This is to certify that the dissertation entitled THE DISCRETIONARY REPORTING OF NONCONTROLLING INTERESTS AND ITS ASSOCIATION WITH THE MARKET ASSESSMENT OF CREDIT RISK presented by _ =93! BEI DONG V -*. LIBRARY Mlchlgan State ! ifl-ix. “1‘37"?“ «.— has been accepted towards fulfillment of the requirements for the Doctoral degree in Accounting and Information System K-K_______\. Major Professor’s Signature LI'IlgIOX Date MSU is an affirmative-action, equal-opportunity employer “'3‘“ PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE CII 7. I; 8g.“ 99 9 Add 5/08 KIProilAccaPreleIRCIDaIeDue.indd THE DISCRETIONARY REPORTING OF NONCONTROLLING INTERESTS AND ITS ASSOCIATION WITH THE MARKET ASSESSMENT OF CREDIT RISK By Bei Dong A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Accounting and Information Systems 2008 ABSTRACT THE DISCRETIONARY REPORTING OF NONCONTROLLING INTERESTS AND ITS ASSOCIATION WITH THE MARKET ASSESSMENT OF CREDIT RISK By Bei Dong This dissertation empirically examines the determinants of firms’ decision to classify noncontrolling interests in the liability versus mezzanine section of the balance sheet, and the implication of the classification decision on the credit market’s assessment of risk. I hypothesize and find that firms with greater exposure to litigation risk from noncontrolling shareholders and firms with more serious information asymmetry problems are more likely to classify noncontrolling interests as a liability to mitigate agency costs. However, consistent with more serious agency conflicts due to controlling ownership, I predict and find that family firms are more likely to adopt the mezzanine classification. After controlling for the cndogeneity of firms’ classification choice and other factors affecting credit ratings, I find that firms choosing to disclose noncontrolling interests in the liability section have higher credit ratings (around one notch) compared with those that disclose noncontrolling interests in the mezzanine section. This finding is consistent with my hypothesis that creditors are more likely to favorably assess the credit risk of companies that conservatively classify claims on their balance sheet. Overall, my results suggest that the classification choice of noncontrolling interests is informative about the market’s assessment of firms” credit risk, and therefore, limiting accounting choices could potentially eliminate relevant information from financial statements in situations where significant information asymmetry and agency costs may be prominent. To my husband, my mother and father. iii ACKNOWLEDGMENTS I would like to thank members of my dissertation committee for their invaluable guidance and support: K. Ramesh (Chair), Ana Herrera, Marilyn Johnson, and Ed Outslay. Without their support, I could not have finished this long journey. I would especially like to thank Ramesh for his motivation, guidance, inspiration, and insight throughout the whole doctoral program. My experience to work with him during the past years has been an extremely valuable and enjoyable one. One thing I learnt through this journey from him is that you must believe in what you are doing and work confidently in order to make things happen. I would also like to thank Donna Coallier of PWC, Chris E. Hogan, John (Xuefeng) Jiang, Gregory Jonas of Moody’s, Yen-Jung Lee, Edward (Xuejun) Li, Kathy Petroni, Min Shen, Mike Shields, and Isabel (Yanyan) Wang and workshop participants at Michigan State University, University of South Florida at Tampa, and Western Michigan University for helpful comments and suggestions. Most importantly, I would like to thank my husband, Jing Wang, my parents, Xiao Bing Yuan and Shou Nian Dong for their love and support. iv TABLE OF CONTENTS LIST OF TABLES--- - - - _ - - -_ - -- . ------- vii LIST OF FIGURES - - - - - -- - - ------ - - ---------------------------- viii CHAPTER 1 INTRODUCTION--- - . ......... - l CHAPTER 2 DETERMINANTS OF THE DECISION TO CLASSIFY NONCONTROLLING INTERESTS AS LIABILITY ........ 7 2.1 Controlling-noncontrolling shareholder agency problem ............... 8 2.1.1 Family firms ................................................................ 10 2.1.2 Exposure to litigation by noncontrolling shareholders ................................................................ 1 l 2.2 Manager-shareholder agency problem .......................................... 14 2.2.1 Alleviate information asymmetry problems in growth firms ................................................................ 14 2.2.2 Alleviate manager-shareholder agency problem in financially-constrained firms ...................................... 15 CHAPTER 3 CREDIT MARKET EFFECTS OF BALANCE-SHEET CLASSIFICATION OF NONCONTROLLING INTERESTS .......... 17 CHAPTER 4 RESEARCH DESIGN -- - - _ -_ - _ 20 4.1 Determinants of the decision to classify noncontrolling interests as liabilities ..................................................................... 20 4.2 Credit market effects of the balance-sheet classification of noncontrolling interests ................................................................. 22 CHAPTER 5 SAMPLE AND UNIVARIATE ANALYSIS ------ -- - - 25 CHAPTER 6 MULTIVARIATE RESULTS - - -- - ----- 27 6.1 Determinants of the reporting decision ......................................... 27 6.2 Credit market analysis of reporting decisions ............................... 29 6.3 Additional analysis ........................................................................ 31 6.3.1 Firm characteristics ..................................................... 31 6.3.2 Earnings quality and liability classification choice 32 CHAPTER 7 CONCLUSION ........................................................................................ 35 APPENDICES- -- - -- -- - -- -_ _ _ - ------ 47 Appendix A: Background ................................................................... 47 Appendix B: Credit ratings’ numerical conversions ........................... 51 Appendix C: Sensitivity analysis ........................................................ 52 Appendix D: Balance-sheet classification examples .......................... 57 BIBLIOGRAPHY ........................................................................................................... 62 vi LIST OF TABLES Table 1 Summary Statistics ........................................................................................... 38 Table 2 Correlation Matrix ............................................................................................ 40 Table 3 Regression results for Eq. (1): Analysis of the determinants of the decision to disclose noncontrolling interests in liability sections ....................................... 44 Table 4 Regression results for Eq. (2): Instrumental variable estimation of credit market effect of disclosing noncontrolling interests as liability ...................... 45 Table 5 Descriptive statistics of 207 S&P 500 firms in fiscal year 2004 across classification choice for noncontrolling interests ............................................. 46 Table C.1 Descriptive statistics of S&P 500 firms in fiscal year 2004 across classification choice for noncontrolling interests ................................................................... 52 Table C.2 Rank regression results for Eq. (1): Analysis of the determinants of the decision to disclose noncontrolling interests in liability sections .................... 53 Table C.3 Regression results for Eq. (1): Analysis of the determinants of the decision to disclose noncontrolling interests in liability sections ....................................... 54 Table C.4 Regression results for Eq. (2) (using RATING_Klock as dependent variable): Instrumental variable estimation of credit market effect of disclosing noncontrolling interests as liability .................................................................. 55 Table C.5 Regression results for Eq. (2) without control: Instrtunental variable estimation of credit market effect of disclosing noncontrolling interests as liability ........ 56 vii LIST OF FIGURES Figure 1 Two types of agency problems: manager-shareholder agency problem and controlling-noncontrolling shareholder agency problem ................................. 37 viii CHAPTER 1 INTRODUCTION I investigate firms’ balance-sheet classification decision of noncontrolling interests in the pre-F AS 160 era, and the association between the classification choice and firms’ credit ratings.l The Financial Accounting Standards Board (hereafter, FASB) defines noncontrolling interests as “the portion of the equity (residual interest) in a subsidiary attributable to the owners of the subsidiary other than the parent and the parent’s affiliates.”2 FAS 160 “Noncontrolling Interests in Consolidated Financial Statements,” which is effective for fiscal periods beginning after 12/ 1 5/2008, views most noncontrolling interests in subsidiaries as equity of the consolidated group.3 However, F AS 160 has raised considerable debate among practitioners and in the business community.4’ 5 Moreover, given that noncontrolling interests fall in a grey area on the I My sample includes S&P 500 firms with noncontrolling interests during the fiscal year 2004 before FAS 160 had taken effect when firms had discretion in classifying noncontrolling as liability, equity, or mezzanine. 2The FASB believes the term noncontrolling interest is more precise than minority interests, because the latter implies that the holder has less than 50 percent of the voting shares. However, in the case of a variable interest entity under Interpretation 46(R), a controlling interest can hold a minority of the voting shares but still exercise control through other means. 3 FAS 150 requires that mandatorily redeemable securities (e.g., mandatorily redeemable noncontrolling interests) that “embody an unconditional obligation requiring the issuer’s redemption by transfer of assets at a specified or determinable date (or dates) or upon an event that is certain to occur" be disclosed in the liability section. I find that firms in my sample did not change their classification of noncontrolling interests during the period 2002 — 2004, suggesting that FAS 150 did not have an effect on my sample firms. In other words, my dissertation focuses on the classification of non-mandatorily redeemable noncontrolling interests for which firms have discretion in classifying them as liabilities, equity, or mezzanine. 4 Based on forty-nine comment letters to the 2005 Exposure Draft, “Consolidated Financial Statements, Including Accounting and Reporting Noncontrolling Interests in Subsidiaries,” (hereafter, 2005 ED), the majority of respondents did not support the classification of noncontrolling interests as equity. Ms. Seidman, one of the FASB members, disagreed with the equity classification because she believed that noncontrolling interests is different from the residual interests of parent firms. For example, noncontrolling interests do not “absorb losses relating to other activities of the parent and thus would be higher in standing than the residual claim of shareholders of the parent entity (pg. 3, FAS I60)” 5 Appendix A provides background information on accounting rules for the balance-sheet classification of noncontrolling interests prior to F AS 160, the debate among practitioners, academics. and standard setters over the proper classification, and the competing theories underlying the debate. liability-equity spectrum, their balance sheet classification was a key issue in the FASB’s broader project on liabilities and equity. Despite the importance of the classification issue, there is no systematic empirical evidence on corporate reporting of noncontrolling interests. For example, I find that in year 2004 roughly a quarter of the S&P 500 firms with noncontrolling interests disclose them in the liability section, with the remainder of the firms reporting in the mezzanine section (i.e., between liability and equity), suggesting significant cross-sectional variation in classification practicesé‘ 7 Given the importance of financial leverage to firm value (e.g., Hamada 1969; Rubinstein 1973; Bowman 1980; Mulford 1985; Choi 1988) and the significant consequences of liability-equity classification to different stakeholders (Hopkins 1996; Engel et al. 1999), I argue that the accounting classification choice for noncontrolling interests is an important financial reporting decision. Extant literatures argue that accounting choices can be used to alleviate various agency problems (Jensen and Meckling 1976; Smith and Warner 1979; Watts and Zimmerman 1986; Fields et al. 2001).8 Further, recent accounting research suggests that accounting conservatism can mitigate agency costs, and thereby, reduce the cost of capital (Ahmed et al. 2002; Francis et al. 2004; Zhang 2006). In my setting, I view the 6 None of my sample firms disclosed noncontrolling interests in the equity section. In its survey of consolidated financial statement, the American Institute of Certified Public Accountants reported that only 3 out of 85 companies display noncontrolling interests in the equity section. Further, AICPA survey “Accounting Trends & Techniques “of 600 public companies during the period 1993 to 2002 also shows that classification variation in their sample exists only between liability and mezzanine choices. Exhibits 1, 2, and 3 provide examples of liability (United Technologies), mezzanine (Borders Group), and equity (from Appendix A of F AS 160) classification of noncontrolling interests. 