探討犯罪公司定罪前後銀行貸款利率之差異 - 政大學術集成

全文

(1)國立政治大學財務管理學(系)研究所碩士學位論文. 探討犯罪公司定罪前後銀行貸款利率之差異 The Impact of Corporate Convictions on Syndicated Bank Loan. 立. 政Prices治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 指導教授：張元晨博士研究生：黃聖雯撰. 中華民國一 ○ 四年七月. i.

(2) 誌謝辭時光匆匆，轉眼間碩士學習階段也來到尾聲，眼下就要與生活了六年的政大告別，心中感觸良多。猶記得兩年前剛推甄上財管所的自己，因大學時期授課重點的差異，對未知的研究所生活是既期待，又怕受傷害，而實際一路走過後，我並不諱言研究所的課程著實富有相當的挑戰性，然而正是要迎接挑戰、正視挫折，我們才得以擁有自我突破的勇氣，擁抱進一步成長的機會，就讀碩士這兩年，說長不長，卻仍帶來滿滿意想不到且無可取代的收穫，如今回首，我真的很慶幸當時的自己選擇進入財管所進修。今日得以順利完成此篇碩士論文，要感謝的人實. 政治大首先，我要感謝我的指導教授張元晨老師悉心的指導與包容，當撰寫論文立. 在太多，此刻心中的喜悅與感恩難以言喻。. ‧ 國. 學. 遇到瓶頸時，老師的建議與指點總是能我感到豁然開朗，一路走來，對於老師的用心與耐心我由衷懷抱萬分感謝；也要感謝口試委員蔡湘萍老師與黃柏凱老. ‧. 師的寶貴意見與指教，使本論文得以更加完備、充實；此外，亦特別感謝之寧. sit. y. Nat. 與依婷兩位學姊，從統計軟體的使用到資料的整理與分析，兩位學姊總是願意. al. er. io. 耐心教導我為我解惑，於此獻上我最誠摯的謝意。也謝謝同窗芝榮在本人準備. v. n. 求職及撰寫論文之時，總是慷慨分享所識、耐心指教，亦謝謝文萱與姿儀耐心. Ch. engchi. i n U. 協助我熟悉統計軟體的語法運用與捕捉原始公司資料以及晧雯與光耀的各類線上即時解惑。另外，也謝謝財管所各位優秀的同學，從你們的身上我真的學到很多，與各位一起出遊、夜聊、吃吃喝喝、運動以及無數在研究室共度歡笑的日子都會是我所珍惜的回憶，兩年的碩士生涯因為有你們更顯可貴。最後，僅以本論文獻給我最親愛的家人，一路走來你們的支持與關懷是對我最大的鼓勵與依靠，謝謝你們的包容讓我無後顧之憂完成學業。黃聖雯謹識於國立政治大學財務管理研究所中華民國一百零四年七月 ii.

(3) 摘要近年來知名公司面臨起訴或遭到定罪的消息層出不窮，然而在公司定罪的相關研究中，多數文獻著重於探究遭受定罪之事實對公司價值的影響，較少研究聚焦於定罪事實對公司所參與之聯貸案契約條件的影響，有鑑於此，本篇文章旨在探討犯罪公司於定罪前後其聯貸借款條件之借款利率是否有顯著差異。本篇研究發現事件前後犯罪公司之借款利率差異並不顯著。當我們試以修正資料庫登錄與犯罪新聞時間差的問題後，結果顯示犯罪公司之借款利率於犯罪相關新聞發布後顯著上升，除此之外，考量交乘作用後發現犯罪公司間，犯罪罪. 政治大類公司更傾向面臨較高的借貸成本。立. 刑為詐欺類以及受判罰金較高的犯罪公司於事件後，相較於其他犯罪公司，該. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. iii. i n U. v.

(4) Abstract A number of well-known corporate prosecutions have been raised in recent years. Previous literature focuses more on the valuation impact of corporate convictions. Empirical evidence on how corporate convictions affect the contract terms of syndicated loan is sparse. We examine how corporate misconducts affect the pricing of bank debt after corporate convictions. We find that the difference of loan spreads between facilities initiated before and after conviction are insignificant. Taking potential lag of initial announcement time into consideration, we find significant evidence that the effect of corporate conviction is positively associated with the. 政治大. increases of loan spreads. Moreover, interaction terms with conviction variable show. 立. that firms with large fines or convicted of fraud-related crimes tend to face higher loan. ‧ 國. 學. spread after convictions.. ‧. Keywords: Corporate Conviction, Corporate Misconduct, Loan Spread, Loan Contract, Database Challenges. n. er. io. sit. y. Nat. al. Ch. engchi. iv. i n U. v.

(5) Content 1. Introduction……………………………………………………………..…….......1 2. Literature Review………………………………………………………..………..6 2.1 Corporate Misconduct………………………………………………………...6 2.2 How Banks Manage Risks Using Loan Contract………………….…….......10 2.3 Hypothesis………………………………………………………….………..11 3. Data and Methodology…………………...……..………….……………..……..13 3.1 Sample Selection……………………………………………………...……..13 3.2 Methodology………………………………………………………...………17 4. Empirical Analysis…………………………………………………….……..….19 4.1 Effect of Corporate Conviction on the Cost of Bank Debt……….…………19 4.2 Specific Conviction Terms on the Cost of Bank Debt….………….………..21 4.2.1 The Impact of Legal Penalty－Fined or not……………………………21 4.2.2 Compliance Program……………………………………………....……22 4.2.3 State of Registry…………………………………………….…………..22 4.2.4 Conviction Type and the cost of Bank Debt………………….……..…..23 5. The Announcement Effects………………..………………...…… ... … ..25 5.1 Effect of Corporate Conviction on the Cost of Bank Debt……………….…26 5.2 Specific Conviction Terms on the Cost of Bank Debt………………..……..27 5.2.1 The Impact of Legal Penalty－Fined or not…………………. .……….27. 立. 政治大. ‧ 國. 學. ‧. 5.2.2 Compliance Program………………………………………....……..…..28 5.2.3 State of Registry…………………………………………………….…..28. sit. y. Nat. 5.2.4 Conviction Type and the cost of Bank Debt…………………..….……..28 5.3 Suggestions on Further Studies……………………………….….……..29. n. al. er. io. 6. Conclusion…………………………………………………………………31 Table 1 Summary Statistics……………………………..…………….….….……..33 Table 2 Summary Statistics (Pre/Post-conviction loans)…………………………33 Table 3 Regression Results (Conviction, Firm Characteristics and Features of Loans) ...………..34 Table 4 Regression Results (Characteristics of Convictions, Firms and Loans) ………………..35 Table 5 Summary Statistics (Adjustment on release time of news)………….……...36 Table 6 Summary Statistics (Pre/Post-conviction loans)…………………………36 Table 7 Regression Results (Conviction, Firm Characteristics and Loan Features) ….......……..37. Ch. engchi. i n U. v. Table 8 Regression Result (Characteristics of Convictions, Firms and Loans) ………………....38 Table 9 Regression Results (Conviction, Firm Characteristics and Loan Features) ….......……..39 Table 10 Regression Result (Characteristics of Convictions, Firms and Loans) ………………..40 References....………………………………………………………………...……..41 Appendix A…………………………………………………………………...……43 Appendix B…………………………………………………………………...……44 Appendix C…………………………………………………………………...……46 Appendix D…………………………………………………………………...……48 v.

