強制採用國際會計準則對銀行聯貸市場的影響 - 政大學術集成

全文

(1)The Effects of Mandatory IFRS Reporting on the Syndicated Loan Market. Chun Yao. 學. ‧ 國. 立. 政治大. Under the supervision of Professor Ting-Wong Cheng. Nat. er. io. sit. y. ‧. and Professor Chen-Lung Chin. al. n. v i n A dissertation C submitted in partialUfulfillment of hengchi. the requirement for the degree of Doctor of Philosophy. Department of Accounting National Chengchi University January, 2012.

(2) 謝辭僅以此篇論文獻給我在天上的爺爺與奶奶，謝謝您們對我的照顧與愛護。這本論文的誕生首先最要感謝的就是我的兩位指導老師：鄭丁旺教授與金成隆教授，感謝您們在我求學的過程中，無論是課業上的指導或生涯規劃上的建議，因為您們，使我得以順利完成我的博士學位。也感謝論文口試老師─劉啟群教授、林修葳教授、周冠男教授與黃瓊慧教授，因為您們在口試時所給予的寶貴意見，使得此論文更趨完整。此外，感謝周玲臺老師在專題討論課程時，對我的教導與叮嚀，學生會銘記在心。. 治政大回首七年的博士生涯，歷歷在目。感謝這一路走來的夥伴─慈青與謙恆，沒立. 有您們的陪伴，我想我無法披荊斬棘、撐到現在。感謝素芳學姐、美珠學姊、心. ‧ 國. 學. 瑤學姐、倩如學姐、昭蓉學姐、麗雯學妹與家璿學妹，謝謝您們在我博士生涯遇. ‧. 到困難時，傾聽我的心事、給予我寶貴的意見。還有我的同門師妹─佩怡，謝謝. y. sit. io. al. er. 樂。. Nat. 您在我口試期間，幫助我順利完成口試，聽我分享寫作論文過程中的悲愁與喜. v. n. 還有，我要感謝這一路求學路走來，陪伴我的許多好友與長輩。李瘋瘋(李. Ch. engchi. i n U. 欣怡)、高老大、富吉、昆豪、秀英、阿潘、欣薇、太后(秀玲)、徐倫倫、鳥政，您們無論在我開心、難過還是沮喪時，謝謝您們一直都陪著我，看我笑、看我瘋、看我哭，少了您們，我人生可能不會這麼精采、這麼有趣。尤其是李瘋瘋與高老大，謝謝您們在我寫作博士論文這段期間，在我最失意沮喪時，義無反顧的陪伴著我，還帶我去我這輩子最不想回憶起的城市─首爾(哈哈哈……)。此外，我還要感謝林穎芬老師，謝謝您在工作上、學業上給我的支持與鼓勵；感謝蕭朝興老師，謝謝您讓我喜愛上學術研究，並鼓勵我去讀博士班；感謝廖珮珊老師、侯家鼎老師、歐惠吉老師，您們是我求學路上亦師亦友的長輩，少了您們對我的教導與付出，我不可能有今天這麼一點點的成就。 i.

(3) 最後，我要感謝我的家人─我的爸爸、媽媽、姑姑、甜甜、蕾蕾，您們永遠都是我最重要、讓我最幸福、也讓我最痛苦的人，謝謝您們養育我、陪著我成長。我更感謝我在天上的爺爺與奶奶，我這個博士學位是獻給您們倆的，可惜我不夠爭氣讓您們早點看到我畢業，我能做的就只是努力寫好這篇論文，並將這篇論文獻給您們。要感謝的人真的太多太多了，真的無法只以「那就謝天吧！」一句帶過。這一路走來，有好多好多我認識的、或我不認識的，我喜歡的、或我不喜歡的，造就了今天的我，我真的很感謝您們，不論是被我提到或沒提到的人，謝謝您們！. 政治大. 我現在的心境正如同李白的詩：「眾鳥高飛盡，孤雲獨去閒。」所有的繁華. 立. 靡麗，將過往雲煙，轉眼成空，只有我心中的敬亭山，是我真正追求的本心，「相. ‧. ‧ 國. 學. io. sit. y. Nat. 姚畯謹致於一月深冬. n. al. er. 看兩不厭」。. Ch. engchi. ii. i n U. v.

(4) The Effects of Mandatory IFRS Reporting on the Syndicated Loan Market. (Dissertation). 立. 治政Abstract 大. ‧ 國. 學. ‧. In this dissertation, I examine how the mandatory adoption of International Financial Reporting Standards (IFRS) affects ownership structure and debt covenants in the syndicated loan market. I hypothesize and document that the proportion of the loan retained by syndicate lead arrangers increases after a borrower adopts mandatory IFRS reporting. Further, I document that foreign lenders are relatively less likely to be. y. Nat. sit. n. al. er. io. involved in syndicated loan deals after the adopting of mandatory IFRS reporting. Finally, I find that syndicate lenders are less likely to use financial covenants in debt agreements after the mandatory IFRS adopting, regardless of income statement-based or balance sheet-based covenants. Overall, these results are in line with the argument by Schipper [2003]. Specifically, the adoption of a principles-based accounting system (e.g., IFRS), characterized by limited interpretation and implementation guidance, increases the difference in professional judgment among debt contracting parties, which in turn reduces lenders’ and borrowers’ demand for accounting information in signing debt contracts. Finally, the negative effect of the mandatory IFRS adoption on the syndicated loan market is weaker in common-law countries (in countries with stricter enforcement regime) than in code-law countries (in countries with weaker enforcement regime).. Key words:. Ch. engchi. i n U. v. International Financial Reporting Standards (IFRS), Syndicated Loans, Debt Contracts, Covenants, Home Bias. iii.

(5) 1. Introduction This dissertation examines whether widespread mandatory adoptions of International Financial Reporting Standards (IFRS) facilitate trading of the syndicated loan market. The extant studies on IFRS suggest that mandatory IFRS reporting around the world enhances financial reporting quality and increases financial statement comparability (Daske et al. [2008], Li [2010], Byard et al. [2011], DeFond et al. [2011], Tan et al. [2011]), and in turn benefits equity markets, in particular international capital markets. In addition to equity market, syndicated loans, however,. 政治大 that financial statement information. are an increasingly important source of corporate finance in the U.S. and most other countries.. 1. Despite the fact. 立. is particularly. important for contracting purposes (Leftwich [1983], Watts and Zimmerman [1986],. ‧ 國. 學. Ball [2001], Holthausen and Watts [2001], Ahmed et al. [2002], Li [2010]),2 little. er. io. sit. y. Nat. syndicate ownership structure and debt contractual terms.. ‧. work to date explores whether and how the mandatory IFRS adoption influences. This dissertation addresses these questions by examining whether the. n. al. introduction of mandatory. v i n C hreporting aroundUthe world IFRS engchi. can facilitate the. syndicated debt market. Specifically, the purpose of this paper is four-fold. First, I would like to investigate whether introduction of mandatory IFRS alters syndicate loans structure. Second, given the prior findings on a phenomenon referred to as. 1. Syndicated loans become a very significant source of financing around the world as well as in the U.S. For example, global syndicated lending has grown strongly from the beginning of the 1990s to date. Signings of new loans totaled $1.6 trillion in 2003, more than three times the 1993 amount (Gadanecz [2004]). In the U.S., over the past decade, there have been $780 billion in new debt securities and only $2 billion for equities (Graham et al. [2008]). And according to the American Banker, syndicated lending generates most part of underwriting revenue for the financial sector (about 51% of total U.S. corporate finance)(Weidner [2000]). 2 In their review article, Armstrong at al. [2010] further conclude that “… financial reporting is useful because efficient contract are possible when contracting parties commit to a more transparent information environment,” and indicate that accounting information plays a more explicit role on debt contracts than on contracting in the minority-interest setting. 1.