7 Based on my analysis of S&P 500 firms’ lO-K forms during the period 1993 — 2004, I find that 4 out of 263 firms with noncontrolling interests in 2004 changed their classification at least once during this period. Of those 4 firms, 1 (2) of them changed from mezzanine (liability) classification to liability (mezzanine) classification, and one firm changed its classification twice. classification of noncontrolling interests in the liabilities section as a conservative financial reporting choice as it leads to understatement of net assets. Unlike most prior research which focuses on owner-manager agency problems, I contend that the classification choice for noncontrolling interests is also influenced, may be to a greater extent, by agency problems between controlling and noncontrolling shareholders. With respect to agency problems arising from control, I hypothesize that: 1) due to better alignment of the interests of managers and family owners, family firms are more likely to report noncontrolling interests as lower priority claims (i.e., mezzanine), possibly reflecting their indifference to protecting the interests of noncontrolling shareholders; and 2) however, firms with higher litigation exposure to noncontrolling shareholders (“deep pocket”) are more likely to choose the liability classification to signal reduced agency conflicts. From the standpoint of manager-shareholder agency problems, I examine whether the managerial decision to classify noncontrolling interests as liability is related to incentives for: 1) alleviating information asymmetries generally; and 2) limiting agency conflicts between managers and shareholders when firms are financially constrained. My empirical analysis using a sample of 128 S&P 500 firms with noncontrolling interests in the fiscal year 2004 largely supports of my predictions. Reflecting the agency conflicts due to control, the likelihood of classifying noncontrolling interests in the 8 I consider the classification of noncontrolling interests an accounting choice which “is a decision whose primary purpose is to influence (either in form or substance) the output of the accounting system in a particular way” (Fields et al. 2001, page 256). liability section decreases by 17.9 percent in family firms.9 However, the probability of choosing the liability classification increases by l 1.3 percent when parent companies with “deep pockets” face potential litigation risk from the controlling relationship (proxied by the ratio of consolidated assets to the claims of noncontrolling shareholders). To mitigate manager-shareholder agency problems, I find that firms with higher information asymmetry (proxied by the extent of growth opportunities) are 10.9 percent more likely to choose the conservative approach of classifying noncontrolling interests as liabilities. I do not, however, find a relationship between financial constrain and the balance sheet classification of noncontrolling interests. Taken together, while my results on the role of accounting classification in manager-shareholder agency conflicts is consistent with prior research, I provide new evidence on how agency conflicts due to controlling ownership influence accounting choice. While my study focuses on a single balance-sheet accounting classification choice, I also provide evidence that firms adopting liability classification choice are more likely to provide higher-quality earnings. This suggests that firms use multiple accounting choices to satisfy the demand for high-quality financial information to address agency problems (Fields et a1. 2001). Considering the prominence of credit-based financing in the capital markets, the relevance of accounting leverage ratios for credit risk assessment (Shi 2003; Ashbaugh et al. 2006; J iang 2005; Lee 2005), and the importance of accounting conservatism to creditors (Ahmed et al. 2002; Beatty et al. 2006; Zhang 2006), I also investigate the 9 The marginal effect of continuous variables (dummy variables) is measured as the change in the probability of adopting liability classification for a change in the independent variable equal to its inter- quartile range (from zero to one) in the sample, holding all other independent variables at their means. potential economic significance of the classification choice from the credit market standpoint. While extant empirical studies have not resolved whether balance sheet classification choices of hybrid securities are relevant to investors,10 the balance-sheet classification of noncontrolling interests provides a unique setting for examining this question because it has no impact on net income, and therefore, the analysis is not affected by considerations other than the placement of information on the balance sheet. “ Extant literature argues that firms with conservative accounting practices are rewarded with lower cost of debt (Zhang 2006) and higher credit ratings (Ahmed et al. 2002). Accordingly, I hypothesize that firms would receive higher credit ratings when they choose to classify noncontrolling interests under liabilities because the classification choice is likely to signal firms’ desire to use balance sheet conservatism to limit agency conflicts. Based on the 107 of my sample firms with credit rating data, I find that on average, firms that classify noncontrolling interests as liabilities have credit ratings that are around one notch higher than those of firms adopting mezzanine section. The result is obtained after controlling for the various determinants of credit risk and the cndogeneity '0 Archival studies find that stock market treats securities with both liability and equity characteristics (e.g., mandatorily redeemable preferred stock) in a fashion that is different from the accounting treatment the securities receive in financial reports, suggesting that liability-equity classification has no discerning effect on investor valuation (Kimmel and Warfield 1995; Cheng et al. 2003). However, experimental research documents that liability-equity classification of hybrid securities matters to equity analysts, an important group of financial statement users (Hopkins I996). ” Lee at al. (2006) find that insurance companies that have used security gains and losses to manage earnings and those that have a reputation for poor financial reporting quality choose to provide information on comprehensive income in a statement of equity rather than a performance statement as recommended by the FASB. In the context of comprehensive income reporting, firms use different levels of reporting saliency as a response to differential costs of transparent disclosures. Given the proximity of the locations for the two classification choices, I do not expect disclosure saliency to be an important issue with respect to noncontrolling interests. However, one must be careful to note that this dissertation does not provide a direct test of the comparison between liability and mezzanine classification on the degree of saliency, thus it is possible that the observed association could be confounded by managers’ concern on and market reaction to the different saliency of different classification choices. of the classification choice (e.g., Fisher 1959; Ziebart and Reiter 1992; Sengupta 1998; Shi 2003; Ashbaugh et al. 2006; Klock et al. 2005). In summary, this dissertation also has implications for the FASB who recently curtailed the discretion in the accounting classification of noncontrolling interests. My credit rating analysis indicates that balance-sheet classification provides useful information for the market’s assessment of firms’ credit risk. This finding implies that limiting accounting choices could potentially eliminate relevant information from financial statements in situations where information asymmetry and agency problems may be prominent (Fields et al. 2001: p. 261). While eliminating accounting discretion is arguably driven by the F ASB’s recent focus on decision usefulness, my analysis suggests that stewardship benefits of accounting discretion may suffer. The remainder of the dissertation proceeds as follows. Sections 2 and 3 discuss various determinants of the balance-sheet classification of noncontrolling interests and the potential impact of the classification on credit markets. Section 4 describes the research design, and Section 5 discusses sample selection and descriptive statistics. Section 6 presents empirical findings and the results of robustness tests. Section 7 concludes. CHAPTER 2 DETERMINANTS OF THE DECISION TO CLASSIFY NONCONTROLLING INTERESTS AS LIABILITY According to Dye (2002 page 1125-1126), “financial reporting is, at its roots, a process of classification: firms are ‘going concerns’ or not; transactions are ‘recognized’ in a firm’s financial statements or not; leases are capital leases or operating leases; expenditures are assets or expenses; financial claims are liabilities or equity, etc.” The classification process underlying financial reporting has a real effect on resource allocation because it helps investors predict firms’ future cash flows, which in turn, affects their investment decisions (Dye 2002). Consistent with this notion, extensive research documents that liability-equity classification has significant consequences for various stakeholders such as buy—side financial analysts (Hopkins 1996) and managers (Engel et al. 1999).12 Moreover, accounting literature argues that accounting choices are affected by and can be used to alleviate various types of agency problems (see Fields et al. 2001 and Watts 2003 for a review). As depicted in Figure 1, there are two types of agency problems in my setting: 1) between controlling and noncontrolling shareholders, and 2) between managers and shareholders. Of the two types, I argue that the agency problem due to control dominates the classification of noncontrolling interests. In a majority- owned subsidiary, managerial decisions about the operating, investing, and financing '2 Hopkins (I996) provides experimental evidence that buy-side financial analysts price the finn’s mandatorily redeemable preferred stock as if the firm had issued debt (equity) when the stocks are classified as liability (equity). Engel et al. (1999) find that firms incur significant costs (i.e., 10 to 43 million) to manage their debt-to-assets ratio by an average of 13 percent through issuing trust preferred activities of subsidiaries and the ensuing economic benefits are often controlled by parent company shareholders. 13 The controlling shareholders often have closer ties to both the parent and subsidiary managers than do noncontrolling shareholders, making opportunistic alliances between managers and controlling shareholders more likely. My objective in this section is to identify how these types of agency problems will influence the noncontrolling interests classification. 2.1 Controlling-noncontrolIing shareholder agency problem The financial economics literature argues that market forces and/or legal statutes cannot completely eliminate the controlling-noncontrolling shareholder agency conflicts, leading to the expropriation of noncontrolling shareholders’ interests by controlling shareholders (Hazen 1977; Brudney and Clark 1981; Bebchuk 1989; Shleifer and Vishny 1997; La Porta et al. 2000).14 In the case of a majority-owned subsidiary, controlling shareholders gain if they cause subsidiary assets to be transferred to the parent corporation without adequate compensation to noncontrolling shareholders. Such wealth transfer mechanisms could include outright theft, dilution of noncontrolling interests by issuing additional shares to the controlling shareholders, and asset sales to parent firms at favorable prices (Shleifer and Vishny 1997; LaPorta et al. 2002). stock. While Hopkins ( I996) examines the behavioral implications of balance sheet classification, Engel et al. (1999) focus on transaction structuring by managers to achieve a specific accounting classification. '3 Officers and directors of the parent typically serve in the same capacity for the subsidiary. Schipper and Smith (1986) document that in 34 of 48 of their sample firms, the President or CEO of the subsidiary is also a parent manager. In l8 cases, one or more of the secretary, treasurer, and corporate counsel is the same person for parent and subsidiary. Also, 56 of 57 subsidiary boards have at least one member who is also a parent director or an officer in the parent firm. '4 Those studies articulate the agency problem between controlling and noncontrolling shareholders but not provide empirical evidence. Direct expropriations related to control are often discussed in the popular press. For instance, Mr. LeBow, who owned 60-percent of the Brooke Group, apparently benefited from favorable loans, and from selling assets to his other controlled interests at non-arms-length prices (Cohen 1993). The Wall Street Journal notes that “a look at Mr. LeBow’s Securities and Exchange Commission filings for the past three years shows how easily a dominant shareholder can tap into the cash stream of a company he controls.” Empirical research also provides evidence on expropriation by controlling shareholders. For example, Maury and Pajuste (2002) find that the dividend payout made to noncontrolling shareholders is lower than that to controlling shareholders. Moreover, . studies linking ownership structure and performance find a “roof-shaped” relationship, which suggests that as the controlling ownership increases beyond a certain level, controlling shareholders begin to retain more private benefits that are not shared with the noncontrolling shareholders (Stulz 1988; Wruck 1989; McConnell and Servaes 1990; Morck et al. 1998).15 In this section, I will investigate the effect of controlling- noncontrolling shareholder agency problem on firms’ classification choice of noncontrolling interests. '5 However, some studies provide competing evidence suggesting no exploitative behavior through controlling ownership of publicly-traded companies. In particular, Denis and Denis (1994) find that majority-owned firms are as profitable as diffusely-held firms. Holdemess and Sheehan (1988) find that majority-owned firms do not invest in overly risky capital projects nor do they provide excessive executive compensation. 