(6) The Impact of Corporate Convictions on Syndicated Bank Loan Prices. 1. Introduction Over the past two decades, with the business expansion of corporations and the maturity of global financial market, the market for syndicated loan has experienced rapid growth. As one of the largest sources of worldwide corporate financing, the. 政治大 institutions under the structure 立 of syndicated loan . It has also inspired large varieties. syndicated loan market helped facilitate the use of capital by lining up with financial 1. ‧ 國. 學. of studies on the effective contracting tool in response to specific conditions or changing environment. The specific conditions including adjustment of borrower’s. ‧. credit worthiness have been extensively reviewed in previous literature. Researchers. sit. y. Nat. also examine how bank reflect borrower’s risk and cost through loan contracting.. n. al. er. io. Graham, Li and Qiu (2008) study the effect of financial restatement on bank loan. i n U. v. contracting, and they found that delinquency and negligence of firms can imply the. Ch. engchi. inaccuracy in the information previously known to the lending bank; therefore, prior beliefs concerning loan risk often need to be reevaluated. That is, a financial restatement, either as result of fraudulent intentions or incautious action, creates uncertainty about the credibility of financial statements, hence, increases the firm’s perceived informational asymmetry from the bank’s perspective. It is the first paper focusing on the cause and effect between corporate misreporting and bank loan contracting.. 1. In a syndicated loan, two or more lenders jointly offer funds to a borrower where the roles of lenders can be further divided into two groups: lead arrangers and participants. The lead arrangers retain part of the loan while, at the same time, act as the agents for the lending syndicate. Lead arrangers are responsible for both ex-ante due diligence and screening and the ex-post monitoring against borrowers. On the other hand, the participant lenders rely on the information provided by lead arrangers, including their assessment of the borrower’s credibility. 1.

(7) However, as financial restatement gains more attention in the research of syndicated loan, loan-contracting issues concerning other corporate misconducts or delinquencies has rarely been explored. In recent years, a number of well-known corporations had been prosecuted as a result of violating US domestic regulations or International Standards. Among those convicted, many have borrowed from syndicated loan markets. For instance, during 2008, the German multinational industrial firm, Siemens AG, was prosecuted for its violation of Foreign Corrupt Practices Act (FCPA) involving payments of $1.4 billion. 政治大 list of those accused, during year 2008 to 2010, other prominent multinational 立 in bribes to government officials in sixty-five countries. Siemens was not alone in the. corporations, such as Japanese electronics giant, Panasonic Co., Sweden polymer-. ‧ 國. 學. engineering whale, Trelleborg Industries, and the world's largest LCD panel maker, LG. ‧. Display Co.(Korea based) were also prosecuted for violating U.S. Antitrust Act.. sit. y. Nat. Similar to the consequences resulted from financial restatement, being prosecuted can. io. er. be costly to firms under investigation. Negative effects resulted from being prosecuted can observed in many forms, ranging from potential obligation to pay substantial fines,. al. n. v i n shaking investors’ confidence inC the credibility of corporate h e n g c h i U disclosure, dampening the demand for a firm’s securities and leading to a substantial loss in the market value of. the firm. Moreover, the records of being prosecuted, like financial misreporting, may potentially induce a change in loan contracting for the following reasons. First, the fact that a firm is convicted may indicate its incapability of abiding by the law, which may signal its deficiency in corporate governance and internal control. This can make the firm a comparatively undesirable borrower to the banks; therefore, in order to control the risks borne, banks may raise the spread charged, reduce debt maturity, and demand more collaterals. Second, if the firm in concern is deemed as capable of abiding the rules, then the fact that it’s convicted would 2. be interpreted as a record of rebel.

(8) behavior or disobedience on purpose, which would discount the firm’s integrity in the eyes of lenders and make it an even worse borrower. Either way, with the record of being prosecuted, it’s expected that there’d be differences in bank contracting before and after corporate litigation. Among the research related to the effect in the level of prosecution and corporate misconduct, most existing literature focuses on the release of related news of the company and how it may influence the firm’s market value in terms of equity market; However, we know very little about how the events of being prosecuted affect specific contract terms in syndicated loan market. This paper attempts. 政治大. to fill this gap by analyzing the impact of corporate convictions on the price of bank debt.. 立. The reason why we explore the impact of convictions using syndicated loan data. ‧ 國. 學. is that they provide multi-dimensional information about corporate loans, and that the. ‧. reactions of lead arrangers and participant banks to the violation record of common. sit. y. Nat. laws can be observed explicitly through various features of loan contracts. These terms. io. er. of contract allow us to explore the effects of corporate being prosecuted on the direct (interest rate) and indirect (loan maturity, covenant restriction, choice of loan origin). al. n. v i n C h allow us to lookUinto how the structure of bank costs of debt. Moreover, loan contracts engchi. loans such as the number and shares of lenders in a syndicate loan and loan transaction fees may be affected by the indictment record of the borrower. Even though similar to the issue of financial restatement, the effect of corporate convictions can be quite different from that of a financial restatement. The issue of how potential consequences from convictions affect firms has been raised in the field of Law. Professor Brandon Garrett of University of Virginia School of Law has pointed out in his book Too Big to Jail that when prosecutors target the Goliaths of the corporate world, they usually find themselves at a huge disadvantage in terms of capital and human resource. To avoid time consuming trial and possible economic clash, American courts 3.

(9) routinely hand down harsh sentences to individual convicts, but consider a very different standard of justice applies to corporations. He reveals a pattern of negotiation and settlement in which prosecutors tends to compromise with the company accused. When some corporations are considered too economically important to fail, they even receive government bailout by agreeing to, structural reforms so that they can avoid negative consequences of criminal convictions. He shows that convicted corporations are not exactly at disadvantage in negotiations. Therefore we examine whether banks respond in the same courtesy like prosecutors?. 政治大 incentives for banks to compromise on loan deals as well. For example, banks may have 立 Even though banks may be different from government and prosecutors, there are. the incentives to settle with less defensive terms in order to attract new clients and. ‧ 國. 學. compete with its opponents; or they may toned to settle with the terms just to hold on. ‧. to its existing clients.. sit. y. Nat. To answer these questions, we begin by examining the effect of corporate. io. er. convictions on loan spreads. We measure loan spread as the amount the borrower pays in basis points over LIBOR (London Interbank Offered Rate). In addition, corporate. al. n. v i n C hstudying the impact convictions vary in severity. When of corporate convictions on engchi U. bank loan contract, we should bear in mind that different types of violations may result in different degree of impact on contract terms. Since some of violations are minor misdemeanors than others, taking lenders’ concern into consideration, we expect that the syndicate loan market will take more precautions against firms that are involved in. fraud-related or other significant violations of laws in terms of a larger increase in loan spread. To summarize, this paper makes three contributions. First, it contributes to the loan contracting literature by studying how corporate litigation may or may not affect the loan spread of bank debt. Second, this paper is the first to connect Loan Pricing 4.

