study under mandatory IFRS adoption
Student number: 2749157
Name: Yang Yang
Study programme: MSc IFM
Supervisor: Mr. Halit Gonenc
Second evaluator: Mr. Mario Hernandez Tinoco
Rijksuniversiteit Groningen Faculty of Economics and Business
Department of Economics, Econometrics and Finance Duisenberg Gebouw
Nettelbosje 2 9747 AE Groningen
Abstract：This thesis empirically investigates the relationship between creditor rights and dividend policies across 40 countries around the world, particularly focusing on how this relationship is affected by the mandatory IFRS adoption. The results show that firms in countries with high creditor rights pay more dividends. Furthermore, the mandatory IFRS adoption has mitigating effects on the positive association between creditor rights and dividend policies. The findings consistently hold when applying an alternative measure of dividend payouts, including voluntarily adopting firms, using an alternative regression model, including legal origin as an additional control variable, and using different sample compositions.
1. Introduction
Dividend policy has received considerable attention and disputation from both business practitioners and academic researchers in the corporate finance. Over the past four decades, although extensive theoretical and empirical research conducted into the motivations of paying dividends attempt to explain the dividend puzzle by Black (1976), there is no consensus model that managers can use to determine their corporation’ optimal dividend policy. This may be due to the fact that these theories about dividend policies differ in their relative emphasis; for instance, the signaling theory highlights information asymmetry between corporate management and outside shareholders, the catering theory emphasizes the demand of investors, the agency theory stresses agency costs between corporate insiders and outside investors, and so on.
Dividend policies are principally involved into the decision to determine the amount and pattern of payments to shareholders or keep the extent of their profits as retained earnings for a stable growth. Managers consider dividend policies as a crucial element of a corporate, and hence pay careful attention to the choice of dividend policy. A large number of prior literature provides substantial evidence on the importance of shareholder protection in the development of equity capital market, investment policies and dividend policies. More specifically, La Porta et al. (2000a) propose two competing agency models for a relationship between minority shareholder protection and dividend policies, namely the outcome and substitution models. The outcome model suggests that strong legal rights will enable minority shareholder to extract high dividend payouts from a firm and dividends are regarded as an outcome of the legal protection of shareholders. On the contrary, the substitute model suggests that weak legal protection will result in high dividend payouts since managers consider dividends as a substitute for weak legal protection of shareholders. According to the empirical analysis, La Porta et al. (2000a) find that the outcome model of dividends can explain connections between agency costs of equity, legal protection of shareholders and dividend payouts and indicate a positive relationship between dividend payouts and shareholder protection.
argues that the stakeholder theory leads managers to take account of the interests of all stakeholders in a corporation including not only shareholders but also creditors when it is considered to be a long-term value creation. Dividend policy is considered as a major source of conflicts arising between debtholders and shareholders. Brockman & Unlu (2009) present that mangers substitute restrictive dividend policies for weak creditor protection in an attempt to decline agency costs of debt, which indicates that the substitute model of dividends can explain the connection between agency costs of debt, creditor rights and dividends payouts. They find that firms in countries with weak creditor rights are less likely to pay dividends and pay lower amounts of dividends. This evidence indicates that creditors prefer that corporations announce less dividends, since dividends enable shareholders to attain benefits at the expense of creditors and increase creditors’ risk of expropriation by shareholders.
Recently, an increasing number of academics and researchers have allocated a large amount of resources and time to examine the economic influence of International Financial Reporting Standards (IFRS). IFRS is developed by the International Accounting Standards Board (IASB), aiming to advance internationally accepted accounting principles used by corporations to prepare their financial statements. Since its establishment in 2001, an increasing number of countries require or permit their domestically public companies to present financial statements in compliance with IFRS, which shows the rapid development of the standards. The mandatory adoption of IFRS reporting around the global is considered as one of the most essential policy issues in corporate operations and financial accounting. Prior researches indicate that the mandatory implementation of IFRS has the potential to improve financial reporting and reduce information asymmetry. The mandatory adoption of IFRS is expected to decrease the information gap between companies and investors by improving information disclosures and declining information asymmetry. These effects would thereby influence strategic corporate decisions.
research of international accounting and corporate dividend policy. Specifically, this thesis first focuses on the relationship between country-level creditor protection and corporate dividend policies, and how this relationship is influenced by the mandatory introduction of IFRS. Unlike prior studies that mainly focused on firms in the European Union (EU) area or a single country, this thesis analyzes corporate decisions around the mandated introduction of IFRS reporting based on a broad and global sample of firms.
This empirical analysis employs 279,784 firm-year observations from 40 countries over the sample period 1991-2013. This thesis documents the following results. First, this thesis finds a significantly positive connection between creditor rights and corporate dividend payouts, consistent with the results in the study of Brockman &Unlu (2009) who indicate that managers conduct more restrictive payout policies as a substitute for weak creditor rights in an effort to reduce the agency conflicts with creditors. Second, this thesis suggests that mandatory IFRS adoption has a negative and significant effects on corporate dividend policies, which is in line with the results in the study of Hail, Tahoun, & Wang (2014) who present that because of the improvement of information environment after mandatory introduction of IFRS, managers are less likely to use dividend payouts as a commitment device. Third, this thesis examines the impact of mandatory IFRS adoption on the association between creditor rights and corporate dividend policies and find a significantly negative relation. The results are robust to an alternative dependent variable, the inclusion of voluntary IFRS firms, an alternative research design, legal origin as an additional control variable, and different sample compositions. The results confirm that after mandatory adoption of IFRS that contributes to the improvement of information environment, creditors are less likely to include dividend restrictions on debt contracts to acquire additional control rights and managers are less likely to substitute restrictive dividend policies with weak creditor protection.
2. Literature review
In this section, this thesis first reviews some important dividend theories briefly. After that this thesis presents the role of creditor protection in the field of corporate finance. Finally, this thesis discusses the development, objective and economic influence of mandatory IFRS adoption.
2.1 Dividend policy theories
Over past a half century, financial researchers have proposed various theories, hypotheses and explanations for paying dividends based on the dividend irrelevance model of the Miller & Modigliani (1961) and the Black (1976)’s dividend puzzle. In their seminal research, Miller & Modigliani (1961) propose that in perfect and complete capital markets, firm value is independent of a firm’s dividend payout policy. The dividend irrelevance model is based on several restrictive assumptions: no taxes, no transaction costs, no agency costs, no bankruptcy, and no asymmetric information. Moreover, Black (1976) investigates why corporations pay dividends and claims that “the harder we look at the dividend picture, the more it seems like a puzzle, with pieces that just don’t fit together” (p. 5). Below are several major theories of corporate dividend literature. Specifically, some traditional theories are involved into market imperfection (information asymmetry and agency costs), and then some recent explanations focus on the incentives and motivations of dividend payouts (catering theory and lifecycle theory).
Signaling theory: Information asymmetry serves as the basis for signaling theory. Firms pay dividends to reduce information asymmetry between management and shareholders, and then convey private information to investors about future prospects of a company (John & Williams, 1985; Baker, Singleton, & Veit, 2011). Booth & Chang (2011) support the information asymmetry explanation and indicate that dividend changes signal future prospects of the firm. Thus, managers could deliver information about future prospects to the investing public through announcements of dividend policies.
theory and show dividend payouts increase with profitability and firm size, but decrease with growth prospects. Thus, firms should retain their earnings in the early lifecycle stage and pay all earnings as dividends at maturity.
