The compensation and power of parent’s CEO and
subsidiary’s Financial Reporting Quality
Bram Collet
Supervisor: Simona Rusanescu
University of Groningen
Faculty of Economics and Business
MSc Accountancy & Controlling
AbstractAmerican Multinational Corporations use Earnings Management for compensation matters. The CEO’s of these MNC’s get paid variable related to their performance in the consolidated statements by bonuses and stock-options and they behave opportunistic to higher their
compensation. Earnings of subsidiaries are material for these consolidated statements, so parent’s manage earnings in subsidiaries, which reduces the Financial Reporting Quality of the subsidiaries, which does not benefit the users of the financial information of the
subsidiaries. If the Corporate Governance mechanisms like the board and Compensation Committee are more independent, the parent’s CEO has less power to influence their own compensation. Their variable compensation has less effect on the discretionary accruals of the subsidiaries. I hand-collect data regarding European material subsidiaries of American
MNC’s and test it for a negative association between percentage of parent’s CEO variable compensation, bonuses and stock-options, on subsidiary’s discretionary accruals. I also test the negative association of fully independent parent’s board and Compensation Committee on this association. I find evidence for bonus schemes, stock-options and the moderating effect of Corporate Governance mechanisms on stock-options. This research contributes to other research because my research is about parent-level CEO compensation and Corporate Governance to subsidiary-level FRQ, other researches are only about parent-level effects. I suggest parent’s board and Compensation Committee to be more fully independent because subsidiary’s FRQ is important for them. Future research can look for effects in other
countries, compare parent’s and domestic/foreign subsidiaries discretionary accruals or use more specified CEO-compensation variables.
Keywords
Subsidiary Financial Reporting Quality, Earnings Management, CEO Compensation, CEO Power, Corporate Governance, Board, Compensation Committee.
B.W.E.Collet@student.rug.nl Student number 2974762
Introduction
Multinationals Corporations (MNC’s) are striking in an integrated world economy
(Beuselinck, Cascino, Deloof & Vanstraelen, 2018). MNC’s all over the world are investing in subsidiaries in foreign countries, there are a lot of foreign direct investments (FDI) (UNCTAD, 2018). The World Investment Report of 2018 states that MNC’s of the United States have the most FDI outflows in 2017 and 2016, American MNC’s invest the most abroad. A lot of these American FDI outflows are going to developed countries like countries in the European Union, FDI inflows in the European Union are about 25% percent in the last few years and adds up to a few hundred billion dollars per year. The Netherlands is one of the favourite FDI inflow countries. In the fourth quarter of 2018, The Statistics Netherlands (CBS) reported that 80 percent of the FDI inflows are directly moved out of Netherlands, so these investments do not benefit the Dutch companies, they are only being used for the
benefits of the MNC’s (CBS.nl, 2018). CEO’s, the chief executive officers, the highest
executives of MNC’s, are getting compensated for the whole business group, the parent-company and their subsidiaries. Earnings Management is being used to increase their compensation, bonuses are higher because of higher earnings, and stock-options are being exercised through lifted share prices (New York Times, 2019). The average raise in
compensation of American CEO’s was 1.1 million dollars and this is a raise of 6.3 percent, which is a lot higher than the average American worker’s raise in compensation (New York Times, 2019). Stock-options are an important and still increasing part of the CEO
compensation (New York Times, 2019). Congress has set more restrictions for CEO-pay according to performance, still compensations keep increasing while scandals and logic would proof the opposite to be fair (New York Times, 2019). The problem is that the Corporate Governance of the company, the board of directors and the Compensation Committee, are supposed to represent the shareholders, but instead they serve the CEO and their management (Houston Chronicle, 2019). The Compensation Committee determines the right mix of salary, bonus and stock options, to stimulate the CEO’s to do what’s best for the company (Houston Chronicle, 2019). In the board of directors and the Compensation Committee there are a lot of retired CEO’s who aren’t fully independent, a recent CEO can influence the Compensation Committee to get the compensation he/she wants (Houston Chronicle, 2019).
Compensation is incentive for Earnings Management (Durnev, Li, & Magnan, 2017). The earnings of the subsidiaries can be material for the earnings in the consolidated statements of the business group that determines the parent’s CEO compensation (Beuselinck, Cascino, Deloof & Vanstraelen, 2018). The MNC-parents use subsidiaries since parents have shown to
substantially influence subsidiary-level decision-making with regards to financing and
investment policies (Bartlett & Ghoshal 1989) (Robinson & Stocken 2013) (Beaver, Cascino,
Correia & McNichols 2019). Dyreng, Hanlon and Maydew (2012) found evidence that
MNC’s use Earnings Management in and between parent-level and subsidiary-level. Earnings Management can be used as proxy for Financial Reporting Quality (FRQ), if there is Earnings Management, the FRQ is lower (Durnev, A., Li, T., & Magnan, M, 2017). FRQ refers to financial statements that provide accurate and fair information about the underlying financial position and economic performance of an entity, it is important for the user’s decision making to have a high FRQ, otherwise they do not make the most optimal decisions and this can be unbeneficial for the company (Herath & Albarqi, 2017). Most prior Earnings Management studies focus almost exclusively on consolidated financial statements (Dechow, Ge & Schrand 2010). There is less research about the specific effect of Earnings Management on MNC’s subsidiaries and their users, while there is Earnings Management in the subsidiaries for the benefit of the consolidated financial statements (Beuselinck, Cascino, Deloof &
Vanstraelen, 2018). This has a negative influence on the FRQ of the subsidiaries and the users who want to make decisions (Herath & Albarqi, 2017)
The aim of this study is to provide an answer to the following research question: What is the
association between the parent’s CEO variable compensation and the Financial Reporting Quality of the subsidiaries and which role can more independent Corporate Governance mechanisms play between them?
I predict that CEO variable compensation at parent-level negatively associates the FRQ of the subsidiaries. Managers such as CEO’s use Earnings Management to raise their compensation (Gong et al, 2019 and Dichev et al, 2013). Healy (1985) showed that variable compensation like bonus schemes and stock options are two highly used executive compensation
components that depend on financial accounting information. The efficient contracting theory shows that financial accounting information has a role to moderate the information
asymmetry between the agent and the principal, to make the contract as optimal as possible and let work the agent in the interest of the principal (Scott, 2014). The positive accounting theory explains the management accounting choices in response of these contracts and why managers act to maximize their own bonuses (Watts & Zimmerman, 1978). Many studies show that the CEO chooses opportunistic accounting procedures that inflate earnings to maximize bonus compensation (Balsam, 1998, Das et al., 2013 and Healy et al., 1987) and increase stock prices, which in turn maximizes equity-based compensation like stock-options
(Beneish & Vargus, 2002 and Sloan, 1996). I predict that a more independent board of directors has a negative association on the association between parent’s CEO variable compensation and the FRQ of the subsidiaries. Bebchuk, Fried & Walker (2002) set the power theory, CEO’s can have certain power to influence the Corporate Governance of a company to change their own compensation and monitor at the expense of shareholder value. The CEO has influence on the appointment of the outside directors in the board or the
Compensation Committee and even if directors are fully independent while others are less independent, fully independent directors will accept certain excessive CEO Compensation, since they are afraid for an anti-management reputation (Bebchuk et al. 2002). Cornett, Marcus & Tehranian (2008), Sun (2012) Brown & Lee (2010) and Bertrand & Mullainathan (2001) support this.