2.1.1 Family firms Recent research finds that a nontrivial fraction of publicly-traded US. companies consists of family firms where founders or descendants control senior management positions (Anderson and Reeb 2003). Even among S&P 500 firms, roughly one third are family firms in which family members control approximately 19% of the shares outstanding (Anderson and Reeb 2003). However, extant literature argues that while family firms in the US. have less severe owner-manager agency problems, they suffer from more severe agency conflicts between family owners and other shareholders (Gilson and Gordon 2003; Fan and Wong 2003).’6 The level of financial reporting quality supplied by family firms may depend on the relative dominance of the manager-shareholder agency problem (the alignment effect) versus the agency problem due to family control (the entrenchment effect). The alignment effect predicts that family firms would provide higher-quality financial information due to: 1) better alignment of interests between shareholders and managers (or firm value); 2) longer investment horizon of family members (Stein 1998 and 1989; Chami 1999; James 1999; Anderson and Reeb 2003); and 3) incentives to preserve family reputation (Anderson et al. 2003). In contrast, the entrenchment effect implies that given founding family members have incentives and opportunity to extract wealth from other shareholders, family firms are more likely to manipulate their financial reports for their private benefits. '6 To my best knowledge, there is no empirical evidence documenting this type of agency problems in US. family firms. Wang (2006) argue that the results of Fan and Wang (2003) may not be generalized to the countries like US. with stronger investor protection. 10 In contrast to the competing theories regarding the effect of concentrated family ownership on the quality of financial reports, empirical studies find that family firms report better quality earnings (i.e., less positive discretionary accruals, greater earnings informativeness, higher earnings persistence, and higher earnings predictability) (Wang 2006; Ali et al. 2007), and are more likely to provide bad news warnings in management earnings forecast (Ali et al. 2007), suggesting the alleviation of the owner-manager agency problem. However, I argue that agency problems that arise due to any controlling ownership would be more serious in family firms when compared to other firms. Because of better alignment of interests between family members and managers, family-owned parent companies are more likely to behave opportunistically when dealing with noncontrolling shareholders of subsidiaries than parent companies that are not controlled by family members. Consistent with this argument, I predict that, when reporting noncontrolling interests in the balance sheet, family firms are more likely to resort to a classification that presents a favorable picture of its financial position from the perspective of stakeholders other than noncontrolling shareholders. Accordingly, I expect a negative relation between family membership and the decision to classify noncontrolling interests as liabilities. H]: Family firms are less likely to classify noncontrolling interests as liabilities. 2.1.2 Exposure to litigation by noncontrolling shareholders Expected litigation risk under the federal securities laws is thought to be an important determinant of public companies’ disclosure decisions (e.g., Francis et al. 1994; Skinner 1994 and 1997; See Healy and Palepu, 2001, for a review). In the setting I 11 examine, firms can face lawsuits from noncontrolling shareholders for any alleged expropriation by controlling shareholders. Both securities regulations and court decisions indicate a concern for protecting the rights of noncontrolling shareholders. For example, the Securities and Exchange Commission (SEC) requires an independent assessment of the “Fairness of the Transaction’ under Rule l3e-3 of Securities and Exchange Act of 1934 when a controlling owner acquires a noncontrolling interest.17 In a similar vein, court decisions on conflict of interests in parent-subsidiary transactions (e.g., assets sales) establish that noncontrolling interests have the right to require that benefits be distributed on a pro-rata basis. '8 The litigation risk faced by controlling shareholders is highlighted in a recent case involving Ford Motor Co. of Canada Ltd. in which a court awarded judgment to its noncontrolling shareholders as the original buyout price paid by the controlling shareholder, Ford US, was not deemed a fair consideration (Transfer Pricing Report 617, 03/ 1 7/2004). Another case in point is the claim brought by the noncontrolling shareholders in the management buyout of Quiznos, where a Denver Judge ruled that the original offer of $8.50 per share substantially underestimated the per-share fair value of at least $32.50.19 Extant research argues that firms with “deep pockets” are exposed to higher litigation risk (Francis et al. 1994, Roger and Stocken 2005). For instance, larger firms ’7 See http://www.sec.gov/about/forms/ruleI3e-3.pdf (accessed on 1 1/23/2006). '8 Johnson et al. (2000) argues that, in common law countries, the duty of loyalty principle, or fiduciary duty principle is specifically employed to deal with the situations of conflict of interests (e.g., controlling and noncontrolling interests). “In the case of fiduciary duty, the very fact that the interests of a director are in conflict with those of the company itself constitutes the basis for liability, and if the interests of the company are prejudiced as a result of such conflict, liability for breach of fiduciary duty arises...”(Shibuya 1972, p. 127). '9 see http://www.forbes.com/business/global/2004/030l/Ol8.html (accessed on 0l/28/2008) l2 tend to have better insurance that provides resources for plaintiff attorney fees as well as paying defense costs, making them an attractive target for lawsuits. The auditing literature also documents that auditors are more exposed to litigation risk when they work for larger clients (Carcello and Palmrose 1994). Consistent with these findings, 1 posit that when parent firms have “deep pockets,” they face a higher risk of litigation from noncontrolling shareholders. Recent studies suggest that firms with higher litigation costs are more likely to engage in economic behavior that would reduce the likelihood of class action lawsuits under securities laws. From a valuation standpoint, Lowry and Shu (2002) provide evidence that firms with higher litigation risk substantially underprice their IPOs to lower expected litigation costs. From a disclosure standpoint, studies show that managers preemptively disclose bad news to prevent large stock price declines on earnings announcement dates, and thereby, reduce the potential costs of shareholder lawsuits, (e.g., Skinner I994; Kasznik and Lev 1995).20 In my setting, this suggests that firms with large litigation exposure from noncontrolling interests are more likely to follow business practices that would be construed as respecting the rights of noncontrolling shareholders. Analogously, from an accounting perspective, I expect these firms will adopt the liability classification for noncontrolling interests as a signal of fair dealing and protection of the rights of noncontrolling shareholders. Therefore, my hypothesis is: 20 Previous studies (Skinner, I994; Kasznik and Lev, 1995) argue that early voluntary disclosure of large negative earnings surprise reduces expected legal costs in two ways. First, if the information is disclosed “voluntarily” prior to the mandated release date, it is difficult for the plaintiff, who does not know for sure when the manager first received the bad news, to argue that the manager withheld information. Second, disclosing early limits the period of nondisclosure, thereby reducing the damages that plaintiffs can claim. 13 H2: Firms with “deep pockets” are more likely to classify noncontrolling interests as liabilities. While prior research views “deep pockets” purely based on the total economic resources at a firm’s disposal, I view “deep pockets” in my context as the total economic resources controlled by parent firms relative to the potential claims by noncontrolling shareholders. I argue that parent firms become more attractive litigation targets when they are more capable of covering potential litigation claims from noncontrolling shareholders. 2.2 Manager-shareholder agency problem Because of the information asymmetry and the conflict of interest between managers and shareholders, managers have incentives and opportunities to expropriate wealth from shareholders, resulting in deadweight losses, and thus, a reduction in firm value (Berle and Means, 1932; Jensen and Meckling, 1976; Watts 2003a). Accounting research suggests that financial disclosures can be used to alleviate owner-manager agency problems, and the consequent reduction in firm value, through providing verifiable information that is useful in monitoring and evaluating managerial decisions (Watts and Zimmerman 1986; Bushman and Smith 2001). In this section, I will investigate the effect of owner-manager agency problem on firms’ classification choice of noncontrolling interests. 2.2.1 Alleviate information asymmetry problems in growth firms Previous studies argue that growth firms are more likely to have pronounced information asymmetry problems that are generated from the nature of the firm’s investment opportunity or growth Options (Smith and Watts 1992). The increases in 14 information asymmetry provide more opportunities to managers to manipulate financial reports to transfer wealth to themselves at the expense of uninformed shareholders (e.g., excessive compensation of managers). Such behavior reduces resources available for positive NPV projects, generating deadweight losses due to foregone projects. Therefore, capital markets tend to charge a higher cost of capital to growth firms to price protect for expected opportunistic behavior on the part of managers (Jensen and Meckling 1976; Watts and Zimmerman 1986). Extant literature argues that accounting conservatism can alleviate information asymmetry problem between managers and shareholders through: I) constraining opportunistic managerial behavior in overstating unverifiable net assets by requiring higher verification standards for gain recognition (Watts 2003); 2) limiting excessive compensation to managers that is costly to firms (Watts 2003); and 3) assisting in identifying negative NPV projects by timely recognition of economic losses, thus playing a monitoring role in firms’ investment policies (Ball 2001). Thus, to alleviate the adverse effects of information asymmetry on the cost of capital, managers of growth firms are more likely to adopt conservative accounting practices. Applied to my setting, growth firms are more likely to disclose noncontrolling interests as liabilities. Accordingly, I expect a positive relation between firms’ growth prospects and the probability of classifying noncontrolling interests as liabilities. H3: Growth firms are more likely to classify noncontrolling interests as liabilities. 2.2.2 Alleviate manager-shareholder agency problem in financially-constrained firms Financial economics literature points out that investors will charge a higher cost of capital to firms with uncontrolled agency problems as compensation for a higher 15 expropriation risk (e.g., F ama and Jensen 1983). Accordingly, firms in this category are unable to find financing for all of their profitable projects (i.e., face a “financially- constrained” optimization problem; Lamont et al. 2000), and therefore, are forced to make suboptimal real decisions, which are reflected in the firms’ stock prices (Sharpe 1964, Lintner 1965, and Merton 1973; Bemanke and Gertler 1989, Gertler and Gilchrist 1994; see Bemanke et al. 1996 for a review).2| Consistent with this argument, recent financial economics research documents that asset pricing factors that proxy for financial constrains are positively associated with equity returns (Campello and Chen 2005; Whited and Wu 2005). To the extent financial constraints are a reflection of uncontrolled agency problems, managers would take actions (including accounting choices) to resolve manager-shareholder agency conflicts. Consistent with my earlier discussions, I predict that firms facing financial constraints are more likely to classify noncontrolling interests as liability as part of a conservative financial reporting strategy (Ball 2001, Watts 2003 and 2006, Ahmed and Duellman 2007), which leads the following hypothesis: H4: Financially-constrained firms are more likely to classify noncontrolling interests as liabilities. 2' Following Lamont et al. (2001), financial constraints are frictions that prevent a firm from firnding all desired investments. This inability to fund investment might be due to various reasons such as credit constraints, inability to issue equity, dependence on bank loans, or illiquidity of assets. 1 do not use “financial constraints” to mean financial distress, although the two constructs are undoubtedly correlated. l6 CHAPTER 3 CREDIT MARKET EFFECTS OF BALANCE-SHEET CLASSIFICATION OF NONCONTROLLING INTERESTS Given the relevance of accounting leverage ratios for credit risk assessment (Shi 2003; Ashbaugh et al. 2006; Jiang 2005; Lee 2005), and the importance of accounting conservatism to creditors (Ahmed et al. 2002; Beatty et al. 2006; Zhang 2006), I consider the following research question in this section: does the balance-sheet classification of noncontrolling interests provide relevant information to the credit market? Most of the prior research tries to demonstrate whether a claim (e. g., unrecorded capital lease commitments or unfunded pension liabilities) is viewed by the market as debt- or equity- like (e.g., Bowman 1980; Dhaliwal 1986; Kimmel and Warfield 1995; Cheng et al. 2003) through examining how the claim affects the equity market’s assessments of the firm’s systematic risk.22 However, prior research has largely overlooked the role of accounting classification for credit market’s assessment of risk. I appeal to the credit market to shed light on the importance of accounting classification for the following two reasons. First, creditors are concerned about the firm’s downside risk, which is generally determined by the ability of firms to pay interest and principal (i.e., creditworthiness). From the creditors’ perspective, accounting leverage ratios are important indicators that help “broadly define a company’s position relative to rating categories” (2002 S&P Corporate Ratings Criteria page 53-55). Thus, creditors appear to consider accounting liability- equity classification in their assessment of credit risk. Secondly, the credit market plays a substantial role in satisfying the external financing needs of corporate America. For 17 example, in 2002 (2000), US. corporations raised more than $1.2 trillion in the bond market compared to $120 ($200) billion in the equity market (Beller 2003). Given the prominence of credit-based financing in the capital markets, the question whether the accounting classification of corporate claims is associated with the credit market’s risk assessment is important. Specifically, I investigate whether firms’ decision on the placement of noncontrolling interests on the balance sheet provides creditors with relevant information about firms’ desire to alleviate various agency conflicts.23 Prior studies argue that firms adopting conservative financial reporting have lower cost of debt, because conservatism helps creditors in terms of: 1) decreasing the downside risk from limiting inappropriate wealth transfer to equity holders (e.g., Ahmed et al. 2002); and 2) providing early signals about the deterioration of firms’ financial performance (Zhang 2006). For example, Ahmed et al. (2002) document that firms employ accounting conservatism to solve the bondholder-shareholder conflicts on dividend distribution, and firms using more conservative reporting have higher credit rating. Zhang (2006) also finds that accounting conservatism benefits firms in terms of lower initial interest rates. I extend the prior research on accruals-based conservatism to conservatism in balance sheet classification. In my setting, I view the classification of noncontrolling 22 Under the assumption of perfect markets. riskless debt and no taxes, several studies show that debt-like obligations affect systematic equity risk through financial leverage. 