Catering theory: Firms emphasize investor demand in decisions about dividend policies (Baker & Wurgler, 2004); that is, managers cater to investor demand by paying dividends when investor sentiment supports dividend payouts and by omitting or cutting dividends when investor sentiment does not support dividend payouts. Ferris, Jayaraman, & Sabherwal (2009) use a cross-country sample and show evidence of catering among firms in common law countries. Thus, managers could attract and capture investors through identifying and satisfying their demands.
Agency theory: Jensen & Meckling (1976) propose the principal-agent relation that indicates two agency conflicts: conflicts of interests between corporate insiders and outside shareholders and conflicts of interests between outside shareholders and debtholders. Firms use dividend payouts to overcome agency problems between corporate insiders and outside minority shareholders (Jensen, 1986; La Porta et al., 2000a). Dividends reduce agency costs of equity by preventing managers from pursuing individual objectives and committing to unprofitable projects. Easterbrook (1984) proposes that dividends serve as a device to take away the free cash flow from the control of management and pay it to shareholders. Additionally, La Porta et al. (2000a) confirm that dividend policies can decrease agency costs of equity. A large number of studies primarily examines the importance of dividend policies in mitigating conflicts of interests between corporate insiders and outside minority shareholders. However, recently an increasing amount of studies investigate the conflicts between the stakeholders (Bøhren, Josefsen, & Steen, 2012). More specifically, more recent research focuses on the relationship between shareholders and creditors. For example, Brockman & Unlu (2009) indicate that in countries with weak creditor protection, managers use restrictive dividend policies to reduce agency costs of debt.
2.2 Creditor rights
may use the prospect of invoking their powers to affect corporate policies. A large amount of early studies emphasizes the powers of creditors during firm bankruptcy or during period of financial distress (Seifert & Gonenc, 2016). More recently, empirical research presents that creditors use their powers to protect themselves under many circumstances where firms do not face bankruptcy or financial distress. These studies demonstrate that legal protection of creditors play an important role in strategic corporate decisions by affecting corporate cash holdings, capital structure choices, corporate investment decisions, innovative activities, and bank loans. Seifert & Gonenc (2016) indicate that strong creditor rights and country governance are significantly related to less corporate cash holdings due to the fact that strong creditor rights result in great supply of funds and strong country governance makes it easy and riskless for capital markets to supply funds. Cho et al. (2014) present the importance of creditor protection in determining corporate capital structure across a large sample of countries. They find that creditor rights negatively influence long-term leverage, consistent with the demand-side view. They suggest that managers and shareholders tend to decrease the likelihood of financial distress by avoiding use of debt financing if strong creditor protection increases the profitability of their losing control in bankruptcy. Similarly, Acharya, Amihud, & Litov (2011) propose that strong creditor rights lead firms to perform risk-reducing activities, such as undertaking diversifying acquisitions or decreasing corporate leverage. Acharya & Subramanian (2009) find that firms in countries with strong creditor rights conduct less innovative operations. Bae & Goyal (2009) investigate whether the legal protection of creditors influence the size, maturity, and interest rate spread on loans and find a significantly negative association between creditor rights and interest rate spreads. Based on these recent studies, it can be concluded that creditors do use their legal protection to affect corporate policies. This thesis contributes to this emerging literature by examining whether creditor rights affect corporate dividend policies.
2.3 The mandatory adoption of IFRS
quality, understandable and enforceable global accounting standards that require high-quality, transparent and comparable information in financial statements and other financial reporting to help participants in the world’s capital markets and other users make economic decisions” (Pacter, 2016, p. 8). According to the IFRS Foundation’s study on 140 jurisdictions, over 120 countries around the global require or permit the use of IFRS by their domestically listed companies for financial reporting. Countries with major capital markets around the world, such as Australia, Brazil, EU Member States, and Hong Kong, require their domestic publicly accountable entities to present financial statements in consistency with IFRS (Horton, Serafeim, & Serafeim, 2013). Other countries, such as United States, Japan, India, and Vietnam, are currently considering adopting IFRS, and specifically Japan have already allowed the voluntary adoption of IFRS by public companies in their consolidated financial statements. Additionally, the Financial Accounting Standards Board (FASB) and the IASB jointly announced the Norwalk Agreement in 2002, which aim to decrease the differences between US generally accepted accounting principles (GAAP) and IFRS (Ball, Li, & Shivakumar, 2015). India permits or requires the use of IFRS for at least some domestically public companies. China has already committed to convergence of their country-specific GAAP to IFRS. Therefore, IFRS has become a global accounting standard for financial reporting by the use of both developed and developing economies.
Zijl (2014) investigate the effect of mandatory IFRS adoption on the information quality of financial reporting in France, Germany and Sweden. They find that both forecast accuracy and forecast dispersion present a significant improvement after mandatory IFRS adoption, showing an increase in information quality. Iatridis (2010) examines how mandatory adoption of IFRS affects in the quality of financial statement information through focusing on the switch from the United Kingdom GAAP to IFRS. This study shows that the mandatory implement of IFRS reduces the scope for earnings management by more timely loss recognition and value relevant accounting measures, indicating a general enhancement of accounting quality. Horton, Serafeim, & Serafeim (2013) investigate the effect of mandatory IFRS adoption on firms’ information environment through cross-country data. They find that mandatory IFRS adoption significantly improves forecast accuracy and other measures of the quality of the information environment. This improvement comes from enhanced information quality and improved accounting comparability, consistent with both information and comparability effects. Li (2010) employs EU countries as a sample and examines the effect of mandatory IFRS adoption on the cost of equity capital. The results suggest that mandatory IFRS adoption results in increased information disclosure and improved information comparability, and then leads to reduction in the cost of equity capital.
As a summary, the mandatory induction of IFRS does have influence on corporate financial decisions by affecting the information environment. Firms mandatorily adopting IFRS would provide more sufficient information in a more useful form for investors. The mandatory introduction of IFRS results in the higher information disclosure and financial reporting quality. These effects convey positive information to investors as information asymmetry and agency costs tend to weaken. Thus, this thesis next examines whether mandatory adoption of IFRS influences the association between creditor rights and dividend payout policies by its effects on the information environment.
3. Hypothesis development
3.1 Creditor rights and dividend policies
creditors losses, mandatory extensive disclosure laws and procedures make it easy and cheap for creditors to recover damage (La Porta, Lopez-De-Silanes, & Shleifer, 2006). In countries with weak creditor protection, a shareholder-friendly manager is more likely to shift the risks to creditors or conduct excessive payouts to shareholders. It is expected that creditors in countries with weak legal protection hardly achieve claims and recover damages even during firms’ bankruptcy or financial distress, and thus they are willing to demand additional control rights. Private credit agreements enable creditors control additional rights when bankruptcy rights fail to protect them (Brockman & Unlu, 2009). When creditors lend to firms, they generally include debt covenants in their debt contracts. Smith & Warner (1979) indicate that creditors regard debt covenants as an important governance mechanism to influence their debtors’ behavior. Particularly, these debt covenants are concerned with the allocation of decision rights under some circumstances where managers fail to act in creditors’ best interest. Creditors would put great importance to the constraints in debt contracts, such as requirements for dividend payouts, which prevent managers of firms from taking actions to expropriate debtholders’ wealth (Hong, Hung, & Zhang, 2016). Such restrictions are used to alleviate conflicts of interests between creditors and shareholders. Consequently, the presence of debt covenants in debt contracts is motivated by their ability to secure creditors’ interest by reducing agency conflicts and transferring control rights (Chava & Roberts, 2008).