(10) Corporation’s Dealscan database and the database of Corporate Convictions, handcollected by Professor Brandon Garrett of University of Virginia School of Law. Through the collaboration of these two distinctive yet potentially relevant database, we expect this paper may provide results that are relevant for many studies in the future. The remainder of the paper is organized as follows. The next chapter presents related literature covering how different corporate misconduct affect the value of the firm and how banks cope with risk and information asymmetry through the use of contracting mechanism, and potential challenges when collaborating corporate. 政治大 Results and implication are summarized in Chapter 4, while Chapter 5 examines 立. misconduct database. Chapter 3 describes the data, summary statistics and methodology.. possible time lag of the conviction database. Chapter 6 concludes.. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 5. i n U. v.

(11) 2. Literature Review Because company wrongdoings can be costly and harmful to its reputation, which may also impose real effects on firm’s cash flow. The payments of fine, restitution and legal fares following the prosecutions and convictions alone can be pecuniarily costly for the firm as well. Moreover, being involved in an allegation could also induce a decrease in the present value of the firm’s cash flows if investors, customers and suppliers change the terms of trade based on a firm’s misconduct. As it is addressed in some of related research (Karpoff et al., 1993, 2005, 2014), some costs or negative effect to the criminal. 政治大 Most of related studies focus 立the effect of the criminal act on equity market; however,. firm, by which legal penalties can hardly explain, are considered as reputational loss.. ‧ 國. 學. taking potential effects mentioned above into consideration, we expect corporate convictions can affect firm’s financing costs as well in at least two ways. First,. ‧. convictions may affect lending banks’ evaluation of a company through revisions in. sit. y. Nat. perceptions about the firm’s business prospect as well as credibility. Second, when. al. er. io. borrowing companies are sanctioned with fine or restitution, their cash flow are directly. v. n. affected, which may be different from lender’s evaluation of the firm before conviction.. Ch. engchi. i n U. In view of these effects, we expect there are adjustments on bank loan terms before and after corporate convictions. Followings are some of the related studies in criminal action, its impact, and possible challenges in database collaboration marked for related studies in corporate misconducts. Based on these results, we may develop more comprehensive perspectives on how the events of being prosecuted may affect bank loan contracting in syndicated loan market.. 2.1 Corporate Misconduct Recent studies on corporate crimes concerns various types of violation, including fraudulent crime (eg. defraud US, false statement, misbranding), environmental act (eg. 6.

(12) clean water act, migration of birds, hazardous material emissions), bribery (illegal payment to foreign government/FCPA), antitrust etc. Alexander (1999) finds that when criminal allegations arise, they are often surrounded by reports of terminated or suspended customer relationships along with potential management or employee turnover. These reports are more frequent if the parties affected are customers, as in fraud, than if they are third parties, such as those in environmental crime. In the research toward fraud-related misconduct, Karpoff and Lott (1993) find that initial press reports of allegations or investigations of corporate fraud against private parties correspond to. 政治大 stock of affected companies. For frauds against government agencies, the loss in value 立 an average decrease of 1.34 percent, or $60.8 million, in the values of the common. is 5.05 percent, or $40 million. Based on a subset of firms with sufficient data indicated. ‧ 國. 學. in this paper, only 6.5 percent of the loss represents court-imposed costs, with penalties. ‧. and criminal fines accounting for 1.4 percent. Even though legal sanctions are imposed. y. Nat. on these criminal firms, the loss in terms of stock value implies that there’s certain. er. io. sit. degree of loss (reputational cost) incurred as the penalty for these firms’ criminal act. Corporate fraud usually impose direct cost on parities with whom the firm does business. al. n. v i n C h and investors. with, such as firm’s suppliers, customers Hence, when a firm commits engchi U. fraudulent misconducts, their counterparties may respond in a way that would compel the firm to lower its output prices or endure higher input prices, affecting its expected future cash flow. Furthermore, as in their studies on environmental violations, Karpoff et al. (2005) develop the measurement of reputational loss as the difference between losses in the market value of the firms violating environmental laws and the legal penalties on which they are imposed. They find that firms investigated or charged with violation of environmental regulations experience statistically significant and economically meaningful decreases in common share values. The loss in share value, while 7.

(13) significant, is not on average larger than the legal penalties imposed on the violating firm, implying that the share value losses are more likely due to prospective legal penalties instead of reputational costs. In addition, Karpoff et al. (2014) find that the costs for firms that are prosecuted for bribery depend on whether the bribery is comingled with charges of financial fraud. The firms facing bribery charges without any sorts of involvement in financial fraud face total costs that averages up to only 2.4% of their market capitalization from the bribery enforcement action, usually in the form of large regulatory fines and penalties.. 政治大 relationship with its customers, suppliers, or investors. Firms do not suffer large 立 On average, however, the bribery charges do little harm to the firm’s business. reputational losses when they are caught bribing; However, when the bribe is. ‧ 國. 學. accompanied by financial fraud, the reputational loss tends to be large. Fraud-related. ‧. bribe-payers tend to face higher future costs or lower revenues as their counterparties. sit. y. Nat. tend to change the terms with which they do business with the firm, resulting in. io. er. reputational loss that averages up to 18.8% of firm equity value, and the ex post NPV of bribery related to financial fraud is negative, -25.2%. Their findings show that in its. al. n. v i n C hof bribery is moreUlike an environmental violation impact on firm reputation, the action engchi and less like consumer fraud. That is, the firms’ counterparties tend to care more for. whether the firms’ financial statement are misrepresented. In general, they don’t vary their willingness to do business with the firm when it is caught bribing. Consistent with results in Karpoff et al (1993), studies show that when companies are found concealing information or making adjustments to the preceding messages, it places uncertainty about the companies’ reported financial information in the perspectives of banks. Graham, Li and Qiu (2008) find that financial restatement implies that the information previously known to the lending banks is imprecise. In order to correct formerly perceived information of the company, lending banks may use 8.

(14) contracting tools such as spread, debt maturity, covenants, collateral requirements etc. as precautions against information asymmetry, showing uncertainty to the credibility of a restating firm. In this study of relationships between restatement effect and the cost of raising bank debt, it is found that loans initiated after restatement show significant increase in loan spread (the increase in spread is significantly larger for fraudulent restating firms) with shorter maturity, higher likelihood of being secured, and more covenant restrictions. Moreover, on average, each loan has fewer lenders after restatement and presumably to compensate for monitoring activities, the upfront and. 政治大 According to these studies, we expect the adjustments coming along with 立. annual fees charged by lenders are higher for restating firms.. corporate convictions to be different in degrees, concerning violation types.. We. ‧ 國. 學. assume that fraud-related crimes are to display more influence on ex post loan term,. ‧. because fraud-related crimes usually demonstrate direct costs on the firms’ stakeholders,. sit. y. Nat. who, in return, could directly affect company’s profit and loss, potentially imposing. io. er. larger loss than other types of crimes. Other types of violations concerning environment or bribery, since they tend to pose costs on indirect third party, are expected to have less. al. n. v i n C hand ex post loan U impact on the difference of ex ante terms compared to fraud-related engchi misconduct.. There are many research papers that examine the causes and effects of financial misconduct. Most of those research papers compile samples using four databases－ SCAC, AAER, AA and GAO. Karpoff et al. (2014) identify four different types of challenges frequently faced when conducting research on financial misconduct using these database. The challenges are late initial revelation dates, scope limitations, potentially extraneous events, and complete and partial data omissions. First of all, the initial public revelations of financial misconduct may have occurred months before the initial coverage in these databases. Moreover, these databases may have omitted other 9.