This research focuses on a sample of 2446 observations of subsidiaries incorporated in the EU who have parents in the United States of America. I use the discretionary accruals of the subsidiary for the subsidiary’s FRQ. For CEO variable compensation I use the percentage of variable compensation of the parent’s CEO versus constant compensation, the presence of bonuses and the value realized by options for parent’s CEO’s. For fully independent
Corporate Governance mechanisms I use the percentage of fully independent directors on the board and full independency of the Compensation Committee. I control for the effects: tenure of the parent’s CEO, presence of the Big Four at parent-level, performance, size, growth and leverage at the subsidiary-level and parent-level and year-, industry- and country effects. My findings suggest that parent’s CEO bonus presence and the value realized by options has a negative association with subsidiary’s FRQ, while the percentage of variable compensation of the parent’s CEO does not have a negative association. Fully independent Corporate
Governance mechanisms do moderate this association for the value realized by options only, so these Corporate Governance mechanisms look more after stock-price then after bonuses. The full independency of the board moderates the value of the options realized by the parent’s CEO and the subsidiary’s discretionary accruals negatively. The full independency of the Compensation Committee moderates the association between value options realized by the parent’s CEO and the negative subsidiary’s discretionary accruals.
Previous literature, like Gong et al., 2019, Dichev et al., 2013, Balsam, 1998, Das et al., 2013, Healy et al., 1987 has found evidence between CEO compensation and Earnings
Management, Marcus & Tehranian (2008), Sun (2012) Brown & Lee (2010) and Bertrand & Mullainathan (2001) have found evidence that Corporate Governance can moderate this
effect. This study is an addition to previous mentioned literature, because it has found
evidence between parent-level CEO variable compensation and subsidiary-level discretionary accruals and the moderating effect that fully independent Corporate Governance has on this association. There is a gap in research about the effect between parent-level and subsidiary-level Earnings Management due to a lack of available data about parents owning subsidiaries, but for this research this data is hand collected and that’s why I can make this contribution. MNC’s use their subsidiaries for Earnings Management so they can use the different characteristics of the countries incorporated subsidiaries (Bushman & Piotroski, 2006) and monitors and auditors have more distance to monitoring if subsidiary-level is used instead of parent-level (Choi, Kim, Qiu & Zang, 2012) (Ayers, Ramalingegowda & Yeung, 2011). Earnings Management in subsidiaries is bad for the FRQ of the subsidiaries and the decision-making of the users (Herath & Albarqi, 2017). This research gives practical implications that companies should make their Corporate Governance fully independent to shrink the
association between parent’s CEO value of exercised options and subsidiary’s FRQ, since users of financial information can make decisions that can benefit the subsidiary and the parent. The American Government wants that financial reporting of their MNC’s is a faithful representation for everybody. They could make a law which makes it mandated to put a worker representation in Corporate Governance mechanisms like the board and/or the Compensation Committee. They are more independent from the CEO and want to decrease the CEO-to-worker-compensation ratio
The next section provides a literature background and develops the hypotheses. Following, there is a section that describes the data and the FRQ, compensation and Corporate
Governance measures. There is also a section that provides results of empirical tests and the last section concludes.
Literature background FRQ subsidiaries
According to the Financial Accounting Standards Board (FASB) and the International
Accounting Standards Board (IASB), Financial Reporting Quality (FRQ) represents financial statements that provide accurate and fair information about the underlying financial position and economic performance of an entity (Herath & Albarqi, 2017). Financial information must be relevant as annual reports have a crucial role in determining the level of relevance by disclosing information about business opportunities and risks and providing feedback on how major market events and significant transactions affected entities (Beest, Braam, & Boelens, 2009). Financial information should be reliable because information should be free from bias and material mistakes (Cheung, Evans & Wright, 2010). Financial information must be comparable so you can compare the changes in reports across time and with other companies (Herath & Albarqi, 2017). Financial information must be understandable since the easier it is for the users to understand the information, the higher the quality that will be achieved (Cheung, Evans, & Wright, 2010). Financial information should be timely, because information must be available to decision makers before losing its powerful and good influences (Herath & Albarqi, 2017). Users like investors and debtholders use the Financial Information of companies for their decision-making which can benefit the company, so FRQ is important for users (Herath & Albarqi, 2017). Financial information must be a faithful representation because it reflects and represents the real economic position of the financial information that has been reported (Herath & Albarqi, 2017).
Financial information isn’t a faithful representation if an entity extensively engages more Earning Management and the FRQ of the entity lowers (Herath & Albarqi, 2017). Healy & Wahlen (1999) states that Earnings Management occurs when managers use judgement in financial reporting and in structuring transactions to alter financial reports to either mislead stakeholders about the underlying economic performance of the company or to influence contractual outcomes that depend on reported accounting numbers. Scott (2014) states that a way to mislead stakeholders is a manager using accounting policies affecting earnings in order to achieve specific earnings objective. Watts (1977) states that one of these specific objectives is to influence and increase the number of bonuses a manager wants by
manipulating accruals and selecting accounting procedures. The manager wants to obtain private gain like more compensation (Schipper, 1989). Managing these earnings isn’t illegal so managers can use Earnings Management without breaking the law (Dechow & Skinner, 2000). Managers can lower accruals one time and another time they can higher accruals for
this compensation (Scott, 2014). Jones (1991) created a model that measures these accruals, these accruals are items like amortization expenses, accounts receivables, inventories,
accounts payable and accrued liabilities, these accruals are used for Earnings Management. If these accruals in the Financial Reporting are changed over time for the compensation of managers, this information is less timely because managers can lower/higher accruals with accounting policies, so they do not reflect their real value through time (Herath & Albarqi, 2017). Financial reporting is not comparable, reliable and relevant since financial information isn’t a faithful representation thanks to Earnings Management, so it is harder to see the real value of these accruals (Herath & Albarqi, 2017). If the accruals are managed, the share price does not either reflect the real value of a company, some users of the financial reports can’t unravel the Earnings Management and so the financial reporting isn’t understandable, reliable and relevant for these users of the firm (Scott, 2014). So, FRQ is less relevant, reliable, comparable, understandable, timely and a less faithful representation since accruals are being manipulated and this is less visible for stakeholders who use the financial reporting (Choi & Pae, 2011). Earnings Management can be used as a proxy for FRQ (Durnev, Li & Magnan, 2017).
There can be several reasons why companies engage in Earnings Management while this isn’t good for their FRQ. Managers can inflate earnings, and accordingly, maximize their
compensation (Gong et al., 2019). Another motivation for Earnings Management can be to avoid debt-covenant violations. Sweeney (1994) and Watts & Zimmerman (1986) state that managers use income maximization to avoid debt covenant violations. Earnings Management can also be used to dodge political costs such as higher taxes or getting less grants, managers are likely to use Earnings Management to avoid these costs. Earnings Management is being used to avoid taxes by shifting earnings to offshore firms where there are weaker legal regimes and limited investor protections (Durnev, Li & Magnan, 2017). Another motivation for Earnings Management is to meet investors’ earnings expectation, this is the threshold management theory (Burgstahler & Dichev, 1997). Studies like Burgstahler & Dichev (1997) and Dechow & Skinner (2000) notice that there are unusually low frequencies of small decreases in earnings and small losses and unusually high frequency in small increases in earnings and small positive income, this happens because the managers smooth and maximize earnings to reach investors’ expectations. If a company plans to issue new or additional shares to the public, managers maximise their earnings in this period because this rises the amount of money received from the share issue (Cohen & Zarowin, 2010).