& Reisel (2012) employ a cross-country sample of restrictive covenants and examine whether the debt contracts substitute for weak country-level creditor protection. They find that creditors are more likely to use covenants in debt contracts in countries with weak creditor protection. It is expected that creditors in countries with weak legal protection tend to require debt covenants to prevent potential expropriation from debtors. Thus, creditors in countries with weak legal protection tend to exercise their additional control over corporate dividend decisions through debt covenants in private credit agreements (Brockman & Unlu, 2009).
La Porta et al. (2000a) suggest that dividends are a substitute for legal protection of minority shareholders. Similarly, it is expected that restrictive dividend policies may be considered as a substitute for weak legal protection of creditors. Managers operating under weak creditor rights are more likely to undertake restrictive dividend policies through formal covenants and informal agreements to establish reputation among creditors, so as to raise external capital on attractive terms and reduce future financing costs (Brockman & Unlu, 2009). Managers tend to choose restrictive dividend policies that reduce potential agency conflicts with creditors. That is, when the potential agency conflicts are likely to increase in firms in countries with weak creditor rights, managers of these firms tend to focus on actual conflicts and decrease them by restrictive dividend policies. Creditors are more willing to supply financing capital to these firms, because they recognize that, with better reputation about restrictive dividend policies, their benefits would not be expropriated by shareholders. Firms in countries with strong creditor protection have weak incentives for reputation-building mechanisms among creditors because creditors would force firms to implement restrictive dividend policies by using their strong creditor rights. Thus, this reputation-building mechanism is more effective in countries with weak legal protection of creditors.
find that creditor rights are an important determinant of private credit development. They suggest that creditors in countries with strong legal protection are more willing provide credit since they can more easily obtain repayment and collateral. Firms operating under countries with strong creditor rights have little need to hold funds for future investments since they recognize that strong creditor rights would result in greater availability of funds because of lower interest rates, higher loan maturities and increased supply of funds (Bae & Goyal, 2009; Houston et al., 2010; Seifert & Gonenc, 2016). Managers are confident with their ability to acquire funds in developed capital markets, and thereby they have less of a need to hold as many funds for future investments. Thus, firms are easier to raise external financing from capital markets in countries with strong creditor rights, so they are more willing to make dividend payouts.
Based on the aforementioned arguments, first of all, creditors in countries with weak legal protection tend to demand additional control rights to protect themselves from the potential expropriation and risk-shifting behavior by shareholders. They frequently use debt covenants to restrict dividend payouts to shareholders under weak creditor rights. Secondly, managers in countries with weak creditor rights are more likely to undertake restrictive dividend policies to establish reputation among creditors in order to raise financing capital among them. Thirdly, strong creditor rights result in greater availability of funds and higher supply of funds in financial markets. Under this condition, managers are less likely to save funds for future investments and more willing to pay dividends. Thus, it is expected that there is positive relationship between creditor rights and corporate dividend policies. The hypothesis is formulated as:
Hypothesis 1a: Country-level creditor rights positively affect corporate dividend policies.
the creditors holding an empty shell” (p. 10). These behaviors lead to the conflicts of interests between creditors and shareholders. An important function of creditor protection is to protect creditors against expropriation by managers and shareholders. Obviously, rational creditors recognize these potential threats and expropriation actions by shareholders. It is expected that creditors in countries with strong creditor rights would use their strong legal powers to prevent managers from high dividend payouts, so as to avoid the potential expropriation by shareholders. When firms pay excessive dividends, they might be involved into high external costs in the future. In such cases, these firms are less likely to pay dividends. Restrictive dividend policies not only protect creditors of firms from potential wealth and risk transfer behavior but also save funds that enable managers to undertake profitable investment projects. Thus, both creditors and managers are willing to employ restrictive dividend policies.
Creditors supply funds to firms since they expect that these firms would generate profits to pay back the interest and principle of the debt. Firms that fail to meet these obligations fall into bankruptcy and financial distress. Creditors have strong powers during bankruptcy and financial distress and they may use these powers to affect corporate policies. Firms in countries with strong creditor rights tend to conduct less investment and lower dividend payouts since creditors prefer low variance of cash flows (Nini, Smith, & Sufi, 2009). Additionally, creditors may reduce management’s decision-making powers, mandate the dismissal of management under bankruptcy and impose private costs on mangers through their legal protection (Acharya, Amihud, & Litov, 2011). According to the catering theory, firms emphasize investor demand in their decisions about dividend policies. Combining with these cases, it is expected that managers tend to cater to creditors’ demand when they make strategic corporate decisions. Strong creditor rights result in high costs for firms in case of default and financial distress, and hence they cater to creditors and keep them satisfied. The greater are the legal protection of creditors, the more likely managers cater to creditors’ demand, which includes restrictive dividend policies.
hand, strong creditor rights lead to high costs for firms in case of default and financial distress. Managers tend to cater to creditors’ demand and conduct restrictive dividend policies. Thus, it is expected that there is a negative relationship between creditor rights and corporate dividend policies. The hypothesis is formulated as:
Hypothesis 1b: Country-level creditor rights negatively affect corporate dividend policies.
3.2 The mandatory IFRS adoption and dividend policies
A growing literature demonstrates an important association between information asymmetry and corporate dividend policies. Particularly, Hail, Tahoun, & Wang (2014) investigate a relationship between the information environment of the firm and its dividend policies in an international setting. They employ the mandatory adoption of IFRS as one of country-level events for a general improvement of the information environment. According to the signaling theory, firms tend to reduce information asymmetry and convey private information to investors about future prospects through dividend payouts. Dividend payouts can be considered as a means of conveying management’s commitment and building a favorable reputation to attract investors. After the mandatory implement of IFRS that reduces information asymmetry and improve reporting transparency, there is no need for dividends to serve as information-conveying and reputation-building mechanisms. Additionally, low information asymmetry decreases the pressure on the management to convey private information through dividend payouts. Thus, it is expected that the mandatory adoption of IFRS negatively affect dividend policies. In contrast, after the mandatory introduction of IFRS that improve investors’ monitoring capabilities, they are able to put more pressure on corporate insiders and demand higher dividends. Thus, it is expected that the mandatory adoption of IFRS positively affect dividend policies. According to the empirical analysis, Hail, Tahoun, & Wang (2014) find that after the mandatory IFRS adoption, firms are less likely to pay dividends, but more likely to cut dividend payouts.
3.3 The role of mandatory IFRS adoption
in their best interests (Guay, 2008). In other words, creditors would do little monitoring by the debt covenants because they believe that managers would make their best effort to meet their obligations. Debtholders that represent an important class of capital providers rely heavily on financial statement information (Wu & Zhang, 2014). Guay (2008) indicates that both financial reporting and debt covenants play an important role in assisting creditors in effectively monitoring their investments. Particularly, financial reporting would provide creditors with reliable information about firms’ business operations, such as assets, leverage, cash flows, liabilities, current financial performance, etc. Debt covenants would provide creditors with control rights that enable creditors to intervene in decisions of the management when financial statement information implies that managers take actions to expropriate from creditors. Smith & Warner (1979) indicate that the major role of debt covenants is to alleviate the costs associated with the conflict of interest between shareholders and creditors. When creditors concern themselves with agency conflicts with firms, a large number of restrictive debt convents are expected to be included in debt contracting.
with firms, thus they are more likely to remove stringent requirements in debt contracts, such as requirements for restrictive dividend payouts. The reduction of information asymmetry and improvement of financial reporting weaken the role of debt covenants in mitigating agency conflicts between creditors and shareholders. Because of the effect of mandatory adoption of IFRS on the information environment, there would be less of a need for creditors to use debt covenants reduce agency conflicts with firms and there would be little reason for creditors to acquire additional control rights through debt covenants.