(15) relevant announcements that can affect a researcher’s use of the events or they may miss large numbers of events they were designed to capture. Lastly, most of the events captured by these databases are unrelated to financial fraud. In this paper, we use Garrett’s corporate conviction database. Although Garrett’s corporate conviction database is not discussed in the research of Karpoff et al. (2014), we apply alternative way as an effort to mitigate the potential biasness mentioned by Karpoff in this paper.. 2.2 How Banks Manage Risks Using Loan Contract There are many studies focusing on general factors as well as environment that. 政治大 (2004) find 立 syndicates are more concentrated when the. would affect financial contracts and the structure of syndicated loans. Lee and Mullineaux. quality of. ‧ 國. 學. information on borrowing firms is worse, and that syndicate structure is more concentrated with fewer lenders when firms have a higher default probability. Sufi. ‧. sit. Nat. borrower requires more intense investigation and monitoring.2. y. (2007) finds that lead arrangers tend to retain a larger share of the loan when the. n. al. er. io. Qian and Strahan (2007) show how financial contracts respond to the legal and. i n U. v. institutional environment. In their study of how external situation in terms of creditor. Ch. engchi. protection shape the terms of bank loans, they show that under strong creditor protection where information sharing system is sound, loans have more concentrated ownership, longer maturities and lower interest rates. Other papers study how lending banks use loan pricing and contracting tools in response to different environment, degrees of information and riskiness of borrower. Freixas and Rochet (1997), the traditional banking literature, suggest that credit risk is the major lending risk faced by banks and is one of the primary determinants of loan pricing. As it is known, greater lending risk leads to higher loan interest rates. 2. As is noted in Sufi (2007)-‘’Anecdotal evidence suggests that nonpecuniary reputation effects or loan officer career concerns can provide high costs of default even if loans are perfectly hedged.” 10.

(16) Diamond’s (1991) theory indicates that debt maturity is a nonmonotonic function of risk rating. While both low and high risk firms use short-term debt, there’re different implications. Low risk firms, compared with high risk firms, are more capable of rolling over their debt; On the other hand, firms with higher risks may be rejected long-term debt because of a high default probability. Scherr and Hulburt’s (2001) find results consistent with Diamond’s (1991) theory.. 2.3 Hypothesis Studies have shed lights on how corporate misconduct demonstrate impacts on the. 政治大 establish bank loan contract 立in response to risks of the borrowers and environment.. market value of the company under different types of violations, and how banks. ‧. ‧ 國. hypothesis:. 學. Based on the results of related literature, in this paper, we have established following. Post-Conviction Effect and the Cost of Debt. sit. y. Nat. To compensate for the risks they’re bearing, lending banks tend to charge higher. n. al. er. io. price of capital when the borrowing firm requires more intense investigation and. i n U. v. monitoring. Moreover, when a company is charged with specific crimes, it usually leads. Ch. engchi. to certain degree of loss either due to court sanctions or reputational loss. These effects could induce a re-evaluation in terms of financial and business outlook, which are what lending banks based on when constructing loan terms. In this paper, we establish our regression model with loan spread as our dependent variable. To see if the cost of debt would increase after company is convicted, our null hypothesis is that the coefficient on the Post-Conviction is equal to zero.. H1: The spread of post-conviction loan contracts is higher compared with that of loan contracts made before conviction.. 11.

(17) Different Types of Violation and the Cost of Debt Following Karpoff et al.(1993, 2005, 2014) and Graham (2008) , we expect there to be difference in the degree of impact based on different types of violation. According to Karpoff et al. (2005, 2014), the additional financial impact (reputational loss) on criminal firms is less obvious when it is environmental law or bribery restriction that they’re disobeying; and when firms are charged with fraudulent action or misreporting, they’re more likely to face additional loss other than legal penalty, hence, more likely to face higher cost of loan. Corresponding to these findings, we predict that, among. 政治大 higher loan spread than those committing non-fraud-related crimes. Our null hypothesis 立 those convicted, those committing fraud-related crimes are likely to be charged with. is that the coefficient on the interaction term Post-Conviction*Fraud-Related is equal. ‧. ‧ 國. 學. to zero.. sit. y. Nat. H2: The change in the spread of post-fraud-related-conviction loan contracts is higher. io. n. al. er. compared with that of loan contracts made after non-fraud-related-conviction.. Ch. engchi. 12. i n U. v.

(18) 3. Data and Methodology 3.1 Sample Selection We use convicted corporations data from Federal Corporate Prosecution and Plea Agreement Database which is hand-collected by Professor Garrett of University of Virginia, School of Law. The database includes 1011 corporate convictions during the decade from 2001 to 2010 along with 203 corporate plea agreements from 1992 to 2011. Twelve percent of the firms were foreign (142 convicted corporations and 35 corporate plea agreement), that is, they were incorporated outside the United States. This database. 政治大 against antitrust act and FCPA, 立 just to name a few, and it also covers information involves different types of violation including false statement, misbranding, offense. ‧ 國. 學. relevant to the disposition of each case of criminal offense. Information ranges from prosecuting offices, agreement date, type of violations to the amount of fines and. ‧. restitution etc. Apart from the disposition of the events, the database also contains. sit. y. Nat. information on the criminal corporations (eg. whether it’s public-listed, foreign or. n. al. er. io. domestic incorporated, with or with no subsidiaries, acquire compliance program or not. i n U. v. etc.) In our sample set, for companies that commit crimes more than once, we focus on. Ch. engchi. the first criminal act. The reason is that the key purpose of this study is to compare the cost and terms of debt before corporate convictions with that after corporate convictions. If we were to keep the second or third criminal convictions, the pre-crime influence window of the second or third criminal act could overlap with the post-crime influence window of the first conviction, this overlap may confound the comparison. We organize our criminal corporation sample based on Garrett’s database and we use the bank loan data from Loan Pricing Corporation’s (LPC) Dealscan database. This database contains detailed information for U.S. and foreign loans made to corporations, starting from 1986 to 2010. The data are gathered from SEC filings with parts derived. 13.