So, there are many reasons why companies use Earnings Management in their advantage. Most of these Earnings Management studies focus almost exclusively on consolidated financial statements (Dechow, Ge & Schrand 2010). Consolidated statements are the assets, liabilities, revenues and expenses of the whole business group, so from the parent and their subsidiaries. They manage earnings in their subsidiaries because they can be material for their consolidated statements and their motivations to practice Earnings Management (Beuselinck, Cascino, Deloof & Vanstraelen, 2018). The FRQ implication of individual financial reports of subsidiaries are largely unexplored (Beuselinck, Cascino, Deloof & Vanstraelen, 2018). While Dyreng, Hanlon & Maydew (2012) has found evidence that MNC’s use Earnings Management in and between parent and subsidiaries, Earnings Management is very high in the MNC’s foreign income. Results of Fan (2008) show that foreign earnings are mostly managed to avoid losses of the parent company by sending earnings to these subsidiaries. Subsidiaries in foreign countries with different legal/judicial systems, security laws and a different political economy, can use those differences to manage earnings a different way (Bushman & Piotroski, 2006). Parents use the subsidiaries to manage earnings as the auditors and monitors are more distant to the subsidiaries so it more difficult to audit and monitor them (Choi, Kim, Qiu & Zang, 2012) (Ayers, Ramalingegowda & Yeung, 2011). So, there are different reasons why parents use subsidiaries to manage earnings and manipulate the FRQ of the subsidiary. MNC-parents can use subsidiaries for these reasons since parents have shown to substantially influence subsidiary-level decision-making with regards to financing and
investment policies (Bartlett & Ghoshal 1989) (Robinson & Stocken 2013) (Beaver, Cascino,
Correia & McNichols 2016). If the earnings of the subsidiaries are managed, the consolidated statements are managed as well and this stimulates reporting objectives like managers’ compensation (Beuselinck, Cascino, Deloof & Vanstraelen, 2018).
Hypotheses development
A company must hire employees. The agency theory explains the system of hiring employees: the principal (the company) hires an agent (the manager) to perform services on their behalf which involves delegating some decision-making authority to the agent (Jensen & Mackling, 1976). Both parties want the best for themselves, so the company must compose a proper contract and make costs to monitor the managers, since the managers will not always act in the best interest of the company (Jensen & Mackling, 1976). The efficient contracting theory expands this agency theory, financial accounting information plays a role to moderate the
information asymmetry between the agent and the principal to make the contract as optimal as possible and let the agent work in the interest of the principal (Scott, 2014). Healy (1985) shows that variable compensation like bonus schemes and stock options are two highly used executive compensation components that depend on financial accounting information. The positive accounting theory explains the managers accounting choices in response to these contracts (Watts & Zimmerman, 1978) (Watts & Zimmerman, 1986). Managers act to maximize their and the companies’ benefits and therefore the decisions based on these
contracts have economic consequences (Watts & Zimmerman, 1986). The positive accounting theory explains bonus schemes, how managers can optimise their bonuses through Earnings Management. The compensation of more than 90% of top executives have an equity-based compensation plan as well as a bonus plan, these top executives are the CEO’s and they have high decision-power (Gaver, Gaver & Austin, 1995). Many studies show that the CEO chooses opportunistic accounting procedures that inflate earnings to (i) maximize bonus compensation (Balsam, 1998, Das et al., 2013 and Healy et al., 1987) and (ii) increase stock price, which in turn maximizes equity-based compensation like stock-options (Beneish & Vargus, 2002 and Sloan, 1996). Therefore, CEO’s might engage in Earnings Management activities to inflate earnings, and accordingly, maximize their compensation (Gong et al., 2019). According to a survey from Dichev et al. (2013), 88.62% of the CEO’s of public firms manage earnings to influence compensations.
For the bonus-plan of the CEO, Healy (1985) has observed that managers like the CEO always have inside information on the firm’s net income, outside parties do not have this information, managers can maximize their bonuses by using this information. Figure 1 shows a graph which illustrates the cap (the maximum of bonus managers can get) and bonus-bogey (a minimum for bonuses) in bonus schemes of managers. Healy (1985) has discovered
that when there is abonus-bogey (a minimum for bonuses) and managers do not have the
reported net income to reach this bogey, managers would use an Earnings Management patterns to lower the net income, since they do not reach their objective to get an increasing bonus. Managers move losses to the period the bogey is not reached, so the net income is higher in future periods and then they reach the bonus-cap. Managers may smooth reported earnings over time because they want to receive relatively constant bonuses. If the income is higher than the bonus-cap in one period, they can use income minimization as they do not get bonus for the reported income above the bonus-cap.
Figure 1 Bonus scheme
In combination with bonus plans, CEO’s and other managers can exercise stock-options as well, a stock-option provides the right to buy shares in the future for a (lower) fixed price which it is set at the grant date (Scott, 2014). CEO’s should raise the market value, that way they can buy shares for less than the market value. The profit a CEO makes with this option to buy shares for a lower fixed price is the intrinsic value of an option (Scott, 2014). Bartov & Mohanram (2004) and Baker, Collins & Reitinga (2009) have found evidence for abnormal high share prices and earnings at the time a stock-option can be exercised and after exercising the stock-options there was an abnormal decrease of earnings and an abnormal share price to compensate the earnings from before the stock-option. Efendi, Srivastava and Swanson (2007) have found evidence that companies where the CEO have stock-options, stock is overvalued. The bonus plans and equity-based compensation are most of the time linked to the consolidated statements of the business group (Scott, 2014).
CEO’s use Earnings Management in material subsidiaries to change the consolidated
statements and to meet compensation objectives (Beuselinck, Cascino, Deloof & Vanstraelen, 2018). Therefore, I formulate the following hypothesis.
H1: Parent’s CEO variable compensation is negatively associated with the Financial Reporting Quality of the subsidiaries.
Bebchuk, Fried & Walker (2002) set the power theory, CEO’s can have certain power to influence the Corporate Governance mechanisms of a company to change their own
compensation at the expense of shareholder value, these Corporate Governance mechanisms can be the board or the Compensation Committee. Boards consist of inside directors
(managers and employees of the firm) and outside directors (no connection to the company) (Hillman, 2003). The role of the board is advising and monitoring (Hillman, 2003). Outside directors, who are independent of the company’s business, are better at monitoring the
company then inside directors (Hillman, 2003). There is evidence that the managers are better monitored and use Earnings Management less if the board contains of more outside directors then inside directors (Klein, 2002). The Compensation Committee is an independent
Corporate Governance mechanism which determines a fair compensation for the CEO (Hermanson et al. 2012). The CEO can influence the appointment of the outside directors in the board or the Compensation Committee and even if directors are fully independent while the others are less independent, fully independent directors will accept certain excessive CEO Compensation, because there are afraid for an anti-management reputation (Bebchuk et al. 2002). Brown & Lee (2010) and Bertrand & Mullainathan (2001) support the power theory, they report a negative relation between Corporate Governance quality and excess of stock-options. Cornett, Marcus & Tehranian (2008) and Sun (2012) found the moderating effect that Corporate Governance independency has a negative association on the association between CEO variable pay and Earnings Management. These two corporate governance mechanisms (Board and Compensation Committee) influence the CEO’s and their compensation. If these corporate governance mechanisms are more independent, the CEO has less reason to
influence subsidiaries’ decision-making for consolidated compensation incentives. Therefore, I formulate the following hypothesis.
H2: More independent parent’s Corporate Governance mechanism are negatively associated with the association between parent’s CEO variable compensation and the Financial
I split this hypothesis in two sub-hypotheses because board and Compensation Committee are two different Corporate Governance mechanisms who influence the CEO in different ways. If the parent’s board is more fully independent, the CEO has less power to influence the
subsidiaries’ decision-making thanks to better monitoring of the board. Therefore, I formulate the first sub-hypothesis.
H2a: A more independent parent’s board is negatively associated with the association between parent’s CEO variable compensation and the Financial Reporting Quality of the subsidiaries.
If the parent’s Compensation Committee is more independent, the CEO’s have less power to influence their own compensation and they will get a fairer compensation. The CEO will have less incentives to manage earnings in subsidiaries to reach consolidated compensation
incentives, so the consolidated compensation incentives are fairer and less manageable.
H2b: A more independent parent’s Compensation Committee is negatively associated with the association between parent’s CEO variable compensation and the Financial Reporting Quality of the subsidiaries.
