MSc A&C-Controlling
Subsidiary Earnings Management and the role of Parent Risk and Institutional
Quality
Supervisor: Simona Rusanescu
Abstract: Using a sample of multinational corporations (MNCs) with parents in the United
States and Subsidiaries in Europe I examine the relationship between Parent Risk and Subsidiary Financial Reporting Quality (FRQ). I argue that Parent Risk causes Earnings Management in Foreign Subsidiaries because risk from investing in R&D triggers the MNC to engage in earnings management, and depending on parent-subsidiary differences, it is more opportune to manage earnings in subsidiaries. Also, I examine the relationship between Subsidiary Country Institutional Quality and Subsidiary FRQ to see what is the influence of subsidiary country level characteristics on earnings management in order to enhance understanding of factors influencing earnings management in subsidiaries. I measured parent risk with the Research and Development (R&D) expenditures, and earnings management with discretionary accruals. I find that Parent Risk is positively associated with subsidiary earnings management suggesting that parent risk is detrimental to subsidiary FRQ. Concerning the second relationship, I employed the World Bank Rule of Law Index estimates as my measure of Subsidiary Institutional Quality. I find mixed results but additional tests indicate that Rule of Law is negatively associated with subsidiary earnings management which suggests that rule of law deters subsidiary EM. My findings provide insight into the relationship between parent risk and subsidiary earnings management and subsidiary country characteristics’ and earnings management.Table of Contents:
- Cover Page - Abstract -1 Introduction
-Research Question
-2 Theoretical Background & Hypotheses Development -FRQ
-Earnings Management -Risk
-Risk & Earnings Management -Hypothesis 1 -Institutional Quality -Hypothesis 2 -Conceptual Model -3 Methodology -Sample Selection -Main Variables -Models -4 Results -Descriptive Statistics -Pearson Correlations - Regression Results -5 Discussion and Conclusion
-Findings -Implications
-Limitations & Further Research -References
1. Introduction
This study examines the relationship between Risk and Financial Reporting Quality (FRQ), and between subsidiary Institutional Quality (IQ) and FRQ, both within the context of Multinational Corporations (MNCs). Multinational Corporations (MNCs) are best defined as broad entities characterized by a high degree of organizational complexity. For conceptualization purposes, MNCs are best represented by one parent company and many subsidiaries across different geographical locations. The rise in popularity of MNCs has been well documented, especially in more recent years as there has been a dramatic increase in MNCs from developing countries. Also, to put into context their relevance, MNCs accounted for one third of global trade in 2010 (World Investment Report, 2010). Given their growing relevance, economic significance, and global extent, MNCs are the focal point of this study.
Risk is a term closely associated with uncertainty and probability of success or failure. According to Sunder (2015) there are two main concepts of risk used in business and economics, risk as hazard or vulnerability, and a more formalized conception of risk as dispersion of outcomes. MNCs have an intricate relationship with risk primarily because of their organizational complexity (operating across many borders), among other factors (La Porta et al., 1998). Risk is a fundamental concept in this paper because it is one of my main independent variables, and I argue that it is detrimental for (Subsidiary) FRQ because in this context risk means exposure to loss or damage, as in Sunder (2015).
FRQ can be best conceived as the quality (or lack thereof) of the accounting numbers reported in financial statements responsible for communicating the financial position and economic performance of a business entity (Herath & Albarqi, 2017). Generally speaking, FRQ depends on reliability and accuracy of financial information. These traits can be compromised by earnings management (EM); Healy & Wahlen (1999) define EM as a practice that implies managerial discretion that alters financial reports.
The general convention is that risk and FRQ are negatively associated. For example, Chu et al. (2018) and Ham et al. (2017) illustrate how short auditor tenure and CEO narcissism, their respective measures of risk, are detrimental to the reporting quality of an organization. In situations of risk, which ultimately have an impact on financial statements, earnings management can attenuate the negative impact by portraying a more favorable economic reality. Hence, risk can
cause earnings management. I argue that laxer consequences from financial misreporting, and distance from regulating bodies incentivizes parents to manage earnings abroad and eventually differences in Institutional Quality, and different corporate tax rates provide the opportunity for parents to manage earnings in foreign subsidiaries (Beuselinck et al., 2019; Dyreng et al., 2012). In light of the above, I consider that A) Parents have the ability to influence subsidiary reporting quality and B) higher levels of risk prompts parents to engage in EM using their foreign subsidiaries to depict a more favorable economic reality. Consistent with my argument, Beuselinck et al. (2019) find that risk is closely associated with earnings management activity in subsidiaries. I build upon this argument but specifically focus on parent risk as the factor influencing the magnitude of subsidiary earnings management.
My study also investigates how the level of IQ in the subsidiary country influences the extent of EM. Subsidiary Institutional Quality is an aggregate concept representative of a subsidiary country’s laws, individual rights, government services, and regulations. In stand-alone firms, IQ influences FRQ because the concepts comprising IQ set the rules upon which economic activity happens and is regulated, including financial reporting. Therefore, it is perceived that there is an inverse relation between country institutional quality and earnings management (Ball et al., 2000; Leuz et al., 2003). In MNCs, the focus rests upon IQ differences between parent and subsidiaries and how these result in more or less earnings management. For example, Beuselinck et al. (2019) find that the degree of subsidiary earnings management is explained by cross country variation in institutional quality rather than only looking at institutional quality in the subsidiary. I tailor my study to focus on subsidiary country institutional quality as the potential explanatory variable for the magnitude of earnings management in subsidiaries. Additionally, I perform the same tests but I segment observations in relation to the parent’s level of IQ, so whether the subsidiary is above or below the parent’s IQ. Overall, I expect that subsidiary country IQ influences the degree of subsidiary EM, and that subsidiary country IQ is negatively related to subsidiary EM. In summary, my main research questions are:
What is the relationship between Parent Risk and Subsidiary FRQ, and between Subsidiary Institutional Quality and Subsidiary FRQ?
My empirical analysis is based on a sample of private European subsidiaries of US MNCs. I obtained financial information of subsidiaries and parent companies from Orbis and Compustat,
respectively. The data spans from 2012-2017. I follow Boubakri et al. (2013), Li et al. (2013) and Shust (2015) and measure parent risk as Research & Development expenditures scaled by total assets. Most studies (Claessens et al., 2000; Laeven & Levine, 2009) measure risk using volatility of ROA, but I use R&D expenditures because it is inherently risky and embodies risk in a straightforward fashion. Subsidiary FRQ is the absolute value of discretionary accruals estimated with the Jones (1991) model. Further, I measure Institutional Quality as Rule of Law using the World Bank Rule of Law Index (Kaufmann et al., 2011).
This study shows that MNC parent risk is positively and significantly related to subsidiary discretionary accruals. This suggests that an increase in parent risk results in an increase in subsidiary discretionary accruals, which in turn means that MNC parent risk diminishes the quality, preciseness, and reliability of the consolidated financial statements. My results also show that the level of rule of law in the subsidiary´s country is positively and significantly related to subsidiary discretionary accruals. This is both unexpected and not consistent with prior literature, as it implies that an increase in subsidiary rule of law results in an increase in subsidiary discretionary accruals. In order to corroborate my findings, I perform additional tests and after segmenting my observations, I find that subsidiary IQ is negatively associated with subsidiary EM. Such relation is stronger in the group with a higher IQ, relative to the parent’s IQ. The latter results demonstrate that subsidiary IQ deters subsidiary EM, and thus favors FRQ of the subsidiary and the MNC.
