.003 1 𝑅𝑖,𝑡 -0.003 0.075** 1 𝐷𝑖,𝑡 -0.032 0.082** 0.665** 1 𝑀/𝐵𝑖,𝑡 0.001 0.075** 0.036 0.024 1
16
5.1. Findings for earnings conservatism and CEO Narcissism
Table 6 presents the results from the panel regression testing the relationship of CEO narcissism with earnings conservatism using a fixed model. The effect of CEO narcissism on earnings conservatism was tested using a fixed regression analysis, with earnings per share as the dependent variable. Model 1 of Table 6 shows the relation between earnings conservatism (asymmetric timeliness) and earnings. Model 1 shows that the coefficient estimate for return and negative returns (β=4.548, p<0.05) and (β=-4.426, p<0.05), respectively. The results indicate, that for S&P 500 firms earnings conservatism exists.
In Model 2 the impact of CEO narcissism on earnings conservatism was tested. I found that the coefficient for the interaction between narcissism and yearly return indicate a negative influence of CEO narcissism and the yearly return. Furthermore, the results show that the interaction of bad news and narcissism is significant at the 5% level, which indicates that narcissism is positively associated with earnings conservatism. Hypothesis 1 is supported and it is found that under efficient contracting theory, earning conservatism is positively associated with narcissism during a period. Model 1 shows a R-Square of 0.78%, which indicates that the asymmetric timeliness of earnings is relatively low. The results of the regression testing in Model 2 show an increase in R-Square (8.3%) of roughly 7.5%. This means that the asymmetric timeliness is significantly higher compared to Model 1, which indicates that when CEO narcissism is included in the model as an explanatory variable, the explanatory power of the model is enhanced. To test the overall significance of the regression models I use the F-values, which are statistically significant at the 5% level in both models.
17
TABLE 6
Earnings conservatism on CEO Narcissism Panel Regression Dependent Variable: 𝑬𝑷𝑺𝒊,𝒕 Model 1 Fixed Effects Model 2 Fixed Effects Constant 6.212 (0.000) 6.395 (0.000) 𝑅𝑖,𝑡 4.548 (0.000) 4.337 (0.000) 𝐷𝑖,𝑡 -0.452 (0.056) -0.389 (0.100) 𝐷 ∗ 𝑅𝑖,𝑡 -4.426 (0.000) -4.483 (0.000) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡 -0.398 (0.449) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝐷𝑖,𝑡 0.729 (0.072) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝑅𝑖,𝑡 -4.110 (0.001) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝐷𝑖,𝑡∗ 𝑅𝑖,𝑡 3.156 (0.025) 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐷𝑢𝑚𝑚𝑖𝑒𝑠 Included Included 𝑌𝑒𝑎𝑟 𝐷𝑢𝑚𝑚𝑖𝑒𝑠 Included Included R-square 0.0078 0.083 F-Value 10.15 (0.000) 9.17 (0.000) p-values are presented in between parenthesis. Variable descriptions can be found in Panel A of Table 3.
5.2. Findings for balance sheet conservatism and CEO Narcissism
Panel A of Table 7 shows the frequency of firms with a market-to-book value higher and lower then 1 (𝑀/𝐵𝑖,𝑡 ≥ 1 and 𝑀/𝐵𝑖,𝑡 < 1, respectively). Balance sheet conservatism is indicated by 𝑀/𝐵𝑖,𝑡 higher than 1. The mean of 𝑀/𝐵𝑖,𝑡 is 3.97*, which means that balance sheet conservatism exists for the S&P 500 firms included in the sample. In every year, the number of firms with 𝑀/𝐵𝑖,𝑡 ≥ 1 is higher relative to firms with 𝑀/𝐵𝑖,𝑡 < 1. The decreasing amount of firms with 𝑀/𝐵𝑖,𝑡 < 1 after the year 2008 can be explained by the share prices being at an all-time low due to the financial crisis in the stock exchange market starting in 2008.