(19) from direct research by LPC through contacts with borrowers, lenders, and the credit industry at large. The basic unit of this empirical analysis is a facility, also known as a loan or a tranche in Dealscan. Facilities are grouped into packages, so a package may involve one or more loans. While each facility has only one borrower, under the structure of syndicated loan, facilities can have multiple lenders. Based on the fact that not every corporation in Garrett’s conviction list has been involved in syndicated loan or attribute their information to Dealscan database, we use the following procedure to form our sample. First, by cross-referencing corporate. 政治大 344 convicted corporations with Dealscan database; and by cross-referencing corporate 立. information between Dealscan and Garrett’s database, we are able to directly identify. information on Bloomberg BusinessWeek and LexisNexis database, we further expand. ‧ 國. 學. our sample with 255 more convicted corporations that have no records of receiving. ‧. syndicated loan but have affiliated corporate units such as parent company or. sit. y. Nat. subsidiaries that are listed in LPC Dealscan database. Combining with Dealscan’s. io. and that of its affiliates (parent company or subsidiaries).. al. er. information of corporations, we are able to connect the information of convicted firm. n. v i n Ch In our sample, the convicted and conviction-related firms can be further divided engchi U. into two groups, the ones that are subsidiaries of the convicted firms and the ones that. are the parents of the convicted firms. However, there’s the problem concerning the timing of ownership transfer when recognizing convicted companies’ affiliates from one another. Take M&A for example, when a corporation is involved in a conviction, sometimes the negative impact can left the company vulnerable for a take-over. Assuming the convicted company is then bought by another corporation, and both its new and previous parent companies have borrowing records in LPC database, since it is unreasonable to link our conviction information to both of the parents (it may demonstrate diverse impacts on loan spread), in this paper, we base our priority of 14.

(20) recognition on its original owner, but if an ownership transfer occurred shortly after the conviction(within two years), then we recognize the new owner as the convicted company’s parent company. For instance, the parent company, Firm A, of the convicted company, Firm a, may have acquired loans before Firm a commits criminal action, but transferred its ownership (of Firm a) to Firm B after the conviction. Since Firm A no longer owns Firm a after the conviction, at first glance, it seems inappropriate to study the impact of having a convicted subsidiary based on the changes in Firm A’s loan contracts after the conviction. While it seems more relevant to observe the changes in. 政治大 law under the ownership of Firm A, however, there could be concerns that Firm A might 立. Firm B’s financial contract before and after it acquires Firm a, since Firm a violates the. be the one to blame for the lack of scrutiny or failing at internal control(upon Firm a).. ‧ 國. 學. As a matter of fact, there are several convicted cases, shown in GAO and SEC, indicate. ‧. that parent companies are responsible for the delinquencies of their subsidiaries. In. y. Nat. many of those cases, the parent companies are asked to pay for the damage caused by. er. io. sit. their subsidiaries as the result of conviction. Hence, it’s rather difficult to assert which related corporation (in this case, Firm A or Firm B) to focus on. To determine which. al. n. v i n relevant,Cwe use LexisNexis U h e n g c h i database. related firm is more. to obtain the timing. information on when a transfer of ownership occurs. While we keep both types of companies (Firm A and Firm B) in our sample, we give our priority to the firm that owns Firm a before Firm a got involved in criminal conviction, however, if another firm acquires Firm a within 2 years after Firm a is convicted, it’s then included in our sample instead of Firm a’s previous owner. That is, if Firm B acquires Firm a within 2 years after Firm a’s conviction, then Firm B’s bank loan contract will be our focus in this study instead of that of Firm A’s; if Firm B acquires Firm a 3 years after the conviction, then we decide that Firm A is more appropriate to represent the affiliate of the convicted firm. Finally, to permit fair comparison of the debt contract before and 15.

(21) after corporate conviction, we remove firms that only have pre-conviction loans or only have post-conviction loans. To include the firm characteristics of the convicted and convicted-related corporations, we also merge our previous sample, constructed with both conviction information from Garrett’s and their bank loan information in LPC Dealscan, with Standard and Poor's Compustat database using link file built by Wharton Research Data Services (WRDS). Corporations are removed if both they and their affiliates have no loan information in Dealscan or have missing Compustat information.. 政治大 of observation window. In this paper, taking both sample size and influence window 立 To study the impact of corporate conviction on loan terms, we avoid overextension. into consideration, we construct an observation window of 5 years. That is, pre-. ‧ 國. 學. conviction loans must be made within 5 years before corporate conviction; likewise,. ‧. post-conviction loans must be made within 5 year after corporate conviction. The. sit. y. Nat. spread (the price of the bank borrowing) is measured as the Dealscan data item all-in. io. er. spread drawn, which is the amount the borrower pays in basis points over LIBOR or LIBOR equivalent for each dollar drawn. In this paper, we focus on facilities with. al. n. v i n spread quoted over LIBOR bothC before and after corporate h e n g c h i U convictions. [Insert Tables 1 and 2 here]. Tables 1 and 2 presents the summary statistics results of loan contract terms for our sample firms. Number of observations, mean and standard deviation of debt contract terms are reported for loans in the full sample, facilities started before conviction, and facilities started after convictions. The details of definitions and measurements of all the variables are reported in Appendix B whereas Appendix A presents the distribution of the types and the primary purposes of loans in our sample.. 16.

(22) 3.2. Methodology We divide our empirical research into two parts and examine the degree of how. corporate convictions affect the pricing of bank loan. We also examine the difference in results where specific conviction condition is taken into concerns, such as types of violation, the size of fine, nationality of convicted corporation or whether criminal firm has obtained compliance program. First, to avoid excessive reduction in the number of observation (number of facilities of our sample), we segregate all types of violations into two groups. We. 政治大 bidding, and product recall立 as fraud-related crime. For those unrelated to any of these recognize crimes related to fraud, financial misrepresentation, product safety, deceptive. ‧ 國. 學. violations, such as environmental violations, antitrust, FCPA and export violations, we categorize them as nonfraud-related crime. To see if fraud-related crimes bring more. ‧. incentives for banks to charge higher cost of capital, we set a dummy variable equal to. sit. y. Nat. 1 if corporate conviction is derived from fraud-related misconduct, 0 otherwise.. n. al. er. io. The fine sanctioned by court could vary from one firm to the other. In our sample,. i n U. v. not all convicted corporation is required to pay fine. To study how fine may affect the. Ch. engchi. price of debt, we set a dummy variable which is equal to one if the convicted company is sanctioned with fine, and zero otherwise. According to Moot (2008), many corporations facing increasing exposure to corporate crime have responded by strengthening their compliance programs to reduce the likelihood of violations, but these programs also present risks since they tend to inspect and unveil violations that would otherwise stay undiscovered by the government. Recognizing this dilemma, many federal regulators have adopted policies that reduce or eliminate penalties for corporations that adopt effective compliance programs. Hence, a company convicted of a violation may have its penalty reduced if. 17.

(23) it had an effective compliance program in place at the time of the violation. In this paper, to see if undertaking a compliance program would make a difference for the convicted firm on the price of loan, we set a dummy variable indicating one if the convicted firm has applied compliance program at the time of conviction, and zero otherwise. Garrett (2011) indicates that when Federal prosecution is placed, foreign firms on the whole are treated differently than U.S. based multinationals. Big U.S. based corporations tend to obtain deferred and non-prosecution agreements while foreign firms tend to plead guilty to their crimes and pay higher fines. Based on Garrett’s. 政治大 discussing the relationship between corporate conviction and the price of loan. In this 立. findings, we are interested in seeing if the base of corporate nationality matters when. sub-section, we focus on the potential effect of corporate nationality by dividing our. ‧ 國. 學. sample firms into two groups－domestic firms (if the convicted firm is incorporated in. ‧. the U.S.) and foreign firms (if the convicted firm is incorporated in countries other than. sit. y. Nat. the U.S.) to see if convicted firms incorporated in non-US countries tend to face higher. io. al. n. section.. er. loan spread than those incorporated in US. The results of regressions are shown in next. Ch. engchi. 18. i n U. v.