My study contributes by shedding light into the relation between risk and FRQ from an MNC perspective. This is relevant because MNCs amass a great deal of interest from investors, regulators, and the academic community due to their size, and place in global business. I find that parent risk influences earnings management activity in subsidiaries. My findings are parallel to that of other studies in this line of research like Beuselinck et al. (2019) and Dyreng et al. (2012), but these studies tend focus more on where earnings management takes place rather than looking at parent risk as a determinant of FRQ within MNCs. Furthermore, both studies rely on the modified Jones model whereas I use the Jones model (1991). Also, using R&D expenditures as my measure of risk sheds light into an alternative measure of risk not commonly employed in academic research and much less in a discussion of MNCs and FRQ.
My study also contributes to literature concerning FRQ within MNCs. As in Beuselinck et al. (2019); Dyreng et al. (2012) and Durnev et al. (2017) I also use the Rule of Law index from the
World Bank; however, my final sample of subsidiary countries consisted of only European nations whereas in the aforementioned studies the scope was broader and in some cases was further segmented into the top 15 countries with the strictest Rule of Law estimates. Lastly, my study highlights the importance of considering subsidiary and parent country institutional characteristics when trying to explain the magnitude of earnings management in an MNC setting, as outlined by Beuselinck et al. (2019).
The rest of the paper is structured in the following way: Section 2-Theoretical Background, where I present the theory underlying this research, and develop my hypotheses. Section 3-Research Design, where I describe my sample selection, main variables, and empirical models. Section 4-Findings and lastly Section 5-Discussion and Conclusions, where I discuss my findings, implications, limitations, and avenues for further research.
2. Theoretical Background & Hypotheses Development
Financial Reporting Quality (FRQ)
FRQ is an accounting topic which focuses on the quality of financial reports, including the processes leading up to the production of these reports. FRQ can be defined as, financial statements providing accurate and fair information about the economic performance and financial position of an entity (Herath & Albarqi, 2017). FRQ has received a lot of academic attention and has a vast number of proxies and determinants. Most determinants fall under the umbrella of corporate governance, some examples are: the composition of the board of directors, ownership structure, or CEO characteristics. FRQ is significant for investors and other users of financial reports because financial reports guide decision making (Herath & Albarqi, 2017). For example, a bank can determine an organizations’ interest rate on a bank note by looking at its leverage ratio1, accompanied by other metrics, and comparing it against industry peers or other predetermined levels. Using the leverage ratio as a benchmark for interest rates demonstrates that financial statements provide important information for users of financial reports which eventually have an impact on the organization. Thus the quality of reporting or lack thereof in terms of accuracy and reliability has an impact.
1 Leverage Ratio: Financial ratio that determines how much capital comes in as debt and assesses the firm’s ability
Earnings Management is a managerial practice which can compromise the accuracy and reliability of FRQ. Leuz et al. (2003) define earnings management as the alteration of firms’ reported economic performance by insiders to either mislead some stakeholders or to influence contractual outcomes. In general terms, the alteration of financial reports results in either increasing or decreasing revenues. Some accrual-based earnings management practices include: the change of accounting methods or assumptions, big baths, goodwill impairment, and the determination of timing of the revenue and expense recognitions (Fatma & Hidayat, 2019).
Earnings management can be utilized when the firm is fundamentally underperforming yet it seeks to maintain an earnings growth trend, and in order to meet or beat analyst expectations (Yao, 2017). Otherwise, investors might be reluctant to invest in companies neither meeting or beating analyst expectations as it signals a deviation from forecasted performance, which may reflect poorly on upper management. Another incentive highlighting the importance of income increasing EM is the avoidance of underwhelming equity valuations by analysts, and avoid debt covenant violation (Healy & Wahlen, 1999).
Other incentives for income increasing EM involve CEO compensation and CEO turnover. Often times CEO compensation packages carry stipulations in which payouts are determined by company performance tied to a specific metric, for example Net Working Capital or Organic Sales. Therefore, CEOs may use EM to boost figures in order to maximize their earnings (Dechow et al., 2010). Such measure also ensures job security for upper management, and thus is another appeal behind EM (Healy & Wahlen, 1999). The examples above highlight that managers are highly sensitive to earnings disappointment. Such concept is representative of a situation that occurs when an organization fails to meet internal earnings expectations. These consequences have a material effect on the organization in the form of higher interest costs, lower credit ratings, or lower valuations.
Alternatively, there are incentives for income-decreasing EM. For example, decreasing earnings can reduce an organizations’ corporate tax payments. Income-decreasing EM underestimates reported earnings and consequently reduces the amount of corporate taxes to be paid. Such strategy also allows the organization to use the remaining earnings as a buffer for future periods, especially in periods of poor performance.
There are many incentives for managers to use EM; however, there are negative consequences associated with EM too. EM causes a lot of reversing entries which if not properly
managed, and correctly flagged by accounting software can lead to errors in the accounting records capable of overstating or understating specific account balances. A great number of errors due to mismanaged reversals can result in mispricing of firms by equity and bond investors, eventually generating reluctance amongst investors and downgrading the organization’s corporate reputation (Dechow et al., 2010; Rodriguez-Ariza et al., 2016). Furthermore, EM causes information asymmetry between managers and external stakeholders, such as investors and financial analysts (Herath & Albaqi, 2017; Ge et al., 2016). The failure to disclose information relevant for investors can be considered fraudulent reporting, and aside from violating a number of ethics codes in accounting, it can lead to litigation such as a class action suit (Cutler et al., 2019).
Financial Reporting Quality of Multinational Corporations
Influence of the Parent
Parent corporation influence over subsidiaries has captivated a considerable degree of academic interest. For example, Beuselinck et al. (2019) argue that MNC parents exercise a great deal of influence over its subsidiaries operating, reporting, investing, and financing activities. The degree of parent influence is related to the level of autonomy of the subsidiary, and one way of measuring this is through ownership percentage. A parent achieves control when it owns at least 50.1% of the subsidiary voting rights. Furthermore, in their study of family controlled firms and earnings management in Italy, Bonacchi et al. (2018) found that when parents exercise control over subsidiaries, subsidiary managers use EM to align with parents’ goals. This statement demonstrates parent influence over subsidiary reporting matters. Control also allows parent corporations the right to appoint board members to a subsidiary and often times these can be shared directors. The latter are board members who hold a seat in the parent and in the subsidiary (Beuselinck et al., 2019). Bonacchi et al. (2018) found that shared directors coordinate earnings management in the subsidiaries. In this context, shared directors are a result of control and reflect parent influence over subsidiaries, including EM.