(6) “home bias”, I further examine whether foreign lenders are more likely to participate in syndicated loan deals after the introduction of mandatory IFRS reporting than before. 3 Third, I examine how the extent to which financial covenants used in syndicate loan systematically varies following the mandatory IFRS reporting. Finally, I examine whether the extent to which the adoption of IFRS is associated with the syndicate loan structure and the use of financial covenants is conditional on the institutional environment, including (1) a country’s legal original and (2) the country-level enforcement regimes.. 政治大 The first question to be addressed is to examine whether the adoption of IFRS 立. reporting alters the ownership structure of syndicated loan deals. To address this. ‧ 國. 學. question, I test how the proportion of the loan retained by the lead arranger is. ‧. influenced before and after introduction of mandatory IFRS. The reason for focusing. sit. y. Nat. on the lead arranger is that the lead arranger is better informed than other participant. io. er. banks (Dennis and Mullineaux [2000], Bharath et al. [2007], Sufi [2007], Ball et al. [2008]). Prior to syndication, the lead arranger establishes a relationship with the. al. n. v i n C h memorandumUto potential buyers, determines borrowing firm, presents a confidential engchi the syndicate composition, negotiates the pricing, and administers the facility. After. syndication, the lead arranger is in charge of monitoring the borrower. Thus, the unobservability of the lead arranger’s due diligence and monitoring effort results in severe adverse selection and moral hazard problems to which syndicate participants are exposed (Sufi [2007], Ball et al. [2008], Ivashina [2009]).4 Thus, the fraction of 3. In the context of syndicated market, Carey and Nini [2006] reveal that foreign lenders tend to be reluctant to participate in loans due to a phenomenon referred to as “home bias.” 4 The key distinction between the adverse selection and moral hazard hypotheses is the assumption of where information asymmetry lies. In the adverse selection hypothesis, the lead arranger has private information on the firm that is unknown to participant lenders. In the moral hazard hypothesis, all lenders are unfanmiliar with the borrower and the moral hazard problem is most serve when the lead arranger must learn about the firm (Sufi [2007]). 2.

(7) loan retained by the lead arranger could convey a lead lender’s resolution to exert their effort in due diligence and monitoring. However, previous research finds that when the borrowing firm provides more transparent information (e.g., financial statement), the more percentage of the loan can be syndicated by the lead arranger (Dennis and Mullineaux [2000], Sufi [2007], Ball et al. [2008]). Accordingly, to the extent that mandatory IFRS reporting influences accounting quality via financial reporting transparency and comparability (Bae et al. [2008], Barth et al. [2008], DeFond et al. [2011], Tan et al. [2011]), I expect the proportion of the loan retained by. 政治大. the lead arranger systematically varies after firms adopt mandatorily IFRS reporting.. 立. However, financial reporting (e.g., IFRS) can trigger economic consequences. ‧ 國. 學. through its informational and contracting (or stewardship) roles (Beaver [2002]),5 and. ‧. the two roles are not necessarily compatible with each other. For example, Gassen. y. Nat. [2008] reveals that stewardship-related demand for accounting information varies. er. io. sit. inversely with the valuation usefulness of that information.6 As a result, it is possible in our setting that there are two distinct perspectives about how the mandatory IFRS. n. al. 5. Ch. engchi. i n U. v. The information (contracting) role of financial reporting can also be referred to as the ex-ante or valuation role (ex-post or stewardship) role of accounting information, which is mainly intended to address adverse selection (moral hazard) problems (e.g., Beyer et al. [2010]). 6 In the pre-contract syndication process, debt holders, as well as shareholders, demand financial statements that supply information about the amounts, timing, and uncertainty of firms’ prospective cash receipts. Shareholders seek such information that helps them either perform equity valuation or compensate management for current period performance. Debt holders view this information as an indicator of the firm’s ability to service the debt and avoid a flow-based insolvency (Ross [2002]). However, in the post-contract syndication process, debt creates an agency relationship between shareholders and debt holders (Jensen and Meckling [1976], Smith and Warner [1979]). Studies (Myers [1977], Smith and Warner [1979], Kothari et al. [2010]) indicate that the assets substitution and underinvestment problems drive debt holders’ demand for information about (1) the value of the firm’s nets assets in the event of liquidation and (2) the extent of other claims on those assets. Because debt holders expect these agency problems, they can mitigate these agency problems by relating debt covenants to accounting numbers in the debt contract. This contracting demand from a GAAP can be only effective under the situation in which a GAAP has verifiable and conservative features (Kothari et al. [2010]). Beatty et al. [2008a] find that debt contracts have a tendency to include income escalator when the agency costs of debt are likely to be higher. Similarly, Li [2010] shows that debt contracts tend to eliminate transitory components from GAAP earnings in order to enhance contracting usefulness. To the extent that GAAP earnings mainly reflect informational role of accounting, these adjustments are indicative of incompatibilities between the informational and contracting roles of accounting. From this perspective, the adoption of mandatory IFRS could have some adverse effects on the debt contracting (discussed later). As a result, the positive effects of mandatory IFRS reporting on the equity market could not be necessarily observed in the debt market. 3.

(8) adoption impacts syndicate structure.. First, from the information perspective, enhanced financial reporting quality and comparability due to the mandatory IFRS adoption will mitigate the adverse-selection problems between lend arrangers and non-lead lenders, and in turn reduce the potential participant lenders’ (non-lead lenders in general and foreign lenders in particular) demand for the lead arranger to hold a larger proportion of syndicated loan.. 政治大 characterized by limited interpretation and guidance, will increase the noise of 立 Second, from the contracting perspective, the adoption of mandatory IFRS,. financial reporting and the difference in subjective judgment among parties to debt. ‧ 國. 學. contracts (Schipper [2003]). This will increase pre-contract negotiation cost to lenders. ‧. and borrowers and thereby reduce the likelihood of contracting between lead. sit. y. Nat. arrangers and potential participant lenders. Anticipating the information disadvantage,. io. er. potential lenders would require the lead arranger to retain a larger proportion of syndicated loan. Overall, due to the two distinct perspectives, it is thus an empirical. n. al. question whether and how ownership structure.. v i n C hmandatory IFRS U the adoption engchi. affects the syndicate. In addition to the lead arranger’s holding, I further explore the effect of mandatory IFRS adoption on foreign lenders. Prior studies show that investors are reluctant to make cross-border investments due to a phenomenon referred to as “home bias”, and the main factor contributing to home bias is the high cost of information about foreign investments (Kang and Stulz [1997]). The extant research indicates that in the context of voluntarily IFRS adoption that foreign mutual fund ownership is higher among IFRS adopters compared to firms using local accounting standards 4.

(9) (Covrig et al. [2007]). Kim et al. [2011] also find that voluntary IFRS adopters attract more foreign lenders participant in loan syndicate than non-adaptors. In this paper, I extend extant studies and explore whether these conclusions can be generalized into the setting of mandatory IFRS adoption.. Proponents of accounting standards harmonization argue that mandatory IFRS adoption can enhance comparability of financial reporting and reduce the foreign lenders’ information costs about syndicated loan. However, the allegedly increased. 政治大 comparability” (Schipper [2003]). That is, although “one-size fits all” IFRS provide 立. comparability due to mandatory IFRS adoption might be so-called “surface. one set of uniform accounting standards across firms and countries, lack of detailed. ‧ 國. 學. guidance and need for greater subjective judgment will lead to surface comparability,. er. io. sit. y. Nat. for foreign lenders to participate in a syndicated loan.. ‧. rather than real comparability. In this case, IFRS adoption will reduce the incentive. The second question to be addressed is to examine whether financial reporting. al. n. v i n Cthe quality of IFRS adopters affects of debt contractual terms. Prior research h echoice ngchi U. indicates that in the context of debt covenants, lenders rely explicitly on financial statement numbers (Leftwich [1983], Asquith et al. [2005], Li [2010], Costello and Wittenberg-Moerman [2011]). For example, some studies reveal that lenders impose more stringent terms when financial reporting quality is low (e.g., Gigler et al. [2009], Costello and Wittenberg-Moerman [2011]). When introduction of mandatory IFRS can alter financial statement quality and information transparency, an important question naturally arises as to whether lenders make increased or decreased use of contractual terms based on accounting numbers. In this dissertation, I will explore whether lenders of syndicated loan deals decrease or increase their use of 5.