In the context of debt financing, the conflicts of interests between creditors and shareholders arise because managers of firms acting in the interests of shareholders might take certain actions that expropriate wealth from creditors to shareholders, such as distributing the firm’s cash to shareholders through dividend payouts (Shivakumar, 2013;
Christensen, Nikolaev, & Wittenberg-Moerman, 2016). Creditors prevent shareholders from acquiring benefits at their expense through forcing firms to undertake restrictive dividend policies. The mandatory adoption of IFRS would reduce information asymmetry and enable investors to obtain better monitoring capabilities and more sufficient information. The improvement of information environment also makes creditors more confident to evaluate firms’ financial performance. As a result, it could be expected that creditors would be more active in disciplining corporate policies and monitoring the behavior of managers after mandatory IFRS adoption. Therefore, managers are less likely to conduct the expropriation at the expense of creditors since they recognize that creditors have stronger monitoring abilities and better quality of financial reporting information. Creditors would easily observe the potential misbehavior from managers and effectively examine whether they are expropriated when they could acquire sufficient information. After the mandatory implement of IFRS, managers have less need to substitute restrictive dividend policies with weak creditor protection and creditors have little demand restrictive dividend payouts as an alternative governance mechanism.
information asymmetry, managers are less likely to conduct the expropriation at the expense of creditors. They also would be less likely to substitute restrictive dividend policies with weak creditor protection. Thus, it is expected that the mandatory adoption of IFRS decreases positive effects of country-level creditor rights on corporate dividend policies. The hypothesis is formulated as:
Hypothesis 2a: The positive effect of country-level creditor rights on corporate
dividend policies decreases with the mandatory adoption of IFRS.
Hypothesis 2b: The positive effect of country-level creditor rights on corporate
dividend policies increases with the mandatory adoption of IFRS.
high quality of accounting information, and hence assist creditors in informed and unbiased judgements about whether they are expropriated by the management (Iatridis, 2010). Creditors also rely on declined information asymmetry and improved financial reporting information to assess the ability of firms to conduct repayments of debts. Thus, managers are less likely to conduct expropriation and risk-shifting behaviors since they recognize that they would be caught easily if they misbehave. The mandatory implement of IFRS would lower creditors’ reliance on restrictive dividend policies as a disciplinary mechanism on the potential expropriation by managers and shareholders. After the mandatory adoption of IFRS that improves financial reporting and declines information asymmetry, creditors would be less concerned about agency conflicts with firms, thus they are less likely to force managers to pay low dividends by using their legal or market powers.
In sum, the mandatory adoption of IFRS enables creditors to rely more on improved financial reporting and declined information asymmetry, and less on restrictive dividend policies in mitigating the agency conflicts with firms. These would reduce the demand for creditors to use their legal protection to demand restrictive dividend policies. Thus, it is expected that the mandatory adoption of IFRS decreases the negative effect of country-level creditor rights on corporate dividend policies. The hypothesis is formulated as:
Hypothesis 3a: The negative effect of country-level creditor rights on corporate
dividend policies decreases with the mandatory adoption of IFRS.
Hypothesis 3b: The negative effect of country-level creditor rights on corporate
dividend policies increases with the mandatory adoption of IFRS. 4. Research methodology
4.1 Data and sample selection
of industrial firms. In certain circumstances, mandatory IFRS adoption might have no influence on firms that have already voluntarily switched to IFRS reporting before it was mandated, which would also might result in the improvement of financial reporting information and the reduction of information asymmetry (Daske et al., 2013). Following Hail, Tahoun, & Wang (2014), this thesis has excluded firms that voluntarily implement IFRS before the mandatory adoption. This thesis eliminates all firm-year observations with missing or abnormal necessary financial data, such as negative total assets and negative cash flows. After imposing these data requirements, this thesis obtains a sample of 279,784 firm-year observations from 40 countries over the period 1991-2013. The firm-level accounting and financial data come from DataStream. All of the accounting and financial data are winsorized at the 1st and 99th percentile to reduce the influence of potential outliers and occasional reporting errors. The measurements and sources of all the variables are listed in Table 1.
4.2 Regression variables
4.2.1 Dependent variables
This thesis examines the impact of mandatory IFRS adoption on the connection between legal protection of creditors and corporate dividend policies using the tobit regression. In the tobit regression, following the study of Brockman & Unlu (2009), this thesis measures dividend payouts, dividend to sales ratio (Div/sales), by scaling total cash dividends by net sales. As a robustness check, this thesis, following Byrne & O’Connor (2012), considers the dividend to earnings ratio (Div/earnings), defined as dividends per share divided by earnings per share. Brockman & Unlu (2009) present two dividend policy variable: the amount of dividend payouts and the probability of paying dividends. This thesis also tests the robustness of the results by using the logit regression. In the logit regression, this thesis forms a dividend dummy variable, Payer, which equals one if total cash dividends paid are positive, and zero otherwise. This thesis follows the definition of total cash dividends by La Porta et al. (2000a) and measures total dividends as cash dividends paid to common and preferred shareholders.
4.2.2 Independent variables
Table 1
Definitions of variables
Variables Measurements
Dependent variables
Div/sales Ratio of total cash dividends paid to net sales.
Div/earnings Ratio of dividends per share divided by earnings per share.
Payer Equal one if total cash dividends paid are positive, and zero otherwise.
Independent variables Country-level variables
IFRS adoption (IFRS) Equal one if a firm operates in a country that mandatorily adopts IFRS, and zero otherwise.
Creditor rights (CR) Creditor rights index from Djankov, McLiesh, & Shleifer (2007). Shareholder rights (SR) Revised anti-director index from Djankov et al. (2008).
Legal origin (Law) Equal one if a firm is originated from a common law country, and zero otherwise.
Private credit (PC) Total private credit by deposit money banks and other financial institutions over GDP from World Development Indicators of World Bank.
Firm-level variables
Growth opportunity (Grow) Annual growth rate in sales.
Profitability (Prof) Ratio of EBITDA (earnings before interest, tax, depreciation and amortization) to book value of total assets.
Firm size (Size) Natural logarithm of net sales expressed in U.S. dollars.
Share repurchases (SRE) Equal one if the amount of funds used to decrease the outstanding shares of common and/or preferred stock is positive, and zero otherwise.
Dividend payoutst-1(Divt-1) One year lagged ratio of total cash dividends to net sales.
Cash flow variability (CFV) Standard deviation of cash flow over the last three years.
Cash holdings (Cash) Ratio of cash and short-term investments to book value of total assets.
Leverage (Lev) Ratio of total debt (both short-term and long-term) to the book value of total assets.
whether there is no automatic stay or asset freeze imposed by the court; (3) whether secured creditors are paid first out of the proceeds of liquidating a bankrupt firm; and (4) whether an administrator, and not management, is responsible for running the business during the reorganization - each of which equals one if a country’s law and regulations provide the specific type of power to creditors, and zero otherwise” (Djankov, McLiesh, & Shleifer, 2007, p. 302). Thus, this index ranges from zero (weak creditor rights) to four (strong creditor rights). The creditor rights index from Djankov, McLiesh, & Shleifer (2007) is available for the period 1978-2003. Seifert & Gonenc (2016) indicate that the creditor rights index as a time-invariant country-level measure presents little time series variation over time. In addition, Hong, Hung, & Zhang (2016) show that changing country-level institutions is a slow process and thus the index of creditor rights is extremely steady over time. Thus, this thesis determines the research period between 1991-2013 and assumes that the creditor rights index remains unchanged from 2003. As mentioned in the introduction section, La Porta et al. (2000a) find a positive and significant relationship between shareholder protection and dividend policies. This thesis also employs a measure of shareholder rights from Djankov et al. (2008), namely the revised anti-director index, which measures the strength of control rights to minority shareholders against expropriation by the controlling shareholders. A high value of this index implies strong protection granted by a country’s law and regulations to minority shareholders.