2’ For example, discussions with a senior executive of a major credit-rating agency suggest that the agency’s analysts do not consider noncontrolling interests as equity in their financial analysis if: I) there is put option available for noncontrolling shareholders; 2) if there is a plan to sell the subsidiary; or 3) if noncontrolling interests have a unilateral right to block distributions from subsidiary to parent firm. However, I argue that the classification choice of noncontrolling interests still matters to creditors in the 18 interests in the liabilities section as a conservative financial reporting choice as it leads to understatement of net assets. Consequently, I posit that creditors are more likely to favorably assess the credit risk of companies choosing the liability classification, as firms have signaled their intent to alleviate various agency problems by adopting a conservative financial reporting practice. Overall, I expect a positive relationship between firms’ liability classification choice of noncontrolling interests and creditors’ evaluation. H5: Firms’ decision to classify noncontrolling interests in the liability section of the balance sheet will be positively associated with the credit market’s assessment of firm risk. An alternative prediction is that the accounting liability-equity classification does not convey relevant information to the market. For example, Kimmel and Warfield (1995) and Cheng et al. (2003) find that investors view mandatorily-redeemable preferred stock as equity, and non-redeemable preferred stock as debt, indicating that securities are treated by the stock market in a fashion which is different from the accounting treatment the securities receive in financial reports.24 This implies that liability-equity classification has no effect on investors’ valuation of firms. Both of these papers conclude that liability- equity classification “may lack representational faithfulness.” sense that it reflects firms’ desire to use balance sheet conservatism to limit various agency conflicts rather than merely reflect the perceived liability/equity nature of the item. 2" Following ASR268, “Presentation in financial statements of ‘redeemable preferred stocks”, RPF D should not be classified as equity, suggesting that at least RPFD is regarded as non-equity. Later, SFAS 150, “Accounting for certain financial instruments with characteristics of both liabilities and equity”, requires that mandatorily redeemable preferred securities be classified as liability. 19 CHAPTER 4 RESEARCH DESIGN 4.1 Determinants of the decision to classify noncontrolling interests as liabilities I use a probit regression to model the probability of disclosing noncontrolling interests in the liability versus mezzanine section as a function of the firm characteristics predicted by the hypotheses in section 3 and various controls: Prob (LIABILITYi) = F660 + ,8] FAMIL YFIRMI- + fi2 NONCONTROLLING,- +fi3INV_GROWTHi +fl4C0NSTRAINi +fl5SIZEi (1) + fl6 VOLATILITYI- + [97Audit0rl- + 81-), where LIABILITYl- = one if firm i disclosed noncontrolling interests in the liability section and zero if in the mezzanine section, FAMILYFIRM i = one if firm i is a founding family firm as identified by Business Week, and zero otherwise, NONCONTROLLING,- = natural logarithm of the ratio of total assets to noncontrolling interests of firm i (proxy for litigation risk), IN V_ GROWTH i = ratio of book value to market value of common equity of firm i, C ONS TRAIN i = one if firm i had negative average free cash flows calculated over the preceding five years, and zero other otherwise,25 SIZE,- = natural logarithm of the beginning-of-year total assets of firm i, 25 Following Standard and Poor’s Corporate Ratings Criteria (2002 S&P Corporate Ratings Criteria), free cash flows are defined as cash flows from operating activities (Compustat #308) minus capital expenditures (Compustat #128). 20 VOLA TILITYi = firm i’s 2-digit SIC decile rank of the standard deviation of noncontrolling interests scaled by the standard deviation of total liabilities in the preceding five years, Auditor, = auditor dummies. Based on H1, I predict a negative coefficient on FAMIL YFIRM. I use NONC ONT ROLLING as a proxy for firms’ litigation risk from noncontrolling shareholders and predict a positive coefficient for it. I use the book-to-market ratio as a proxy for the inverse of firm growth prospects, and consistent with H3, expect a negative coefficient on IN V_GROWT H. Consistent with H4, I expect the coefficient of CONSTRAIN to be positive. I also include two control variables. Because perceptions of excess volatility may have negative implications for firm values, firms with noncontrolling interests that are significantly more volatile than the firms’ existing liabilities may be reluctant to disclose noncontrolling interests in the liability section. Therefore, I expect a negative coefficient for the VOLA TILITY variable, consistent with the notion that firms with more volatile noncontrolling interests than liabilities are more likely to disclose noncontrolling interests in the mezzanine section. I also include the beginning-of-year total assets as a control for the effect of firm size on firms’ classification choice. In addition, auditor dummies are included to control for the possible influence of auditors’ opinion on firms’ classification choice (Antle and Nalebuff 1991).26 26 Literatures argue that published financial statements should be read as “a joint statement from the auditor and manager” (Antle and Nalebuff I991), indicating that auditors have significant influence on firms’ financial reporting. 21 4.2 Credit market effects of the balance-sheet classification of noncontrolling interests I use credit rating to proxy for the market assessment of firms’ credit risk. Previous studies find that a firm’s credit rating is positively associated with the probability that its future cash flows will be sufficient to pay interest and principal in a timely manner. For example, Altman (1992) documents that more than 90 percent of defaulting issues are rated by S&P below investment grade (BBB) at least one year prior to default. Hand et al. (1992) and Liu et al. (1999) find that bond return is positively related to upgrades in bond rating. Together, these studies suggest that credit ratings are an appropriate proxy for firms’ credit risk. To test H5, I therefore regress credit rating on the balance sheet classification of noncontrolling interests (LIABILITY) and other control variables: RATINGi = yo + ylLIABILITY+ ngONTROLSi + 1),, (2) where RATING is the long-term issuer credit rating assigned by Standard & Poor’s six months after the fiscal year end. Note that the preceding discussion suggests that there are differences in characteristics between firms that choose to disclose noncontrolling interests in the liability section versus the mezzanine section. This suggests that LIABILITY is endogenous and ordinary least-squares (OLS) estimation of Eq. (2) will not provide consistent parameter estimates given LIABILITY is likely correlated with the error term t). A common approach to deal with cndogeneity is to use an instrumental variables estimator. Accordingly, I employ fitted probabilities from Eq. (1) as an instrument for LIABILITY (Wooldridge, 2002, pp. 623—625). 22 I allow a six-month lag between the close of the fiscal year and the production of credit ratings to ensure that rating agencies have sufficient time to incorporate the information disclosed in firms’ annual reports into their credit rating decisions. The ratings range from AAA (highest rating) to D (lowest rating — payment in default), reflecting S&P’s opinion of an issuer's overall creditworthiness. Following Ashbaugh et al. (2006), I recode the credit rating data (Compustat Quarter data SPDRC) to remove unassigned codes and collapse the ratings into seven categories, with a higher value of RATING indicating a better credit rating.27 Ordered probit and logit models are used to estimate Eq. (2) because I cannot assume a constant difference in credit risk between adjacent rating categories. Consistent with H5, I expect a positive coefficient for LIABILITY, suggesting that creditors assign high ratings to firms that classify noncontrolling interests as liabilities. Following prior literature (e.g., Fisher 1959; Ziebart and Reiter 1992; Sengupta 1998; Shi 2003; Ashbaugh et al. 2006), I identify the following control variables that are associated with credit ratings: FAMILYFIRM as defined above, which is expected to be positively correlated with credit ratings (Anderson et al., 2003), LEV = ratio of total debt to total assets. The higher this ratio, the higher the firm’s default risk and the lower will be its credit rating, SIZE = natural logarithm of beginning-of-the year total assets. Larger firms are expected to have lower default risk, and hence, higher rating, 27 Appendix A demonstrates the recoding methods employed by Ashbaugh et al. (2006) and Klock et al. (2005). As a sensitivity check, I also recode the credit ratings following Klock et al. (2005) to allow credit ratings to range from I (D ratings) to 22 (AAA ratings). Table C5 of Appendix C shows that the results are qualitatively similar to those presented in Table 4. 23 CAP_INT EN = property, plant and equipment scaled by total assets. Capital intensity is expected to be positively correlated with credit ratings, INT_COV = operating income before depreciation divided by interest expense. This ratio is expected to be positively associated with credit ratings, ROA = income before extraordinary items divided by total assets. ROA is expected to be positively correlated with credit ratings. 24 CHAPTER 5 SAMPLE AND UNIVARIATE ANALYSIS For my determinants analysis, my sample includes 128 S&P 500 firms during fiscal year 2004 with noncontrolling interests that have COMPUSTAT data necessary for my analysis.28 Firms’ balance-sheet classification choice of noncontrolling interests is manually collected from Form lO-K. Exhibit 1 (Exhibit 2) contains an illustrative example of a firm that discloses noncontrolling interest in the liability (mezzanine) section. Following the 2003 Business Week family firm classification, I identify 48 (37.5%) family firms in my sample.29 Of these 128 firms, 3] (24%) reported noncontrolling interests in the liability section, while 97 (76 %) reported noncontrolling interests in the mezzanine section. In my analysis on the association between credit rating and the classification decision, the sample includes 106 of the 128 firms with long-term domestic issuer credit rating data (COMPUSTAT Quarterly data SPDRC). Comparable to the full sample, I find that 26 of the 106 firms (25%) reported noncontrolling interests in the liability section while 80 (75%) reported noncontrolling interests in the mezzanine section. Panel A of Table 1 presents descriptive statistics on the two groups of firms as well as p-values from tests of whether the means and medians of the two groups differ. Although I use log-transformed values for total assets in my primary analysis, I report 2’ I do not model the decision of firms to have controlling interests in other companies. However, I did compare S&P 500 firms with and without noncontrolling interests in terms of various firm characteristics (i.e., FAMIL YFIRM, CONSTRAIN, IN V_GROW TH, SIZE, and ROA). Table CI of Appendix C shows that these two groups of firms are not significantly different from each other in terms of the firm characteristics I examine. 29 In their November 10, 2003 issue, Business Week identifies 177 (35.4%) S&P 500 firms as “family firms”. Business Week classification is based on Anderson and Reeb’s (2003) methodology. Given the 25 raw values in this table for ease of interpretation. In univariate comparisons of the two groups of firms, I find that FAMIL YFIRM, NONCONTROLLING, IN V_GROWT H, and total assets are significantly different. The evidence suggests that family firms are more likely to disclose noncontrolling interests in the mezzanine section, while firms with higher litigation exposure to noncontrolling shareholders, larger firms, and growth firms are more likely to disclose in the liability section. Panel B of Table 1 presents descriptive statistics on the mean and median credit ratings of the two groups of firms. Univariate tests provide some evidence that firms classifying noncontrolling interests in the liability section have higher credit ratings than firms classifying noncontrolling interests in the mezzanine section. stickiness in family ownership (Ali et al. 2007), I feel that family ownership in 2003 would be a good proxy for that in 2004. 