(24) 4. Empirical Analysis 4.1 Effect of Corporate Conviction on the Cost of Bank Debt Table 2 presents the means of the differences between the variables representing bank contract terms before and after corporate conviction. The results show that after sample firms are convicted, loan spread, loan maturity and the number of covenants increase at significance level of 1%, 5% and 1%, respectively; On the other hand, the appliance of performance pricing and the number of lenders decrease at significance level of 1%. The results show that loan size, number of security and lead bank share of. 政治大. post-conviction facilities are not significantly different from that of pre-conviction. 立. facilities.. ‧ 國. 學. In this section, we use regression analysis to examine the effect of corporate conviction on the cost of bank debt. The main empirical model is as follows:. y. sit. io. er. Industry fixed effect, Year fixed effect). ‧. Nat. Loan Spread=f (Post-conviction indicator, Firm characteristics, Loan characteristics,. Each observation in our regression represents a single facility (loan). The. al. n. v i n Cdebt, dependent variable is the cost of spread. To capture the effect of corporate h eloan ngchi U. conviction, we define a dummy variable, post-conviction, which is equal to one if the loan is activated after criminal conviction and zero otherwise. We control for firm characteristics, loan characteristics (loan size, loan maturity and a dummy equal to one if performance pricing is applied in the loan), along with industry and year effects. Detailed definition for all the variables are reported in Appendix B. The regression results are reported in Table 3. Column 1 analyze the cost of debt with post-conviction dummy as the only independent variable. The estimated coefficient of post-conviction, is -34.24, which is significant at the 5% level. This results implies that for the conviction firms in our sample, loan spreads tend to decrease 19.

(25) after corporate convictions. The regression in column 2 of Table 3 includes firm characteristics that could influence the cost of bank loans. These variables include Ln(assets), the natural logarithm of a firm’s total assets, to measure corporate size. Larger firms have easier access to external financing. They are also hypothesized to have less information asymmetry and are associated with smaller monitoring cost. Therefore, in general, larger firms are likely to borrow from banks on better terms. We also control for the ratio of long-term debt to total assets, also known as leverage ratio. Firms with higher. 政治大 a higher cost of bank borrowing. Moreover, since profitable firms generally have low 立 leverage ratios, all else equal, have higher default risk and thus we expect them to face. default risk and thus, they can borrow at a lower cost, we also include the profitability. ‧ 國. 學. of firms as one of our control variables which is measured in the ratio of earnings before. ‧. interest, taxes, depreciation, and amortization (EBITDA) to total assets. Apart from that,. sit. y. Nat. we use Market-to-Book as a proxy of a firm’s growth opportunities, calculated by the. io. er. ratio of market value of the company to book value of the company. The results in column 2 of Table 3 shows that small, highly levered, distressed firms with few growth. al. n. v i n Cahhigher cost of debt.UHowever, after controlling for opportunities are associated with engchi firm characteristics, the effects of corporate convictions on the loan spread is no longer significant. Column 3 of Table 3 shows results of regression with additional control for loan characteristics that might be associated with the price of debt. In this regression model, we control for loan maturity, the natural logarithm of loan maturity in months. Since lending banks requires a liquidity premium for longer-term debt, we expect there to be a higher loan spread when the maturity of debt is longer. Aside from the maturity of loan, we also include Ln(loan size), the natural logarithm of the amount of a loan, which, according to Graham et al.(2008), may capture economies of scale in bank lending and 20.

(26) thus is expected to be inversely related to the loan spread. The dummy variable performancepricing is equal to one if a loan contract has applied mechanism of performance pricing and zero otherwise. The purpose of this dummy is to control for the possibility that lenders price loans differently if they contain performance pricing. The regression results show that larger loans, loans with shorter maturity and performance pricing clause tend to have smaller spreads while corporate conviction is not significantly related to the cost of bank debt.. 4.2 Specific Conviction Terms on the Cost of Bank Debt. 政治大 to draw the influence of corporate 立 conviction on the change of loan spread. If any, they Judging from the results shown in Table 3, there are no significant evidence for us. ‧ 國. 學. are inversely related, which is quite different from what we expect. In this section, we further look into some specific terms obtained by or condition of the convicted. ‧. companies in order to see if there are significant differences in the results where the. sit. y. Nat. change of loan spread can be explained by particular conviction characteristics instead. er. io. of the sheer fact of being convicted.. n. a. v. l CPenalty－Fined or not 4.2.1 The Impact of Legal ni. hengchi U. When a company is convicted of crime, the payments regarding fine, restitution and legal fares following the sanctions can be pecuniarily costly for the firm. Arlen (2012) finds that average fines ranges from $5.7 to $17.3 million from 2006 to 2008 for firms paying non-zero fines. Alexander et al.(1999) presents empirical data showing higher total sanctions after the Organizational Sentencing Guidelines were adopted in 1991. To find out the potential impact of fine, we divide our samples into two groups －fined and not fined, and assign a dummy variable fined equal to one if the company is sanctioned with fines or restitutions, and zero otherwise. We focus on PC*fined, which is the interaction term of post-conviction and fined measuring the difference 21.

(27) between the effects displayed by firms sanctioned with and without fines. Column 1 of Table 4 shows the results of our main regression with an additional variable PC*fined added in the model. Decrease of 0.01% in adjusted R square shows a slight slide in the explanatory power of our model. The coefficient of Post-Conviction remain insignificant with t-value of -1.3; even though the coefficient of PC*Fined is positive, it is not statistically significant. Hence, among our sample of convicted corporations, we are unable to reject the hypothesis that there’s no difference between the cost of bank faced firms sanctioned with fines and those without.. 政治大 As discussed earlier, the 立adoption of compliance program could potentially affect. 4.2.2 Compliance Program. ‧ 國. 學. how prosecutors lay down the sanction. Here, we define a dummy variable withCompliance to indicate whether the convicted firm has obtained compliance. ‧. program at the time of violation agreement, if it does, then the dummy is equal to one,. sit. y. Nat. if not, then zero. To observe the intra-sample difference, we focus on. n. al. er. io. PC*withCompliance, the interaction term of post-conviction and withCompliance.. i n U. v. With this additional variable in our model, the results are shown in column 2 of Table. Ch. engchi. 4. The coefficient of PC* withCompliance is negative, meaning that if a convicted firm has adopted compliance program at the time of violation, it tends to face a lower spread than those without compliance program. However, neither the coefficient of the interaction term nor that of post-conviction is significant.. 4.2.3 State of Registry Studies show that firms incorporated outside of US tend to face larger size of legal penalty when breaking the laws. Garrett (2011) indicates that when Federal prosecution is placed, foreign firms on the whole are treated differently than U.S. based multinationals. To see if the indirect effect of the difference in the degree of sanction 22.