18
In Panel B of Table 7 a two group mean difference test is shown for firms with a market-to-book value lower than one and higher than one. As with Panel A of Table 7, 𝑀/𝐵𝑖,𝑡 ≥ 1 indicates balance sheet conservatism. The results indicate that firms with balance sheet conservatism experience higher CEO narcissism relative to others. However, the mean difference for narcissism is not statistically significant. Additionally, the test is repeated for several other variables, and the results indicate that only ROA differs significantly for the two groups of firms. This means that firms with balance sheet conservatism experience a statistically significant higher return on assets than firms with less balance sheet conservatism. According to the test results, there is no significant evidence to support Hypothesis 2 and it is found that CEO narcissism is not significantly associated with balance sheet conservatism.
n=1.280 firm-year observations between the years 2008-2014. Variable descriptions can be found in Table 3. * Significant at the 0.05 level.
TABLE 7 Panel A: Frequency table for balance sheet conservatism
2008 2009 2010 2011 2012 2013 2014 𝑀/𝐵𝑖,𝑡 < 1 24 4 5 11 10 4 7 (15.58) (2.47) (2.70) (5.91) (5.08) (2.01) (3.55) 𝑀/𝐵𝑖,𝑡 ≥ 1 130 158 180 175 187 195 190 (84.42) (97.53) (97.30) (94.09) (94.92) (97.99) (96.45) Total 154 162 185 186 197 199 197
The percentages were given in parentheses below the number of firms. * Significant at the 0.05 level.
Panel B: Balance sheet conservatism on CEO Narcissism
𝑴/𝑩𝒊,𝒕< 𝟏 𝑴/𝑩𝒊,𝒕≥ 𝟏 Mean Difference Significance
(2-tailed) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖 -0.062 0.002 -0.064 0.407 𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡 60.88 60.68 0.196 0.813 𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡 0.68 0.68 -0.008 0.895 𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 7.86 8.14 -0.282 0.238 𝐺𝑒𝑛𝑑𝑒𝑟𝑖 0.97 0.96 0.005 0.844 𝐿𝑛𝐴𝑇𝑖,𝑡 9.161 9.334 -0.176 0.238 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 -3.852 -9.376 5.524 0.920 𝑅𝑂𝐴𝑖,𝑡 0.047 0.077 -0.030 0.002*
19
6. Robustness
I test whether my results are robust and not excessively affected by other factors from model assumptions. In particular, I perform a Heckman two stage approach to test whether my regression analysis model experiences sample selection bias problems. I find very similar results, which can be found in Table 8, and the conclusions regarding the hypotheses remain unchanged.
There are some implications I encountered when performing the regression analyses. First, due to the low number of firms with a ‘lower’ balance sheet conservatism as seen in Panel A of Table 7 a proper regression analysis could not be performed for balance sheet conservatism and narcissism. Second, Table 6 shows the panel regression analysis of earnings conservatism and narcissism. When I included firm controls to the regression analysis the variable results were not significant. When included, the CEO control variables did give significant results.
I encounter a selection bias problem when performing the hypotheses tests, as CEO’s are not randomly appointed to firms. On the contrary, CEO’s are selected by firms based on their abilities and how they will be a good fit for the firm. The relation between CEO narcissism and accounting conservatism might therefore be biased. To control for variables that are expected to influence the narcissism score, I model CEO narcissism as a function of accounting conservatism and characteristics of the CEO (age, tenure, chair, gender) and firm controls (such as size and leverage).