(10) accounting-based contractual terms when borrowers adopt mandatory IFRS. A financial covenant is a monitoring mechanism which is designed to protect lenders’ interest and maintain close scrutiny over the performance of borrowers (Dichev and Skinner [2002]). Therefore, lenders tend to use accounting-based debt covenants as an early warning signal to handle borrowers’ post deteriorations in credit risk. 7 To address this question, I will examine how lenders of syndicated loan deals alter the intensity of financial covenants following the adoption of mandatory IFRS.. 政治大 mandatory IFRS adoption on the choice of debt contractual terms. On the one hand, 立 In the same vein, there are two distinct perspectives on the effect of the. from the information perspective, increased accounting quality due to the IFRS. ‧ 國. 學. adoption will limit management’s opportunistic discretion in manipulating accounting. ‧. numbers to satisfy requirements of financial debt covenants (Barth et al. [2008]).8 As. io. al. er. post-IFRS period than for prior-IFRS period.. sit. y. Nat. a result, lenders are more likely to use financial covenant in debt agreements for. n. v i n C contracting On the other hand, from the the adoption of mandatory h e n g cperspective, hi U. IFRS will lower ex post debt-contracting value of accounting information. As mentioned above, the difference in professional judgment among practitioners about accounting recognition and measurement will increase after IFRS adoption. This. implies that financial covenants are less efficient in conveying changes in a borrower’s creditworthiness after the IFRS adoption. In addition, prior studies 7. An early warning sign of violations from borrowers gives lenders the chance to renegotiations (DeAngelo et al. [1994], Dichev and Skinner [2002]). Renegotiations usually accompany negative effects on borrowers such as an increase in interest rates, restrictions on the firms’ ability to access credit markets, a reduction in borrowers’ access to lines of credit and so on (Sufi [2009], Roberts and Sufi [2009]). 8 IFRS can increase a borrower’s accounting quality by restricting earnings management, requiring more timely loss recognition, and increasing higher value relevance disclosures (Barth et al. [2008]). 6.

(11) indicate that, due to very few scope exceptions (or treatment exceptions) and increased use of fair value measurement, income volatility in countries adopting IFRS increases significantly (Schipper [2003], Barth et al. [2008]). These will decrease lenders’ use of financial covenants as an ex post monitoring tool after a borrower adopts IFRS financial reporting. Accordingly, based on preceding discussion, I do not predict the sign of the effect of the IFRS adoption on the choice of financial covenants.. 政治大 countries which include 22 mandatory IFRS 立. For my empirical tests, I construct a sample of non-U.S. borrowers from 2,219 loans from 35. countries and 13. non-mandatory IFRS countries during the period from 2000 to 2009. Empirical. ‧ 國. 學. analyses use a difference-in-difference approach to examine my research questions. ‧. because this allows me to control possible changes for the IFRS adopters by. sit. y. Nat. concurrent changes that non-IFRS adopters also experience. Most countries, except. io. er. for Singapore, in my sample do not mandatorily adopt IFRS until 2005, when it becomes mandatory to do so. More specifically, all hypotheses are investigated by. al. n. v i n C hloan variable on aUdummy variable for the type of regressing the interested syndicated engchi adopter (mandatory versus non-mandatory adopters), a dummy variable for the time. period (preversus post-mandatory adoption period), the interaction between these two dummy variables, and a set of control variables.. My empirical results can be summarized as follows. First, I find that syndicate lead arrangers hold a higher proportion of loan after a borrower adopts mandatory IFRS reporting. To check my robustness, I also use two alternative measures, a Herfindahal index and the number of lenders, to investigate how the effect of mandatory IFRS reporting on the change of the loan syndicate structure. The results 7.

(12) show similar findings: the loan syndicate structure becomes more concentrated and the number of lenders becomes smaller after a borrower adopts IFRS reporting. Second, the empirical evidence also shows that potential foreign lenders are less likely to participate in a syndicate loan deal following a borrower’s adoption of mandatory IFRS reporting.9 Third, the empirical results indicate that a syndicate loan contract is less likely to include accounting-based financial covenants after mandatory IFRS adopting, regardless of balance sheet-based covenants or income statement-based covenants.. 政治大 In addition, I examine two additional questions to make my analysis complete. 立. First, a loan contract usually comprises n-contractual terms which could be based on. ‧ 國. 學. either accounting variables or non-accounting variables, and lenders would trade off. ‧. these terms in debt contracts to compensate their risk (Melnik and Plaut [1986]). If. sit. y. Nat. IFRS reporting reduces both lenders’ or borrowers’ incentives to sign debt contracts. io. er. based on accounting information, lenders would substitute accounting-based terms with non accounting-based terms when debt contracts are signed. In the same sample. al. n. v i n C hare unwilling to give of 2,219 loans, I find that lenders more credit after a borrower engchi U. mandatorily adopts IFRS. A debt contract offered by lenders following mandatory IFRS reporting becomes small loans, higher loan rates, shorter-maturity loans, more subject to collateral requirements, and less use of the performance pricing provision based on an accounting ratio.. My final question to be addressed in this dissertation is to examine whether the effects of the introduction of mandatory IFRS reporting on syndicated loan deals are 9. An international syndicated loan is defined by the Bank for International Settlements (BIS) as a facility for which there is at least one lender present in the syndicate whose nationality is different from that of the borrower. 8.

(13) not even distributed across countries. This question is motivated by recent IFRS literatures suggesting heterogeneity in capital market effects around the IFRS mandate (Daske, et al. [2008], Ball et al. [2003], Ball [2006]).10 These papers point to a limited role of standards alone in determining reporting quality and, in contrast, highlight the importance of firms’ reporting incentives, which are shaped by many institutional factors. In the same vein, I predict in my setting that the negative relation between the IFRS adoption and the syndicate structure and the use of financial covenant is conditional on such institutional factors. In this dissertation, I focus. 政治大 country-level enforcement regimes. 立. primarily on the following two institutional factors: a country’s legal original, and the. ‧ 國. 學. In common-law countries where corporate disclosure is the solution for. ‧. mitigating information asymmetries and accounting information is perceived to have. sit. y. Nat. higher quality with more timely loss recognition (Ball et al. [2000]). As a result,. io. er. increased information transparency and quality due to the IFRS adoption are more effective in conveying a borrower’s creditworthiness and lead arrangers’ monitoring. n. al. Ch. and due diligence effort. In contrast, firms in. engchi. v i n code-law U. countries emphasize. relationship with banks, so information asymmetries usually are resolved via insider communication. This will reduce the demand for contracting role of accounting. Hence, I hypothesize that the negative relation between the IFRS adoption, the syndicate structure, and the use of financial covenants is weaker in common-law countries than in code-law countries. As predicted, my empirical results provide supporting evidence of this prediction. Specifically, I find that lead arrangers are required to retain a relatively less proportion of syndicated loan, foreign lenders are 10. For example, extant studies argue that IFRS implementation is likely to be heterogeneous across countries (Ball [2006]); in addition, firms’ reporting incentives are shaped by markets and countries’ institutional environment, and play a key role for reporting outcomes (Ball et al. [2003], Ball [2006]). 9.

(14) relatively more likely to participate in a syndicated loan, and lenders make relatively increased use of financial covenants when a borrower in common-law country adopts mandatory IFRS.. Regarding the country-level enforcement regime, Leuz [2010] indicates that convergence of reporting practices is unlikely due to persistent enforcement differences around the world. Daske et al. [2208], Li [2010], Armstrong et al. [2010], and DeFond [2011] all find that the documented capital-market effects of IFRS. 政治大 debt market, prior study also finds that when lending to a company, lenders should 立. reporting are relatively pronounced in countries with the stronger enforcement. In the. assess not only the credit quality of the borrower but also the risk due to weak laws or. ‧ 國. 學. institutions. In addition, strong creditor rights seem to enhance loan availability as. ‧. lenders are more willing to provide credit on favorable terms, consistent with. sit. y. Nat. enforcement regime argument (Qian and Strahan [2007]). I thus also argue that the. io. er. negative effects of the IFRS adoption on the syndicate structure and the use of financial covenants are relatively weaker in countries with stronger enforcement. As. n. al. i n Csupports predicted, the empirical evidence my argument. hengchi U. v. My dissertation is closely related to Kim et al. [2011] (hereafter KTY). My paper is different from KTY in several dimensions. First, my study concentrates on the setting of mandatory IFRS, while KTY focus on voluntary IFRS. Although KTY perform several sensitivity tests to address the self-selection bias issue, it is possible that KTY can not completely rule out the possibility that the positive economic consequences from voluntary IFRS are driven by the changes in firms’ incentive and the economic environment, rather than the change in the financial reporting system. Second, I find that lead arrangers retain a larger share of syndicated loan, and foreign 10.