(28) would affect convicted corporation’s cost of bank debt, we define a dummy variable ForeignCriminalCorp (FCC), which is equal to one if the convicted firm is incorporated outside of US, and zero otherwise (US domestic). We focus on the relationship. between. the. interaction. term. of. post-conviction. and. ForeignCriminalCorp(FCC), that is, PC*FCC, and the spread of bank loan. The results is shown in column 3 of Table 4. Even though the explanatory power of the model slightly increases, the coefficient of post-conviction and PC*FCC are both insignificant.. 4.2.4 Conviction Type and the cost of Bank Debt. 政治大 on the cost of bank debt, we立 divide our sample into two groups based on (1).the severity To examine whether different types of violations have different degree of impact. ‧ 國. 學. of crime and (2).whether the violation is fraud-related.. Under our first rule of classification, if corporation is convicted with crimes that. ‧. tend to directly pose offenses against government, international laws, or lead to rather. sit. y. Nat. strict legal penalties, they are considered and classified as Type1 of convicted firm. This. n. al. er. io. type includes well-known violations of laws such as Antitrust, FCPA, or Conspiracy to. i n U. v. defraud US etc. More details on the classification rules are listed in Appendix C. When. Ch. engchi. company is convicted with Type1 of violation, we say the corporate misconduct is more severe and can potentially cause more damages to the reputation of the firm than other types of violation. Hence, we add a dummy variable CrimeSeverity in our main regression － when a firm is classified as Type1 convicted firm, the dummy CrimeSeverity is equal to one, and zero otherwise. While post-convictions is inversely related to loan spread at significance level of 5%, column 4 of Table 4 shows that the severity of crime is positively associated with the cost of bank loan at significance level of 5%. That is, compared to convictions of minor violations, corporations convicted of Type 1 violations tend to face higher loan spreads after convictions. With this. 23.

(29) interaction term added in our regression, the explanatory power of the model slightly increases by 0.15%, from 75.54% to 75.69%. Studies have shown that reputational loss tend to appear when criminal firms are involved in fraud-related crimes. In accordance with these findings, we expect fraudrelated convictions to be more influential to the change in loan spread compared to nonfraud-related convictions, with the assumption that fraud-related convictions would increase the cost of firm’s bank debt. Hence, under our second rule of classification, we group our samples into fraud-related and nonfraud-related convicted firms. To study. 政治大 if the loan is activated after fraud-related convictions and zero otherwise. Here, we 立. the effect of violation types, we define a dummy variable, FRC, which is equal to one. focus on the interaction term of post-convictions and fraud-related convictions. The. ‧ 國. 學. regression results are reported in column 5 of Table 4. While the effect of conviction. ‧. is still significantly and inversely related to the spread of bank loan, there’s no. y. Nat. significant evidence for the interaction term PC*FRC, which measures the impact of. er. io. sit. fraud-related convictions on the cost of loan, to reject null hypothesis that there’s no difference in the impact between fraud-related and nonfraud-related convictions on the. al. n. v i n C h this variable increase change in loan spreads. While adding the explanatory power of engchi U. the model, the increased margin is relatively small compared to the variable concerning severity of crimes. In chapter 5, taking lag in initial announcement into consideration, we take a further look into the potential effect of conviction on the cost of loan.. 24.

(30) 5. The Announcement Effects In the previous chapter, we examine the impact of corporate conviction and some specific criminal conditions (fines, compliance program, the country where convicted firm is incorporated, the severity and types of violation) on the cost of debt. However, with both year and industry effects controlled, most of our results show that there’s no significant evidence to reject our null hypothesis. Among the results including interaction terms shown in Table 4, only when the severity and the types of crime is taken into account does the conviction effect appear to be significant. Even so, the. 政治大 loan. These findings are quite 立on the contrary to both our expectation and the related coefficient indicates that the conviction effect is inversely associated with the cost of. ‧ 國. 學. studies on the effect of corporate violation.. Among related studies of criminal effects, Karpoff et al.(2014) identify late initial. ‧. revelation dates as one of the challenges frequently faced when conducting research on. sit. y. Nat. financial misconduct. That is, the initial public revelations of financial misconduct may. n. al. er. io. have occurred months before the initial coverage in the database. Even though Garrett’s. i n U. v. database is not discussed in Karpoff et al.(2014), by cross-referencing news on. Ch. engchi. LexisNexis, we discovered that the convictions dates recognized in Garrett’s database, like the databases pointed out in Karpoff et al. (2014), also face the problem of late initial revelation dates. In our sample, some related news covering the corporate convictions has been announced months or years before the agreement date of the conviction logged in US Department of Justice. Taking this condition into consideration, there’s possibility that the lending banks have adjusted their evaluation toward convicted firms after related news releases (instead of after the agreement date of the conviction). This may help us explain why the results based on agreement dates would derail from relevant research. Hence, in this. 25.

(31) section, we intend to catch a glimpse of the relationship between the announcement effect of corporate conviction and the cost of loan based on sketchy adjustments on the turning point recognized. In the previous chapter, we base our turning point in years. That is, we distinguish facilities initiated before and after convictions based on the year difference between facility start year and agreement year. Therefore, the lag in the recognition of the release of corporate conviction make little difference to our previous results if it’s within 1 year. In our sample, there’s around 1/3 of the agreement date of convictions that are set one. 政治大 results comparing previous chapter, we shift each agreement year one year backward 立. year later than the release of its news. To gain a glimpse of the potential difference in. as the proxy of announcement year of the news. For example, if the agreement year of. ‧ 國. 學. the conviction is 2001, here we assume year 2000 is the year when the news of. ‧. prosecution releases, and so on. Based on this sketchy adjustment of turning point. sit. y. Nat. recognized, the following shows the results of our empirical model.. io. er. 5.1 Effect of Corporate Conviction on the Cost of Bank Debt Tables 5 and 6 presents the summary statistics results of loan contract terms for. al. n. v i n C h mean and standard convicted firms. Number of observations, deviation of debt contract engchi U. terms are reported for loans in the full sample, facilities started before and facilities started after the proxy year for announcement. Table 6 presents the means of the differences between the variables representing bank contract terms before and after proxy year for criminal news. The results show that after conviction news releases, loan spread, loan maturity and the number of covenants increase at significance level of 1%; On the other hand, the appliance of performance pricing and the number of lenders decrease at significance level of 1%. The results show that the coefficients of loan size, number of security and lead bank share of ex post facilities increase, but not significantly different from that of ex ante facilities. 26.