Research design Sample selection
My sample consist of 2446 observations, 1685 privately held, non-financial subsidiaries of 356 US MNCs from the years 2015, 2016 and 2017. The first step in the sample collection was to use Compustat database to obtain a list of non-financial listed MNCs from the US. Next, for each MNC and year, I collected the list of material subsidiaries from the Edgar Portal (Securities and Exchange Commission). Finally, based on the information on their jurisdiction, I selected only subsidiaries that are in the EU 28, Norway or Switzerland. The financial data of the subsidiaries is collected from the Orbis database. I downloaded information regarding the parent MNC’s from the MSCI GMI Ratings datasets in the WRDS database. This database contains information about the independency of Corporate
Governance mechanisms and how CEO’s get compensated. For the financial information of the MNC’s I used the Compustat database. Table 1 contains the sample selection, I deleted companies with missing information and financial companies because it is difficult to define abnormal accruals for financial services firms (Klein, 2002).
Table 1 Sample used in analyses
Firms
Initial Sample 31.065
Missing Corporate Governance information
(4.511)
Missing discretionary accruals information
(22.940)
Missing CEO compensation information
Financial companies
Missing financial information parents
(906)
(38)
(224)
Final sample 2.446
Variables
Earnings Management can be used as a proxy for the FRQ (Durnev, Li & Magnan, 2017). There are several models to detect Earnings Management. Dechow, Sloan & Sweeney (1995) has modified the Jones model and has showed that their modified model exhibits the most power in detecting Earnings Management, since the normal Jones model measures revenues instead of account receivables, which are easier manipulated by the management. The modified Jones model I use for Earnings Management is shown in figure 2.
Figure 2 Modified Jones model
= k1t + k2 + k3+ εit
TA are the total accruals and this is the change in current assets minus change in cash minus change in current liabilities plus change in short term loans minus deprecation
Assets i,t-1 is the lagged total assets
∆ Revenues is the change in revenues compared with last year ∆ Receivables is the change in receivables compared with last year PPE is the Property, Plant and Equipment
The ε is the error term of the Modified Jones Model from Figure 2 and is the proxy for Earnings Management, these are the discretionary accruals, in Figure 3 you can see how I measured the error term. I cross-sectional estimate this error term with years and industry.
Figure 3 Error term of modified Jones Model
εit = - k1t + k2 + k3
I winsorize this variable and all other continuous variables between 1% and 99% to get rid of outliers. I make the discretionary accruals an absolute value because earnings are being managed upwards or downwards. I call this variable discretionary accruals subsidiary (DACSUB).
I measure CEO variable compensation with three variables from the MSCI GMI Ratings dataset. I measure CEO variable compensation parent (CEOVCPAR) with one variable from the MSCI GMI Ratings dataset. Ceovariabletpm measures which percentage of the total CEO compensation is variable. This contains bonus compensation and equity-based compensation for the CEO. This variable is between 0 and 100. I specific look for the bonus compensation and equity-based compensation.
I look for the bonus compensation with a dummy variable to find out if a CEO gets a bonus or not. If the CEO compensation contains a number of bonuses, the dummy variable contains a 1 and if the CEO compensation does not contain a bonus the dummy contains a 0. I call this variable CEO bonus parent (CEOBPAR). This is a dummy since I do not have information about bonus schemes and their bogey and cap, so I look for an effect if a bonus is included in CEO compensation or not.
I look for the equity-based compensation at the value of options of the CEO that has been realized. I divided this value by the total assets of the company to control that CEO’s from bigger companies get higher options. I call this variable CEO option parent (CEOOPAR.)
Full independency Corporate Governance mechanisms
I measure full independency of the board with two variables from the MSCI GMI Ratings dataset. The first one is DirectorsOutside, calculated as the total number of fully independent directors on the parent’s board and I divide it to the DirectorsTotal variable which is the total amount of directors on the parent’s board. This value is between 0 and 1, where 1 would be a fully independent board and 0 a fully dependent board. I call this variable fully independent board parent (FIBPAR).
The second one is a dummy of the full independency of the Compensation Committee. If the Compensation Committee is fully independent the dummy variable contains a 1 and if the Compensation Committee isn’t fully independent the dummy variable contains a 0. I call this variable FICCPAR (fully independent Compensation Committee parent).
Ali & Zhang (2015) and Hou, Priem, Garanova (2014) have found evidence that if an CEO is longer tenured, there is less Earnings Management. Early CEO’s want to build up a reputation and use Earnings Management as a tool for this. This variable is called CEOTPAR.
The presence of Big four at the parent’s audit have a negative effect on Earnings Management since they are likely to be more experienced, have more resources to detect Earnings
Management and have more reputation risks (DeFond & Jiambalvo 1991, 1993; Becker et al. 1998; DeAngelo, 1981). I call this variable big four parent BIG4PAR.
BeDechow (1994), DeFond & Jimbalvo (1994), Dechow & Dichev (2002), Hribar & Nichols (2007), Roychowdhury (2006), Zang (2011) and Beuselinck et al. (2018) show factors that affect Earnings Management, specific for the parent or subsidiary. They found evidence that size, performance, growth and leverage of entities influences Earnings Management as smaller, loss-making, high-growth, leveraged entities exhibit higher values of absolute discretionary accruals. Size (SIZESUB) is the logarithm of total assets of the subsidiary. Performance (PERFSUB) is a dummy, if the subsidiary makes a profit it is a 1 and if it is loss-making it is a 0. Growth (GROWTHSUB) is the change in sales in the subsidiaries. Leverage (LEVSUB) is the total short-term and long-term debt of the subsidiary divided to the book value of the total assets of the subsidiary. Size (SIZEPAR) is the logarithm of total assets of the parent. Performance (PERFPAR) a dummy, if the subsidiary makes a profit it is a 1 and if it is loss-making it is a 0. Growth (GROWTHPAR) is the market to book value of parent’s equity. Leverage (LEVPAR) is the total short-term and long-term debt of the parent divided to the book value of the total assets of the parent.
I include year, country and industry effects because I use panel data and there can be different effects between years, industries and countries. The years 2017, 2016 and 2015 are dummy variables. The countries are dummy variables as well, included are Italy, Belgium, Norway, Denmark, Greece, Poland, Czechia, Germany, Romania, Spain, Finland, the Netherlands, Bulgaria, Austria, Portugal, Hungary, Sweden, Luxembourg, Estonia, Croatia and France. The industry-dummies are the GICSectors, they are Energy, Materials, Industrials, Consumer Discretionary, Consumers Staples, Health Care, Information Technology and Communication Services.
Empirical models
These are the empirical models. See appendix A for the definitions of the variables. To test the control variables, I use this model:
DACSUBi,t=β0+β1CEOTPAR i,t+ β2SIZESUBi,t + β3PERFSUBi,t + β4GROWTHSUBi,t +
β5LEVSUB + β6BIG4PARi,t+ β7SIZEPARi,t + β8PERFPARi,t + β9GROWTHPARi,t +
β10LEVPAR+ β11YEAR+ … + β13YEAR + β14COUNTRY+ … + β35COUNTRY+
β36INDUSTRY+…+β44INDUSTRY+ εi,t
To test Hypothesis 1, I use this model:
DACSUBi,t=β0+ β1CEOVCPAR i,t /CEOBPAR i,t /CEOOPAR i,t +β2CEOTPAR i,t+
β3SIZESUBi,t + β4PERFSUBi,t + β5GROWTHSUBi,t + β6LEVSUB + β7BIG4PARi,t+
β8SIZEPARi,t + β9PERFPARi,t + β10GROWTHPARi,t + β11LEVPAR+ β12YEAR+ … +
β14YEAR + β15COUNTRY+ … + β36COUNTRY+ β37INDUSTRY+…+β45INDUSTRY+ εi,t
To test Hypothesis 2, I use this model:
DACSUBi,t=β0+β1CEOVCPARi,t/CEOBPARi,t/CEOOPARi,t. +β2CEOVCPARi,t ×
FIBPARi,t/CEOVCPARi,t × FICCPARi,t/CEOBPARi,t × FIBPARi,t /CEOBPARi,t ×
FICCPARi,t,/CEOOPARi,t × FIBPARi,t /CEOOPARi,t × FICCPARi,t,+ β3CEOTPAR i,t+
β4SIZESUBi,t + β5PERFSUBi,t + β6GROWTHSUBi,t + β7LEVSUB + β8BIG4PARi,t+
β9SIZEPARi,t + β10PERFPARi,t + β11GROWTHPARi,t + β12LEVSUB + β13YEAR+ … +
Results
Descriptive statistics
In table 2 you can see the descriptive statistics of my sample.