(15) lenders are less likely to participate in syndicate loan after mandatory IFRS adoption; however, KTY do not examine the effect of IFRS adoption on the syndicate ownership structure, but find foreign lenders are more likely to participate in syndicate loans after voluntary adoption. Third, my results further reveal that optimal choice of n-contractual terms exist in a debt contract after mandatory IFRS reporting while KTY don’t. Finally, KTY conclude positive effects of the voluntary IFRS adoption from the argument of accounting informational roles. In contrast, I argue that the adopting of mandatory IFRS can trigger economic consequences through its informational and. 政治大. contracting roles, and document a net negative effect of IFRS adopting on the syndicated loan market.. 立. ‧ 國. 學. This dissertation could make contributions to the literature in several ways. First,. ‧. I contribute to the literature on the role of the adoption of mandatory IFRS on. sit. y. Nat. syndicated loan. While prior studies on economic consequence of mandatory IFRS. io. er. (e.g., Daske et al. [2008], Li [2010], Byard et al. [2011], DeFond et al. [2011], Tan et al. [2011]) focus primarily on equity market, my paper concentrates exclusively on the. al. n. v i n economic effects of mandatoryC IFRS reporting on debt h e n g c h i U markets. Second, the extant. research on mandatory IFRS finds that the adopting of mandatory IFRS on average has a positive effect on equity market. In contrast, I find that from the contracting role of accounting, the IFRS adoption increases information asymmetries between lead arrangers and non-lead lenders and between lenders and borrowers by documenting that lead arrangers (foreign lenders) retain a larger (smaller) share of syndicated loan, and lenders decrease their use of financial covenants.. Third, I contribute to the literature on cross-border investment by documenting that foreign potential lenders are less likely to participate in a syndicated loan when a 11.

(16) borrower adopts mandatory IFRS. In contrast to prior research on home bias, which focuses mainly on equity market (Florou and Pope [2009], Bruggemann et al. [2009]), I document that foreign lenders are less likely to participate in syndicated loan deals after the introduction of mandatory IFRS reporting.. Fourth, I extend and complement prior research on debt covenants by documenting a significant relation between financial reporting quality (e.g., mandatory IFRS adoption) and debt contractual terms. Prior literature on debt. 政治大 restatement (Graham et al. [2008]) and internal control weaknesses (Costello and 立. covenants finds that accounting information quality, as measured by the incidence of. Wittenberg-Moerman [2011]), influences intensity of financial covenants used in. ‧ 國. 學. contractual terms. In this dissertation, I argue that there are two distinct perspectives. ‧. on the effects of introduction of mandatory IFRS reporting on a syndicated loan, and. sit. y. Nat. find that whether to adopt mandatory IFRS is an important determinant of intensity of. io. er. financial covenants employed in syndicated loan deals. Finally, while Watts [2003] suggests that debtholders’ information need is an important force that shapes the. al. n. v i n C my properties of accounting number, have policy implication for regulators, h efindings ngchi U. in particular US regulators.11. 2. Literature 2.1 IFRS v.s. reporting quality and comparability Proponents of IFRS adoption argue that IFRS reporting can improve financial reporting quality and enhance their comparability across firms and countries. 11. The FASB and the IASB are now working on a joint project to develop an improved common conceptual framework. However, they drop the stewardship sub objective and subsume it under the “decision usefulness”. In other words, they no longer view stewardship and valuation usefulness as alternative sub-objectives. This could “blurs an important difference in the channels by which accounting information affects economics relationships (Gassen [2008]).” 12.

(17) Literature on IFRS indicates that IFRS are capital-market oriented and are more comprehensive than most local GAAP (Ding et al. [2007], Bae et al. [2008]). Using a sample of firms from 21 countries, Barth et al. [2008] find that firms adopting IFRS exhibit less earnings smoothing, less earnings management towards a target, more timely losses recognition, and more value relevance of accounting amounts. However, the sample period of Barth et al. [2008] precedes the mandatory IFRS reporting for most sample firms and, hence, one can’t attribute their results to changes in a firm’s reporting incentive or changes in financial reporting system.. 政治大 Comparability is also the driver for demanding one set of uniform accounting 立. standards across firms and countries (Barth [2006], Leuz [2010]). 12 The reason. ‧ 國. 學. behind this argument is that the same accounting reporting can reduce information. ‧. processing costs, thus inducing cross-border investments and the integration of global. sit. y. Nat. financial markets (Armstrong et al. [2010], Covrig et al. [2007]). For cross-border. io. er. investors, the problems with information uncertainty drive their portfolios overweighting on firms domiciled in their home countries. Harmonizing accounting. al. n. v i n theseCinformation problems. h e n g c h i U The. globally helps to reduce. direct effect on IFRS. worldwide reporting is to decrease costs of acquiring foreign expertise to analyze accounting information (Barth et al. [1997]). Therefore, research finds that if the degree of conformity in accounting methods is high, foreign investors more incline to hold cross-border investment (Bradshaw et al. [2004], Covrig et al. [2007], Beneish et al. [2010], DeFond et al. [2011]).. Additionally, differences between local GAAP and IFRS affect foreign analysts 12. IASB [2008] and FASB [2008] define comparability in the Conceptual Framework Exposure Draft as “the quality of information that enables users to identify similarities and differences between sets of economic phenomena.” 13.

(18) following and their forecast accuracy (Ashbaugh and Pincus [2001], Bae et al. [2008]). De Franco et al. [2011] directly develop measures of financial statement comparability. They find that comparability leads to greater analyst following, to more accurate forecast, to a reduction in forecast optimism, and to less dispersion in earnings forecasts.. IFRS reporting can bring forth improved reporting and disclosure quality and, hence, lower information asymmetries among investors, which in turn is expected to. 政治大 [2000], Verrecchia [2001], Botosan [1997], Hail and Leuz [2006]). For example, 立. reduce cost of equity capital, and to increase market liquidity (Leuz and Verrecchia. Daske et al. (2008) examine the capital-market effects around the mandatory. ‧ 國. 學. introduction of IFRS reporting in 26 countries using a variety of proxies. They. ‧. document an increase in market liquidity, a decrease in firms’ costs of capital, and a. io. er. adopting IFRS experience reduction in costs of equity.. sit. y. Nat. corresponding increase in equity valuation. Li [2010] also finds that EU firms. al. n. v i n Byard et al. [2011], HortonCet al. [2008] and Wang h e n g c h i U et al. [2008] find that analyst. properties such as forecast accuracy, analyst following and forecast dispersion improve after the mandatory adoption of IFRS. They argue that IFRS adoption may improve analysts' information environment by enhancing disclosure and transparency or by increasing the comparability of financial reports.. 2.2 IFRS v.s. reporting incentive Research on the international accounting indicates that accounting standards per se do not determine financial reporting quality (Ball et al. [2000, 2003], Leuz et al. [2003], Hail et al. [2010a, 2010b]). The reason behind this argument is that because 14.