26 CHAPTER 6 MULTIVARIATE RESULTS 6.1 Determinants of the reporting decision In Panel A of Table 2, I report the correlations among the variables in Eq. (1). Pearson (Spearman) correlations are in the upper right (lower left). Consistent with my univariate descriptive statistics, LIABILITY is negatively correlated with FAMIL YFIRM, and positively correlated with NONC ONT ROLLING and SIZE. The results of the probit, logit, and OLS estimations are in Table 3.30 The negative coefficient on FAMILYFIRM suggests that family firms are more likely to adopt mezzanine classification. This is consistent with the argument that founding family members are more likely to adopt classification that presents a favorable picture of its financial position from the perspective of stakeholders other than noncontrolling shareholders. The coefficient on NONCONTROLLING is positive and significant, suggesting that firms with more economic resources controlled by parent firms relative to potential claims by noncontrolling shareholders (i.e., “deep pocket”) are more likely to disclose in the liability section. This is consistent with my argument that such firms face higher litigation risk from noncontrolling shareholders, and thus are more likely to adopt the liability classification to signal fair dealing and protection of the rights of noncontrolling shareholders. The coefficient on IN V_GROWTH is negative and significant, demonstrating that growth (value) firms are more likely to include noncontrolling interests in the liability 3° 1 estimate Eq. ( I) in both raw values as well as ranked values for the independent variables. Table C2 in Appendix C shows that the results are qualitatively similar. 27 (mezzanine) section. This result supports H3 that growth firms are more likely to adopt a conservative accounting classification as part of a strategy to alleviate information asymmetry problems between managers and shareholders, and thus lower their cost of capital. The insignificant coefficient on CONSTRAIN does not support H4. With respect to control variables, the coefficient on SIZE is positive and significant, indicating that larger firms are likely to adOpt conservative accounting treatment. I do not find a significant coefficient on VOLA T ILITY . I use coefficient estimates to calculate the marginal effect of each independent variables on LIABILITY to evaluate its economic significance (Wooldridge 2002, pp. 458—9). The marginal effect of FAMIL YFIRM indicates that family firms are 17.9 percentage points less likely to classify noncontrolling interests as liabilities, holding all other variables at their mean values. In addition, the probability of liability classification increases by 11.3 percent and decreases by 10.3 percent when moving from the 25th percentile to the 75‘h percentile of NONCONTROLLING and IN V_GROW T H in the sample. Overall, the marginal effects of family control, litigation risk, and information risk from growth opportunities appear to be economically important. To sum up, the empirical evidence in this section supports the hypotheses that firms with greater exposure to litigation risk from noncontrolling shareholders and growth firms are more likely to classify noncontrolling interests as liability. However, family firms are more likely to adopt the mezzanine classification, reflecting more serious controlling-noncontrolling shareholder agency conflicts due to family control. 28 6.2 Credit market analysis of reporting decisions In this section, I examine whether the balance sheet classification of noncontrolling interests is associated with credit ratings after controlling for the endogeneity of firms’ classification choices and other factors affecting credit ratings. The empirical analyses in this subsection are based on the instrumental variable methods described earlier. In Panel B of Table 2, I report the correlations among the variables in Eq. (2). Pearson (Spearman) correlations are in the upper right (lower left). Consistent with H5 , LIABILITY is positively correlated with RATING. Table 4 presents formal tests of H5. The coefficient on LIABILITY is significant and positive, suggesting that firms classifying noncontrolling interests in the liability section receive a higher credit rating. This finding is consistent with my hypothesis that creditors are more likely to favorably assess the credit risk of companies that conservatively classify claims on their balance sheet. This is because conservatism in balance sheet classification may help alleviate downside risks through signaling restraint on the part of controlling shareholders as well as providing early signs of deterioration in firm performance. As expected, credit ratings are positively associated with firms’ size, ability to cover interest (IN T_COV), and profitability (ROA). I do not find FAMIL YFIRM, LEV, and CAP_IN TEN have significant impact on RATING in my sample. Table 4 shows the results obtained from running ordered probit and ordered logit regressions using fitted values from the probit and logit estimation, respectively, of Eq. (1) as an instrument. Two-sided p-values are calculated based on bootstrap standard errors using 1000 replications and are adjusted for heteroskedasticity (White 1980). By 29 assuming Eq.(l) and Eq.(2) are both linear, I also employ a ZSLS to estimate Eq. (2) as a sensitivity check. One common reservation about using linear probability models in bond rating studies is that the approach assumes the average credit rating difference between any two adjacent ratings is the same. Several studies that compare the performance of the linear probability model and the ordered probit model (e. g., Kaplan and Urwitz 1979; Noreen 198 8) find that the conceptually-superior ordered probit model performs no better than the simplistic linear probability model in terms of rating prediction. In addition, Angrist and Krueger (2001) argue that 2SLS can typically capture an average economic effect even for dummy endogenous regressors. Table 4 shows that the tenor of my results does not change across the different estimation methods.31 While previous literatures argue that liability-equity classification has no effect on stockholders’ valuation of firms (Kimmel and Warfield 1995; Cheng et al. 2003), I find that firms’ classification choice matters to creditors. I think this is because prior studies focus on whether the classification provides stock market with information about the liability/equity nature of a corporate claim. However, my study focuses on the value relevance of classification choice to creditors in the sense that it reflects firms’ desire to use balance sheet conservatism to limit various agency conflicts. 3 ' To ensure that my results are not sensitive to the choice of control variables included in the credit rating model, I also perform the estimation with no control variables in the credit rating model. Table C6 in Appendix C shows that the results are similar. 30 6.3 Additional analysis 6.3.1 Firm characteristics Consistent with prior research on conservatism (Givoly et al. 2007), I find that the balance-sheet placement of noncontrolling interests is highly sticky over time (see footnote 7).32 However, the measurement date of my independent variables so far is driven by the calendar time (i.e., 2004) that I chose to collect information on my dependent variable. To examine the sensitivity of my results to the measurement period of the independent variables, I re-measure NONC ON TROLLING, IN V_GROWT H and CONSTRAIN to proxy for firms’ litigation risk, growth prospectus, and financial constrain status in a longer time horizon. Specifically, NONC ON T ROLLING = natural logarithm of the average ratio of total assets to noncontrolling interests calculated over the preceding five years for firm i, IN V__GROWTH= one if firm i’s ratio of book value to market value of common equity calculated over the preceding five years is lower than industry mean, and zero other otherwise. CONSTRAIN = one if firm i’s average free cash flows calculated over the preceding five years is lower than industry mean, and zero other otherwise, Table C3 of Appendix C shows that the tenor of my results is unaffected, indicating that firms’ such consistent classification choice is not sensitive to the measurement period of the independent variables. More important, this suggests that the determinants I investigate represent certain stable firm characteristics that affect firms’ 31 “sticky” classification choice. Such results are also consistent with the argument that the degree of firms’ conservatism shouldn’t vary significantly from period to period because it is determined by firm characteristics that are fairly stable over time (Givoly et al. 2007) 6.3.2 Earnings quality and liability classification choice To the best of my knowledge, this is the first paper documenting that firms employ liability classification of hybrid securities as part of their conservative financial reporting strategies. Fields et al. (2001) argue that management will adopt a multi- dimensional financial reporting strategy for the firm by jointly choosing a set of accounting choices (Hagerman and Zmijewski 1979; Zmijewski and Hagerman 1981).33 In other words, given a single accounting choice is just one element of the optimal set of financial reporting choices, I expect different accounting choices to be highly correlated. In this section, I investigate relationship between firms’ classification choice of noncontrolling interests and earnings quality.34 I predict that firms adopting liability classification are more likely to provide higher-quality earnings. Given liability classification of noncontrolling interests is considered as one of firms’ conservative financial choices, firms’ doing so may reflect a 32 Givoly et al. (2007) argues that the firm’s degree of reporting conservatism is stable because it is mostly affected by firm characteristics that would not fluctuate significant from period to period. 33 Most of the extant literatures focus on one accounting choice at a time (Fields et al. 2001). Exceptions are Hagerman and Zmijewski (1979), and Zmijewski and Hagerman (1981). Both papers simultaneously consider four specific accounting choices (LIFO vs. FIFO, straight-line vs. accelerated depreciation, the amortization period of past service pension costs, and flow-through vs. deferral method for the investment tax credit) and analyze underlying managers’ incentives. However, they don’t provide any evidence on the relationship among those accounting choices and only focus on income-statement accounting choices. 3" My measure of earnings quality is based on firms’ discretionary accrual which is considered to be a summary measure of timing differences of all accounting choices (Watts and Zimmerman, 1990, p. I38). Thus, I am investigating the relationship between firms’ balance-sheet classification choice and the accounting choices affecting income statement. 32 demand for higher-quality accounting information as a whole for alleviating various types of agency problem concerns. Consistent with this notion, empirical studies document that different attributes of firms’ financial accounting information are highly correlated. For example, Francis et al. (2004) investigates seven attributes of earnings (i.e., accrual quality, persistence, predictability, smoothness, value relevance, timeliness, and conservatism) and find that correlations among those attributes are positive and significant.3'5 To examine the association between firms’ classification choice of noncontrolling interests and earnings quality, I consider the two measures of earnings quality widely used in the previous literature. Following Ashbaugh et al. (2006) and Kothari et al. (2005), I alternatively measure Earnings Quality as the standard deviation of finn- specific performance-matched discretionary accruals from the prior five years, where discretionary accruals are calculated from a cross-sectional estimation of Dechow and Dichev’s (2002) model modified by McNichols (2002), or from a cross-sectional estimation of the Dechow et al. (1995) model. I estimate the regression model using the population of COMPUSTAT firms at the 2-digit SIC code level for each year. Table 5 presents descriptive statistics on the two groups of firms as well as p- values from tests of whether the means and medians between them differ. In the univariate comparisons, I find that Earnings Quality is significantly different, suggesting that firms adopting liability classification have earnings of higher quality. This indicates 35 In stead of examining specific accounting choices, Francis et al. (2004) focus on summary measures of different attributes of earnings. However, Fields et al. (2001) argue that both accounting choices and earning management (or earnings quality) are used to influence the outcome of accounting information system to achieve certain goals (e.g., alleviate agency problems). 33 that firms use multiple accounting choices to satisfy the demand for hi gher—quality financial information to address agency problems (Fields et al. 2001). 