Table 2 Descriptive statistics
Mean Median Std. Dev 25% 75% Min Max
DACSUB .078 .068 .055 .032 .120 0.001 0.195 CEOVCPAR 73.179 78.152 19.105 62.831 86.885 12.054 96.155 CEOBPAR .253 0 .435 0 1 0 1 CEOOPAR 244.636 0 596.715 0 77.472 0 2329.642 FIBPAR .808 .846 .118 .75 .9 .285 1 FICCPAR .887 1 .315 1 1 0 1 CEOTPAR 6.767 6 5.090 3 10 0 18 GROWTHSUB .811 .024 2.866 -.0741 .142 -.983 11.041 SIZESUB 10.007 9.948 1.845 8.723 11.196 3.895 16.442 PERFSUB 1 .848 .358 1 1 0 1 LEVSUB .270 .234 .221 .072 .424 0 .734 BIG4PAR .968 1 .174 1 1 0 1 GROWTHPAR 3.989 3.190 2.856 2.134 4.911 .942 12.026 SIZEPAR 8.763 8.596 1.543 7.711 9.665 4.386 12.309 PERFPAR .883 1 .320 1 1 0 1 LEVPAR .612 .620 .164 .493 .720 .289 .905
See appendix A for the definitions of the variables. On average the discretionary accruals in subsidiaries managed is 7.8%. The minimum and maximum is close to 0 and 0.2, it is close to zero through this variable is an absolute value and the maximum means that earnings are not managed more than 19.5% in subsidiaries.
CEOVCPAR is the percentage of the variable compensation divided by the total
compensation of parent’s CEO, the mean is 73% so on average 73% of the CEO’s get their compensation not as a constant compensation but as variable compensation related to their performance. The mean of CEOBPAR is 0.253, so 24% of the parent’s CEO’s get bonuses. The median of CEOOPAR is 0 so more than 50% of the CEO’s do not realized stock-options. FIBPAR is the percentage of directors in the parent’s board who are fully independent, you can see that the mean and the median are higher than 80%, so on average boards have more
than 80% fully independent directors. The mean of FICCPAR is 0.89, so 89% of the Compensation Committees in companies are fully independent.
The tenure of the CEO’s is on average 6.8 year and 50% percent of the CEO’s is tenured 6 years or less in the company. The subsidiaries are growing on average as the mean of GROWTHSUB is positive, so on average there is more growing than shrinking. The parents are growing as well on average as the mean is above 1, 1 means no change in the market value to book value for the parents. The performance of the subsidiaries and parents are positive on average. The PERFSUB’s and PERFPAR’s means are rather high (0.85 and 0.88), so on average most companies make profit instead of losses. The mean of LEVSUB, the leverage of the subsidiaries (0.27) is much lower than the mean of LEVPAR, the leverage of the parents (0.61), because subsidiaries are owned by the parents and do not need much long-term debts.
Almost every parent is audited by the Big Four because the mean of BIG4PAR is 0.96, this means that 96% of the parent-companies are audited by the Big Four.
Correlations
On the next page in Table 3 you see the correlations of the variables. See appendix A for the definitions of the variables. There are no significant strong correlations between the variables. Most significant correlations are around zero which means that the variables correlating to each other are none or very weak. There are a few significant moderate correlations, these correlation coefficients are between 0.3 and 0.5. FICCPAR, the independency of the parent’s Compensation Committee has a significant moderate positive correlation with FIBPAR, the independency of the parent’s board. So, if one Corporate Governance mechanism is more independent, another one will be more independent as well. SIZEPAR, the amount of assets of the parent has a significant positive moderate correlation with SIZESUB, the amount of assets of the subsidiary, so big parents and big subsidiaries correlate with each other. LEVPAR, the leverage ratio of the parents, has a significant positive moderate correlation with GROWTHPAR, the market to book value of the parent and SIZEPAR, the amount of assets of the parent. Leveraged parents correlate positive with growing and size.
Table 3 Correlations 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 (1) DACSUB 1 (2) CEOVCPAR -0.003 1 (3) CEOBPAR 0.056* 0.223* 1 (4) CEOOPAR 0.017 0.258* -0.081* 1 (5) FIBPAR -0.008* 0.166* 0.079* 0.067* 1 (6) FICCPAR -0.017 0.049* 0.081* 0.073* 0.385* 1 (7) CEOTPAR 0.030 0.217* 0.157* 0.224* -0.016 0.0245 1 (8) GROWTHSUB 0.013 -0.004 0.001 0.028 0.018 0.048* 0.004 1 (9) SIZESUB -0.111* 0.035 0.017 -0.089* 0.137* 0.048* -0.116* -0.018 1 (10) PERFSUB -0.008 0.046* 0.015 0.056* 0.039* 0.023 -0.001 0.002 0.049* 1 (11) LEVSUB -0.027 0.011 0.007 -0.014 0.001 -0.053* 0.027 0.017 -0.184* 0.004 1 (12) BIG4PAR -0.044* 0.067* -0.002* 0.021 0.079* -0.026 -0.126* 0.009 0.105* 0.002 -0.018 1 (13) GROWTHPAR -0.021 0.235* 0.073* 0.167* 0.147* 0.048* -0.001 0.019 0.056* 0.085* 0.020 0.099* 1 (14) SIZEPAR -0.025 0.220* 0.116* -0.096* 0.289* 0.121* -0.170* -0.017 0.426* 0.032 -0.052* 0.243* 0.167* 1 (15) PERFPAR -0.025 0.124* 0.065* -0.003 0.068* 0.073* 0.001 0.004 0.111* 0.059* 0.051* 0.087* 0.147* 0.207* 1 (16) LEVPAR -0.019 0.118* 0.055* -0.007 0.096* -0.038 -0.126* -0.018 0.165* -0.057* 0.039* 0.227* 0.344* 0.301* -0.027 1 *significance on a 0.05-level
Regression
In table 4 and 5 you can see the regressions for hypotheses 1 and 2.