Parent-Subsidiary relationships are subject to the agency problem, also known as principal-agent dilemma. This is part of the broader Agency Theory, which addresses the dynamics in the relationship between principals and agents. Such theory underlines that problems between principals and agents arise when there is information asymmetry, goal divergence, and a pursuit of self-interests, among other causes. In stand-alone firms, this conflict involves managers and
shareholders yet in MNCs we are confronted with many managers operating across many different countries, under many different cultural norms, and often times pursuing different goals; thus making it difficult for shareholders to monitor managers. Also, increasing monitoring costs such as the timely access to information for shareholders. Parent control can effectively solve the agency problem and provide theoretical support concerning the extent of parent influence over the MNC. The degree of ownership, reflected by controlling rights, permits MNC parents to maintain and enhance fluid communication with subsidiaries and further facilitate the implementation of entity wide programs, rules, and regulations.
Both U.S GAAP and IFRS contain the accounting standards and principles needed when measuring and recognizing items and drawing up financial statements. Drawing up financial statements in MNCs is more complex than in stand-alone firms because statements must be consolidated. The consolidation process is an obligation undertaken by the MNC parent to present financial results as a single entity (Beuselinck et al., 2019). This process begins with reviewing all subsidiaries in which an MNC-parent has rights over the subsidiary, under IFRS 10 rights refer to voting rights, right to a return from the subsidiary, and the right to influence the amount of returns from the subsidiary (IAS Plus). If these conditions are met, then the parent must incorporate the subsidiaries into its financial position when submitting financial reports. This step is closely followed by adjustments in order to translate foreign currencies into the reporting currency of the parent. Lastly, all intercompany transactions2 are ruled out from the consolidated balances and ownership from entities other than the parent are outlined (Beuselinck et al., 2019). In summary, the FRQ of the consolidated financial statements is the result and reflection of the FRQ of the respective subsidiaries.
MNC-parents are incentivized to manage earnings in foreign subsidiaries due to laxer consequences from engaging in EM. Financial misreporting, including earnings management, can result in less punitive legal penalties for the MNC if this is done at the foreign subsidiary rather than at the parent. Put differently, the risk of engaging in earnings management is much lower for the MNC when such action is done at the foreign subsidiary. Legal enforcement falls upon the responsibility of members affiliated to government, these range from police to financial regulating bodies such as the SEC in the United States. Therefore, if the foreign subsidiary is of lower
institutional quality3 relative to the MNC parent, engaging in EM in the subsidiary can result in less punitive consequences (legal and financial) than if such action takes place at the MNC-parent (Beuselinck et al., 2019. In such case parent directors could even claim having no knowledge of such occurrence, and shield themselves behind the pretext that any financial misreporting was the result of organizational complexity, and goal divergence.
Another incentive for parents to manage earnings in foreign subsidiaries deals with distance. More specifically with distance between the parent financial regulating body and subsidiaries. Dyreng et al. (2012) noticed that, firms farther from SEC offices manage earnings more than those closer to SEC offices. The same argument holds for auditors. It has been found that monitoring subsidiary financials by MNC parent auditors is considerably much more difficult (Dyreng et al., 2012). For example, take the Big4 (PwC, EY, Deloitte, and KPMG) they are located in several countries, yet auditor quality amongst countries is not uniform mainly due to different academic backgrounds and different qualifications for exercising auditing duties. Local enforcement also plays a role in explaining the disparity amongst auditing quality. Hence the quality of the audit performed at the foreign subsidiary level is not par to the quality of the audit at the parent level, even though the same firm carries both out. Thus it is reasonable to assume that EM occurs more often in foreign subsidiaries because parents are incentivized by laxer consequences and distance from monitoring bodies, among other reasons.
There are many opportunities through which MNCs can manage earnings abroad, some are due to differences in institutional quality, and different corporate tax rates. Beuselinck et al. (2019) and Bonacchi et al. (2018) focus on the opportunities stemming from institutional quality differences; MNCs based in countries with strong institutions find it less costly to manage earnings abroad in subsidiary countries where institutional quality is lower. Moreover, subsidiary managers in countries with weak investor protection, a reflection of institutional quality or lack thereof, have been found to manage earnings to meet parent goals. Strong institutions represent a hurdle for MNCs wishing to engage in EM. For example, MNC parents in the US comply with the Securities and Exchange Commission (SEC) regulations, this body has more resources and qualified personnel than a financial regulating body in Latin America or the Caribbean. Therefore, it is easier
3 Institutional Quality: Broad concept encompassing a country’s laws, individual rights, and government services
for parents to manage earnings in a location where institutional quality, and as a result oversight, is lower and perhaps of lower quality.
Dyreng et al. (2012) bring forth that different corporate tax rates also provide MNCs with opportunities to manage earnings abroad. Dyreng et al. (2012) argue that MNCs can reduce corporate taxes by managing earnings in tax havens. Such statement assumes that the foreign subsidiary corporate tax rate is considerably less than that at the parent country, hence managing earnings in such jurisdiction with a favorable corporate tax rate is less of a strain on the MNC’s resources.
Risk
Risk has developed as a formal avenue for academic research across many fields, and on a general level risk is evaluated in terms of possibility of loss or other adverse scenarios (Chenhall, 2003). Firms are subject to many kinds of risks responsible for compromising the firm’s objectives which are divided into operations, reporting, and compliance, and across different levels: entity, division, operating unit, and function (COSO, 2013). Some examples of accounting risks capable of affecting a firm’s realization of its goals are: deficient internal controls, lack of compliance with financial regulating bodies, and managerial underperformance.
Firms operate within predetermined and agreed upon levels of risk, outlined by the firm’s risk appetite. Risk appetite is the amount of risk an organization willingly takes upon in order to achieve objectives (Barfield, 2013). The Risk Return Trade Off explains, from a financial perspective, the benefit behind embracing risk. This concept states that higher risk is linked with the probability of a higher return, whereas lower risk is associated with a greater probability of a smaller return. Aside from benefits, risk also has its drawbacks. For example, managing risk constrains the firm’s financial and human resources. In its simplest form, hedging is a form insurance against adverse outcomes (Cohen et al., 2017). These insurance or hedging costs constitute a considerable portion of risk management, like insurance policies. Similarly, implementing Enterprise Risk Management (ERM) frameworks such as COSO-ERM represent an expense in the form of salaries for qualified personnel and fees for compliance experts. Ultimately, acceptance or rejection (of risk) depends upon the firm’s resources and its risk appetite; if risk management costs prove unfeasible then the firm must retain that risk. Similarly, if the cost of bearing risk is significantly low and within its risk appetite, then the firm will accept it.
In accounting, the study of risk has led to the development of numerous measures and determinants of this concept. One of these measures is research and development (R&D) expenditures. Cherensky (1994) defines R&D as a complex and inherently risky activity comprised of two forms of research (basic and applied), the latter with a more commercial focus than basic, and development entails the translation of knowledge into products and processes. There are benefits behind R&D such as the development of new products or the improvement of current processes, however, its drawbacks are numerous. For example, R&D is commonly associated with a low probability of success and its benefits are often times distant and uncertain (Li et al., 2013). Cherensky (1994) summarizes the drawbacks behind R&D: the time window for realizing gains is very long, realizing benefits is much more difficult, and the technology is much more complex in terms of understanding.