34 CHAPTER 7 CONCLUSION In this dissertation, I examine firms’ balance-sheet classification decision of noncontrolling interests, and the economic significance of the classification choice from the credit market standpoint. I find that firms with greater exposure to litigation risk from noncontrolling shareholders (i.e., firms with “deep pockets”) and firms with more serious information asymmetry problem (proxied by the extent of growth opportunities) are more likely to classify noncontrolling interests as liabilities. However, consistent with more serious agency conflicts due to controlling ownership, I find that family firms are more likely to adopt the mezzanine classification. Contrary to expectations, my analyses did not find that financially—constrained firms are more likely to adopt liability classification. Moreover, I demonstrate that firms adopting liability classification choice have higher- quality earnings, suggesting that it is employed together with other accounting choices to satisfy the demand for higher-quality financial information. After controlling for the endogeneity of the classification choice and various determinants of credit risk, I find that on average, firms that classify noncontrolling interests as liabilities have credit ratings that are around one notch higher than those of firms that include them in the mezzanine section. This is consistent with my hypothesis that creditors are more likely to favorably assess the credit risk of companies adopting conservative financial reporting. My results should be of interest to both standard setters and academia. First, my study addresses the current debate about the classification of noncontrolling interests. To my best knowledge, this is the first paper providing consistent evidence on the 35 determinants and informativeness of firms’ balance sheet classification choices if noncontrolling interests. Second, my dissertation provides evidence for the debate over whether the balance-sheet placement of information is relevant to investors. My credit rating analysis indicates that accounting balance-sheet classification provides information about the market’s assessment of firms’ credit risk. Thus, limiting accounting choices could potentially eliminate relevant information from financial statements in situations in which significant information asymmetry and agency costs may be present. Third, this study contributes to the accounting choice literature. Fields et al. (2001) note that some methodological limitations in previous work have hampered progress, for example, failing to correct for endogeneity problems and failure to model multiple incentives driving a reporting choice. My analyses and results provide support for their assertion that resolving these issues can yield value to researchers studying firms’ accounting choices and the consequences of those choices. 36 o o o o O 4 . I EoEommSNE o e o o o o o o - EoEowmcd—z «xmiommfizohaoox o o o o o o o o o o o NoLEoO 8th SHE Evan " Ea shaman . p n O O .5395 Duane. n hoe—eaouaamfiowaaa—Z o O O O O .5395 Gnome. " mam:oh:3:oZ-u=:-ob=o0 meoEonoumnm 358m 0 whogonoumzm .. fa. .. .. . , ,. . o o o o o o O . _ . I . .. Gaming—Em, w£=obzoOV. .wE obzoocoz. . _ ,. 3: _ ._ : meow—.onosamwhgmfiosm 8038a xoaowm 822085 we:Bungaocmfizobcoo ES :8an Gnome uoEononefiéomeSwE “mEoan twosome mo 39$ 03% aeaa 37 $3.8 03.0 38.8 R _ .o Owe. 83 x92 coxoozav “we: 0388-93 Sm o:_m>-m BEES/H 83: :52 80.: Re: 83 :3 83. 8:. 382 as: 5E: as: Aownzv a Gmuzvii i i i RENEE: TEENS: Goom are x85: «8E DEE seem a s @333 ems-3:. BEE moEmE> $8.6 ooHnZv 385:: mam—35:85: How @2050 coceoEmmEo $085 voom So» Roma 5. mafia com gym 2: mo moumuflm oZEtomoQ ”GEES mcsfi H683 m 3.8m 98.8 58 mood 89¢ 53 3.3 ES: 39“ 93.8 2:... an: $98 08.2 8&8 25:25 .aoa< $2.8 Ed c Sec 0 mad EwEmwSo 82.8 ms... 82 0:3 22 $3 SEES E 92.8 5:. mm? 33. 83. 5% ezedofizoozoz 55.8 E; o 32 o 63 55: SEE 9mg gm an GOXOOZBV CE o 50 Cd— 0 ~80 mm Nam 58; 2983625 pom 2.37% $66-9: b 2 E b 2 2 3 . > I I - Imonzv I, IAEWEI I enigma: 33”:qu AWE wmfiuzv $332: man—82828: 8m 853:5 bafifiam — finch. 86:0 cosmofimmflo mmouoa voom 80% Roam 3 menu oom mfim wfi mo moumufim 0351809 ”638% aommmoou coteofimmflov < 353 38 .Awmi 2339808 3:56:35 1338 SEE Awoma Hmwmzafioov 3:333 message Eofi $5: 5.8 3 amazon 0.8 $5: 58 BE .Amcofiu mwccmm 888.50 macm moomv £5ch mwcsmm 38380 {com 28 Banzai wEBozom a“ .Amoomv .3 Ho xoog mo 225:8 565250 2: no 3me 022.43” 8 BEE/coo mm Omamm .AOMQmma Samsafioov can 89» 30mm 05 Stu 25:08 xmm mmoom ow pudendum 3 wocwmmmm worse :35 53mm atofiwaé H «oSMIDZEEN .688 a s SEER .6 2328 sausage as 8 Ban BEE 3 causes a 95% .Gmfima 833808 use :8» Roma 2: Ste 262: 5m mzoom a. nausea 3 cocmmmmm marsh :38 53mm EoYmGE u awzeeiwIOZEE .38» 26 @6805 05 E 83:52 :39 mo com§>ow Havana 05 E 838. memouowfi wfizobcooco: mo count/2o 3.858% 05 mo x52 26% Dam SEEN mm SE "ESE VINO; ..~ SE Sm .338 38 83¢?waafiwon 06. mo :3an2 Rea-m: "MNNW 4 SE no.“ 333 .8888 .«o 0:?» 98.38 9 029» xoon mo 0:8 ”ESQfivIE/Q 3.339850 5:8 83 use £80» 02% wcmwoooa 05 ~26 wean—3:8 950$ :98 RE owfiocé 33me me: .~ 8.6 b 28 HEEEWZQU .N SE Sm Ewen—BE mam—35:85: 9 Somme :33 we 2:2 05 mo 85:..wa ESE: HOZNNNQKNZQUZQZ .ommEofio 83 use .939: ummfiwzm 3 REESE mm SE 333% wcmcgom a mm .N SE a use ”EEK; SVEQ .couoom vague 05 E .2 23 can couoom bane: 2: E 385:: mes—868:8 tome—“En .~ 85 a use “\ENNEV: 355.52. 033:; ~ 935—. 39 $8.8 $88 88.8 288$ :88 88.8 83 88 28.8- mod- 82.. ES- 88.8 83:33 £88 £88 888 92.88 3:88 88.8 m :8- 83 m2... 2:. mm. .o- a _ .8. 8:. 28% 63.8 288$ 28.3 68.8 88.8 82.8 088- a8... 88. _ 8a... 8:. 83- 38. 2302281822 388 38.8 2888 $8.8 $8.8 $2.8 M28- 888 82 83 8 8 38.8- 88- 282828 288$ $8.8 $8.8 $8.8 $8.8 88.8 82.. N88- :2 88 83 88.8 m8... 02382228202 9:8 €88 $8.8 $8.8 $8.8 €88 N88- 88. 82.- «8.8. 88.8 88. _ 82.. 3.2223382 88.8 :8... 88.8 9.28 68.8 98.8 :88 8:. 88.8- 8 8- m8... 88- 83 22283 2.2333 2N8 5885 822 2828209 958328202 5833i 3883 a $56 «NTZV mozafia> 8098885 Ea 802308 05 mo anE coca—880 A2958 586$ 838$mean < Edam 5.5a: noun—«tau N oEaH 40 $8.8 $8.8 83.8 $38 23.3 32.- 32.- 83.. Sod 8m... $62 $8.3 $2.8 $3.8 $2.8 285$ u 8.....- 5.? mod :3 83. \59 ES $8.8 $8.8 $3.8 $2.8 E88 I 22. 85.8 E; $3- 83. 55: $0 :8.ch $8.8 $38 82.8 $223 2m... 2 3 83 $2.- $2.. :3 9?.8 $8.3 $8.8 638 3:8 ego- $2- 85¢. :38 ”8.8 ESESE 2: _.8 EH _ .8 $8.8 $3.8 83.8 82 £3- £3- E; 8:. mma. 38:3 $8.8 $3.8 $2.8 :8.ch I mm»... SN... 92...- 33. En..- FESG ES $5.8 $8.8 $8.8 28:3 83 SE SE. 8.2.. $3- 335.29g 35.8 $3.8 $8.8 $2.8 ~25 83 35¢ 83 ~28- 82335209202 $58 $2.8 $8.8 638 m3? 838 83 SS- 83- ESE SEE $2.8 $8.8 $2.8 $2.88 32.. 8.3 $2.. 83 3:. Egg: 38:2 $3.8 $8.8 $5.8 En..- $3- 8 S- E... 83 fizufiawawéwz 2~§5w2©9 UZSNQKNZQUZQZ 353m EEVR \ENSMSQ «wzcfiww DZNE a $85 oSHZV 835E? 8095385 98 #89698 05 80 x59: noun—8.80 A0383 mass“ “688 m 35$ $888 N 28? 41 28:3 238 2......3 88.8 28.8 28:3 8.... at... .8... mm...- 8...... SN...- amv...- .8... 28:3 888 28....v. 23.8 88.8 28:3 E... 8.... 8.....- m:...- 2...... $2.- 83...- Agni-.2. $3.8 $38 2.3.... $8.8 88.8 238 u 8...... 8.....- 8.... 8:... 5.... $8... 8...... 2882. m5 $8.8 28:3 82.8 $3.8 28.8 23.8 2.2.- 8.2.- 23 8.... .8...- owo... OS... 83 $8.8 $3.8 $8.8 .838 23.8 23.8 8...... .8... E... 8.....- 8.... 3......- E....- 8-8-5308 2. .8 88.8 23.8 5...... 22.8 83.8 8:.- ~.~...- 2...... E... 5.....- 8.... «fl... 8N8 2.....3 28.8 at... 28.8 2.3.... 88.8 I 8.....- ma...- .8... 08:. Q8...- :~... 8.... 583%. ES $....8 23.8 88.... 28:3 23.8 2:8 28:3 88...- ..n~.... «.3... 22. NB...- mm; 3.... 2.883208 2.....3 $3.8 8%.... $2.8 28:3 88.8 28.8 «3.... 2 2.- 88:.- ..m... 8.2.- ..m....- 2:... 823388209262 2:8 :38 2%.... .58 $88 288 2.8.8 3.....- ....:. m3... m8...- ow.....- 2:.- mb....- 88:33:... 22.8 228 2.8.8 2. .8 $8.8 $3.8 2....8 8;... m8... 8:...- mm....- 2.... N3... 8.....- .3338: 2......3 2......3 2.88 2......3 $8.8 $8.8 28:3 3.... N3... 8......- Sm...- 8.....- m2... 83....- Eaéfiwnbéwx .68 \BD .82. 252. $6 83 ES: 59. MN;- 58.8.6 82. $.88. N 2...; 42 .. 0.8.08 8. 00800.. 0.0 003080., 808.0 ..< .6... 80.088800. 808.0 .80. .3 000...... Q E. 888800. 080.. 808.800....0 0.0.0.. 08008. H 80% Am... .80888000 080880 80.0.8. .3 .0003... Am 3... 88.88000 82.800800 0.0.0.. 08008. 880.080 H AQ0IE<~ .3... 80-88800. 8080 .80. .2. 00.08 S... 88.8800. 80888.00 .80 .80... £80808. .0 Z M 2m 8 2.1880 .6... 8088800. 80000 .80. 0. .00.. 888800 + 9.. 888800. 500 .80. -.0 0.8.8 M 808 "080.850.. 0.8.0.80.» 80000 .80. 8008-08.00-w8.88.w0n .0 8:...0w0. .8308 0.... 0. 08m A0280»... .0>0. o . .o 0... .0 800....8w.m -.. 00.0.00 0. 83.0.0800 0088.80.08 8. 0.0 38.0.78 000.003. 8.3 880.0508 82.0.0800 8.80-0.80. 808.00% 80. .030. 0... 080 82.0.0800 80808808008. 80800.. 8.080. Em... .0888 08-.- 0 90.800. N 030,—. 43 3.3. 0.8.3. 0...0..m0000.wo.0.00 .0. 00.08.00 8080 08.0880 80 00000 00.0.8200 0.0 88.9-.. 000.0030 0 .8...0> 8008 ..0... .0 8.00..0> 80080800... .080 :0 0.8.0.0.. 0.8800 08 8. .080 0. 0.00 80... 0.00..0> 08 .0 0.880808 8mm 0... 0. 0.8.00.2. 8mm 08 80.. 0.00..0> 80080800... 08 8. 08000 0 .0. 800000.880 30.00.. 0.8.8.000 .0 b...0000... 08 8. 08000 08 00 008.808 0. 90.0088. >888... 8.00..0> 8088.800 .0 800.0 .08....08. 00F 0 .. 01.0.; 8. 008.00 0.0 8.00..0> ..< a 0mm. .0 0.. 00m... $8.... NM 00800.. mam... NS..- vmmd 03.0 3.0... mam... 03... N00... - 8.3.5.308 N3... $0... .30... 000... «um... M00... .3... N0 . .0 + MNE R . .o .30- mom... own..- mum. .- 0mm... 3.0- M00... + 2.5.0.0200 03... 0mm..- 8.... 2.....- mbN.N- .3... 2.....- 0mm..- - 08.80.5182. 02.... 5...... 8...... m3... mum... 02.... 0...... m5... + 023008.200202 as... 8.....- .00... me... «an..- 03... 05...- 000...- - 3.5.... 0.... S. 0020?... 80.0.0000 U08.0?8. .MMVMMM“. 80.0....000 008.0?8. .MMMHWA 8060.000 80 :03 .38. 0080.). 80808.80”. 4 89m 8.0080 > . . _ 00.0.08. 0 . as... 8."... 080.800 30.00.. 8. 0.8.0.8. 8808800808 08.8.0 0. 808.000 08 .0 8808.88.00 0... .0 0.08.08... .9. ..0m. .0. 8.88. 80.0883. m 0.00.. 44 .§§ .353: bsuméemeefi 8: :83 Em 08 ES 326232 8:: mafia 28.5 @3983. 92580: no :83 3818—8 08 $33-: 3:663: o .38: 58 2. 3 2a 3 E 3833 B s: 2 3% AN: .:m :0 message sz 35:8 2.: E 52%waqu 8.: 6083ng 5 mm com: mm C: .:m :0 coumfismo :wE 80¢ 829’ 32.: AN: .3. :0 559550 :38: 3on 05 E \EEEV: .8: :8ng ca mm. :8: mm 2: .cm :0 552580 :38: EB: 32.? BEL .N 29¢ Ea ~ 2.5 a 8&2. as 85%? =< 45 2:6 mm :mmm: 33: -~ 353.— vm:.: Gmé vm:.: 3.92 :m:.: 23.5 + ".62 b::.: 3:: 9:: :m:.: «3.: 3:: + kQUIKZN m2: wow: .3:: m::.: wmw: Q ~.: + ZED/filmy»: 3:: «MN: m::.: mg: N::.: :3: + MN? wmm: 0:2- Sm: :mmN- cam: m3.7 - Am: 2m: mmm: NS: 2:: :2: Eq.: + EEKNNNEVK mm:.: :NNA ~::.: ~ 3.: ~::.: and + xhflm 3% o “SEEN 3205000 cos—gé. E20580 ”029$ E20580 c mqmm 0 two: :08qu a :5on @2030 8502 Sumecmm =me . m93.2.5 :33:on $8.6 8T2: .3253 mm 336:: wfizobcooao: wfimofimmv :0 Soto gov—.88 :38 Mo 53850 28%? 3583me Am: .Um go: 3152 :Bmmouwom v «Bah. .325 58 an S 55 E .55 55:5me 3 a: 5 58mm .652 338 865538532 8: 35:85 58 :8 58:82:58 83 :85: 88.5 :8355 @8580: 8 355: 3535—55 58 5525?: 325-03.: 5 46 AN: .:m .8 8:58:55 awn: 35:8 55 8 KEN‘NEV: 8: 858858 :5 mm 35.: mm 2: :m .8 8:58:55 2:2 88: 5589/ 38: N Am: 6m .8 8:58:55 :88 35:8 56 8 8;:me 8: :58:me :5 mm :55: mm C: .:m .8 8:58:55 :88 80.: 553? 3:5 2 .m 535:. ::5 ~ 535:. 8 38.5: 55 55358; =< a m? _ .o N: :mmm: 33: NM 8:35.: 3:: Gm: 3:: $.82 :m:.: :3: + VQK 3:: 3:: mg: :m:: 2:: as: + :9: ES m2: vow: «mo: 8:: wmw: Q g: + ZMRZNIKVU 2:: «MN: m::.: 3:: N::.: :mm: + MNE. wmm: :3. T Sm: :mmd- :3: ~62- - Am: 2m: mmm: NC: 2:: :2: :3: + EEKxSEVK mm:.: :34 3:: 3m: 5:: mg: + x5335: 5 5235‘ 8555500 55238. 855580 55:_5>-m 8555500 5 mqmm 5 two: 3580 .158: 35:8: 3.3.2 838:5 $6 a 553583, 3:335 588.: :oTZ: 523:: 55 $5558: was—0885.8: 85258: .8 55:3 5538 :38 .8 8:58:55 5388; 5:58:88: Am: :m 8.: 52:55: 855585”: m 53.; APPENDICES Appendix A: Background Pre—F AS 160 GAAP FAS 160 requires that noncontrolling interests in subsidiaries be presented in the consolidated balance sheet within the equity section, but separate from the parent shareholders’ equity. As stated in F AS 160, the objective is to “improve the comparability by eliminating that diversity” of noncontrolling interests classification in the practice. Prior to F AS 160, there is little official guidance on how to classify noncontrolling interests: “ARB No. 51, Consolidated Financial Statements, and F ASB Statement No. 94... are the prevailing authoritative GAAP rules on accounting and reporting standards for consolidated financial statements.” (FASB, 1991, par. 14). However, neither pronouncement offers a definition of noncontrolling interests nor prescribes how to classify noncontrolling interest in financial statements. In addition to those accounting guidelines, the SEC and FASB have promulgated standards (i.e., ASR 268 and FAS 150) that shed some light on the disclosure of mandatorily redeemable noncontrolling interests. In the pre-F AS 1 50 period, noncontrolling interests that are subject to mandatory redemption or whose redemption is outside the control of the issuer were required to be disclosed outside of permanent equity. In the post-FAS 150 period, noncontrolling interests that “embody an unconditional obligation requiring the issuer’s redemption by transfer of assets at a specified or determinable date (or dates) or upon an event that is certain to occur” are required to be disclosed in the liability section (F AS 150). However, 47 noncontrolling interests that are outside the purview of FAS 1 50 could be classified in the liability, mezzanine or equity section. Debate on the balance sheet classification of noncontrolling interests The FASB received forty-nine comment letters in response to the 2005 ED. The majority of respondents favored the parent company approach to financial reporting over the economic unit approach, and therefore, did not support the classification of noncontrolling interests as equity. Specifically, they argued that since noncontrolling interests do not have an ownership interest in the equity of the parent company, they should not be classified as equity in the consolidated balance sheet. They also argued that classifying noncontrolling interests in the consolidated equity section would affect the calculation of financial and performance ratios, which could confuse users. In contrast, the few respondents who agreed with the equity-classification of noncontrolling interests (i.e., academics, accounting societies, certain preparers, and one accounting firm) supported the economic unit concept as more appropriate than the parent company concept. They pointed out that there is no support in the conceptual framework for classifying noncontrolling interests as liabilities or mezzanine items and that the holders of noncontrolling interests are owners of a residual interest in a component of the consolidated entity. Theories underlying classification of noncontrolling interests The lack of agreement on a theory of consolidation and a clear definition of liabilities and owners’ equity has led to variations in the reporting of noncontrolling interests. The debate as to whether the appropriate basis of accounting should rely upon 48 the economic entity concept or the parent company concept has been flourishing for a long time.” 38 At one end of the spectrum, the economic entity theory views the subsidiary as part of the consolidated firm, and given that the investors representing the noncontrolling interests are equity holders of the subsidiary, they are also deemed equity holders of the consolidated firm (Paton 1922; Moonitz 1942 and 1960; Sapienza 1960; Hendriksen 1970). At the other end, the parent company theory evolved from the representative viewpoint proposed by Husband (1938 and 1954), who argues that for the consolidated entity, corporate entrepreneurs rather than the holders of noncontrolling interests are the parent company’s equity claimants (Kohler 1938; AACS 1955). Hence, a noncontrolling interest is an outside interest and should not be reported as an element of stockholders’ equity. Proponents have used this theory to argue that noncontrolling interest is a liability, or should be presented in its own special category separate from the parent’s equity. Standard setters’ views on the adoption of the economic entity concept versus the parent company concept have been mixed. ARB 43 (1953) provides support for the economic entity theory. Chapter 7 of ARB 43 states that “the income of the corporat[ion] 37 The earliest reference is a presentation made by William M. Lybrand at the annual meeting of the American Association of Public Accountants in October 1908 which was published in two parts in the Journal of Accountancy in November 1908 and December 1908. Lybrand depicts “common Stock of Subsidiary Companies Not Owned by the Holding Corp.” under a general heading of “Liabilities,” following “Common Stock of the Holding Corp.” [November 1908, p. 40]. In part 11, Lybrand states that “under capital stocks will be included the stock issues of the holding company and separately stated, such part of the stocks of the subsidiary companies as are not owned by the holding company” [December 1908, . 