Other independent variables
(constructed on 2-digit SIC codes) to control for the fixed effects of year and industry, respectively.
According to the lifecycle theory, this thesis first controls firm growth opportunity, profitability and firm size. Firm growth opportunity is annual growth rate in sales (La Porta et al., 2000a). The relationship between firm growth rate and dividend policies might be ambiguous. On the one hand, firms with good investment opportunities and growth prospects have less incentives to pay dividends, because they need internally generated funds for their profitable projects. Meanwhile, these firms might have potential needs for external funds, so they choose low dividend payouts to establish a reputation among debtholders. Their connection could be negative. On the other hand, firms with good investment opportunities and growth prospects may pay high dividends since they need to establish a reputation among shareholders to acquire their financing. Their relationship could be positive. Profitability is the ratio of EBITDA to book value of total assets, which is expected to be positively associated with corporate dividend policies (Mitton, 2004; Brockman & Unlu, 2009; O’Connor, 2013). Profitability of a firm as a source of internal funds implies the amount of earnings available to be retained. Due to the fact that profitable firms generate more internal funds, they have sufficient cash available for dividend payouts. Firm size is natural logarithm of net sales expressed in U.S. dollars, which is expected to be positively related to dividend payouts (Brockman & Unlu, 2009; Byrne & O’Connor, 2012; Shao, Kwok, & Guedhami, 2013). Large firms tend to have better access to capital markets, which lead them easier to raise funds on attractive terms such as low costs and few constraints. Thus, large firms are more likely to pay dividends.
otherwise, which is positively associated with dividend payouts. Shares repurchasing is considered as a complement of dividend payouts (Hail, Tahoun, & Wang, 2014). Following Hail, Tahoun, & Wang (2014), this thesis also adopts the dividend payouts lagged by one year to measure a firm’s payout history. Firms are more likely to pay dividends if they did before, which indicates a positive relationship between lagged dividend payouts and dividend payouts.
In addition to variables mentioned above, cash flow variability is measured as the standard deviation of cash flow over the last three years, as a proxy for firms’ cash flow uncertainty. Firms with high cash flow variability are less likely to pay dividends since they tend to worry about a lack of cash in the future, which present a negative association between cash flow variability and dividend payouts. Cash holding is the ratio of cash and short-term investments to total assets. The relationship between cash holdings and dividend policies is ambiguous (Shao, Kwok, & Guedhami, 2013; O’Connor, 2013). On the one hand, firms with high cash holding have sufficient funds to afford dividend payouts when they have no need for external financing. On the other hand, when firms have profitable projects or future growth opportunities, they need to finance them by their cash holdings. This result contributes to low dividend payouts. Leverage is the ratio of total debt (both short-term and long-term) to the book value of total assets, which is expected to be negatively associated with dividend payouts (Hail, Tahoun, & Wang, 2014). Bankruptcy might happen when a firm is not able to generate profits to pay back the interest and principle of the debt. Thus, firms tend to maintain internal cash flow to meet their obligations and hence a high level of financial leverage leads to low dividend payouts.
4.3 Regression model
To examine the hypothesis 1, the following regression equation (1) is the basic model to analyze the connection between creditor rights and dividend payouts.
𝐷𝑖𝑣$% = 𝑏(+ 𝑏*𝐶𝑅-%+ 𝑏.𝑆𝑅-%+ 𝑏0𝑃𝐶-%+ 𝑏2𝐺𝑟𝑜𝑤$% + 𝑏7𝑃𝑟𝑜𝑓$%+ 𝑏9𝑆𝑖𝑧𝑒$%+ 𝑏<𝑆𝑅𝐸$% + 𝑏>𝐷𝑖𝑣$%?*+ 𝑏@𝐶𝐹𝑉$%+ 𝑏*(𝐶𝑎𝑠ℎ$% + 𝑏**𝐿𝑒𝑣$%
+ 𝑏G𝑌𝑒𝑎𝑟_𝐷𝑢𝑚𝑚𝑖𝑒𝑠G+ 𝑏L𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦_𝐷𝑢𝑚𝑚𝑖𝑒𝑠L
To investigate the hypothesis 2 and 3, the mandatory adoption of IFRS and the interaction term of the mandatory adoption of IFRS and creditor rights are added to the model (1), so as to examine impact of mandatory IFRS adoption on the relationship between creditor rights and dividend policies.
𝐷𝑖𝑣$% = 𝑐(+ 𝑐*𝐼𝐹𝑅𝑆-%+ 𝑐.𝐶𝑅-%+ 𝑐0𝐼𝐹𝑅𝑆-%∗ 𝐶𝑅-%+ 𝑐2𝑆𝑅-%+ 𝑐7𝑃𝐶-%+ 𝑐9𝐺𝑟𝑜𝑤$% + 𝑐<𝑃𝑟𝑜𝑓$%+ 𝑐>𝑆𝑖𝑧𝑒$%+ 𝑐@𝑆𝑅𝐸$%+ 𝑐*(𝐷𝑖𝑣$%?*+ 𝑐**𝐶𝐹𝑉$%+ 𝑐*.𝐶𝑎𝑠ℎ$%
+ 𝑐*0𝐿𝑒𝑣$%+ 𝑐G𝑌𝑒𝑎𝑟_𝐷𝑢𝑚𝑚𝑖𝑒𝑠G+ 𝑐L𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦_𝐷𝑢𝑚𝑚𝑖𝑒𝑠L
+ 𝜇$% 2 This thesis employs the interactive term of mandatory IFRS adoption and creditor rights to examine the effect of mandatory IFRS adoption on the relationship between creditor rights and dividend policies. In addition, this thesis not only includes a set of firm-specific control variables but also include a set of country-level control variables that may affect corporate dividends. The coefficient 𝑐0 presents the extent of influence of mandatory IFRS adoption on the effect of creditor rights on dividend policies.