Furthermore, Cherensky (1994) outlines three underlying risks behind R&D, “(1) There is the technology risk that a particular R&D project will not be technically successful; (2) There is the market risk that a particular R&D project will be technically successful but will not result in a profit for any firm; and (3) There is the appropriability risk that a particular R&D project will result in a profit, but not for the sponsoring firm” (pg. 323). In this context, appropriability refers to the firm’s ability to extract revenues from its R&D activities. Risks associated with R&D are varied, as shown by Cherensky (1994) these range from technological success/failure, to lack of profitability on the market. In addition, it is noteworthy to mention that R&D projects are relatively worthless until fully realized. In the case of liquidation or project failure, R&D projects have low to negligible collateral values because of the limited alternatives for these (Kothari et al., 2002). In comparison, tangible assets such as buildings and equipment are less risky in terms of collateral value. Kothari et al. (2002) found that if a firm replaces capital expenditures (money destined to buy, maintain, or improve fixed assets) with R&D expenditures, the firm’s earnings variance increase anywhere between 30-70%. This suggests that the difference between expected or budgeted earnings and actual earnings is enhanced by R&D expenditures.
Risk in MNCs
Conducting business across borders adds a layer of complexity to organizations and this is reflected in the risks affecting MNCs. Risks can flow from MNC parent to subsidiary and vice versa. Academic literature has focused on addressing determinants of risk originating both from
MNC parent and foreign subsidiaries. Some determinants of MNC risks originating in the subsidiary are related to differences in tax systems, regulations of labor, and foreign currency risk (La Porta et al., 1998). Meanwhile, some determinants of MNC risk originating in the MNC parent revolve around ownership structure (ex. family owned), CEO characteristics (such as age, background, and even marital status), and corporate governance (ex. size of audit committee and amount of outside directors). In this paper I structure my main argument around the perspective that parent risk has a great degree of influence over the MNC. I posit that my main measure of risk originates in the MNC parent and is capable of threatening the entire MNC. Investing risk measured by R&D expenditures can affect subsidiaries, and the MNC by: constraining access to internal capital and limiting access to external capital and thus threaten the viability of positive NPV projects, increasing earnings volatility of the entire MNC, negatively altering performance metrics which can have consequences for bonus payments and for financial analysts and outside investors (for ex. R&D margin4 and the price-to-research ratio5), and lastly the almost negligible collateral value of R&D expenditures can result in the liquidation6 of a subsidiary or the spin-off of a business unit in order to reduce liabilities (Cherensky, 1994; Kothari et al., 2002). In light of this, parent companies can function as both a source of risk and a bridge through which risk travels through an MNC.
Risk and Earnings Management in MNCs
Generally speaking, risk causes earnings management and earnings management causes risks. I focus on the former, and tailor my argument around risk from investing in R&D. Investing in R&D represents investing in projects whose benefits are uncertain; additionally, investing in R&D over much less uncertain categories such as Plant, Property, and Equipment, generates a lot of earnings volatility (Cherensky, 1994; Kothari et al., 2002; Pandit et al., 2011). Correspondingly, managers in organizations with considerable R&D expenditures such as Tech companies, have a great willingness to engage in EM to compensate for R&D expenditures, especially in cases in which the realization of gains seems uncertain, distant, or even negligible (Pandit et al., 2011;
4 R&D Margin: Measures the percentage of sales allocated to R&D expenditures, also known as R&D to revenue. 5 Price to Research Ratio (PRR): Measures the relationship between a company’s market capitalization and its R&D
expenditures.
6 Liquidation: In this context it refers to the process of selling all of a subsidiary’s assets to generate cash to pay off
Shust, 2015). In other words, managers can compensate R&D losses by engaging in EM, which compromises FRQ.
Prior literature documents that risk is negatively associated FRQ (Chu et al., 2018; Ham et al., 2017). Similarly, prior literature documents that parents influence the entire MNC (Beuselinck et al., 2019; Bonacchi et al., 2018). Thus, risk at the parent level also influences the MNC. The consolidation process reflects this; more concretely, the FRQ of the MNC is the result of the FRQ of individual subsidiaries along with that of the parent. In light of the above, I presume that conditions affecting the MNC parent must have an influence over foreign subsidiaries and the entire MNC. More concretely, risk surrounding the MNC parent incentivizes the parent to manage earnings. Parents finds it opportune to manage earnings away from home; thus prompting subsidiaries to engage in EM. This means that risk originating from the parent influences EM at the subsidiary level. Lastly, the consolidation of financial statements reflects how the FRQ of the MNC is compromised as this is a product of the subsidiaries’ FRQ.
Hypotheses Development
Most of the incentives for parents to manage earnings in foreign subsidiaries are the result of differences between the parent´s home country and the countries where subsidiaries are located. More concretely, most of the incentives and opportunities are due to: differences in institutional quality, different corporate taxes, distance from parent level auditors and parent level financial regulating bodies, and lastly, less burdensome financial and legal penalties in foreign subsidiary countries. Furthermore, I evaluate the linkage between risk and earnings management and as in Pandit et al. (2011) and Shust (2015), I argue that risk incentivizes EM in general and in MNCs because EM is an effective tool to attenuate and mask the impact of risks affecting the economic reality of the organization. In other words, managers rely on EM as a response to risks, such as losses from R&D expenditures, which negatively impact the financial position of an organization. Lastly, I elaborate on the relationship between parent risk and subsidiary FRQ. More specifically, I bring forth how parents influence the MNC, including foreign subsidiaries. These arguments are central to this research paper and motivate the following: parent risk influences subsidiary earnings management. I argue that the relationship between MNC parent risk and subsidiary EM displays a positive association. Therefore, a higher degree of risk occurring within the MNC parent results in higher manipulation of earnings in foreign subsidiaries. My first hypothesis states the following:
H1: MNC parent risk is positively related to subsidiary earnings management
Institutional quality and EM in foreign subsidiaries
Subsidiary country level differences can either deter or enhance EM activity. In cases where IQ is higher in the subsidiary country relative to the parent, EM is deterred as it might not be that opportune for the parent to manage earnings abroad. In contrast, when IQ is lower in the subsidiary country relative to the parent, it might prove easier for the parent to engage in EM. In this context, country level differences typically refer to broad differences between subsidiary countries revolving around institutional factors ranging from taxation, to rule of law, and even investor protection (Durnev et al., 2017). For example, the US is regarded as a nation with relatively more stringent oversight from financial regulators (the SEC) than many Latin American countries. In the context of MNCs and EM, Beuselinck et al. (2019) found that differences in financial reporting oversight, prompts parents from countries with stronger institutions to manage earnings in subsidiary countries with weaker institutions, or less oversight over financial reporting. Not only is the lack of oversight more opportune for parents to engage in EM in a foreign subsidiary, but consequences can also be less severe.
Country differences are also embedded in the effectiveness of courts, which are a reflection of a country’s rule of law. Rule of law is defined as, “the extent of compliance with laws and regulations that, by affecting investor protection, and audit-quality, influence managerial reporting behavior” (Beuselinck et al., 2019; pg. 64). This is a fundamental pillar behind institutional quality and is also relevant when evaluating EM in an MNC setting. Dyreng et al. (2012) find that if EM results in misreporting penalties (ex. litigation costs) these are less if it happened in a foreign subsidiary whose country’s rule of law is less than that of the MNC parent. The critical factor determining the severity of the penalty rests upon the subsidiary country’s rule of law relative to the parent.