120]. E38 Differences of opinion were evident from the start. Newlowe (1948) examined 150 journal articles and books from 1908 through 1945. Four authors preferred that noncontrolling interest be placed among liabilities, and 28 preferred to classify noncontrolling interest as an element of stockholders’ equity. The other 34 sources cited did not address the nature of noncontrolling interests. 49 is determined as that of a separate entity without regard to the equity of the respective shareholders in such income.” (Section B, par. 6). However, ARB 51 states that “. . .the purpose of consolidated statements is to present, primarily for the benefit of the shareholders and creditors of the parent company, the results of operations and the financial position of a parent company and its subsidiaries essentially as if the group were a single company with one or more branches or divisions.” Even though ARB51 does not discuss where to place noncontrolling interests on the balance sheet, the above statement provides support for the parent company theory of equity. While the FASB’s stated preference in the F AS 160 for classification under the equity section reflects its leaning toward the economic entity theory, it also proposes that EPS calculations be based on parent firm information, consistent with the parent company theory. Furthermore, the F ASB’s Statement of Financial Accounting Concepts No. 6 does not give clear guidance on the nature of noncontrolling interests. In sum, there is no clear theoretical consensus that has emerged on the balance sheet classification of noncontrolling interests. 50 Appendix B: Credit ratings’ numerical conversions S&P Letter Compustat Quarter Conversion Conversion Rating Data SPDRC Number (Ashbaugh Number et al. 2006) (Klock et a1. 2005) AAA 2 7 22 AA+ 4 6 21 AA 5 6 20 AA- 6 6 19 A+ 7 5 18 A 8 5 17 A- 9 5 16 BBB+ 10 4 15 BBB 1 1 4 14 BBB- 12 4 13 BB+ 13 3 12 BB 14 3 11 BB- 15 3 10 B+ 16 2 9 B 17 2 8 B- 18 2 7 CCC+ 19 1 6 CCC 20 l 5 CCC- 21 1 4 CC 23 1 3 C 24 1 2 D 27 1 1 51 ._ 29$ 5 Eamon 2“ 85%? E. _. smug €75 32.8 83 33 83 mod ~86 wow 8?qu ca TzC Game mm; 485 a? 83.0 $3 mma samué a TB 88.8 32 c on; o 88 >23:sz Ammmnzv @Sué I 938 8.8 $2 $8 335 82c >355 Es Gmmuzv A3 2qu its $3 0 83 o 93 SKEESE 93“ 8% VHS canoe—$6 H83 0383 5602 :82 SE52 =32 «mo—nmgxw -95 com o=_m>-m quméBH 385:: wadzobcoocoq 32E? 335:: wcfiobcoonoc 5:5 $852: wanbaooaoq 8m 88:0 consommmmfio mmobm voom so» Roma E wee 8m mfim mo mosmcfim ozatomom ~ .0 035—. mania—5 bgmmuom "U fleuoand‘ 52 .Swa 823v gosmmwoxmotgo: com tom: .8“ Echo wawcfim no woman Bum—:38 8m moigd 32993. n ._ mam/2. E pounce 2a 8323:? a 8%... N: £3... 38... N: 355 med de 30o vwod ovfio owed - \ENNEEQ; wmmd mmod homd 3N6 ode Efio + MNE. :86 Sad- om fl .o N _ h. T cmcd mead- + ZNSCWZIQU v2.0 Rod- mad mmfid- wmfio wood- - IEQMU ES mac... baa... $9: $5.: mweé baa... + UZ~um HGOMOWMDOU PMS—grim “COMOWMOOU map—grim «$205000 30 z. 3 505 $6 Bea: Smaacmm 328$ smsfifi> anew wm Tzv 228% base: E 385:: was—838:0: 328% 8 commmooc 2t .3 mgqmcmcbfiov 65 mo £9392 A: .cm How 3168 ~868ch 018% N .U 03.3. 53 Aowo. 8.23. 5638:3858: Ho. 865.? £88 28ch co comm: 3.2523 08 $39»-.. 3.66-03... n .. 2%... a. 88:0: 2a mozmflg .550 =< 3.3550 $50 83 can .538 @535. SE. 832 m. memo» 02.. winoooa 2.. .26 38.3.3 v.30... fine 02.. owfiog m... at: e: 28 H ZNEEMZQD 6.0.3328 88o Sow can .538 .5265 SE. 532 m. may. 02.. wfiwoooa on. 85 38.523 3.3.6 .8888 .o 02$ 86.38 3 033 xoo: mo 2.8 m... 8.5 .3 one HEEQNmuIAZN A 82.. 8.. may. 02.. "mafia—Eon. 2:525» 3m . .: Lfl $2.: mew... wk 359mm :9... 3...: mg... 0.5.: vow... mmod - NEE... .30; m5: 5:: 3:: vum: 3:: m:..: + MNR. :5: God- :3... EN... 3:... 2.....- + Z~§RM>SU :8... $2 3:. RS- 23. :2 - :SQmwnsE 3:: gm: 3:: 3m. mm:.: 3:: + UZEQQKNZQUZQZ 3:: am _ .:- m3: 2:. T 3.: :wmé- - 35%. $3. 3. 002?-.. £06580 son—gd Sumo—too“. 0029.: E06500“. mAO z. 0.. .3on 8:82 5385mm. :w_m $3.253 . . @8068: a . GEE 3H7: Soto 8.63 8.. mam—05:8 30.6% .3233. a. $35.5 958.880: owe—ommv o. “SE8: 2.. .o 3.85.580: on. :o mix—«:4 H... . m .U 03¢... cm 8.. 9.38 commmohwom wfiwooofi 2.. H26 wows—:28 385.5 mafiobcoococ 9 38mm .83 :0 0:8 aways on. :o gimme. .858: H UZNQQQMRZQDZOZ w 54 3:2 853: 3653:5856: Ho: 3533 03 can mcoumozmou 0:3 wEms Bot» @5983 92308. no :83 683528 2m 8:15-: 3:66er o .38: £8 2“ a: Ea 5 am mesa: B S: 2 3% u AN: .5: mo :oflmfiumo :wS @8090 2: S. KEENMEN 5m EoEHEmE cm mm com: fl AC .:m :0 :oumfiumo :wS 80¢ 82$ BE“: 0 AN: .3: mo noumEcmo :5on @2238 on: 5 >583ng .8: :58:me an 3 :3: mm C: .3: we sausage :9on 8on 33g :35: a .N DES Ea _ 22¢ a 8&2. em mosfig 3. a 0mm H .o N: 33.: 82.: mg 359?: 3:: 3.0m: : S: :5: :3: ::m: + VO% 3:: :v:.: 33: 3:: m3: 3:: + :09 ES 1mm: mm: mm:.: 3?: wmm: :: m: + 29;: KN: N::.: 3»: ~::.:v v9: ~:::v mg: + NEW :8: m2}- :3: mSN- m2: wmvg- - ANN SN: 3:: v3: 3:: vm:: mm: + 35.5: SEN: :3: 3:: 3:: :v:.m ~::.:v :3: + x5853: oofigd Eamofiooo cos—ST: E06580 eon—g: 3205000 o mamm u awe: 8.530 a :5on @2090 @0502 cocmaumm $5 a 333:; :82on $85.: 372: 3253: 3 9.8.5:: wE=obqooaoa mammofioflc mo “coho 6me8 565 :0 832530 033:; 358:me ”Au—numb; E0253: 3 «89:196ka mama: AN: em 8.: $152 mommmouwom v .U 935. 55 Sm comics 2m 98 228838 83 mi? Echo waccfim @5308 no woman 85:28 8m weigh onméz/w .38: 58 as S Ea E E macaw“ 3 S: 2 3% .ANV .wm mo aoumfismo :wB @0590 05 E \bbflmw: Ho.“ :58:me nu 8 com: mm 2v .wm mo :ozmficwo «ES 8on 338, ESE .va .Um mo :oflaESmo :9on @8090 05 E NENEQVE Sm “58ng 5 mm wow: mm A: .v@ mo cocflbflmo H503 Boa 82g REE A 035 a 8539 a ozacg oi .SwS 8:53 bfismacoxmobg chobcooao: mafia—8% mo Soto H858 :88 .«o sausage oEuflg 3585me ”65:8 8923 AS .cm 8% 33%.: nommmohwom bass: 8 335%: m .0 933—. N: 823 98.0 885 «a £525 33. E: 3.2V £3 3...? 33. + $333: 333:; Eucaoc 3 «8% b2: $3 386 NM £85 28o N: 25:.— 33. E2 SSV 23 33V 3% + his: AOEEE “5225““ a ESE? 32E 3: 0 263-9 E20580 fl “weigh 2865000 _ Doggd _ E30580 c 3% _ 0 a3 8520 _ 3 BE 8520 8502 8:355 gym a 83m?» BEBE Gee 8T6 0'5 :3 .D 56 momnv mcowmwzno 50:02 #5883058 98 :ommcom 835m 5th Bot SEVwcoA nvmamfi 85:531— thFU Each ow 26 32:53 3% govwcoq ~23 35:55: 3803‘ oatm Bambwa 3583‘. ommJ mwcrsobon 9:8-tonm bEwm whogoefim 98 82—534 m Show 3354 EB .H 2&6 38mm 550 EQN $0me Bflwcfifi : 2: =E€oc fimmnm mwomma taxi VN: mavens 53 088E 235m coo; mwommm wfiocacm 58830 mmflfl 333‘ 2.2.50 .33. m3 33mm 3830 $50 3%; wwwocon 53 2:02: 835 wood 38on E 385:8 98 3:8235 :3 Ea $3 06 35.88 25:8 E m ~ m6 85352—5 wo 85 oEmZooB 3558< mwmflm Eco—«>33 :23 was :96 Bomm< voom 80am 852mm 3328980 900 mflwo—ocnooh BED 5.53m bang: 2: 5 38.3.5 mum—3.5.395:— w:_mc_om_c ma «En—Sn...— muEEaxu Eats—Emma? nooaméoaa—am "G flan—25¢. 57 mmodv N3.0.0... 300300035 .50 00.3...005 .30... .50... 5.5.0m 30:30ouaam .30 H 5.03. 0. mowcwnu Amm .. .v Eogooamsméo Z 0050 030.0058». .30 ... mm. 0050 . 8%. 0 0:52 8.82. 8:832 SN 00.32833 .0503 $.20... ”3.0.00 :. mowcmno 3003003500: .850 0030.=§00< 6va $5.... 0000 8:380 $0: 3:05.. 8593. €50. 500 E 885 006800 033.”. 0:0 NNm.mm . 0.005 >803... 02... 023.0 20.03 .00.N 80600 000mm. 303% ooodoqm 63.00530. m020> an. .m J03m 008800 m:.0033:0 00 000mm. 0007. $085 ooo.omm-00N..~0€0< .00.? .3. .m £005 02.00.05 0.85 .330 5.3m. 300300.80m nmodm 8.3...005 .30... 0.0 3.0.2.800 50.0.33 0. 38030. 5.80.3. Am. .000 v 83on 30:30.. “cognac“. .000 3025.880“. $3. 30:53 €3-32 550 3.2.00. 00.33. 3:30.. 2.» a. 38.53:. an...0.=:00:0: 3:00.00... .0 0.0833 58 gamma— EEEE 3.; S: 8020 mflocomctcoo van 3898:8800 2..ch 33:52— ::2-w8o_ 580 3.? Bow 8.83-80; coda fl . H 85.5.:— 82.89 .83. 8.2 $53 0888 wobomom om“: _ $83 @888 menus—08 .8me $.ch 8::sz 580 can Evian 388< o_.m G 0333 $8508 068,—. 5% 8.8.82 mo cm. s: .8an 8085 28 $830.88 88-825 ”SE—Ba: 88.80 bgam .muoEoAv—oem 88 «3.88— bflcE—z £25531— ow.wmo,m 30mm.“ .33. 8.3— 53880 3.3 853 @888 @28on 33% $0me 850 86$ 6: 808853 28 35895 Simon; 30mm.“ 82.89 .85. onw: 38mm 8885 $50 EB 035382 mausooo< 3.8m; 380838 8888802 3.3 $828ng owévm 3:0?353 8me can :80 ”gamma 82.80 383‘ «com Hoonm 8.83m BEE—8800 .2: £580 828cm 85on 08.8808 95 8 $8835 mam—.2888:— wfimfiummc a: 29883.— 59 owdmoflm m 3.5.00 0001.05.00: .80 000.58. N0.0088 .00....50: .80H 00.30“. 3850 80205.83 .30h 8.0mm $8800 0080.3. 3.3 08008 03080808800 0080 00.088306V 82: 8030800800 008003. 20% 230838 .38 “mm 093 0% 88 .2 hang a .0 8.088080 0:0 .0030. 0082..m .thng .80 5065.? 60889.80 0089.0 ooodoodom a0.00.0 808800 238.00 800.08.008m 00.0mm. 30.58. margin .80 00.5.00: .30... Ex. .0208. bugs». 2:800. 80.0000 08:35.8 0... 8 300808. w:...0.::00:0: $800.00... .0 0.980mm 60 Example of disclosing noncontrolling interests in the equity section (from Appendix Assets Liabilities: Equity: A of FAS 160) ABC Co. Consolidated Statement of Financial Position As of December 31 Cash Accounts receivable Available-for-sale securities Plant and equipment Total assets Total Liabilities ABC Co. shareholders' equity: Common stock, $1 par Paid-in capital Retained earnings Accumulated other comprehensive income Total ABC Co. shareholders' equity Noncontrolling interests Total equity Total Liabilities and equity 61 20x3 20X2 $ $ 570,000 475,000 125,000 1 10,000 125,000 120,000 220,000 235,000 $ 1,040,000 940,?)00 $ $ 555,000 459,000 200,000 200,000 42,000 50,000 194,500 167,000 22,500 16,000 459,000 433,000 26,000 48,000 485,000 481,000 $ 1,040,000 940,%00 BIBLIOGRAPHY Accountants International Study Group. 1972. Consolidated financial statements. Current Recommended Practices in Canada, the United Kingdom and the United States. Ahmed, A., B. Billings, R. Morton, and M. Harris. 2002. The role of accounting conservatism in mitigating bondholder-shareholder conflict over dividend policy and in reducing debt cost. The Accounting Review 77: 867-890. Ali, Ashiq, T.Y. Chen, and S. Radhakrishnan. 2007. Corporate disclosures by family firms. Journal of Financial Economics (forthcoming). Almeida, H., and M. Campello. 2002. Financial constraints and investment-cash flow sensitivities: new research directions. New York University and University of Illinois, Working Paper. Altman, E. 1992. Revising the high-yield bond market. Financial Management 21 (2): 89-92. American Accounting Association. 1957. “Accounting and reporting standards for corporate financial statements: 1957 Revision”. The Accounting Review (October): 536-546. -------- . Committee on Concepts and Standards. 1955. “Consolidated financial statements, supplementary statement No. 7”. The Accounting Review (April): 194-197. -------- . 1964 Concepts and Standards Research Study Committee — The Business Entity Concept. 1965. “The entity concept”. The Accounting Review (April): 358-367. -------- . Financial Accounting Standards Committee. 1999. Liabilities and equity. Accounting Horizons 13:305-307. ------- . 2001. Evaluation of the FASB’s proposed accounting for financial instruments with characteristics of liabilities, equity, or both. Accounting Horizons 15:387- 400. American Institute of Certified Public Accountants. 1956. “A survey of consolidated financial statement practices”. P. 18. 62 ------ . 1959. “Consolidated financial statements”. Accounting Research Bulletin No. 51. -------- . 1994-2003. Accounting Trends & Techniques. New York, NY: AICPA. -------- . 1953. “Reinstatement and revision of accounting research bulletins”. Accounting Research Bulletin No. 43 Anderson, R.C., S. A. Mansi, and D. M. Reeb. 2003. Founding family ownership and the agency cost of debt. Journal of Financial Economics 68 (2003): 263-85. -------- and D.M. Reeb. 2003. Founding-family ownership and firm performance: evidence from the S&P 500. The Journal of Finance LVIII (3): 1301 — 1328. Angrist, J .D. and A. B. Krueger. 2001. Instrumental variables and the search for identification: From supply and demand to natural experiments. The Journal of Economic Perspectives (15) 4: 69-85. Ashbaugh, H., D. Collins, and R. LaFond. 2006. The effects of corporate governance on firms’ credit ratings. Journal of A ccounting and Economics 42: 203 - 243. Aslan, H., D. Easley, S. Hvidkjaer, and M. O’Hara. 2007. Firm characteristics and informed trading: Implications for Asset Pricing. Working Paper. University Houston, Cornell University, and University of Maryland. Ball, R. 2001. Infrastructure requirements for an economically efficient system of public financial reporting and disclosure. Brookings-Wharton Papers on Financial Services: 127-182. Barclay, MJ. and CG. Holdemess. 1989. Private benefits from control of public corporations. Journal of Financial Economics 25 (2): 371-395. Barry, CB. and S]. Brown. 1984. Differential information and the small firm effect, Journal of Financial Economics 90, 629-650. Basu, S. 1997. The conservatism principle and the asymmetric timeliness of earnings. Journal of Accounting and Economics 24: 3-3 7. Beatty, Anne, J Weber, and J. Yu. 2006. Conservatism and Debt. Working Paper. The Ohio State University, and Massachusetts Institute of Technology. Bebchuk, L. 1989. Limiting contractual freedom in corporate law: the desirable constraints on charter amendments. Harvard Law Review 102: 1820-1860. 63 Bemanke, B. and M. Gertler. 