20
TABLE 8
Earnings conservatism on CEO Narcissism Heckman Selection Model (two-stage)
Dependent Variable: 𝑬𝑷𝑺𝒊,𝒕 Model 1 Model 2 Fixed Effects Constant 10.943 (0.000) 6.395 (0.000) Inverse Mills -6.192 (0.000) 𝑅𝑖,𝑡 4.410 (0.002) 4.521 (0.000) 𝐷𝑖,𝑡 0.011 (0.980) -0.397 (0.085 𝐷 ∗ 𝑅𝑖,𝑡 -4.824 (0.002) -4.800 (0.000) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡 -0.090 (0.876) -0.260 (0.612) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝐷𝑖,𝑡 0.056 (0.938) 0.731 (0.065) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝑅𝑖,𝑡 -0.824 (0.746) -4.133 (0.001) 𝑁𝑎𝑟𝑐𝑖𝑠𝑠𝑖𝑠𝑚𝑖,𝑡∗ 𝐷𝑖,𝑡∗ 𝑅𝑖,𝑡 1.331 (0.634) 3.258 (0.018) 𝐶𝐸𝑂𝑎𝑔𝑒𝑖,𝑡 0.027 (0.000) 𝐶𝐸𝑂𝑐ℎ𝑎𝑖𝑟𝑖,𝑡 0.442 (0.000) 𝐶𝐸𝑂𝑡𝑒𝑛𝑢𝑟𝑒𝑖,𝑡 -0.041 (0.021) 𝐺𝑒𝑛𝑑𝑒𝑟𝑖 0.013 (0.944) 𝐿𝑛𝐴𝑇𝑖,𝑡 0.026 (0.382) 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 -0.772e-04 (0.526) 𝑅𝑂𝐴𝑖,𝑡 2.484 (0.000) 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐷𝑢𝑚𝑚𝑖𝑒𝑠 Included Included 𝑌𝑒𝑎𝑟 𝐷𝑢𝑚𝑚𝑖𝑒𝑠 Included Included Wald Chi-Square 73.86 (0.000) 9.17 (0.000) p-values are presented in between parenthesis. Variable descriptions can be found in Panel A of Table 3.
21
7. Conclusion and Discussion
Accounting conservatism is one of the most influential principles of accounting and it reflects a manager’s attitude toward risk (Sterling, 1970; Francis, Hasan, Park, & Wu, 2015). The large part of literature focuses on accounting conservatism in relation to earnings quality and firm performance. Little research is done about the effects of certain CEO characteristics on accounting conservatism.
Recent studies in accounting and management investigate the relation between narcissism and company and corporate strategy, firm performance, as well as fraud and tax sheltering. I contribute to this literature by providing evidence on the effects of CEO narcissism on balance sheet conservatism and earnings conservatism. I predict that, under efficient contracting theory, CEO narcissism is positively associated with accounting conservatism, as it prevents managers´ opportunistic behavior. I also predict that, under rent extraction theory, CEO narcissism is negatively associated to accounting conservatism, as CEO’s with narcissistic tendencies want to receive praise for their reported results. Using 1.280 firm-years from 2005 to 2014, I find evidence that CEO narcissism is positively associated to earnings conservatism. My findings support efficient contracting theory, as it shows that narcissistic CEO´s use more earnings conservatism compared to CEO´s with low narcissism, because they risk losing personal reputation. I do not find evidence that CEO narcissism is associated with balance sheet conservatism. My results are robust to the Heckman robustness test.
Prior work by Ham, Lang, Seybert &, Wang (2015) show that CEO narcissism is not associated with accruals management and conditional conservatism (earnings conservatism). To my knowledge this is the first study which provides evidence for a positive relation between CEO narcissism and earnings conservatism. My findings can be used by firms when appointing CEO´s to , as they manage the firms´ activities and accounting policies. Furthermore, it shows that narcissistic CEO´s will align their personal interest with the interests of the firm and will use more earnings conservatism compared to low-narcissistic CEO’s.
The issue remains of how narcissistic tendencies in CEO’s exactly effect accounting policies. Further research can examine the effects of CEO narcissism on balance sheet conservatism, including firm controls.
22
8. Limitations
As in every study there are several limitations to this paper. First, I base the results of my paper on archival data which might not be representative for the current setting. Furthermore, my data sample consists of S&P 500 firms, which means that smaller firms and private firms are not included in my results. Second, to measure narcissism I draw on the unobtrusive measures from Olsen et al. (2014), which might not be the best way to measure narcissism, furthermore my results may be driven by an unidentified factor besides the included control variables. I do however, exclude the requirement from Olsen et al. (2014) that the CEO must began his/her tenure after 1991, since I start my data from the year 2004.