In MNCs, institutional differences between subsidiary countries create opportunities for EM in the subsidiaries located in jurisdictions with lower quality institutions. In summary, a country´s institutional quality determines the extent of EM practices. I hypothesize that subsidiary Institutional Quality is negatively associated with subsidiary discretionary accruals. My second hypothesis predicts the following:
Conceptual Model H1 (
+
) H2 (-
) 3. Methodology Sample Selection:My sample selection process began by identifying all listed firms from the US using Compustat. I discarded financial firms (SIC Codes 6000-6999) because these are highly regulated industries. Similar to Bonacchi et al. (2018), I excluded parent firms without consolidated financial statements, parent firms with total assets or sales less than 1 million USD, and parent firms with necessary financial information that is not available. I replicate some of the exclusion criteria utilized in Bonacchi et al. (2018) because their study also investigates the effect of parent incentives on subsidiary EM.
Further, I hand-collected the names and jurisdiction of all material subsidiaries included in exhibit 21.1 of 10-K fillings from Edgar for the non-financial US firms. Then, I matched the names and countries of the material subsidiaries with those of the firms covered by the Orbis database. Given Orbis’ coverage, I only selected subsidiaries from 30 European countries, EU 28 + Switzerland and Norway. Please note that there are no observations in the final sample for: Switzerland, Slovenia, Slovakia, Cyprus, Estonia, Latvia, Lithuania, Luxembourg, and Malta. I discarded subsidiaries operating in the financial industry because this industry is highly regulated and financial performance is not comparable with firms in other industries. I also discarded subsidiaries with necessary financial information not available. After discarding firms with unavailable financial information my final sample consists of 7,199 observations and spans from 2012 to 2017. The financial information of the subsidiaries is collected from the Orbis database.
Subsidiary Earnings Management MNC Parent Risk
Subsidiary Institutional Quality
In Table 1, I report the number of observations arranged by year, in Table 2, I report the number of observations arranged by Subsidiary Industry (2 Digit SIC Code), and lastly in Table 3, I report the number of observations arranged by Subsidiary Home Country.
Table 1-Sample Distribution by Year
Year MNC Parent-Subsidiary (# of Observations)
2012 832 2013 997 2014 1162 2015 1290 2016 1382 2017 1536 Total 7199
Table 2-Sample Distribution by Subsidiary Industry SIC Code (Two Digits) MNC Parent-Subsidiary (# of Observations) 07 15 13 17 17 37 20 242 21 25 22 17 26 128 27 80 28 683 30 175 32 45 33 50 34 257 35 604 36 297 37 221
38 261 39 95 42 23 47 19 48 16 49 25 50 1250 51 993 55 38 56 19 57 16 58 17 59 37 72 24 73 882 76 107 80 34 87 450 Total 7199
Table 3-Sample distribution by Subsidiary Home Country Country Avg. Rule of Law Index
(estimate score) # of Observations Austria 1.85 142 Belgium 1.43 535 Bulgaria -0.07 37 Croatia 0.31 31 Czech Republic 1.09 319 Denmark 1.94 36 Finland 2.01 73 France 1.43 1243 Germany 1.69 511 Greece 0.30 60
Hungary 0.52 100 Ireland 1.65 166 Italy 0.36 1017 Netherlands 1.89 29 Norway 2.00 245 Poland 0.70 367 Portugal 1.10 146 Romania 0.22 116 Spain 1.00 487 Sweden 1.98 68 United Kingdom 1.74 1471 Total 7199
*The Average Rule of Law Index is a measure developed by Kaufman et al. (2011) which reflects agents’ perceptions relating to confidence in and adherence to rules, quality of contract enforcement, property rights, and the courts among other aspects. It can either be depicted as a percentage or as an estimate (as above) ranging from -2.5 to 2.5; in units of a standard normal distribution. Moreover, the figures in the second column of the table above are the average of all Rule of Law estimates arranged by subsidiary country.
Variables
Dependent Variable –Subsidiary Earnings Management
In this research study Subsidiary FRQ (SUB_FRQ) is proxied by subsidiary earnings management. My study focuses on accrual based EM. I use the Jones Model (1991) to estimate discretionary accruals of subsidiaries. The model for accruals is estimated for every year and industry combination. The Jones Model separates total accruals into normal accruals (non-discretionary) and abnormal accruals ((non-discretionary). The Jones Model is presented in equation (1), discretionary accruals are represented by the error term ε (Jones 1991; Dechow et al., 1995; Dechow et al., 2010):
TA
i, t /A
i ,t-1= α
1[1
/
A
i,t-1] + β
1i[ΔRev
i,t /A
i, t-1] + β
2i[PPE
i,t/
A
i, t-1] +
ε
i,t (1)Where, TA i, t refers to total accruals and i is firm, t is year, Ai,t-1 is total assets at year t-1 for firm i, ΔRev is change in revenue for firm i in year t, PPE is gross property plant and equipment for firm
Discretionary accruals (DA), also referred to as abnormal accruals, are the residuals from equation (1) (Bonacchi et al., 2018; Dechow et al., 2010; Jones, 1991). Two digit SIC codes are used to define industries and industry-years with less than 15 observations are disregarded from the sample to ensure generalizability. Lastly, I use absolute values of discretionary accruals to measure SUB_FRQ.
First Independent Variable-Parent Risk
I use R&D expenditures to measure MNC parent risk in this paper. R&D expenditures is a straightforward proxy for investment risk based on costs directly associated with a company's plans to innovate. R&D effectively encompasses risk because it is an inherently risky activity with long term gains, uncertain benefits, and complex technological processes. There are several measures of risk more popular than R&D expenditures. For example, volatility of return on assets (ROA). Volatility of ROA has been widely used as a measure of performance risk (Claessens et al., 2000; Laeven & Levine, 2009) and is an indicator of profitability focusing on assets’ efficiency at generating earnings. It is calculated by scaling operating income by total assets and then calculating the volatility. Such measure can indicate risk when it continuously falls as it signals that revenues have not increase or it has over invested in assets, depending per industry. Nevertheless, I find that R&D expenditures is a more fitting measure of risk given that its benefits are much more uncertain and distant in terms of time.
In summary, I use R&D expenditures divided by total assets like in Dechow et al. (2010), as my measure of MNC parent risk (MNC_ParentRisk). Such measure of risk has been utilized before by Boubakri et al. (2013) and Li et al. (2013) and by Shust (2015). To my knowledge only the last mentioned author has paired this measure of risk with discretionary accruals, but did not focus on MNCs.
Second Independent Variable-Subsidiary Country Institutional Quality
Subsidiary Country Institutional Quality is proxied by (Rule of Law). Rule of Law provides an assessment of a country’s compliance with laws and regulations, that among other things, affect investor protection and audit-quality. In turn, the lack of these prompts parent companies to engage in EM through their respective subsidiaries (Beuselinck et al., 2019; Durnev et al., 2017; Dyreng et al., 2012). I use the World Bank Rule of Law index similar to Beuselinck et al. (2019); Durnev
et al. (2017); and Dyreng et al. (2012). Such indicator developed by Kaufmann et al. (2011) reflects agents’ perceptions relating to confidence in and adherence to rules, quality of contract enforcement, property rights, and the courts among other aspects.