1989. Agency costs, net worth, and business fluctuations. The American Economic Review 79 (1): 14-31. , , and S. Gilchrist. 1996. The financial accelerator and flight to quality. Review of Economics and Statistics 78:1-15 Botosan, C. A. and M. A. Plurnlee. 2005. Assessing alternative proxies for the expected risk premium. The Accounting Review (January): 21-53. Brudney, V. and R. Clark. 1981. A new look at corporate opportunities. Harvard Law Review 94: 997-1062. BusinessWeek, 2003. Family, Inc. November 10. Campello, M. and L. Chen. 2005. Are financial constraints priced? Evidence from firm fundamentals, stocks, and bonds. University of Illinois and Michigan State University, Working Paper. -------- . Chen and L. Zhang. 2005. Expected returns, yield spreads, and asset pricing tests. University of Rochester, Working Paper. Carlstrom, C. T. and T. S. Fuerst. 1997. Agency costs, net worth, and business fluctuations: A computable general equilibrium analysis. The American Economic Review 87 (5): 893-910. Chami, R. 1999. What’s different about family business? Working paper, University of Notre Dame and the International Monetary Fund. Cheng, Q., P. Frischmann, and T. Warfield. 2003. the market perception of corporate claims. Research in Accounting Regulation 16: 3-28. Choi, D.S., Levich, R.M., 1990, “The capital market effects of international accounting diversity”, Irwin Professional Pub Cohen, L. 1993. Head of Brooke group draws on its coffers to tune of millions. The Wall Street Journal (July 30): p. A — 2. Cooper, 1. and S. Davydenko. 2004. Using yield spreads to estimate expected returns on debt and equity. London Business School, Working Paper. Cooper, W. W., and Y. Ijiri, eds. 1983. Kohler ’s Dictionary for Accountants. Sixth edition. Englewood Cliffs, NH: Prentice Hall, Inc. 64 DeAngelo, H., L. DeAngelo and E. M. Rice. 1984. Going private: minority freezeouts and stockholder wealth. Journal of Law and Economics 27 (2): 367-401. DeFond, M. L. and C. W. Park. 2001. The reversal of abnormal accruals and the market valuation of earnings surprises. The Accounting Review 76 (3): 375-404. Denis, D. and D. Denis. 1994. Majority owner-managers and organizational efficiency. Journal of Corporate Finance 1 (1): 91-118. Dhaliwal, D. 1986. Measurement of financial leverage in the presence of unfunded pension obligations. The Accounting Review 61: 651-661. Dye, R., 2002. Classifications Manipulation and Nash Accounting Standards. Journal of Accounting Research 40 (2002): 1125—62 Duke, J ., and G. Hunt. 1990. An empirical examination of debt covenant restrictions and accounting-related debt proxies. Journal of Accounting and Economics 12: 45-63 Elton, E. J. 1999. Expected return, realized return, and asset pricing tests. Journal of Finance 54: 1199-1220. -------- , J. J. Gruber, D. Agrawal and C. Mann. 2001. Explaining the rate spread on corporate bonds. Journal of Finance 56: 247-277. Engel, E., M. Erickson and E. Maydew. 1999. Debt-equity hybrid securities. Journal of Accounting Research 37: 249-274. Fama, E. and MC. Jensen. 1983. Separation of ownership and control. Journal of Law and Economics 26: 301-326. Fazzari, S. M., R. G. Hubbard, B. C. Petersen, A. S. Blinder and J. M. Poterba. 1988. Financing constraints and corporate investment. Brookings Papers on Economic Activity 1: 141-206. Fields, T.D., T.Z. Lys, and L. Vincent. 2001. Empirical research on accounting choice. Journal of Accounting and Economics 31: 255-307. Financial Accounting Standards Board (FASB). 1985. “Elements of Financial Statements Concepts Statement No. 6. Stamford, CTzFASB. -------- . 1987. “Consolidation of all majority-owned subsidiaries”. Statement of Financial Accounting Standards No. 94 . 65 -------- . 1990. “Distinguishing between liability and equity instruments and accounting for instruments with characteristics of both”. Discussion Memorandum. -------- . 1991. “An analysis of issues related to consolidation policy and procedures”. Discussion Memorandum. -------- . 2003. “Accounting for certain financial instruments with characteristics of both liabilities and equity”. Norwalk, CT: FASB. -------- . 2004. FASB Project Update: Liabilities and Equity. www.fasb.org -------- . 2005. “Consolidated financial statements, including accounting and reporting of noncontrolling interests in subsidiaries Exposure Draft. Norwalk, CT: F ASB. -------- . 2007. “Noncontrolling interests in consolidated financial statements Statement of Financial Accounting Standards No. 160 Fisher, L., 1959. Determinants of risk premiums on corporate bonds. The Journal of Political Economy 67: 217 — 237. Francis, J ., R. LaFond, P. M. Olsson, and K. Schipper. 2004. Cost of equity and earnings attributes. The Accounting Review 79 (4): 967-1010. Gilson, .R. J. and J. Gordon. 2003. Controlling controlling shareholders. Working Paper #228. Columbia law School, The Center for Law and Economic Studies, New York. Givoly, Dan, C. Hayn and A. Natarajan. 2007. Measuring reporting conservatism. The Accounting Review 82 (1): 65-106 -------- , and --------. 2000. The changing time-series properties of earnings, cash flows and accruals: Has financial reporting become more conservative? Journal of Accounting and Economics 29: 287-320. Gomes, J .F., A. Yaron, and L. Zhang. 2004. Asset pricing implications of firms’ financing constraints. University of Pennsylvania, Working Paper. Graham, RC. and CE. Lefanowicz. 1999. Majority and minority ownership of publicly- traded firms: a test of the value of control using market multiples. Journal of Business Finance and Accounting 26 (1) & (2): 171-198. Greene, W., 1998. LIMDEP, Version 7.0: User’s Manual. Econometric Software Inc, Plainview, NY. 66 -------- . 2000. Econometric analysis, 4th Edition. Prentice Hall, Saddle River, NJ. Hamada, R. 1969. Portfolio analysis, market equilibrium, and corporate finance. The Journal of finance 24: 13-31 ------- . 1972. The effect of the firm’s capital structure on the systematic risk of common stocks. The Journal of Finance (May): 435-452. Hazen, Thomas, 1977, Transfer of corporate control and duties of controlling shareholders — common law , tender offers, investment companies-and a proposal for reform, University of Pennsylvania Law Reuiew 125, 1023-1067. Hand, J. R. M., R. Holthausen, and R. Leftwich. 1992. The effect of bond rating agency announcement on bond and stock prices. Journal of Finance 47 (2): 733-753. Healy, P., and K. Palepu. 1993. The effect of firms’ financial disclosure strategies on stock prices. Accounting Horizons (March): 1-11. Hendriksen, E. S. 1970. “Accounting theory”. Homewood: Richard D. Irwin. Holdemess, C. D. and D. P. Sheehan. 1988. The role of majority shareholders in publicly-held corporations: an exploratory analysis. Journal of Financial Economics 20 (1/2): 317-340. Husband, G. R., 1938, “The corporate-entity fiction and accounting theory”. The Accounting Review: 241-253. -------- . 1954. “The entity concept of accounting”. The Accounting Review (October): 552-563. Jaffe, T. 1996. New issues in reverse. Forbes (June 17): 158-160. Jensen, M.C., and W.H. Meckling. 1976. Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3: 305- 360. Johnson, M. F ., R. Kasznik, and K. K. Nelson. 2001. The impact of securities litigation reform on the disclosure of forward-looking information by high technology firms. Journal of Accounting Research 39 (2): 297-327. Johnson, S, R LaPorta, F. Lopez-de-Silanes, A. Shleifer. 2000. Tunnelling. Discussion Paper. Harvard University. 67 Kaplan, R., and G. Urwitz. 1979. Statistical models of bond ratings: a methodological inquiry. Journal of Business 52: 231-261. Kasznik. R., B. Lev , 1995, "To warn or not to warn: management disclosures in the face of an earnings surprise," The Accounting Review, 113-134. Kellogg, R.L. 1984. Accounting activities, security prices, and class action lawsuits. Journal of Accounting and Economics 6: 185-204. Kennedy, P., 1998. A guide to econometrics, 4th Edition. MIT Press, Cambridge, MA. Kimmel, P. and T. Warfield. 1995. The usefulness of hybrid security classifications: Evidence from redeemable preferred stock. The Accounting Review: 151-167 Klock, M., S. Mansi, and W. Maxwell. 2005. Does corporate governance matter to bondholders. Journal of Financial and Quantitative Analysis (40) 4: 693-719 Kohler, E. L. 1938. “Some tentative propositions underlying consolidated reports”. The Accounting review (March): 63-77. Kothari, S.P., Lys, T., Smith, C.W., Watts, R.L., 1989. Auditor liability and information disclosure. Journal of A ccounting, Auditing and Finance 4, 307-339. LaFond, R. and R. L. Watts. 2006. “The information role of conservative financial statements”. Working Paper. Massachusetts Institute of Technology. Lamont, 0., C. Polk and J. Saa-Requejo. 2001. Financial constraint and stock returns. Review of Financial Studies 14:529-554. La Porta, R., F. Lopez-de-silanes, A. Shleifer and R. W. Vishny. 2000a. Agency problems and dividend policies around the world. Journal of Finance 55: 1-33. -------- . 2000b. Investor protection and corporate governance. Journal of Financial Economics 58: 3-27. Lee, Y.J. 2005. The market perception of employee stock options. Working Paper. Georgia State University. Linsmeier, T., C. Shakespeare, and T. Sougiannis. 2004. Liability/equity classifications and shareholder valuation. Working Paper. Michigan State University. 68 Liu, P., F. Seyyed, and S. Smith. 1999. The independent impact of credit rating changes: The case of Moody’s rating refinement on yield premiums. Journal of Business Finance & Accounting 26 (3-4): 337-363. Loss. L., J. Seligman, 1991, “Securities Regulations”, Aspen Law & Business Maddala, G., 1986. Limited dependent and qualitative variables in econometrics. Econometric Society Monographs. Marquardt, C. and C. Wiedman. 2003. Earnings management through transaction structuring: contingent convertible debt and diluted EPS. New York University, Working Paper. Maury, CB. and A. Pajuste. 2002. Controlling shareholders, agency problems, and dividend policy in Finland. The Finnish Journal of Business Economics 1: 15-45. Merton, R. C. 1973. An intertemporal capital asset pricing model. Econometrica 41: 867- 887. -------- , 1987, “A simple model of capital market equilibrium with incomplete information”, The Journal of Finance 42, 483-510. Moonitz, M. 1960. “The changing concept of liabilities”. The Accounting Review (May): 41-46. -------- . 1942. “The entity approach to consolidated statements”. The Accounting Review (July):236-242. Morck, R. A. Shleifer, and R. Vishny. 1998. Management ownership and market valuation: An empirical analysis. Journal of Financial Economics 20: 293-315. Murphy, R. 1956. Corporate divisions vs. subsidiaries. Harvard Business Review 34: 83- 92. Myers, S. C. and N. S. Majluf. 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13 (2): 187-221. Newlowe, G. H. 1948. “Consolidated statements: including mergers and acquisitions”. Boston: D. C. Health and co. Noreen, E., 1988. An empirical comparison of probit and OLS regression hypothesis tests. Journal of Accounting Research 26: 119 — 133. 69 Pacter, P. 1991. “Consolidation Policy and Procedures”. Discussion memorandum. Norwalk, CT: FASB. Paton, W. A. 1922. “Accounting theory”. New York: The Ronald Press. Qiang, Xinrong. 2007. The effects of contracting, litigation, regulation, and tax costs on conditional and unconditional conservatism: Cross-sectional Evidence at the firm level. The Accounting Review 82 (3): 759-796. Rogers, J. L. and P. C. Stocken. 2002. Credibility of management forecasts. Working Paper. University of Pennsylvania. Rubin, S. 1988. “Applications in accounting”. Journal of Accountancy 166: 123-126. Sapienza, S. R. 1960. “The divided house of consolidations”. The Accounting Review (July): 750-763. Schipper, K. and A. Smith. 1996. A comparison of equity carve-outs and seasoned equity offerings. Journal of Financial Economics 15 (1/2): 153-186. Securities and Exchange Commission. 1979. “Presentation in financial statements of ‘redeemable preferred stocks”. Rule 5-02.28 of Regulation S-Xl Sengupta, P. 1998. Corporate disclosure quality and the cost of debt. The Accounting Review 73: 459 - 474. Sharpe, W. F. 1964. Capital asset prices: A theory of market equilibrium under conditions of risk. Journal of Finance 19: 425-442. Shleifer, A. and R.W. Vishny. 1997. A survey on corporate governance. Journal of Finance 52 (2): 737-783. Skinner, D., 1994, “Why firms voluntarily disclose bad news”, Journal of Accounting Research 32, 38-61. Slovin, MB. and ME Sushka. 1997. The implications of equity issuance decisions within a parent-subsidiary governance structure. The Journal of Finance LII (2): 841-857. Stein, J. 1988. Takeover threats and managerial myopia. Journal of Political Economy 96: 61-80. - 70 Stein, J. 1989. Efficient capital markets, inefficient firms: A model of myopic corporate behavior. Quarterly Journal of Economics 103: 655-669. Stulz, R. 1988. managerial control of voting rights. Journal of Financial Economics 20: 25-59. Verrecchia, R. 1983. Discretionary disclosure. Journal of Accounting and Economics mecember): 179 — 194. Wang, D. 2006. Founding family ownership and earnings quality? Journal of Financial Economics 44 (3): 619-655. Watts, R.L., 1993 A proposal for research on conservatism. Working Paper. University of Rochester. -------- . 2003a. Conservatism in accounting Part I: explanations and implications. Accounting Horizons 17: 207-221. -------- . 2003b. Conservatism in accounting Part II: evidence and research opportunities. Accounting Horizons 17: 287-301. -------- , and J. Zimmerman. 1986. Positive accounting theory. Chapter 8. Prentice Hall, Englewood cliffs, NJ. -------- . 1990. Positive accounting theory: a ten year perspective. The Accounting Review 65: 131-156. Whited, T. M. and G. Wu. 2005. Financial constraints risk. University of Wisconsin and University of Michigan. Working Paper Wooldridge J. 2002. Econometric analysis of cross section and panel data. Cambridge, Massachusetts: The MIT Press. Zhang, J. Y. 2006. Efficiency gains from accounting conservatism: Benefits to lenders and borrowers. Working paper. University of South California. Ziebart, D., and S. Reiter. 1992. Bond rating, bond yields and financial information. Contemporary Accounting Research 9: 252 -282. 71 .............. utjmtjijjjjj«ErrI