5. Empirical results
5.1 Descriptive statistics
Table 2
Dividend payouts by country
Country # of Obs. Div/sales Payer Creditor rights Shareholder rights Mandatory IFRS adoption
Mean Median Number Mean
Argentina 738 0.0197 0.0000 336 45.53% 1.00 2.00 2012 Australia 11,562 0.0269 0.0000 5,541 47.92% 3.00 4.00 2005 Austria 274 0.0139 0.0092 194 70.80% 3.00 2.50 2005 Belgium 821 0.0219 0.0065 525 63.95% 2.00 3.00 2005 Brazil 1,868 0.0281 0.0129 1,303 69.75% 1.00 5.00 2010 Canada 9,842 0.0194 0.0000 3,803 38.64% 1.00 4.00 2011 Chile 1,243 0.0476 0.0247 1,014 81.58% 2.00 4.00 2010 China 4,179 0.0364 0.0174 2,828 67.67% 2.00 1.00 n.a Colombia 317 0.0507 0.0297 237 74.76% 0.00 3.00 n.a Denmark 1,973 0.0163 0.0055 1,242 62.95% 3.00 4.00 2005 Finland 1,914 0.0258 0.0140 1,508 78.79% 1.00 3.50 2005 France 8,414 0.0139 0.0058 5,397 64.14% 0.00 3.50 2005 Germany 4,814 0.0140 0.0029 2,650 55.05% 3.00 3.50 2005 Greece 1,520 0.0169 0.0016 943 62.04% 1.00 2.00 2005 Hungary 180 0.0167 0.0000 64 35.56% 1.00 2.00 2005 India 14,579 0.0151 0.0056 9,841 67.50% 2.00 5.00 n.a Indonesia 4,059 0.0185 0.0000 2,025 49.89% 2.00 4.00 n.a Ireland 1,028 0.0130 0.0062 596 57.98% 1.00 5.00 2005 Israel 1,255 0.0257 0.0000 478 38.09% 3.00 4.00 2008 Italy 3,442 0.0194 0.0073 2,207 64.12% 2.00 2.00 2005 Japan 37,737 0.0084 0.0056 32,167 85.24% 2.00 4.50 n.a Malaysia 11,063 0.0259 0.0080 6,774 61.23% 3.00 5.00 n.a Mexico 1,442 0.0171 0.0000 648 44.94% 0.00 3.00 2012 Netherlands 2,395 0.0148 0.0068 1,706 71.23% 3.00 2.50 2005 New Zealand 818 0.0429 0.0210 583 71.27% 4.00 4.00 2007 Norway 1,500 0.0194 0.0033 793 52.87% 2.00 3.50 2005 Pakistan 1,499 0.0275 0.0080 1,033 68.91% 1.00 4.00 2007 Peru 80 0.0272 0.0000 32 40.00% 0.00 4.50 2012 Philippines 1,791 0.0240 0.0000 734 40.98% 1.00 4.00 2005 Poland 1,470 0.0116 0.0000 581 39.52% 1.00 2.00 2005 Portugal 760 0.0218 0.0059 505 66.45% 1.00 2.50 2005 Singapore 7,079 0.0278 0.0092 4,528 63.96% 3.00 5.00 2003 South Africa 3,794 0.0273 0.0113 2,718 71.64% 3.00 5.00 2005 Spain 1,498 0.0326 0.0146 1,038 69.29% 2.00 5.00 2005 Sweden 4,151 0.0193 0.0078 2,410 58.06% 1.00 3.50 2005 Switzerland 1,638 0.0195 0.0103 1,174 71.67% 1.00 3.00 2005 Thailand 6,248 0.0347 0.0154 4,050 64.82% 2.00 4.00 n.a Turkey 1,492 0.0275 0.0000 736 49.33% 2.00 3.00 2006 United Kingdom 25,987 0.0201 0.0112 16,989 65.37% 4.00 5.00 2005
United States 93,320 0.0087 0.0000 27,223 29.17% 1.00 3.00 n.a
Total 279,784 0.0158 0.0019 149,154 53.31% 1.84 3.80
Table 3
Dividend payouts by year
Year # of Obs. Div/sales Payer
Mean Median Number Mean
1991 5,375 0.0193 0.0095 3,911 72.76% 1992 5,933 0.0193 0.0082 4,111 69.29% 1993 6,267 0.0184 0.0075 4,260 67.98% 1994 6,577 0.0180 0.0073 4,392 66.78% 1995 7,698 0.0166 0.0056 4,719 61.30% 1996 8,684 0.0160 0.0052 5,160 59.42% 1997 9,474 0.0151 0.0037 5,309 56.04% 1998 9,868 0.0140 0.0000 4,835 49.00% 1999 11,132 0.0125 0.0000 4,631 41.60% 2000 12,867 0.0119 0.0000 5,881 45.71% 2001 13,774 0.0123 0.0000 6,228 45.22% 2002 14,376 0.0118 0.0000 6,311 43.90% 2003 15,115 0.0125 0.0000 6,897 45.63% 2004 15,873 0.0140 0.0000 7,551 47.57% 2005 16,611 0.0157 0.0000 8,284 49.87% 2006 16,673 0.0162 0.0008 8,511 51.05% 2007 17,194 0.0170 0.0015 8,975 52.20% 2008 16,472 0.0172 0.0022 8,843 53.69% 2009 14,502 0.0159 0.0018 7,734 53.33% 2010 15,171 0.0155 0.0014 8,097 53.37% 2011 14,583 0.0166 0.0028 8,324 57.08% 2012 12,147 0.0222 0.0055 7,428 61.15% 2013 13,418 0.0220 0.0061 8,762 65.30% Total 279,784 0.0158 0.0019 149,154 53.31%
This table presents summary statistics for the firm-year observations of each year, mean and median of dividend to sales ratio and number and mean of payer dummy variable over the sample period 1991-2013. Exact measurements of the variables are given in Table 1. The sample consists of 279,784 firm-year observations from 40 countries over the period 1991-2013.
0 to 4 with an average of 1.84. The shareholder rights index in the sample county range from 1 to 5 with a mean of 3.80. The last column of this table presents the year of mandatory IFRS adoption in the sample period.
variable declines over time, reaches the bottom in 1999, and generally increases subsequently.
Table 4 presents summary statistics for the firm-level variables and country-level variables in the regression model for three different samples: the whole sample, the sample with low creditor rights and the sample with high creditor rights. Following the study of Seifert & Gonenc (2016), this thesis, based on the sample median of creditor rights, divides the whole sample into the sample with high creditor rights (creditor rights > 2) and the sample with low creditor rights (creditor rights ≤ 2). This thesis also makes statistical comparisons of means and medians of those variables between two subsamples. Panel A shows descriptive statistics for the country-level variables and Panel B presents descriptive statistics for the firm-level variables. Based on the results of the t-test, the mean dividend to sales ratio (0.0229) of firms in countries with high creditor rights is significantly higher than that (0.0134) of firms in countries with low creditor rights. Accounting to the Wilcoxon rank-sum test/ Mann-Whitney U test, the median dividend to sales ratio of 0.0081 for firms in countries with high creditor rights is significantly higher than 0.0001 for firms in countries with low creditor rights. These difference-in-difference results from univariate analyses provide preliminary evidence that creditor protection results in an increase in dividend payouts, which is consistent with Hypothesis 1a that indicates a positive relationship between creditor rights and dividend payouts.
5.2 Correlation results
Table 6
The effects of creditor rights on dividend payouts
Panel A: All firms (Div/sales)
Panel B: Alternative dependent variable
(Div/earnings) Model (1) Model (2) Model (3) Model (4) Model (5) Creditor rights 0.0091*** 0.0064*** 0.0048*** 0.0036*** 0.0673*** (0.0001) (0.0001) (0.0001) (0.0001) (0.0016) Shareholder rights 0.0042*** 0.0023*** 0.0792*** (0.0002) (0.0001) (0.0018) Private credit -0.0107*** -0.0058*** -0.0798*** (0.0002) (0.0002) (0.0031) Growth -0.0062*** -0.0064*** -0.0624*** (0.0001) (0.0001) (0.0028) Profitability 0.0351*** 0.0327*** 1.3709*** (0.0005) (0.0005) (0.0161) Size 0.0036*** 0.0040*** 0.0728*** (0.0000) (0.0000) (0.0009) Share repurchases 0.0005** 0.0018*** 0.0104*** (0.0002) (0.0002) (0.0035) Dividend payoutst-1 0.8395*** 0.8325*** 0.6562*** (0.0021) (0.0021) (0.0031) Cash flow variability -0.0219*** -0.0202*** -2.1381***
(0.0008) (0.0008) (0.0291) Cash holdings -0.0040*** -0.0013** -0.1508*** (0.0005) (0.0005) (0.0106) Leverage -0.0161*** -0.0165*** -0.3201*** (0.0003) (0.0003) (0.0065) Intercept 0.0064*** 0.0047** -0.0509*** -0.0568*** -1.1783*** (0.0019) (0.0020) (0.0014) (0.0015) (0.0261)
Year dummies Yes Yes Yes Yes Yes
Industry dummies Yes Yes Yes Yes Yes Left censored 130,630 130,630 130,630 130,630 142,286 Observations 279,784 279,784 279,784 279,784 262,819
This table reports tobit regression results for the impact of creditor rights on dividend payouts. Exact measurements of the variables are given in Table 1. All of the regressions include industry and year fixed effects. The standard errors are presented below the estimated coefficients. The symbol ***, ** and * indicate statistical significance at the 1%, 5% and 10% levels, respectively.
as the accepted limit of VIF. The results present that all of the VIF are below 2, indicating that there are no problems of multicollinearity in the regression model.