I focus primarily on subsidiary country rule of law and bring forth the following example to illustrate my argument: if we have two subsidiaries, one in Germany and one in Venezuela, and they are both implicated in earnings management. It is fair to assume that a German court would opt for more punitive penalties than a Venezuelan court. The latter example can be corroborated by each nation’s position in the Rule of Law Index developed by the World Bank. This index captures and ranks countries in terms of confidence towards authorities and adherence to rules (World Bank Rule of Lax Index). Germany consistently hovers around 1.9 whereas Venezuela scores -2.1. These differences frame the Venezuelan subsidiary as being much more prone to earnings management than its German counterpart thus demonstrating how a country’s rule of law could facilitate earnings management. In light of the above, the lack of rule of law can prompt higher degrees of subsidiary discretionary accruals and vice versa, a higher degree of rule of law can induce a lower degree of subsidiary discretionary accruals (Beuselinck et al., 2019; Durnev et al., 2017; Dyreng et al., 2012).
Empirical Models
Equation 5 presents the empirical model I used to investigate the association between parent risk (MNC_ParentRisk) and subsidiary FRQ (SUB_FRQ).
SUB_FRQ i, t = α + β1 MNC_ParentRiski, t
+ β2 Firm_Size_Parenti, t + β3 Firm_Size_Subi, t + β4 Leverage_Parenti, t +
β5 Leverage_Subi, t + β6 Performance_Parenti, t + β7 Performance_Subi, t +
β8 SalesGrowthSubi, t + β9 Year_Fixed_Effects + β10 Country_Sub+ β11
Industry_Sub + εi, t (5)
The equation above examines the association, or lack thereof, between MNC Parent Risk and Subsidiary FRQ, and includes control variables commonly utilized in similar studies. The dependent variable is SUB_FRQ which is the absolute value of discretionary accruals estimated with the Jones model (Jones, 1991). The main independent variable is MNC_ParentRisk, defined
as R&D expenditures divided by total assets (Boubakri et al., 2013; Li et al., 2013; Shust, 2015). I expect MNC_ParentRisk to have a positive sign.
The control variables employed are all commonly found in the literature and their inclusion allows for a better evaluation of the relationship amongst parent risk and subsidiary FRQ. In this context, firm size is measured as the natural logarithm of total assets, this control measure is included for both subsidiaries and MNC parents. Beuselinck et al. (2019); and Guenther et al. (2017) found that firm size is negatively correlated with discretionary accruals and volatility of earnings; according to Poli (2013) this is because smaller firms are less monitored and restricted than large publicly listed firms. Leverage is included as a control variable for parents and subsidiaries in order to control for the level of debt. In this context, leverage is a determinant of bankruptcy risk and Beuselinck et al. (2019) find that risky subsidiaries exhibit higher discretionary accruals. More specifically, Dechow et al. (2010) find that firms with higher leverage ratios engage in more EM and are more prone to violate debt covenants. Firms with higher leverage ratios engage in more EM in order to maintain creditors’ support through good financial results in order to potentially refinance the firm’s debt (Marcel et al., 2012).
In addition to size and leverage, I also include corporate performance as a control variable. Such control variable is used in my model to control for both MNC-parent and subsidiary performance and profitability. I included this control variable because corporate performance falls under the umbrella of determinants of EM. Corporate performance is negatively related with volatility, and more volatility has been shown to result in more discretionary accrual activity (Beuselinck et al., 2019; Guenther et al., 2017). Like in Marcel et al. (2012) underwhelming financial results tend to make investors reluctant, hence organizations with lower performance will engage in EM to maintain investor support. Corporate performance is measured as return on assets (ROA), this percentage expresses how effectively a company uses assets to generate profits. Also, I control for sales growth in subsidiaries as this is an indicator of overall growth, and high growth is positively related with discretionary accruals (Beuselinck et al., 2019: Dyreng et al., 2012). The rationale behind this is that managers are inclined to reduce the amount of corporate taxes to be paid, and in a period of high sales growth an effective way to do so is by engaging in EM. Thus, a decrease in earnings results in less corporate taxes (Al Najjar & Riahi-Belkaoui, 2001). I measure sales growth as the change in sales scaled by total assets. Lastly, I also control for year, industry, and country effects.
Model 2-Equation 6
SUB_FRQ i, t = α + β1 RuleofLawi,t
+ β2 Firm_Size_Parenti, t + β3 Firm_Size_Subi, t + β4 Leverage_Parenti, t +
β5 Leverage_Subi, t + β6 Performance_Parenti, t + β7 Performance_Subi, t +
β8 SalesGrowthSubi, t + β9 Year_Fixed_Effects + β10 Country_Sub+ β11
Industry_Sub + εi, t (6)
Equation 6 depicts the second model which tests H2. This replaces parent risk with subsidiary Institutional Quality (RuleofLaw) as the main IV, whilst maintaining the same control variables as in Equation 5. I expect RuleofLaw to have a negative sign.
Summary
In line with my argument for H1, I expect that the coefficient of MNC_ParentRisk is
positive. Such would imply that MNC parent risk is positively related with discretionary accruals. Likewise, in line with my argument for H2, where Institutional Quality in the subsidiary country influences the degree to which parents engage in EM in foreign subsidiaries, I expect that the coefficient of RuleofLaw is negative. This suggests that subsidiary Rule of Law would be negatively associated with discretionary accruals, and thus an increase in Rule of Law decreases discretionary accrual activity.
Table 4-Variable Definitions
Variable Definition
SUB_FRQ Subsidiary FRQ measured by absolute value of discretionary accruals using Jones Model (1991)
MNC_ParentRisk Parent Risk measured by Research and Development expenditures scaled by total assets
RuleofLaw Rule of law index estimate of the subsidiary country
Firm_Size_Parent Size of total assets of Parents, measured as the natural logarithm of total assets of Parents
Firm_Size_Sub Size of total assets of Subsidiaries, measured as the natural logarithm of total assets of Subsidiaries
Leverage_Parent Total debt of Parents scaled by total assets of Parents Leverage_Sub Amount of debt relative to assets. Total debt scaled by total
Performance_Parent Performance relates to a company’s utilization of assets to generate returns and is measured as ROA (Return on Assets); Parent Net Income scaled by Parent total assets.
Performance_Sub ROA (Return on Assets); Subsidiary Net Income scaled by Subsidiary total assets.
SalesGrowth Sub Change in sales scaled by total assets for Subsidiaries.
4. Results
Descriptive Statistics
The descriptive statistics of my final sample are shown below in Table 5 and defined in Table 4 (above, in the methodology section). SUB_FRQ the dependent variable is measured by the magnitude of discretionary accruals and its mean is 2%, a figure slightly less than in similar studies like Beuselinck et al. (2019), such difference perhaps lies in the fact that this author used a performance-adjusted modified Jones model which has some improvements in capturing revenue manipulation, or EM. MNC_ParentRisk the main independent variable for H1 displays a mean value of 4%, this figure is considerably higher than in Boubakri et al. (2013) 0.2%. Thus in comparison, R&D expenditures represent a sizable percentage of total assets in MNC Parents.