Table 4 Regression hypothesis 1
VARIABLE/MODEL (1) (2) (3) (4) CONSTANT .118*** .119*** .120*** .120*** CEOTPAR .001 .001 .001 .001 GROWTHSUB .002 .002 .002 .002 SIZESUB -.004*** -.004*** -.004*** -.004*** PERFSUB -.001 -.001 -.001 -.001 LEVSUB -.012** -.012** -.012** -.012** BIG4PAR -.012* -.012* -.012* -.012* GROWTHPAR -.001 -.001 -.001 -.001 SIZEPAR .002* .002* .002* .002* PERFPAR -.002 -.002 -.002 -.002 LEVPAR -.003 -.003 -.003 -.003 CEOVCPAR -.001 CEOBPAR .007*** CEOOPAR .001* ADJUSTED R-SQUARED 0.014 0.014 0.016 0.016 F-VALUE 4.44*** 4.04*** 4.71*** 4.37*** ROOT MSE .055 .055 .054 .054 OBSERVATIONS 2,446 2,446 2,446 2,446 *Significance on a 0.10 level **Significance on a 0.05 level *** Significance on a 0.01 level
Table 5 Regression hypothesis 2 VARIABLE/MODEL (5) (6) (7) (8) (9) (10) CONSTANT .119*** .124*** .118*** .123*** .114*** .121*** CEOTPAR .002 .002 .001 .001 .002 .001 GROWTHSUB .003 .003 .003 .003 .003 .003 SIZESUB -.004*** -.004*** -.004*** -.004*** -.004*** -.004*** PERFSUB -.001 -.001 -.001 -.001 -.001 -.001 LEVSUB -.012** -.012** -.012** -.012** -.012** -.012** BIG4PAR -.012* -.012* -.012* -.012* -.012* -.012* GROWTHPAR -.001 -.001 -.001 -.001 -.001 -.001 SIZEPAR .002* .002* .002* .002* .002* .002* PERFPAR -.002 -.002 -.002 -.002 -.002 -.002 LEVPAR .003 .003 .003 .002 .003 .002 CEOVCPAR -.001 -.001 CEOBPAR .020 .008
CEOOPAR .001* .001 FIBPAR .001 .004 0.006 FICCPAR -.004 -.004 -.003 CEOVCPAR x FIBPAR .001 CEOVCPAR x FICCPAR .001 CEOBPAR x FIBPAR -.016 CEOBPAR x FICCPAR -.001 CEOOPAR x FIBPAR -.001* CEOOPAR x FICCPAR -0.001 ADJUSTED R-SQUARED 0.013 0.013 0.016 0.016 0.014 0.013 F 3.42*** 3.52*** 3.84*** 4.08*** 3.66*** 3.54*** ROOT MSE 0.054 0.054 .0544 .0544 .054 .054 OBSERVATIONS 2,446 2,446 2,446 2,446 2,446 2,446 *Significance on a 0.10 level **Significance on a 0.05 level *** Significance on a 0.01 level
See appendix A for the definitions of the variables. Country, industry and year effects are included in both table 4 and 5. Table 4 shows the coefficients in a linear regression for the control variables and hypothesis 1. Model 1 contains the regression of the control variables on the subsidiary’s discretionary accruals estimated with the modified Jones model. Model 2, 3 and 4 show the results of the hypothesis 1 for the effect of CEO variable compensation and their components bonus compensation and equity-based compensation of the parent’s CEO on the discretionary accruals of the subsidiary. The results of model 1 indicate that the size of the subsidiary is significant negatively associated with the discretionary accruals reported by subsidiaries. Leveraged subsidiaries are significant negatively associated with the
discretionary accruals reported by subsidiaries. Presence of the Big Four as an auditor of the parent is suggestive significant negatively associated with the discretionary accruals reported by subsidiaries If the subsidiaries are bigger, more leveraged and the parent is being audited by the big Four, the subsidiary’s discretionary accruals are decreasing, so the Earnings Management is lower and the FRQ of the subsidiaries is higher. The size of the parents is suggestive significant positive associated with the discretionary accruals reported by
subsidiaries. If the parent is bigger, the subsidiary’s discretionary accruals are increasing, so the Earnings Management is higher and the FRQ of the subsidiaries is lower. Model 2 does not give a significant association between the CEO variable compensation and subsidiary’s discretionary accruals. Model 3 gives highly significant positive association between presence of a bonus scheme for a CEO and subsidiary’s discretionary accruals. If the parent’s CEO gets a bonus, there are more discretionary accruals in the subsidiary since parents use subsidiaries for Earnings Management and if there is more Earnings Management in subsidiaries, the FRQ of the subsidiaries becomes lower. Model 4 gives a suggestive significant positive association between presence of a bonus scheme for a CEO and subsidiary’s discretionary accruals. If the parent’s CEO exercised value for stock-options, there are more discretionary accruals in the subsidiary because parents use subsidiaries for Earnings Management, if there is more
Earnings Management in subsidiaries, the FRQ of the subsidiaries will be lower. The adjusted R-squared in the models is approximately 1.5 percent of all models and that means the
variables explains 1.5% the change of subsidiary’s discretionary accruals. The high
significance of F means that the models are significant and the independent variables reliably predict the dependent variable. The ROOT MSE measures the accuracy of the model, 0 means a perfect model, it observes if the data points are equal to the model’s predicted values. 0.55 means the model isn’t bad predictive, however it isn’t a very good predictive model either.
Hypothesis 1, Parent’s CEO variable compensation is negatively associated with the Financial Reporting Quality of the subsidiaries can’t be fully accepted. CEOVCPAR, the percentage of variable compensation of the parent’s CEO does not have a significant negative association with the FRQ of the subsidiaries. CEOBPAR, the bonus-component of the parent’s CEO compensation, if there is a bonus scheme or not, is negatively associated with the FRQ of the subsidiaries. If the parent’s CEO compensation contains a bonus scheme, there are more discretionary accruals in the subsidiary, which means more Earnings Management, which means less FRQ in the subsidiary. There is a negative association between parent’s CEO bonus scheme and subsidiary FRQ. CEOOPAR, the equity-based component of the parent’s CEO compensation, if the parent’s CEO exercised stock-options and their value, is suggestive negatively associated with the FRQ of the subsidiaries. If the parent’s CEO exercised more value for stock-options, there are more discretionary accruals in the subsidiary, which means more Earnings Management, which means less FRQ in the subsidiary. There is a negative association between parent’s CEO bonus and equity-based component and subsidiary FRQ. These models only have an adjusted R-squared of 0.016, so the independent variables do not influence the dependent variable (FRQ of the subsidiaries) a lot, however they are negative associations. Hypothesis 1 can be partially accepted for the components of parent’s CEO compensation, not for the percentage of parent’s CEO compensation.
Table 5 shows the linear regression for hypothesis 2. Models 5 and 6 shows the results for hypothesis 2 for parent’s CEO variable compensation and the effect of fully independent Corporate Governance mechanisms like the board and the Compensation Committee. Model 7 and 8 do the same for the parent’s CEO bonus dummy and model 9 and 10 do this for the value of options exercised by the parent’s CEO.
Only model 9, the full independency of the board members, has a suggestive significant negative association on the association between the value of option exercised by the parent’s CEO and the subsidiary’s discretionary accruals. If the board is more fully independent, the association the value of exercised options, the equity-based component of the parent’s CEO compensation, has on subsidiary’s discretionary accruals are lower, the Earnings Management at subsidiary-level is lower and the FRQ at subsidiary-level is higher.
Hypothesis 2, which predicted that more independent Corporate Governance mechanisms are negatively associated with the association between parent’s CEO variable compensation and the Financial Reporting Quality of the subsidiaries cannot be accepted. Just in model 9 is CEOOPAR x FIBPAR suggestive significant negatively associated, a more fully independent board has a suggestive negative association on the association between the equity-based component of parent’s CEO variable compensation and the discretionary accruals of the subsidiary. If this Corporate Governance mechanism is more independent, the association between parent’s CEO variable compensation and subsidiary’s discretionary accruals is lower, which means that there is less Earnings Management at subsidiary-level and a higher FRQ at subsidiary-level. This model has an adjusted R-squared of 0.013 so the independent variables do not influence the dependent variable (FRQ of the subsidiaries) a lot, however it still is an association. Hypothesis 2a ‘A more independent parent’s board is negatively associated with the association between parent’s CEO variable compensation and the Financial Reporting Quality of the subsidiaries’ can only be partly accepted for the equity-based CEO’s compensation. Hypothesis 2b ‘A more independent parent’s Compensation Committee is negatively associated with the association between parent’s CEO variable compensation and the Financial Reporting Quality of the subsidiaries’ can’t be accepted.
Additional test
I use the absolute value of subsidiary’s discretionary accruals, so it does not matter if Earnings Management is downwards or upwards. For these additional tests I check the difference between upwards and downwards Earnings Management of foreign subsidiaries. I use the same models with the same variables and in table 6 you can see the coefficients of the most important variables for the two new samples containing only the subsidiaries with upwards Earnings Management and downwards Earnings Management.