(19) standards give managers considerable judgment and substantial discretion, 13 the observed financial reporting information is the outcome of managers’ trade-off between costs and benefits. As a result, firms’ reporting incentives determine observed financial reporting quality (Ball et al. [2000, 2003], Leuz et al. [2003], Ball and Shivakumar [2005], Burgstahler et al. [2006]).14. Discovered factors shaping financial reporting outcome include managers’ incentives (Ball et al. [2000, 2003], Leuz et al. [2003], Ball and Shivakumar [2005],. 政治大 Burgstahler et al. [2006], Lang et al. [2006], Hail and Leuz [2007], Daske et al. 立. Burgstahler et al. [2006]), legal enforcement (Ball et al. [2003], Leuz et al. [2003],. [2008]), ownership structure (Ball and Shivakumar [2005], Burgstahler et al. [2006],. ‧ 國. 學. Fan and Wong [2002]), tax system (Guenther and Young [2000], Haw et al. [2004]),. ‧. legal origin (Ball et al. [2000]), and other institutional features of the economy. Thus,. y. sit. io. al. v i n C argument, above evidence h e n g cempirical hi U. n Consistent with the. er. countries.. Nat. the discernible effect on adoption IFRS should be heterogeneous across firms and. shows that capital. market effects around the introduction of mandatory IFRS reporting are not evenly distributed across countries. Recent governance reforms in European Union (EU) countries around the introduction of mandatory IFRS reporting provide a good laboratory to examine the role of enforcement.15 For example, Daske et al. [2008]. 13. Empirical evidence (Ding et al. [2007]) finds that IFRS reduce the amount of reporting discretion, but IFRS still like many other accounting standards provide firms with considerable judgment and substantial discretion. 14 This argument does not imply that causality exists between reporting incentives and financial reporting quality. Holthausen [2009] points out that we don’t have decisive evidence of main factors that determine financial reporting quality. 15 The EU has mandated IFRS reporting in 2005. Prior to 2005, the EU made much efforts to tighten their enforcements regimes around mandatory IFRS reporting (refer to studies by Hail and Leuz [2007], Daske et al. [2008], Armstrong et al. [2010]). 15.

(20) study the economic consequences around the introduction of mandatory IFRS reporting in 26 countries and document that the capital-market effects are more pronounced in member states of the European Union (EU), possibly reflecting its simultaneous efforts to improve enforcement. Moreover, similar results can be found in the EU sample. Li [2010] investigates if the mandatory adoption of IFRS reduces the cost of equity capital and indicates that the reduction in the cost of equity is mainly driven by firms in countries with strong legal enforcement. Armstrong et al. [2010] investigate equity market reaction to 16 events between 2002 and 2005. 政治大 domiciled in common law countries react more positively than firms domiciled in 立 associated with the adoption of IFRS in the EU. Their results indicate that firms. code law countries. DeFond et al. [2011] further document that although mandatory. ‧ 國. 學. IFRS adoption results in a large improvement in comparability, US mutual fund. 2.3 Accounting information v.s. debt contracting. al. er. io. sit. y. Nat. credible.16. ‧. ownership increases only among adopters in countries where implementation is more. n. v i n C has a direct contracting Accounting information serves role in reducing the agency engchi U. costs arising from debt contracting process (Watts and Zimmerman [1986, 1990], Ball [2001], and Holthausen and Watts [2001]). Literature attributes two important characteristics to this argument. First, before providing access to capital, lenders require reliable information to assess the borrower’s credit quality, as well as information to evaluate the timing and riskiness of the borrower’s expected future cash flows. Second, after debt contracts are signed, debt contractual terms based on accounting information (i.e., covenants, performance pricing) can be regarded as a. 16. Setting in firms cross-listed in the United States also stresses the importance of enforcement for financial reporting quality (Lang et al. [2006], Leuz [2006]). 16.

(21) warning. mechanism. protecting. the. lender’s. wealth. from the. borrower’s. value-reducing decisions. Therefore, I will review literature related to the two researches in the following two sub-sections.. 2.3.1 Roles of accounting information in mitigating information asymmetry Extant studies (Bharath et al. [2008], Beatty et al. [2008b], and Dhaliwal et al. [2010]) argue that the types of information demanded at loan inception determine the choice to obtain public or private debt. Bharath et al. [2008] use factor analysis to. 政治大 accounting quality are more likely to borrow from banks. Moreover, they also find 立. develop a comprehensive accounting quality measure and find that firms with lower. that public debts are more sensitive to accounting quality than private debts. These. ‧ 國. 學. findings are consistent with either private lenders’ access to proprietary information. ‧. beyond accounting reports, or private lenders’ ability to monitor the borrower through. sit. y. Nat. covenants. Beatty et al. [2008b] further examine whether the borrower’s accounting. io. er. quality affects the decision to lease versus to use equity or debt financing. Dhaliwal et al. [2010] investigate whether disclosure costs influence the choices of public versus. al. n. v i n private debts. Their findings areCgenerally consistent U h e n g c h i with the results of Bharath et al. [2008].. Compared to the aforementioned papers, which focus mainly on accounting information for public debts, recent studies examine the role of information and accounting quality in mitigating information asymmetry among lenders in the syndicated loan (Dennis and Mullineaux [2000], Sufi [2007], Ball et al. [2008]). Sufi [2007] argue that because lead arrangers ex ante possess private information about the borrower, participant lenders expect their active monitor on the borrower, which, in turn, induces the lead bank to retain a large share of the loan and forms a more 17.

(22) concentrated syndicate. Ball et al. [2008] extend the evidence in Sufi [2007] and develop a new debt-specific measure of accounting quality, i.e., the debt-contracting value (DCV) measure.17 They find that accounting information (i.e. DCV) conveys more information about a borrower’s credit quality and, hence, mitigate the potential adverse selection and moral hazard problems within the syndicate. This leads to a lower proportion of the loan retained by the lead arranger.. Wittenberg-Moreman [2008] utilizes the setting of secondary syndicated loan. 政治大 between investors. She finds that timely loss recognition is negatively associated the 立 market to examine the effect of accounting quality on information asymmetries. size of bid-ask spreads. Her findings are consistent with Watts [2003] and Ball [2001]. ‧ 國. 學. which argue that lenders assess a potential loan and certain the borrower’s. er. io. sit. y. Nat. assets.. ‧. creditworthiness by knowing the borrower’s verifiable lower bound measures of net. 2.3.2 Roles of accounting information in the design of the debt contract. al. n. v i n Debt agreements consist ofCdifferent aspects of contractual terms such as interest hengchi U. rates, maturities, performance-pricing provisions,18 loan sizes, financial covenants,. dividend restrictions, and income escalators.19 Prior accounting studies focus mainly on how the quality of a firm’s financial reporting is related to the cost of debt. These studies often use the credit rating, interest rates, or the ratio of interest expense to interest bearing debt as a proxy for costs of debt. Ahmed et al. [2002] and Zhang 17. The DCV measure captures the ex ante information known by the lead lender about the credit quality of the borrower at the inception of the loan. DCV is derived by estimating a model of credit rating downgrades as a function of lagged, seasonally adjusted changes in accounting earnings. 18 A typical performance-pricing provision defines that interest rates relate to borrowers’ specific performance measure (e.g., Debt/EBITDA or leverage). As the borrowers’ specific performance measure changes, the interest rates reset based on the terms of the pricing grid. 19 Income escalators define as systematic adjustments to net worth covenants that exclude a percentage of positive net income from covenant calculations (Beatty et al. [2008a]). 18.

(23) [2008] find that firms that report more conservatively have lower costs of debt. Bharath et al. [2008], Ashbaugh et al. [2006], and Francis et al. [2005] use various measures of accounting quality and conclude the same findings. However, there is no direct evidence on why debt holders value high quality accounting information (Armstrong et al. [2010]).. Most literature on conservatism indicates that accounting conservatism plays a vital role during debt contracting. Watts [2003] and Ball [2001] argue lenders assess a. 政治大 verifiable lower bound measures of net assets. Moreover, conservatism is affected by 立 potential loan and certain the borrower’s creditworthiness by knowing the borrower’s. the threat of litigation (Basu [1997]). Besides, conservative reporting constrains. ‧ 國. 學. managers’ opportunistic behavior. Empirical literature finds that conservative. ‧. reporting leads to lower costs of debt (Ahmed et al. [2002] and Zhang [2208]), to an. io. er. asymmetries (Wittenberg-Moerman [2008]).. al. v i n includingCfinancial covenants,Ucan inform hengchi n. Debt contracts,. sit. y. Nat. early transfer of control rights (Zhang [2008]), and to the decreased information. lenders an early. warning signal of borrowers’ deteriorations in credit risk by triggering violations of debt covenants in a timely manner (Dichev and Skinner [2002]). Using a sample of public debt contracts, Nikolaev [2010] finds a positive association between timely loss recognition and the use of debt covenants. Moreover, he also finds that the likelihood of using debt covenants is affected by firms’ accounting quality. Graham et al. [2008] indicate that loans initiated after restatement have more covenant restrictions, while Costello and Wittenberg-Moerman [2011] find that when a firm experience a material internal control weakness, lenders are less likely to use financial covenant in debt agreements. 19.