RuleofLaw, the main independent variable in H2 displays a mean value of 1.27, which indicates
that on average subsidiary countries have a relatively high level of Institutional Quality. Mean
Firm_Size_Sub is 10.21 while mean Firm_Size_Parent is 9.11. In order to put this into context,
please note that firm size observations for subsidiaries have a standard deviation of 1.86 while firm size for parents have a standard deviation of 1.46. Additionally, subsidiaries have a minimum of 3.6 and a maximum of 17.7. Whereas, parents have a minimum of 5.96 and a maximum of 12.5. Altogether, this indicates that subsidiary range of observations of firm size is broader than that of the parent. Moreover, parents (27%) display a higher degree of leverage than subsidiaries (6%). Similarly, parents (14%) have a higher degree of performance compared to subsidiaries (7%). which suggests that Parents are more effective at using assets to generate returns. Lastly.
Sales_Growth_Sub display a -3% mean, the negative percentage suggests that in general sales year
Table 5-Descriptive Statistics
Variable N Mean Std. Dev. Minimum Maximum
SUB_FRQ 7199 0.02 0.03 0.00 0.69 MNC_ParentRisk 7199 0.04 0.04 0.00 0.58 RuleofLaw 7199 1.27 0.55 -0.11 2.1 Firm_Size_Parent 7199 9.11 1.46 5.96 12.49 Firm_Size_Sub 7199 10.21 1.86 3.61 17.73 Leverage_Parent 7199 0.27 0.15 0.00 1.78 Leverage_Sub 7199 0.06 0.31 0.00 10.58 Performance_Parent 7199 0.14 0.06 -0.16 0.58 Performance_Sub 7199 0.07 0.15 -0.99 0.95 SalesGrowth Sub 7199 -0.03 4.90 -412.83 5.54 N=7199 *Variable definitions can be found in Table 4 in the Methodology Section
Table 6 –Pearson Correlation
Correlations significant at the 0.01 level (2-tailed) & at the 0.05 level (2-tailed) are shown in boldface. N=7199, Variable definitions are given in Table 4 in the Methodology Section.
1 2 3 4 5 6 7 8 9 10 1. SUB_FRQ 1 2. MNC_ParentRisk .08 1 3. RuleofLaw -.14 .04 1 4. Firm_Size_Parent -0.02 .04 -.08 1 5. Firm_Size_Sub -0.02 0.00 .12 .38 1 6. LeverageParent -0.01 -.34 -.02 -.10 -.05 1 7. LeverageSub .05 .05 .08 -.03 0.01 0.01 1 8. PerformanceParent -0.02 .09 -.04 -0.01 .05 .08 -0.02 1 9. PerformanceSub -.14 .03 .02 -.04 -.03 0.01 -.2 .09 1 10. SalesGrowthSub 0.01 -0.02 -.01 0.01 .04 0.00 0.00 .03 0.02 1
Correlation Results
The coefficient of the correlation between MNC_ParentRisk and SUB_FRQ (.08) indicates that subsidiary discretionary accruals increase with the parent risk. Such association is as expected in H1. The correlation between RuleofLaw and SUB_FRQ is negative (-.14), nonetheless the correlation between RuleofLaw and SUB_FRQ is not significant, indicating that there is no association between them. This is not consistent with H2, which predicted that institutional quality is negatively and significantly related to subsidiary discretionary accruals. Additionally, correlations suggest that subsidiary discretionary accruals increase with subsidiary leverage and decrease with subsidiary performance. The correlations between subsidiary discretionary accruals and the other controls are not significant.
Table 7-Regression Results
Subsidiary Discretionary Accruals
1 2 MNC_ParentRisk 0.05 ** (0.013) RuleofLaw 0.00*** (0.007) Firm_Size_Parent 0.00 (0.00) 0.00 (0.00) Firm_Size_Sub 0.00 (0.00) 0.00 (0.00) LeverageParent 0.00*** (0.003) 0.00*** (0.003) LeverageSub 0.00*** (0.001) 0.00*** (0.001) PerformanceParent 0.00*** (0.007) 0.00*** (0.007) PerformanceSub -0.03*** (0.003) -0.03*** (0.003) SalesGrowthSub 0.00 (0.00) 0.00 (0.000) R Squared 0.1 0.1 Adjusted R Squared 0.097 0.095 Observations (N=) N= 7199 N=7199
Year Yes Yes
Industry Yes Yes
Country Fixed Effects Yes Yes
Variable definitions are given in Table 4 in the Methodology Section. Standard errors given in parenthesis * p<0.1, ** p<0.05, ***p<0.01.
Regression Results
Table 7 shows the results of the estimation of Model 1(Equation 5) and Model 2 (Equation 6). The coefficient of MNC_ParentRisk is positive (0.05) and significant at the 5% level. This is consistent with H1 which predicted that parent risk increases EM activity in subsidiaries. In terms of economic significance, this result implies that every 1 percentage point increase in R&D expenses in parents increases discretionary accruals in subsidiaries by 0.05 percentage points.
Concerning results reported in Column (2), Table 7 shows the coefficient of RuleofLaw which is positive (0.00) and significant at the 1% level, suggesting that subsidiary institutional quality increases EM activity. This is not consistent with H2 which predicted that RuleofLaw is negatively related with EM activity in subsidiaries. My result is opposite to prior studies depicting an inverse relationship between rule of law and subsidiary discretionary accruals (Ball et al. 2000; Leuz et al. 2003). Given my unexpected results, I performed additional tests for H2 using more granular information by organizing observations into two groups-HIGH (observations above the level of parent rule of law) and LOW (observations below the level of parent rule of law). These are displayed and explained below (Table 8). Lastly concerning controls, the coefficient of
PerformanceSub is (-.03) and significant at the 1% level in both column (1) and (2).
Table 8-Additional Tests
Subsidiary Discretionary Accruals
1 2 RuleofLaw HIGH -0.02*** (0.006) RuleofLaw LOW -0.00*** (0.001) Firm_Size_Parent 0.00*** (0.001) 0.00 (0.000) Firm_Size_Sub -0.00 (0.000) 0.00 (0.000) LeverageParent -0.01*** (0.005) 0.00*** (0.003) LeverageSub 0.01*** (0.002) -0.00*** (0.002) PerformanceParent 0.01** (0.012) -0.01*** (0.008) PerformanceSub -0.03*** (0.005) -0.04*** (0.003) SalesGrowthSub 0.00 0.01***
(0.000) (0.001)
R Squared 0.016 0.033
Adjusted R Squared 0.013 0.032
Observations (N=) N= 2666 N= 4533
Year Yes Yes
Industry Yes Yes
Country Fixed Effects Yes Yes
Variable definitions are given in Table 4 in the Methodology Section. Standard errors given in parenthesis * p<0.1, ** p<0.05, ***p<0.01. Column (1) displays figures for Rule of Law group High and Column (2) for Rule of Law group Low.
Parent rule of law estimate is 1.6 (United States), so observations in column (1) are greater than 1.6 and less than 1.6 in column (2). Table 8 column (1) shows the coefficient of RuleofLaw HIGH which is negative -(0.02) and significant at the 1% level, suggesting that subsidiary institutional quality decreases EM activity. This is consistent with H2. Concerning column (2), the coefficient of RuleofLaw LOW is also negative -(0.004)7 and significant at the 1% level. Thus, also suggesting that subsidiary institutional quality decreases EM activity.