Table 6 Regression upwards and downwards Earnings Management Upwards Downwards (2) CEOVCPAR -.001 .001 (3) CEOBPAR .005* .008* (4) CEOOPAR .001 .001 (5) CEOVCPAR x FIBPAR .001 -.001 (6) CEOVCPAR x FICCPAR -.001 -.001 (7) CEOBPAR x FIBPAR -.047 .025 (8) CEOBPAR x FICCPAR -.005 .003 (9) CEOOPAR x FIBPAR -.001 -.001 (10) CEOOPAR x FICCPAR .001 -.001* Observations 1,252 1,194 *Significance on a 0.10 level **Significance on a 0.05 level *** Significance on a 0.01 level
The significant positive association we find for CEOBPAR at subsidiary’s discretionary accruals, is as well suggestive significant for the positive association for upwards and downwards Earnings Management. CEOOPAR is not suggestive significant anymore for upwards and downwards Earnings Management. This confirms the partially acceptance of hypothesis 1 for at least the bonus component. At the regression with upwards and
downwards Earnings Management combined, I have found in model 9 that CEOOPAR has a suggestive positive association with subsidiary’s accruals and CEOOPAR x FIBPAR is significant negatively associated with subsidiary’s accruals, a more fully independent board has a negative association on the association between the equity-based component of parent’s CEO variable compensation and the discretionary accruals of the subsidiary. For downwards Earnings Management we find a suggestive negative moderating effect for the Corporate Governance mechanisms Compensation Committee instead of the board. CEOOPAR is significant suggestive positive associated with subsidiary’s accruals and CEOOPAR x FICCPAR is significant suggestive negatively associated with subsidiary’s accruals. A more fully independent board has a negative association on the association between the
equity-based component of parent’s CEO variable compensation and the negative discretionary accruals of the subsidiary.
Discussion and Conclusion
The aim of this study has been to research the CEO compensation of MNC’s and how this CEO compensation at parent-level has an association on the Financial Reporting Quality of subsidiaries. I predicted that variable compensation of the CEO at the parent-level would have a negative association with the subsidiary’s FRQ. MNC’s use their subsidiaries to manage earnings which negatively affects the FRQ. The CEO manages earnings as he/she gets paid by performance. I predicted that if the parent’s Corporate Governance mechanisms such as the board and Compensation Committee are more (fully) independent, it would have a negative association on the negative association between variable compensation of the CEO at the parent-level and the subsidiary’s FRQ. The CEO has less power when Corporate Governance is more (fully) independent as it is more difficult for the CEO to influence fully independent Corporate Governance mechanisms.
Findings
Hypotheses 1 and 2 are partly accepted. Hypothesis 1 ‘Parent’s CEO variable compensation is negatively associated with the Financial Reporting Quality of the subsidiaries’ is only
accepted for the separate components of variable compensation and not for the percentage of variable compensation. The presence of a bonus and the value of exercised stock-options in the parent’s CEO compensation has a significant positive association at subsidiary’s
discretionary accruals. If a parent’s CEO has a bonus scheme or exercised stock-options in their compensation, the discretionary accruals are higher, which means there is more Earnings Management in the subsidiary, which results in less FRQ of the subsidiary. I found a
significant suggestive positive association for the presence of a bonus scheme in the parent’s CEO compensation for both upwards and downwards Earnings Management. Parent’s CEO‘s suggestively manage the earnings upwards and downwards to reach the bonus gap.
Hypothesis 2 ‘Fully Independent Corporate Governance is negatively associated with the association between parent’s CEO variable compensation and the Financial Reporting Quality of the subsidiaries’ is also partly accepted as the full independency of the board has an
negative association between the equity-based component of parent’s CEO variable compensation and the subsidiary’s discretionary accruals. If the board becomes more fully independent, the association that the parent’s CEO equity-based compensation has on subsidiary’s discretionary accruals is smaller and the discretionary accruals in the subsidiary are less, which means there is less Earnings Management and a higher FRQ of subsidiaries. I
also found a significant suggestive negative association for the full independency of the Compensation Committee on the negative association between the equity-based component of Parent’s CEO variable compensation and the subsidiary’s negative discretionary accruals. The board and Compensation Committee have a moderating effect only for the equity-based component of the parent’s CEO compensation.
Hypothesis 2a containing the board and hypothesis 2b containing the Compensation
Committee, are both partly accepted only for the equity-based compensation based on stock-price. Although the hypotheses are partly accepted, the regressions in this research have a very low adjusted R-squared. The independent variables in my regressions explain around 1.5% of the change of subsidiary’s discretionary accruals. The independent variables may have the right association on the FRQ of subsidiaries, they do not explain the change in subsidiary’s discretionary accruals a lot.
Theoretical implications
For the theoretical implications of this research I expanded previous research and made an extended contribution. The results for hypothesis 1 are in line with existing results of previous literature. Gong et al (2019) and Dichev et al. (2013) have found evidence that CEO’s manage earnings for compensation matters as well. Harakeh et al (2019) also has found evidence for overall CEO variable compensation. Balsam (1998), Das et al. (2013) and Healy et al. (1987) have all found evidence for the bonus compensation. Beneish & Vargus (2002) and Sloan (1996) have found evidence for equity-based compensation like stock-options. The results of hypothesis 2, the moderating association of the fully independent Corporate Governance between CEO variable compensation and discretionary accruals, are in line with Marcus & Tehranian (2008) and Sun (2012). More specific, the moderating effect of full independency of Corporate Governance mechanisms on stock-options and Earnings Management, are in line with Brown & Lee (2010) and Bertrand & Mullainathan (2001). This research contributes to previous literature, I have found evidence between parent-level CEO compensation and subsidiary-level discretionary accruals. These subsidiary discretionary accruals are important for the consolidated statements for parent’s CEO compensation. MNC’s use their subsidiaries for Earnings Management since they can use the different characteristics of the countries incorporated subsidiaries (Bushman & Piotroski, 2006) and monitors and auditors are more distant in monitoring if subsidiary-level is used instead of parent-level (Choi, Kim, Qiu & Zang, 2012) (Ayers, Ramalingegowda & Yeung, 2011).
Practical implications
This research has practical implications. If the parent’s CEO gets no bonus or exercised less value in options, the discretionary accruals of the subsidiary are lower and the FRQ of the subsidiaries is higher. Unfortunately, companies need variable compensation for their managers because of the efficient contracting theory, variable compensation works with financial accounting information to make a contract as optimal as possible, so the CEO works in the best interest of the company (Scott, 2014). It is hard to implicate practical changes in variable compensation, otherwise the CEO does not have incentives to work in the best interest of the company. A more efficient practical implication to reduce Earnings Management for the equity-based compensation is to make the Corporate Governance
mechanisms more fully independent. The independent directors of the board and the members of the Compensation Committee do not have any ties to the company or the CEO, otherwise the CEO gets more power according to the power theory (Bebchuk, Fried & Walker, 2002). The more powerful the CEO is, the more he/she can influence the way he/she gets
compensated with a less fully independent Compensation Governance mechanisms, the CEO uses Earnings Management as stock price and so options can be manipulated. It is important to get a fully independent Corporate Governance mechanisms which can judge CEO
compensations fairly, this negatively associates the negative association between parent’s CEO value of options exercised and Earnings Management, which is the subsidiary’s FRQ. Parents and their subsidiaries should make the independent directors on the parent’s board and the members on the parent’s Compensation Committee 100% independent to raise the FRQ of the subsidiaries. This is important for the subsidiary and for the parent who owns the
subsidiary, since the decisions of users of financial information of the subsidiaries can benefit the subsidiary and indirectly the parent (Herath & Albarqi, 2017). The American Government wants that financial reporting of their MNC’s is a faithful representation for everybody. They could make a law which makes it mandated to put a worker representation in Corporate Governance mechanisms like the board and/or the Compensation Committee. They are more independent from the CEO and want to decrease the CEO-to-worker-compensation ratio.