(24) Another stream of literature investigates how debt contracts adjust accounting numbers to satisfy the debt contracting demand. Li [2010] finds that contractual definitions of net income are different from GAAP when the loan maturity is longer and when transitory earnings are less useful for debt contacting; however, conservative adjustment does not seem to be a major concern in measuring net income or net worth. Beatty et al. [2008a] examine the use of income escalators and tangible net worth covenants in debt contracts and argue that these modifications make. 政治大 more conservative reports are more likely to have conservative modifications (e.g., 立 contract calculations more conservative. They further find that firms that provide. income escalators) in their debt covenants. They suggest that covenant modifications. ‧ 國. 學. are unlikely to fulfill the lender’s demands for conservative financial statements.. ‧. Frankel et al. [2208] investigate that whether changes in accounting rules for goodwill. sit. y. Nat. (i.e., SFAS 141 and SFAS 142) lead to covenant modifications. They find that firms. io. er. with goodwill on the balance sheet are less likely to have covenant modifications and conclude that creditors view a borrower’s intangible assets as an important source of. al. n. v i n C h health and as itsU ability to pay future principal information about the firm’s financial engchi and interest on the debt.. 3. Background and Hypothesis Development 3.1 Institutional Background Syndicated loans are an increasingly significant source of firm finance and have recently lead to more underwriting revenue than either the equity or the bond market (Altunbas et al. [2006]; Ball et al. [2008]). Syndicated loans are credits granted by more than one financial institution to a borrower. In the syndication process, the syndicate members can be divided into two groups, namely, lead arrangers and 20.

(25) participant lenders. Once the lead arranger signs a mandate20 with a borrower, the syndication process begins. The lead arrangers recruit potential participant lenders to fund part of the loan. They provides information memo (IM) to potential buyers, summarizing the executive summary, investment considerations, a list of terms and conditions, an industry overview, and a financial model. After a syndicated loan agreement is signed by all parties, the lead bank is responsible for monitoring the borrower and deals with necessary amendments which include a covenant waiver or a change in the collateral (Standard & Poor’s [2007], Dennis and Mullineaux [2000], Lee and Mullineaux [2004]).. 立. 政治大. Thus, information asymmetries and the need for monitoring lead to an adverse. ‧ 國. 學. selection problem and a moral hazard problem (Sufi [2007], Ball et al. [2008],. ‧. Ivashina [2009]). Prior to the loan syndicated, because the lead arranger is able to. sit. y. Nat. collect information and establish a relationship with the borrower in the first place, the. io. er. lead bank has an incentive to retain loans of higher quality. Post the loan syndicated, due to the reduced portion of loans retained by the lead arranger, the lead arranger’s. n. al. unobservable effort causes. v i n theCmoral hazard problem. h e n g c h i U Literature. argues that the. information and monitoring need can result in the above moral hazard problems; monitoring by multiple creditors, however, could result in costly duplication of monitoring efforts and inefficient free-riding (Leland and Pyle [1977], Diamond [1984]). For this reason, creditors want to delegate monitoring to one financial intermediary and thus the delegated monitor role of the lead arranger is derived (Sufi [2007]).. 20. A mandate is a preliminary loan agreement which outlines the lead’s syndication strategy, the specification of debt covenant, fees, and collateral as well as the loan amount and a range for the interest rate. 21.

(26) However, if participant lenders delegate direct monitoring to lead arrangers, lead arrangers are less likely to credibly commit to doing proper due diligence because they no longer invest their own money. If participant lenders want to avoid shirking by the lead arranger, a safer way is to hold less percentage of the loan. In this article, I posit that publicly available accounting information not only resolves information asymmetries among contracting parties (e.g., the lead arranger, the participant banker, and the borrower) but also mitigating moral hazard problem between the lead arranger and the participant banker. The main reason is that the debt contracting role of. 政治大. accounting information can be a mechanism of reducing the cost of contracting and monitoring.. 學. ‧ 國. 立. 3.2 Hypotheses Development. ‧. In section 3.2.1~3.2.2 I discuss the competing views on whether and how. sit. y. Nat. mandatory IFRS reporting affects the loan syndicate structure and uses of financial. io. er. covenants in debt contracts. Then, I develop three hypotheses pertaining to how (1) the loan retained by syndicate lead arrangers, (2) the presence of a foreign lender in a. al. n. v i n usesCof financial covenants h e n g c h i U in debt. syndicated loan, and (3). contracts might be. associated with a borrower’s adoption of mandatory IFRS reporting. Finally, in section 3.2.3 I develop how the institutional effects around the introduction of mandatory IFRS reporting on the syndicated loan market.. 3.2.1 Mandatory IFRS reporting versus syndicate structure The informational effect of IFRS reporting on syndicate structure Extant literature posits that mandatory IFRS adoption influences financial reporting quality and transparency (Daske et al. [2008], Hail et al. [2010a, 2010b], Armstrong et al. [2010], Li [2010]). In this dissertation, I first posit that IFRS 22.

(27) adoption may play an important role in the loan syndication process by altering information asymmetry between lead arrangers and other participant lenders.. As mentioned earlier, in the context of a syndicate loan, there exit information asymmetries between a borrower and lenders as well as among lenders themselves. Prior to the signing of syndicate loan contact, potential participant lenders require the lead arranger to retain a larger proportion of loans to mitigate the information asymmetries between lead arrangers and other participant lenders and the associated. 政治大 the signing of loan contract, the lead arrangers also are required to hold a larger share 立 adverse selection problems inherent in a syndicated loan. In addition, subsequent to. of loans to ameliorate potential moral hazard problems.. ‧ 國. 學 ‧. In this section, I address one question as to whether and how the adoption of. sit. y. Nat. mandatory IFRS affects syndicate loan structure. As mentioned above, there are two. io. er. distinct perspectives about this question. From the information perspective, extant research reveals that IFRS adoption can enhance financial reporting quality and. al. n. v i n C h accounting numbers transparency. In this case, high-quality enable potential lenders to engchi U more accurately assess the borrower’s credit quality independently of the lead. arranges and in turn reduce the information asymmetry between the lead arranger and participant lenders. As a result, I can argue that the lead arrangers’ are required to hold a smaller share of syndicate loan after a borrower adopts IFRS reporting.. Similarly, high-quality IFRS also can lower the cost associated with post-contract due diligence and monitoring efforts of the lead arrangers, but also facilitate other participant lenders monitoring the lead arrangers’ contractual duties. Thus, IFRS adoption could help mitigate the moral hazard problem in context of syndicated loan. 23.

(28) Again, this enables the lead arrangers to hold a smaller proportion of syndicate loan.. Besides, in the area of cross-border capital flows, the home-bias literature posits that familiarity with the accounting standards matters for portfolio holdings. Another argument in favor of IFRS adoption is that greater comparability due to IFRS adoption makes it less costly for foreign investors to compare firms across market and countries (Covrig et al. [2007], Byard et al. [2011], Tan et al. [2011], DeFond et al. [2011]). As a result, the global movement towards IFRS reporting may facilitate. 政治大 syndicate loan, greater comparability due to the IFRS adoption makes it is easier for 立. cross-board investment and the integration of capital markets. In the context of. foreign lenders to compare firms among countries, and thus, ameliorate the. ‧ 國. 學. information asymmetry with respect to the borrowers’ credit quality. This would. ‧. enhance the foreign lenders’ incentive to participate in a syndicate loan.. sit. y. Nat. io. er. Overall, from the information perspective, I expect that the adoption of IFRS leads to a lower proportion of loan retained by syndicated lead arranges, and a higher. al. n. v i n C hinvolved in syndicated likelihood of a foreign lender being loan. engchi U. The contracting effect of IFRS reporting on syndicate structure On the other hand, from the contracting perspective, one may argue that IFRS adoption would not mitigate the information asymmetry and the associated adverse selection and moral hazard problems inherent in syndicated loan. First, the current allegedly “principles-based” standards (i.e., IFRS) are characterized by less scope exceptions, less treatment exceptions, and the lack of (or less) detailed implementation guidance (Schipper [2003], Maines et al. [2003]). Second, IFRS has increasingly called for use of fair value in the financial statements. These judgment 24.