Additionally, Table 8 shows that control variables have mixed results in their association with subsidiary discretionary accruals. For example, correlations in column (1) suggest that subsidiary discretionary accruals increase with parent size, subsidiary leverage, and parent performance. According to column (2), subsidiary discretionary accruals increase with parent leverage and subsidiary sales growth. Despite the differences above, correlations from both columns suggest that subsidiary discretionary accruals decrease with subsidiary performance. Lastly, the correlations between subsidiary discretionary accruals and subsidiary firm size are not significant.
5. Discussion & Conclusions
Findings
The focal point of this research study is whether the level of parent risk and subsidiary IQ are associated with subsidiary EM practices. My expectation is that more risk at the parent level creates incentives for the parent to manage earnings to make up for potential loses or unrealized benefits
7 Please note that I included three decimal when displaying the coefficient for this variable to emphasize my results,
however, for consistency I did not do so in any of the tables. I report two decimal places for coefficients and up to three for significance level.
expected from investing in R&D. Regardless of the level of risk surrounding parents, MNC parents are naturally keen on managing earnings abroad mainly due to differences in institutional quality and different corporate taxes. Additionally, they are incentivized to manage earnings abroad by distance from parent level auditors and from the parent level financial regulating bodies, and less burdensome financial and legal penalties in foreign subsidiary countries. EM is done locally, in other words local managers are the ones that carry out such task in their respective locations. Therefore, when the parent is engulfed by risk it will promptly scheme for subsidiaries to engage in EM. In turn, this compromises the FRQ of subsidiaries and due to the consolidation of financial statements, the FRQ of the MNC is compromised too.
Higher levels of risk at the parent level can become detrimental for the realization of organizational objectives. When an MNC invests more resources into projects that are inherently risky, technically complex, and with uncertain benefits, like those from research and development, managers make up for the consequences associated with investing in R&D by engaging in EM. Ultimately, depicting a more favorable economic outlook. For example, Dechow et al. (2010); Hribar & Nichols (2007); Beuselinck et al. (2019) and Dyreng et al. (2012) found that firms displaying more risk tend to engage in earnings management as an attempt to hedge against any detrimental outcomes affecting the financial position. Additionally, parents are tempted to manage earnings away from the home country in countries with lower levels of institutional quality and with tax benefits in order to lower the MNC corporate taxes. In this context lower institutional quality can result in less oversight from financial regulating bodies, and lower penalties for engaging in earnings management.
Consistent with my arguments, I find a positive and significant relationship between MNC Parent Risk and the magnitude of subsidiary discretionary accruals. This suggests that MNC Parent Risk is detrimental to Subsidiary FRQ. Coincidentally, this also means that parent risk is detrimental for the FRQ of the entire MNC, because MNC financial statements are an aggregation of subsidiary financial statements. My results are in line with previous literature, such as Chu et al. (2018) and Ham et al. (2017), who find that risk negatively affects FRQ.
In addition, I focus on the relationship between subsidiary institutional quality, proxied by the Rule of Law Index from the World Bank, and subsidiary discretionary accruals. My expectation is that a high rule of law would be associated with less earnings management activity at the subsidiary level, hence Rule of Law and discretionary accruals are expected to be negatively
related. Contrary to risk, higher levels of Institutional Quality make earnings management less attractive due to effectiveness of courts, stringent oversight from financial regulating bodies, higher misreporting penalties (in fees or litigation costs), and a less lenient attitude towards malpractice (Beuselinck et al., 2019; Dyreng et al., 2012). This suggests that subsidiary level characteristics, such as institutional quality, are expected to explain for the level of earnings management in subsidiaries.
I find a positive and significant relationship between subsidiary institutional quality and subsidiary discretionary accruals. This implies that higher levels of Rule of Law do not deter earnings management from happening, rather the results indicate that Rule of Law is associated with higher discretionary accruals. Given the unexpected results I performed additional tests where I segmented observations into two groups: high and low. My additional tests confirm, Ball et al. (2000) and Leuz et al. (2003) inverse relationship between institutional quality and earnings management. I find a negative and significant relationship between subsidiary institutional quality and subsidiary discretionary accruals. There is a difference in magnitude between the high group and the low; put differently, the negative association between institutional quality (group high) and subsidiary discretionary accruals is stronger than that between institutional quality (group low) and subsidiary discretionary accruals. This suggests that subsidiary countries with a higher rule of law relative to the MNC parent are more effective at deterring subsidiary discretionary accruals than subsidiary countries with a lower rule of law relative to the MNC parent.
The results from the additional test can be contextualized by the findings in Beuselinck et al. (2019) who argue that when parents come from high IQ countries EM occurs in subsidiaries from countries with lower IQ. My results hint at this by demonstrating that subsidiary countries with a lower rule of law, relative to the parent, are less effective at deterring EM and thus more prone to engage in subsidiary discretionary accruals. In summary, my findings provide mixed results but the additional test suggests that subsidiary institutional quality limits EM and thus favors subsidiary and MNC FRQ.
Implications
The finding that parent risk is positively associated with subsidiary earnings management is consistent with prior literature and confirms the widely conceived positive relationship between risk and earnings management (Chu et al., 2018; Ham et al., 2017). Going more in depth, my
finding contributes to the literature on risk and EM by focusing on a measure of risk not used before to measure investment risk within MNCs. R&D is a novel measure of risk for evaluating the relationship between Parent Risk and Subsidiary FRQ. Authors such as Boubakri et al. (2013), Li et al. (2013) and Shust (2015) have used this measure, but not for assessing parent risk and its association with subsidiary discretionary accruals. In this respect, my finding demonstrates that parent risk, measured as R&D expenditures scaled by total assets, causes EM and thus compromises the FRQ of the subsidiary and eventually of the MNC. This finding can aid with the identification and understanding of factors that can lead to fraudulent financial reporting. Studies such as this one can assist different stakeholders within the accounting community (regulators, analysts, and investors among others) in developing a mapping of factors empirically proven to compromise FRQ within MNCs. Also, this finding can assist with the design and implementation of internal controls that grant assurance regarding FRQ.
In addition, my findings concerning the association between subsidiary country institutional quality and subsidiary earnings management emphasize the fact that IQ influences (and is negatively associated with) EM as outlined in Beuselinck et al. (2019); Ball et al. (2000) and Leuz et al. (2003). In addition, comparing U.S parents against a sample of European subsidiaries with higher and lower IQ, highlights the role of IQ as a deterrent against EM, even within the low IQ group. This implies that IQ can be a used as a predictive tool for determining potential EM activity in MNCs and thus provide an idea of the FRQ of the MNC. Likewise, decreases in IQ for either the parent or subsidiary can become a Key Risk Indicator (KRI), especially in the context of MNCs. A decrease in IQ reflects broad changes in a country’s compliance with laws, regulations, and of special concern, investor-protection; therefore, changes in IQ can be monitored to warn against potential issues compromising the realization of objectives for the MNC (Gray & Debreceny, 2014). Such as a lack of transparency in court proceedings surrounding corporate law.
Limitations & Avenues for Future Research
My study on parent risk, IQ, and subsidiary FRQ is subject to limitations. First, my final sample consisted on parent companies solely from the US and subsidiaries only from Europe. This level of specificity behind the final sample limits the generalizability behind my findings and the external validity of my study. Moreover, because R&D remains on the periphery as a measure of