Third, as discussed in section 7, results in the regression analysis were not significant when I included firm controls. I also found no significant results when performing the mean difference test on balance sheet conservatism. As with the measurement for narcissism, the measurement from Basu (1997) might not be the best way to measure accounting conservatism. I planned on using the measurement from Khan & Watts (2009), but I found this to be too complicated for this study, although it might have been a more complete way of measuring accounting conservatism.
23
References
Adams, R. B., & Funk, P. (2012). Beyond the Glass Ceiling: Does Gender Matter?. Management
Science, 58(2), 219-235.
Ahmed, A. S., & Duellman, S. 2013. Managerial Overconfidence and Accounting Conservatism. Journal Of Accounting Research, 51(1), 1-30.
Basu, S. 1997. The conservatism principle and the asymmetric timeliness of earnings. Journal Of
Accounting & Economics, 24(1), 3.
Beaver, W. H., & Ryan, S. G. 2000. Biases and Lags in Book Value and Their Effects on the Ability of the Book-to-Market Ratio to Predict Book Return on Equity. Journal Of Accounting
Research, 38(1), 127-148.
Brown, W. D., He, H., & Teitel, K. (2006). Conditional Conservatism and the Value Relevance of Accounting Earnings: An International Study. European Accounting Review, 15(4), 605-626. Bushman, R., and J. Piotroski. 2006. Financial reporting incentives for conservative accounting: The influence of legal and political institutions. Journal of Accounting and Economics, 42 (1): 107–48.
Campbell, W. K., Bush, C. P., Brunell, A. B., & Shelton, J. (2005). Understanding the social costs of narcissism: The case of the tragedy of the commons. Personality and Social Psychology
Bulletin, 31, 1358 – 1368.
Chalmers, K., Koh, P., & Stapledon, G. (2006). The determinants of CEO compensation: Rent extraction or labour demand?. British Accounting Review, 38(3), 259-275.
Chatterjee, A., & Hambrick, D. C. 2007. It's All about Me: Narcissistic Chief Executive Officers and Their Effects on Company Strategy and Performance. Administrative Science
Quarterly, 52(3), 351-386.
Croson, R., & Gneezy, U. (2009). Gender Differences in Preferences. Journal Of Economic
Literature, 47(2), 448-474.
Deo, M., & Sundar, V. (2015). Gender Difference: Investment Behavior and Risk Taking. SCMS
Journal Of Indian Management, 12(3), 74-81.
Desai, M., Dharmapala, D., 2006. Corporate tax avoidance and high-powered incentives. Journal
of Financial Economics, 79, 145-179.
Desai, M., Dharmapala, D., 2009. Corporate tax avoidance and firm value. The Review of
Economics and Statistics, 91, 537-546.
Feltham, G. A., & Ohlson, J. A. (1995). Valuation and Clean Surplus Accounting for Operating and Financial Activities. Contemporary Accounting Research, 11(2), 689-731.
24
Francis, B., Hasan, I., Park, J. C., & Wu, Q. 2015. Gender Differences in Financial Reporting Decision Making: Evidence from Accounting Conservatism. Contemporary Accounting
Research, 32(3), 1285-1318.
Francis, J., Huang, A. H., Rajgopal, S., & Zang, A. Y. (2008). CEO Reputation and Earnings Quality. Contemporary Accounting Research, 25(1), 109-147.
Freud, S. 1914. On Narcissism
Hales, J., J. L. Hobson, and R. J. Resutek. 2012. The Dark Side of Socially Mediated Rewards: How Narcissism and Social Status Affect Managerial Reporting. Available at:
http://papers.ssm.com/sol3/papers.cfm?abstract_id=2021889
Ham, C. C., Lang, M. H., Seybert, N., & Wang, S. (2015). CFO Narcissism and Financial Reporting Quality. Robert H. Smith School Research Paper No. RHS, 2581157.