(29) and estimates have the potential to limit the contracting usefulness of accounting information by introducing bias and error into reported numbers. Thus, one may argue that potentially participant lenders may be faced with more severe information asymmetry with respect to borrowers’ credit quality and due diligence and monitoring efforts of the lead arrangers. In this case, participant lenders may factor these into the initial design and subsequent renegotiation of the loan. This will require lead arrangers to retain a higher proportion of syndicate loans.. 政治大 few scope and treatment exceptions that are intended to reduce volatility (smooth 立. In addition, to the extent that adoption of principles-based standards implied very. income), one effect would be increased volatility in reported earnings (Schipper. ‧ 國. 學. [2003]). The extant empirical studies provide supporting evidence on increased. ‧. income volatility in countries adopting IFRS (Barth et al. [2008]). In such a case, the. sit. y. Nat. increased contract costs, including costs associated with writing, negotiating, and. io. er. enforcing debt contract, due to IFRS adoption may not mitigate, even exacerbate, the information asymmetry between lead arrangers and participant lenders and the. al. n. v i n Cmoral associated adverse selection and problems. It is therefore possible that U h e nhazard i h gc. potential participant lenders would require lead arranges to retain a higher proportion of syndicated loan.. Another important dimension of IFRS reporting is its greater comparability among firms. However, it is possible that greater comparability due to principle-based system (including IFRS) would be what Schipper [2003] calls “surface comparability”, rather than “real comparability”. Specifically, under rule-based system (e.g., U.S. GAAP), specific guidance on how to apply a standard reduces the effects of the differences in professional judgment, and the resulting comparability 25.

(30) would be more effective. In contrast, under the more principle-based system (including IFRS), the same accounting treatment may lead to dissimilar arrangements due to more judgment and less guidance, thereby make financial reporting less informative. In such a case, foreign lenders will be more reluctant to participate in a syndicate loan due to higher cost of information about foreign investments.. Given the two distinct perspectives on how mandatory IFRS adoption influences syndicate ownership structure, my first and second hypotheses are stated in null form:. 立. 政治大. H1: The proportion of the loan retained by syndicate lead arrangers has an. ‧ 國. 學. association with a borrower adopting mandatory IFRS reporting.. ‧. H2: Whether a foreign lender is involved in a syndicated loan deal has an association with a borrower adopting mandatory IFRS reporting.. sit. y. Nat. n. al. er. io. 3.2.2 Mandatory IFRS reporting versus debt contractual terms. i n U. v. Syndicate participants can assess whether the lead arranger is misleading them. Ch. engchi. about true quality of borrowers via accounting reports (Ball et al. [2008]). To the extent that lenders can relate debt contractual terms to accounting reports in their debt agreements, they can mitigate ex post moral hazard problems arising from expropriation by managers and shareholders (e.g., well-known agency problems such as claim dilution (Smith and Warner [1979]), asset substitution (Jensen and Meckling [1976]), debt overhang (Myers [1977]).. In this section, I focus my analyses on financial covenants in debt contracts to examine the effect of IFRS on syndicated loan deals for the following reasons. First, 26.

(31) covenants can be viewed as an early warning signal to handle borrowers’ post deteriorations in credit risk (Dichev and Skinner [2002]). Second, many covenants using to bind managers’ opportunistic actions are accounting-based (the evidence in Nikolaev’s [2010] research sample can be found). Finally, higher quality of financial reporting limits managers’ ability to make accounting choice to avoid the likelihood that their firm violates accounting-based debt covenants. The setting of mandatory IFRS reporting provides an opportunity to examine whether changes of accounting standards are an ex ante important factor in the debt contracting process.. 治政 The informational effect of IFRS reporting on debt contractual 大 terms Prior studies indicate立 that higher quality of financial reporting can mitigate ‧ 國. 學. agency costs and that borrowers, in turn, face lower cost of debt charged on the loan (Ahmed et al. [2202], Ashbaugh et al. [2006], Ball et al. [2008], Bharath et al. [2008],. ‧. Zhang [2008]). These studies highlight that higher accounting reports mitigate moral. y. Nat. n. al. er. io. value.. sit. hazard problems, which makes it possible for lenders to efficiently monitor their debt. Ch. engchi. i n U. v. In the context of my dissertation, as mentioned earlier, IFRS can increase firms accounting quality by restricting earnings management, requiring more timely loss recognition, and increasing higher value relevance disclosures (Barth et al. [2008]). The will limit management’s opportunistic discretion in manipulating accounting numbers to satisfy requirements of financial debt covenants. According to this argument, I thus expect that lenders are more likely to use financial covenant in debt agreements for post-IFRS period than for prior-IFRS period.. The contracting effect of mandatory IFRS reporting on debt contractual terms 27.

(32) As described above, the adoption of a principles-based accounting system (e.g., IFRS), characterized by limited interpretation and implementation guidance, will increase the difference in professional judgment among practitioners about accounting recognition and measurement. In addition, anecdotal evidence argues that less specificity in accounting standards would lead to an increase in manipulation of financial results (e.g., Intel and former FASB member David Mosso). In the syndicated loan setting, these arguments imply that the IFRS adoption will result in increased disagreements among debt contracting parties. This will increase. 政治大 borrowers’ demand for accounting information in signing debt contracts. 立. renegotiation cost between lenders and borrowers, which in turn reduces lenders’ and. ‧ 國. 學. Smith [1993] indicates that private lenders are to set debt constraints just below. ‧. the actual current value and the lenders could immediately recall their loan by. sit. y. Nat. covenant violations. However, increased judgment and the heavy use of fair value in. io. er. financial reporting by IFRS will provide unreliable signals of the borrower’s liquidation value, and reduce the contracting usefulness of accounting numbers in. al. n. v i n debt contracts (Holthausen andCWatts [2001], Kothari h e n g c h i U et al. [2010]). For example,. Demerjian [2011] finds that a shift by U.S. standards setters towards the “balance sheet approach” reduces the usefulness of balance sheet numbers for contracting. Watts [2006] also points to recent changes accounting standards-especially an increased emphasis on firm valuation-to argue that GAAP-based numbers have become less useful for contracting. Therefore, debt contracting parties will in turn make decreased use of the balance sheet-based financial covenants.. Besides, I argue that debt contracting parties would also make decreased use of the earnings-based financial covenants. This is because, as mentioned earlier, the 28.

(33) higher income volatility could lead to the likelihood of violating debt covenants based on income numbers and in turn increase renegotiation cost between lenders and borrowers. For example, renegotiation costs typically include an increase in interest rates, restrictions on the firms’ ability to access credit markets, a reduction in borrowers’ access to lines of credit and so on (Sufi [2009], Roberts and Sufi [2009]).. I also expect that the increased income volatility due to the IFRS adoption also will increase renegotiation cost between lenders and borrowers, thereby reducing the. 政治大 contracts. As a result, I posit that, from the perspective of contracting role, the IFRS 立. incentives for lenders and borrowers to use accounting information in signing debt. adoption will result in lender making decreased use of financial covenants in. ‧ 國. 學. syndicated loan market. ‧. sit. y. Nat. Overall, the information perspective is in favor of the positive relation between. io. er. the IFRS adoption and the use of financial covenants, while contracting perspective provides the opposite arguments. As a result, I cannot predict how the adoption of. al. n. v i n Cfinancial mandatory IFRS reporting affects in a syndicated loan market. My U h e n gcovenants i h c third hypothesis is thus stated in null form:. H3: Whether syndicate lenders use financial covenants based on accounting numbers in debt agreements is associated with a borrower adopting mandatory IFRS reporting.. 3.2.3 The effect of institutional environment In this section, I examine whether the effects of mandatory IFRS reporting on syndicated loan market, as discussed in H1 through 3, are conditional on certain 29.