5.3 Creditor rights and dividend payouts
In Panel A, Model 1 includes creditor rights and shows that the coefficient on creditor rights is significantly positive at the 1% level, implying a positive relationship between creditor rights and dividend payouts. In Model 2, the other country-level variables are added to the tobit regression model, namely shareholder rights and private credit. There is similar evidence that creditor protection is positively related to dividend payouts. Model 3 removes the country-level control variables and instead puts in the firm-level control variables, namely growth opportunity, profitability, firm size, share repurchases, lagged dividend payouts, cash flow variability, cash holdings, and leverage. The results present that there is a significantly positive relationship between creditor rights and dividend payouts. In Model 4, this thesis includes both the country-level and the firm-level control variables in the tobit regression model and the coefficient on creditor rights is 0.0036, which is significant at the 1% level. This result continues to show that creditor protection is positively associated with dividend payouts. In Panel B, this thesis examines whether the positive impact of creditor rights on dividend payouts persists when applying an alternative measure of dividend payouts, namely the dividend to earnings ratio in Model 5 of Panel B. The coefficient on creditor rights is positive and statistically significant at the 1% level, which is consistent with previous four models. Thus, the main finding of a positive effect of creditor rights on dividend payouts is robust to the choice of a dividend payout measure.
Taken together, these findings in Table 6 point out that creditor rights have significantly positive effects on dividend payouts, consistent with Hypothesis 1a. This evidence also is in support of the substitution model of dividends from the study of Brockman & Unlu (2009), which points out that managers of firms substitute low dividends for weak legal protection of creditors.
5.4 The role of mandatory IFRS adoption 5.4.1 Main results
Table 7
The joint effects of mandatory IFRS adoption and creditor rights on dividend payouts
Panel A: All firms (Div/sales)
Panel B: Alternative dependent variable
(Div/earnings) Model (1) Model (2) Model (3) Model (4) IFRS adoption -0.0011*** -0.0035*** 0.0061*** 0.1929*** (0.0002) (0.0002) (0.0005) (0.0084) Creditor rights 0.0040*** 0.0052*** 0.0959*** (0.0001) (0.0001) (0.0019) IFRS adoption*CR -0.0042*** -0.0932*** (0.0002) (0.0032) Shareholder rights 0.0047*** 0.0021*** 0.0019*** 0.0746*** (0.0001) (0.0001) (0.0001) (0.0018) Private credit -0.0057*** -0.0063*** -0.0053*** -0.0588*** (0.0002) (0.0002) (0.0002) (0.0033) Growth -0.0064*** -0.0064*** -0.0063*** -0.0602*** (0.0001) (0.0001) (0.0001) (0.0028) Profitability 0.0329*** 0.0327*** 0.0330*** 1.3894*** (0.0005) (0.0005) (0.0005) (0.0162) Size 0.0040*** 0.0041*** 0.0040*** 0.0722*** (0.0000) (0.0000) (0.0000) (0.0009) Share repurchases 0.0009*** 0.0018*** 0.0019*** 0.0127*** (0.0002) (0.0002) (0.0002) (0.0035) Dividend payoutst-1 0.8401*** 0.8336*** 0.8332*** 0.6522*** (0.0021) (0.0021) (0.0021) (0.0031) Cash flow variability -0.0196*** -0.0199*** -0.0193*** -2.0978***
(0.0008) (0.0008) (0.0008) (0.0291) Cash holdings -0.0022*** -0.0014** -0.0015*** -0.1593*** (0.0005) (0.0005) (0.0005) (0.0106) Leverage -0.0172*** -0.0165*** -0.0165*** -0.3178*** (0.0003) (0.0003) (0.0003) (0.0065) Intercept -0.0569*** -0.0568*** -0.0592*** -1.2363*** (0.0015) (0.0015) (0.0015) (0.0261)
Year dummies Yes Yes Yes Yes
Industry dummies Yes Yes Yes Yes
Left censored 130,630 130,630 130,630 142,286 Observations 279,784 279,784 279,784 262,819
rights on dividend payouts, which supports Hypothesis 2a.
In Panel A, Model 1 shows that the coefficient on mandatory IFRS adoption is significantly negative at the 1% level, implying a negative association between mandatory IFRS adoption and dividend payouts. In Model 2, this thesis adds creditor rights into the regression model and find similar evidence that there is significantly negative connection between mandatory IFRS adoption and dividend payouts. The results in Model 1 and Model 2 confirm the results in the study of Hail, Tahoun, & Wang (2014), which indicates that there is a significantly negative relationship between mandatory IFRS adoption and dividend policies. Model 3 includes the interaction variable between mandatory IFRS adoption and creditor rights and then investigates the effect of mandatory IFRS adoption on the relationship between creditor rights and dividend payouts. The results present that the coefficient on the interaction variable between mandatory IFRS adoption and creditor rights is negative and significant at the 1% level, which indicates that mandatory IFRS adoption significantly reduces the positive effect of creditor rights on dividend payouts. In Panel B, this thesis examines whether the negative impact of mandatory IFRS adoption on the relationship between creditor rights and dividend payouts persists when applying an alternative measure of dividend payouts, namely the dividend to earnings ratio in Model 4. The coefficient on the interaction term between mandatory adoption of IFRS and creditor rights is negative and statistically significant at the 1% level, which is consistent with the results in Model 3. Thus, the main finding of a negative effect of mandatory IFRS adoption on the relationship between creditor rights and dividend payouts is robust to the choice of a dividend payout measure. Overall, these findings in Table 7 indicate that mandatory IFRS adoption significantly reduces the positive effects of creditor rights on dividend payouts, consistent with Hypothesis 2a. This thesis also finds evidence consistent with the findings from the study of Hail, Tahoun, & Wang (2014) that mandatory IFRS adoption negatively affects dividend policies.
La Porta et al. (2000a). Private credit is negatively related to dividend payouts, presenting that more debts with dividend restrictions contribute to low dividend payouts.
The results on the firm-level variables are generally consistent with prior empirical studies. Large and profitable firms pay higher amount of dividends while firms with high growth opportunities pay lower amount of dividends. These results are in line with the lifecycle theory and the findings in Brockman & Unlu (2009). The positive and significant relation between share purchases and dividend payouts is consistent with the study of Hail, Tahoun, & Wang (2014), indicating that share purchases are complementary to dividend payouts. The positive and significant coefficient on lagged dividend payouts suggests that firms with a dividend payout history could continue to pay dividends. This finding is also in line with the study of Hail, Tahoun, & Wang (2014). The coefficient on cash flow variability is significantly negative, which suggests that firms with high cash flow variability pay lower amount of dividends because they might need cash in the future. Cash holdings have a negative and significant effect on dividend payouts, which indicate that firms tend to finance future growth opportunities by cash holdings rather than pay dividends. This result is consistent with the study of Brockman & Unlu (2009). Similarly, leverage has a significantly negative coefficient on dividend payouts, which is in line with the study of Hail, Tahoun, & Wang (2014). One reason for these lower dividends might be that firms with high level of financial leverage are more likely to meet their obligations instead of paying dividends.