Hennes, K. M., A. J. Leone, and B. P. Miller. 2008. The importance of distinguishing errors from irregularities in restatement research: The case of restatements and CEO/CFO turnover. The
Accounting Review, 83 (6): 1487–519.
Judge, T. A., LePine, J. A., & Rich, B. L. (2006). Loving yourself abundantly: Relationship of the narcissistic personality to self- and other perceptions of workplace deviance, leadership, and task and contextual performance. Journal of Applied Psychology, 91, 762–776
Khan, M., & Watts, R. L. (2009). Estimation and empirical properties of a firm-year measure of accounting conservatism. Journal Of Accounting & Economics, 48(2/3), 132-150.
Kothari, S. P., S. Shu, and P. D. Wysocki. 2009. Do managers withhold bad news? Journal of
Accounting Research, 47 (1): 241–276.
Maccoby, M. 2000. ‘Narcissistic leaders’. Harvard Business Review, 78, 68–77.
O'Boyle Jr., E. H., Forsyth, D. R., Banks, G. C., & McDaniel, M. A. (2012). A Meta-Analysis of the Dark Triad and Work Behavior: A Social Exchange Perspective. Journal Of Applied
Psychology, 97(3), 557-579.
Olsen, K. J., Dworkis, K. K., & Young, S. M. 2014. CEO Narcissism and Accounting: A Picture of Profits. Journal Of Management Accounting Research, 26(2), 243-267.
Olsen, K. J., & Stekelberg, J. 2016. CEO Narcissism and Corporate Tax Sheltering. Journal Of
The American Taxation Association, 38(1).
O'Reilly, C. A., Doerr, B., Caldwell, D. F., & Chatman, J. A. 2014. Narcissistic CEOs and executive compensation. Leadership Quarterly, 25(2), 218-231.
Pinsky, D., and S. M. Young. 2009. The Mirror Effect. New York, NY: HarperCollins.
Resick, C. J., D. S. Whitman, S. M. Weingarden, and N. J. Hiller. 2009. The bright side and the dark side of CEO personality: Examining core self-evaluations, narcissism, transformational leadership, and strategic influence. Journal of Applied Psychology: 1365-1381.
25
Rijsenbilt, A., & Commandeur, H. 2013. Narcissus Enters the Courtroom: CEO Narcissism and Fraud. Journal Of Business Ethics, 117(2), 413-429.
Ryan, S. G. (2006). Identifying Conditional Conservatism. European Accounting Review, 15(4), 511-525.
Schmidt, A. A. (2008). Development and validation of the Toxic Leadership Scale (Doctoral dissertation). Available from ProQuest Dissertations and Theses database. (UMI No. 1453699) Soyer, R. B., Rovenpor, J. L., Kopelman, R. E., Mullins, L. S., & Watson, P. J. (2001). Further assessment of the construct validity of four measures of narcissism: Replication and extension.
Journal of Psychology: Interdisciplinary and Applied, 135, 245–258.
Srindhi, B., Gul, F. A., & Tsui, J. (2011). Female Directors and Earnings Quality. Contemporary
Accounting Research, 28(5), 1610-1644.
Sterling, R. R. 1970. Theory of the measurement of enterprise income. Lawrence, KS: University Press of Kansas.
Wales, W. J., Patel, P. C., & Lumpkin, G. T. 2013. In Pursuit of Greatness: CEO Narcissism, Entrepreneurial Orientation, and Firm Performance Variance. Journal Of Management
Studies, 50(6), 1041-1069.
Westphal, J. D., & Zajac, E. J. 1994. Substance and symbolism in CEOs’ long-term incentive plans. Administrative Science Quarterly, 39: 367–390.
Zhu, D. H., & Chen, G. 2015. CEO Narcissism and the Impact of Prior Board Experience on Corporate Strategy. Administrative Science Quarterly, 60(1), 31-65.
Chalmers, K., Koh, P., & Stapledon, G. (2006). The determinants of CEO compensation: Rent extraction or labour demand?. British Accounting Review, 38(3), 259-275.