(34) institutional factors. The arguments are motivated by recent studies that reporting incentives are shaped by many institutional factors (Ball et al. [2003], Hail et al. [2010a, 2010b]).21 In my dissertation, I focus on the following two institutional factors: (1) a country’s legal original, and (2) the strength of enforcement regime.. The first question I address is whether a country’s legal original affects the association between IFRS adoption and the function of syndicated loan market. It is well-documented that firms in code-law countries tend to resolve information. 政治大 result, one can argue in the debt market setting that, compared to code-law countries, 立. asymmetries via private channel rather than public disclosure (Ball et al. [2000]). As a. accounting income in common-law countries is thus more timely in incorporating. ‧ 國. 學. economic losses that enable lenders to efficiently monitor their financial claims.. ‧. Following the same logic, in the context of syndicated loan, information asymmetries. sit. y. Nat. among lenders and between borrowers and lenders will decrease more in common law. io. er. countries than civil law countries. Besides, the presence of common-law penalties to false signaling, coupled with the fact that International Accounting Standard Board. n. al. Ch. (IASB) adopts a more principles-based approach. engchi. v i n to U develop. IFRS, leads to our. prediction that IFRS adopters in the common-law countries can effectively signal better quality to lenders than those in code law countries, and the negative effects of the adoptions of mandatory IFRS on the syndicated debt could be reduced more in common law countries (Ball et al. [2000]).. Thus, I hypothesize that the negative effects of the adoptions for mandatory IFRS on the syndicated debt market are weaker in common law countries than civil 21. These factors include a country’s legal institutions, the strength of enforcement regime, capital market forces, product market competition, a firm’s compensation, ownership and governance structure and firms’ operating characteristics (refer to Hail et al. [2010a, 2010b]). 30.

(35) law countries.. The second institutional factor I address is the strength of enforcement regime. The motivation for this analysis stems from extant studies documenting that the strength of the enforcement regimes that shapes managers’ reporting incentives is likely to be the primary drivers for the development of standards (Hail et al. [2010a, 2010b], Holthausen [2009], Leuz [2010]). The extant studies find that that the equity market effects of IFRS are relatively more pronounced in countries with strict. 政治大 reporting incentives to firms (Daske et al. [2008], Li [2010], Armstrong et al. [2010], 立 enforcement regimes and in countries where institutional background provides strong. Byard et al. [2011], and DeFond et al. [2011]). The desire for enforcement need is. ‧ 國. 學. more important in a debt market. Cross countries evidences show that the degree of. ‧. creditor protection affects the design of bank loans (Miller and Reisel [2009]). Prior. sit. y. Nat. studies (e.g., Qian and Strahan [2007]) indicate that loans tend to have lower interest. io. er. rates and have longer maturities for firms domiciled in countries with stronger creditor protection. Thus, due to the fact that accounting information plays an. al. n. v i n C hforeign creditors seem important role on contracting need, particularly sensitive to the engchi U. interaction between the strength of enforcement and the accounting quality of mandatory IFRS adoption in borrowers’ environment. Therefore, these arguments yield the following hypotheses:. H4a: The negative effect of the adoption of mandatory IFRS on the syndicated-debt market is less pronounced in common-law countries than in code-law countries. H4b: The negative effect of the adoption of mandatory IFRS on the syndicated-debt market is less pronounced in countries with strong legal enforcement than in 31.

(36) countries with weak legal enforcement.. 4. Research Design 4.1 Empirical model for H1 I begin my first set of analyses about the effect of mandatory IFRS reporting on syndicated loan structures using a difference-in-difference approach. This method is an efficient way to investigate the change in syndicated loan structures in the preversus post-mandatory IFRS reporting period for mandatory adopters and to control possible changes for the IFRS adopters by concurrent changes that non-IFRS adopters also experience.. 立. 政治大. ‧ 國. 學. I proceed in four steps. First, I create two binary indicator variables, Mandatory_Adopter and Post_mandatory, to investigate H1. The first variable,. ‧. Mandatory_Adopter, classifies reporter firm-year observations into “IFRS adopters”. Nat. sit. y. and “non-IFRS adopters,” and takes the value of one for firms that adopted IFRS in. n. al. er. io. countries requiring mandatory IFRS reporting. The second variable, Post_mandatory,. i n U. v. captures whether the firm-year observation falls in the post-mandatory adoption. Ch. engchi. period, and takes on the value of one if firm-year ends on year 2005 or later. Thus, I can define one interaction term, Mandatory_Adopter*Post_mandatory, to capture any incremental effects for borrowers once they mandatorily adopt IFRS relative to the change for non-IFRS borrowers over the same time period. Besides, treatment firms needs to have available deal data for both pre- and post-IFRS adoption date.. Second, I use the dependent variable, Proportion_lead, to investigate whether IFRS reporting borrowers can improve the syndicated debt structure. This variable is measured by the percentage of the loan retained by the lead arranger. Prior studies 32.

(37) have proven that this construct is affected by the quality of information on borrowing firms (Dennis and Mullineaux [2000], Lee and Mullineaux [2004], Jones et al. [2005], Sufi [2007], Ball et al. [2008]). Besides, I also use two alternative constructs to measure the syndicated debt structure: the total number of lenders in the loan syndicate (Num_lead) and a Herfindahal index (Herfindahal_index). Following Sufi [2007], I calculate Herfindahal index by using each syndicate member’s share in the loan; it is sum of the squared individual shares in the loan (this index’ value varies from 0 to 10,000).. 政治大 Finally, my empirical model includes both loan-specific and borrower-specific 立. control variables. Deal-specific control variables are described as follows: (1) Size of. ‧ 國. 學. deal (Log(Loan_Size)): defined as the nature log of the total dollar value of the deal. ‧. converted to US dollar. (2) Loan maturity (Loan_maturity): defines as the number of. sit. y. Nat. years to maturity of the loan. (3) Loan type (Term, Revolver): I use two dummy. io. er. variables to control each of the following loan type: term loans, lines of credit, and all other types of loan (omitted group).. n. al. Ch. engchi. i n U. v. Borrower-specific control variables are described as follows: (1) Firm size (Log(Firm_Size)): measured as the natural log of the book value of the total assets. Ball et al. [2008] argue that lead arrangers could retain less share of the deal when the borrowing firm is larger. (2) Leverage (LEV): measured as the ratio of the borrower’s long-term debt to total assets. (3) Return on assets (ROA): measured as earnings before interests and taxes (EBIT) scaled by average total assets at the time of the deal.. To control for the prior lender-borrower relationship, all empirical models include the variable, Prior_lead, which is a dummy variable taking the value of one if 33.

(38) at least if at least one of the loan’s lead arrangers had been a lead arranger of the borrower’s previous loans over the five-year period preceding the loan’s issuance date, and zero otherwise. Lead arrangers who have previous lending relationships with the borrower could retain a lower proportion of the syndicated loan ownership because the lead bank’s monitoring of the same borrower becomes relatively easier and less costly. Beside, to control for the relationship between lead banks and participant banks, I include Syndicate_relationship in all empirical models: for every syndicate participant, the number of previous relationships between the lead arranger and the. 政治大 estimation is performed over a five-year period preceding the loan’s issuance, and this 立. participant is deflated by the total number of deals syndicated by the arranger (the. measurement refers to Costello and Wittenberg-Moerman [2011]).. ‧ 國. 學 ‧. In addition to control variables for firm characteristics, I also include dummy. sit. y. Nat. variable for firms that trade an ADR in US (US_listing). This empirical model. io. er. includes year-fixed effects to capture common effects on dependent variables in a particular year. The extent of economic development is used to control for country. n. al. Ch. effects and is measured as the natural log of. engchi. v i n the U GDP. per capita in year t. (Log(GDP_per_capita)). Besides, all research models performed in my dissertation are used to examine the lagged relation between a debt related variable in year t and IFRS adoption in year t-1. In this way, the endogeneity concern should be reduced. Appendix A provides specific definitions of each variable.. I combine these variables into the following regression estimated at the facility level (or the tranche level):. Syndicate_Structure=β0＋β1 Mandatory_Adopter＋β2 Post_mandatory 34.