There are a few limitations in this research. There is a possibility that some observations contain errors thanks to measurement issues. I use databases and may exclude some current companies. I used various databases and they do not completely overlap, my sample is only of companies which have overlap between the Orbis, MSCI and Compustat databases. Databases may exclude non-surviving firms, there is a survivorship bias, databases normally omit
merged, acquired and bankrupt firms. Companies can have inconsistent classifications of accounts, companies can have permitted differences in reporting line items. Databases can have recording errors and different classifications for the same items. If you work with time series which my panel data does, there can be structural changes in companies, industries or in accounting methods in time which are hard to analyse. My sample only contains 2015-2017, results can be different in other years since 2015-2017 may have specific characteristics which I cannot analyse. Firms can have one-time significant events which can influence the variables, like goodwill impairment, this is hard to analyse. It is difficult to control these changes. Due to these limitations there is a chance that my sample is unrepresentative.
Future research
This research focuses on American MNC’s and European subsidiaries, further research can focus on the effect of parents and subsidiaries all over the world to check if the outcome of the hypotheses will be different. Future research could look at Earnings Management in the whole concern, at the parent and the subsidiary instead of only the subsidiary. While this research uses foreign Earnings Management in subsidiaries, future research could look for the associations between parent’s CEO variable compensation and discretionary accruals at the subsidiary-level and at the parent-level as well and compare them. Future research can also look for the associations between parent’s CEO variable compensation and discretionary accruals in domestic subsidiaries and compare these results with the foreign subsidiaries. In my research I use a dummy for bonus compensation since I didn’t have the information of bonus bogeys and caps in the bonus schemes of parent’s CEO’s. Future research can try to collect this data and research if Earnings Management is being used to receive the maximum number of bonuses instead of only having a bonus scheme. I use a linear regression for the association between variable compensation of the parent’s CEO and the discretionary accruals of the subsidiary, future research can research if there is an optimal value of parent’s CEO variable compensation that reduces discretionary accruals.
References
Ali, A., & Zhang, W. (2015). Ceo tenure and earnings management. Journal of Accounting
and Economics, 59(1), 60-79. doi:10.1016/j.jacceco.2014.11.004
Armstrong, C.S., Blouin, J.L., Jagolinzer, A.D. and Larcker, D.F., 2015. Corporate
governance, incentives, and tax avoidance. Journal of Accounting and
Economics, 60(1), pp.1-17.
Ayers, B. C., Ramalingegowda, S., & Yeung, P. E. (2011). Hometown advantage: The effects
of monitoring institution location on financial reporting discretion. Journal of
Accounting and Economics, 52(1), 41-61.
Baker, T. A., Collins, D. L., & Reitenga, A. L. (2009). Incentives and opportunities to manage earnings around option grants. Contemporary Accounting Research, 26(3), 649-672. Balsam, S. (1998). Discretionary accounting choices and CEO compensation. Contemporary
Accounting Research, 15(3), 229-252.
Bartlett, C. and S. Ghoshal. 1989. Managing across borders: The transnational solution. Boston, Mass: Harvard Business School Press.
Bartov, E. & Mohanram P., (2004) Private Information, Earnings Manipulations, and Executive Stock‐Option Exercises. The Accounting Review: October 2004, Vol. 79,
No. 4, pp. 889- 920.
Beaver, W., Cascino, S., Correia, M., & McNichols, M. (2019). Group affiliation and default prediction. Management Science, 65(8), 3559-3584. doi:10.1287/mnsc.2018.3128 Bebchuk, L. A., Fried, J. M., & Walker, D. I. (2002). Managerial power and rent extraction
in the design of executive compensation (No. w9068). National bureau of economic
research.
Becker, C. L., DeFond, M. L., Jiambalvo, J., & Subramanyam, K. R. (1998). The effect of audit quality on earnings management. Contemporary accounting research, 15(1), 1-24.
Beest, F., Braam, G., & Boelens, S. (2009). Quality of Financial Reporting: Measuring qualitative characteristics. NICE.
Beneish, M. D., & Vargus, M. E. (2002). Insider trading, earnings quality, and accrual mispricing. The Accounting Review, 77(4), 755-791.
Bertrand, M., & Mullainathan, S. (2001). Are CEOs rewarded for luck? The ones without principals are. The Quarterly Journal of Economics, 116(3), 901-932.
Beuselinck, C., Cascino, S., Deloof, M. and Vanstraelen, A. (2018) Earnings Management Within Multinational Corporations. The Accounting Review
Boffey, D. 2017. EU investigates Ikea after Dutch deals reduce tax bill by €1bn. [ONLINE] Available at: https://www.theguardian.com/business/2017/dec/18/eu-probes-ikea-after-dutch-deals-reduce-tax-bill-by-1bn. [Accessed 5 January 2019].
Brown, L. D., & Lee, Y. J. (2010). The relation between corporate governance and CEOs’ equity grants. Journal of Accounting and Public Policy, 29(6), 533-558.
Burgstahler, D., & Dichev, I. (1997). Earnings management to avoid earnings decreases and losses. Journal of Accounting and Economics, 24(1), 99-126. doi:10.1016/S0165-4101(97)00017-7
Burgstahler, D., Hail, L., & Leuz, C. (2006). The importance of reporting incentives: Earnings management in european private and public firms. The Accounting Review, 81(5), 983- 1016.
Bushman, R. M., & Piotroski, J. D. (2006). Financial reporting incentives for conservative accounting: The influence of legal and political institutions. Journal of accounting and
economics, 42(1-2), 107-148.
CBS. (2018). 80 percent of inward investments channelled out directly. [ONLINE] Available at:
https://www.cbs.nl/en-gb/news/2018/50/80-percent-of-inward-investments-channelled-out-directly. [Accessed 5 January 2019].
Cheung, E., Evans, E., & Wright, S. (2010). An historical review of quality in financial reporting in Australia. Pacific Accounting Review, (0114-0582).
Choi, J. H., Kim, J. B., Qiu, A. A., & Zang, Y. (2012). Geographic proximity between auditor and client: How does it impact audit quality?. Auditing: A Journal of Practice &
Choi, T., & Pae, J. (2011). Business Ethics and Financial Reporting Quality: Evidence from Korea. Journal of Business Ethics, (10551-011-0871-4).
Cohen, D. A. and Zarowin, P. (2010) Accrual-Based and Real Earnings Management Activities Around Seasoned Equity Offerings, Journal of Accounting and Economics,
50(1), pp. 2–19.
Cohen, J. R., Gaynor, L. M., Krishnamoorthy, G., & Wright, A. M. (2011). The impact on auditor judgments of CEO influence on audit committee independence. Auditing: A
Journal of Practice & Theory, 30(4), 129-147.
Congress.gov. 2017. H.R.1 - An Act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018. Available at:
http://ec.europa.eu/competition/elojade/isef/case_details.cfm?proc_code=3_SA_46470 [Accessed 19 June 2019].
Cornett, M., Marcus, A., & Tehranian, H. (2008). Corporate governance and pay-for-performance: The impact of earnings management. Journal of Financial
Economics, 87(2), 357-373. doi:10.1016/j.jfineco.2007.03.003
Das, S., Hong, K., & Kim, K. (2013). Earnings smoothing, cash flow volatility, and CEO cash bonus. Financial Review, 48(1), 123-150..
DeAngelo, L. (1981). Auditor Size and Audit Quality. Journal of Accounting & Economics, 183-199.
Dearborn, M. (2009). Enterprise liability: Reviewing and revitalizing liability for corporate groups. California Law Review, 97, 195-261.
Dechow, P. M. (1994). Accounting earnings and cash flows as measures of firm performance: The role of accounting accruals. Journal of accounting and economics, 18(1), 3-42. Dechow, P.M. & Dichev, I., (2002) The quality of Accruals and Earnings: The Role of
Accrual Estimation Errors, The Accounting Review, Vol. 77, No. 1, pp. 51-89.
Dechow, P.M. & Skinner, D.J. (2000), “Earnings management: Reconciling the views of accounting academics, practitioners, and regulators”, Accounting Horizons, Vol. 14 No. 2, pp. 235-250.
Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1995). Detecting earnings management. Accounting review, 193-225