5.4.2 Robustness tests
In this subsection, this thesis further validates the main findings by re-estimating the baseline regression specification using the sample including voluntary IFRS firms, an alternative regression model, legal origin as an additional control variable, and different sample compositions. This thesis uses the regression equation (2) as the baseline regression model in these robustness checks. The results presented in Table 8, 9 and 10 lend additional support to Hypothesis 2a, which suggests a negative effect of mandatory IFRS adoption on the relationship between creditor rights and dividend policies.
Including voluntary IFRS firms
Table 8
The results with the sample including voluntary IFRS firms
Panel A: All firms (Div/sales)
Panel B: Alternative dependent variable
(Div/earnings) Model (1) Model (2) Model (3) Model (4) IFRS adoption -0.0010*** -0.0034*** 0.0062*** 0.1906*** (0.0002) (0.0002) (0.0004) (0.0079) Creditor rights 0.0038*** 0.0051*** 0.0976*** (0.0001) (0.0001) (0.0018) IFRS adoption*CR -0.0042*** -0.0924*** (0.0002) (0.0030) Shareholder rights 0.0044*** 0.0021*** 0.0019*** 0.0689*** (0.0001) (0.0001) (0.0001) (0.0016) Private credit -0.0056*** -0.0062*** -0.0051*** -0.0584*** (0.0002) (0.0002) (0.0002) (0.0032) Growth -0.0065*** -0.0065*** -0.0063*** -0.0602*** (0.0001) (0.0001) (0.0001) (0.0027) Profitability 0.0343*** 0.0341*** 0.0344*** 1.3958*** (0.0005) (0.0005) (0.0005) (0.0158) Size 0.0040*** 0.0040*** 0.0040*** 0.0714*** (0.0000) (0.0000) (0.0000) (0.0008) Share repurchases 0.0008*** 0.0018*** 0.0019*** 0.0129*** (0.0002) (0.0002) (0.0002) (0.0034) Dividend payoutst-1 0.8384*** 0.8324*** 0.8317*** 0.6475*** (0.0020) (0.0020) (0.0020) (0.0030) Cash flow variability -0.0201*** -0.0203*** -0.0196*** -2.1219***
(0.0008) (0.0008) (0.0008) (0.0286) Cash holdings -0.0021*** -0.0014** -0.0015*** -0.1564*** (0.0005) (0.0005) (0.0005) (0.0104) Leverage -0.0175*** -0.0169*** -0.0168*** -0.3209*** (0.0003) (0.0003) (0.0003) (0.0063) Intercept -0.0560*** -0.0560*** -0.0586*** -1.2078*** (0.0014) (0.0014) (0.0014) (0.0255)
Year dummies Yes Yes Yes Yes
Industry dummies Yes Yes Yes Yes
Left censored 134,737 134,737 134,737 146,807 Observations 290,917 290,917 290,917 273,642
Table 9
The results with an alternative research design
Panel A: All firms Panel B: Including voluntary IFRS firms Model (1) Model (2) Model (3) Model (4) IFRS adoption -0.1990*** -0.3849*** 0.4291*** 0.4388*** (0.0214) (0.0219) (0.0440) (0.0413) Creditor rights 0.3076*** 0.4158*** 0.4067*** (0.0090) (0.0103) (0.0098) IFRS adoption*CR -0.3580*** -0.3589*** (0.0167) (0.0159) Shareholder rights 0.2855*** 0.0990*** 0.0911*** 0.1030*** (0.0078) (0.0095) (0.0094) (0.0087) Private credit -0.2665*** -0.2985*** -0.1973*** -0.1857*** (0.0161) (0.0161) (0.0168) (0.0163) Growth 0.0021 -0.0018 0.0030 0.0012 (0.0097) (0.0097) (0.0097) (0.0095) Profitability 2.0871*** 2.1160*** 2.1558*** 2.2121*** (0.0525) (0.0530) (0.0532) (0.0525) Size 0.2570*** 0.2652*** 0.2644*** 0.2646*** (0.0045) (0.0045) (0.0045) (0.0044) Share repurchases 0.1263*** 0.2028*** 0.2091*** 0.2089*** (0.0180) (0.0182) (0.0183) (0.0179) Dividend payoutst-1 4.1646*** 4.1530*** 4.1312*** 4.0879*** (0.0147) (0.0147) (0.0147) (0.0143) Cash flow variability -1.3097*** -1.3698*** -1.3387*** -1.3728***
(0.0742) (0.0753) (0.0748) (0.0742) Cash holdings -0.3582*** -0.3479*** -0.3684*** -0.3760*** (0.0476) (0.0478) (0.0479) (0.0469) Leverage -1.1557*** -1.1244*** -1.1397*** -1.1488*** (0.0298) (0.0299) (0.0300) (0.0293) Intercept -4.9244*** -4.9312*** -5.1485*** -5.1519*** (0.1387) (0.1403) (0.1407) (0.1360)
Year dummies Yes Yes Yes Yes
Industry dummies Yes Yes Yes Yes
R-squared 0.6147 0.6177 0.6189 0.6141 Observations 279,784 279,784 279,784 290,917
follow transparent reporting and informative disclosures. Table 8 reports results from the tobit regression for the effect of mandatory IFRS adoption on the relationship between creditor rights and the dividend to sales ratio and the dividend to earnings ratio when adding firms that voluntarily adopt IFRS to the whole sample. In Panel A, Model 1 and 2 show a negative and statistically significant (at the 1% level) coefficient in mandatory IFRS adoption, indicating that mandatory IFRS adoption reduces dividend payouts. The results in Model 3 consistently presents that mandatory IFRS adoption have a negative moderating effect on the relationship between creditor rights and dividend payouts, even when using an alternative dependent variable in Model 4 of Panel B. The negative impact of mandatory IFRS adoption on dividend payouts and the negative effect of mandatory IFRS adoption on the relationship between creditor rights and dividend payouts are unaffected when including voluntarily adopting firms. The results in Table 8 using the sample including voluntary IFRS firms are in general consistent with those reported in Table 7.
Alternative research design: logit regression model
Following the study of Brockman & Unlu (2009), this thesis also applies the logit regression model to check whether the main results are consistent with the alternative research design. Table 9 reports results from the logit regression for the effect of mandatory IFRS adoption on the relationship between creditor rights and the likelihood of paying dividends. This thesis runs the logit regression using a same set of control variables in the tobit regression model. In Panel A, the results in Model 1 and 2 show a significantly negative coefficient on mandatory IFRS adoption, indicating that firms are less likely to pay dividends after mandatory IFRS adoption. Model 3 presents that the coefficient on the interaction term between mandatory IFRS adoption and creditor rights loads negatively at the 1% level, suggesting that mandatory IFRS adoption reduce positive effects of creditor rights on the likelihood to pay dividends, even adding the voluntary IFRS firms to the whole sample in Model 4 of Panel B. These results are consistent with dividend payout results in Table 7, and thus the main finding of negative effect of mandatory IFRS adoption on the association between creditor rights and dividend policies is robust to the choice of the regression model.