(32) The regression results are reported in Table 7. Column 1 analyze the cost of debt with post-news dummy as the only independent variable. The estimated coefficient for post-conviction news is 13.09, with t-value of 1.33. Column 2 and column 3 shows results with firm characteristics and additional loan characteristics taken into concerns. The coefficients of Ln(asset), Market-to-Book Ratio, Profitability, Loan Size and Performance Pricing are negative, while Leverage Ratio and Loan Maturity are positively associated with the spread of facility, which, on the whole, is consistent with previous studies in loan contracting. From column 2 and column 3 of Table 7 we can. 政治大 significant at 1% and 5% level, respectively. This results show that when a company is 立. see that the coefficient of Post-Conviction News is positive under both models,. involved in the news of conviction or prosecution, they tend to face a higher price of. ‧ 國. 學. bank debt than before.. ‧. 5.2 Specific Conviction Terms on the Cost of Bank Debt. sit. y. Nat. Similar to chapter 4.2, in this section, we further look into some specific terms. io. er. obtained by or condition of the convicted companies in order to see if there are significant differences in the results where the change of loan spread can be explained. n. al. Ch. by particular conviction characteristics.. engchi. i n U. v. 5.2.1 The Impact of Legal Penalty－Fined or not Column 1 of Table 8 presents the results of our main empirical model with an additional interaction term PCN*Fined included in the regression. The result, where the coefficient of PCN*Fined is 20.67 at significance level of 10%, shows that when convicted, the companies sanctioned with fine tend to face higher spread in loan than those not-fined. While the coefficient of post-conviction news is still positive but not statistically significant, the addition of this variable slightly increases the adjusted R square of the model, from 72.08% to 72.15%.. 27.

(33) 5.2.2 Compliance Program Column 2 of Table 8 shows the results of our model with an interaction term PCN*withCompliance included in the regression. The coefficient of the interaction term is not insignificant enough to reject our null hypothesis that there’s no difference between the costs of loan faced by companies adopting compliance program at the time of conviction and those don’t. While the conviction news is still positively related to loan spread (here, at significance level of 5%), the adjusted R square slightly decreases by 0.03%.. 政治大 In this section, we examine whether there’s difference in the conviction effect 立. 5.2.3 State of Registry. between domestic-incorporated and foreign-incorporated criminal firms. Same as. ‧ 國. 學. section 4.2.3., we set a dummy variable FCC (which is equal to one if the convicted. ‧. firm is incorporated outside of US, and zero otherwise), and focus on its interaction. sit. y. Nat. term with conviction effect, PCN*FCC. The results are shown in column 3 of Table 8.. io. er. While the coefficient of Post-Conviction News is 25.43, significant at 1% level, the coefficient of interaction term PC*FCC is -80.81, at the significance level of 5%. The. al. n. v i n to those incorporatedUin the US, firms incorporated result indicates that, comparedC hengchi outside of US tend to face lower cost of spread after the conviction.. 5.2.4 Conviction Type and the cost of Bank Debt Same as section 4.2.4., we divide our sample into two groups based on (1).the severity of crime and (2).whether the violation is fraud-related. We assign 2 dummy variables indicating whether it is Type1 conviction and whether it is fraud-related. Column 4 and column 5 of Table 8 presents the results with these two additional interaction term included in the model. Under both models, the coefficients of PostConviction News are positive. In column 4, Post-Conviction News is at significance level of 5%, however, its interaction with CrimeSeverity is not significant. Column 5 28.

(34) shows the result of regression model including the interaction term PCN*FRC. At significance level of 5%, the coefficient of PCN*FRC is 24.29, positively associated with the cost of bank debt. This indicates that firms involved in fraud-related convictions, compared to other types of violation, tend to face higher spread of loan after conviction news releases.. 5.3 Suggestions on Further Studies The results based on adjustment of turning point (shifting one year back as proxy. 政治大 only does it indicates that conviction effect exists in loan market and is reflected in the 立. for release time of conviction news) are more consistent with our expectations. Not. spread of bank debt, but it also implies that the news of corporate conviction is more. ‧ 國. 學. influential, not only in equity market but in loan market as well, than the actual time of. ‧. court conviction. However, since this results are based on sketchy adjustments on the. sit. y. Nat. turning point recognized, further accurate information from LexisNexis and Factiva. io. er. database is required so as to update and confirm each of the release time of related news. Even though the sorting process for precise time of related news has yet to be completed,. al. n. v i n C h mentioned earlier we regress loan spreads on variables using 56 criminal companies engchi U. preliminarily sorted with precise news release year along with 890 facilities. The results are presented in table 9 and 10. Although the effect of Post-conviction news are similar to the results presented in table 8, the effects are less significant with accurate criminal. news. Nonetheless, the results shows that the coefficient of PCN*FRC is positively significant, implying that the type of crimes (fraud-related crime) lays certain degree of impact on loan spreads between criminal corporations.. In this paper, we differentiate ex ante and post facilities based on facility starting years and conviction (or conviction-related news released years) instead of starting dates (news-released dates). And we consider the first conviction only, if there’re more 29.

(35) than one. Moreover, we focus on the conviction effect on loan spread. These may explain why some of the results derail from our expectations. If the criminal companies are convicted of more serious crimes after the first one, by excluding the effects of subsequent convictions, the results based on the first conviction may be misinterpreted. Apart from that, there’re possibilities that after the convictions, the convicted companies are not able to, or not welling to, obtain syndicated loan. In this situation, the impact of conviction is more serious and decisive in loan market, but by focusing on the change of loan spread, this effect cannot be captured since the companies in. 政治大 from resolving limitations mentioned above, by using linkage between the databases in 立. question have no ex post facilities for comparison. For potential further studies, apart. this paper, we hope to expand research upon other loan contract terms to see if banks. ‧ 國. 學. stretch their influence in contract terms other than loan spread. This may help us. ‧. understand the cross effect of loan contracts when facing borrowers with conviction. sit. y. Nat. history. In the conviction and agreement database constructed by Professor Garrett,. io. er. more details on the crime and convicted companies are listed for further studies. Potentially, by combining the details with existing prestigious corporate database such. al. n. v i n C h etc., we expect toUunderstand different aspects of as AAER, GAO, LexisNexis, Factiva engchi. conviction effect on loan markets in more angles.. 30.

(36) 6. Conclusion Recently, a number of well-known corporations had been prosecuted as a result of violating US domestic regulations or International Standards. Among the research related to the effect in the level of prosecution and corporate misconduct, most existing literatures focus on the release of news (including the news of being prosecuted or the involvement of lawsuits) of company and how it may influence the firm’s market value in terms of equity market. So far we know little about how the events of being convicted affect specific contract terms such as loan spread in syndicated loan market. This paper. 政治大. attempts to fill this gap by analyzing the impact of corporate convictions on the price. 立. of bank debt.. ‧ 國. 學. In this thesis, we examine corporate misconduct from the debt holder’s perspective by investigating how the price of bank debt changes as a result of corporate conviction.. ‧. Using conviction year as the turning point to distinguish ex ante and ex post facilities,. sit. y. Nat. we find that there's no significant differences on the price of loan between facilities. n. al. er. io. initiated before and those initiated after the conviction. Even as we take some of the. i n U. v. conviction terms and characteristics into account, we find that among all the interaction. Ch. engchi. terms concerning corporate conviction only the coefficient of PC*CrimeSeverity is significant and positive, meaning that compared to firms committing minor violations, firms committing severe violations tend to face higher cost of debt after corporate conviction. Taking the potential lag between media coverage and the time recognized in conviction database into consideration, we further apply a sketchy adjustment on the turning point of time, recognizing the year before conviction as the year when media coverage of the conviction on the corporate misconduct first show up. Under this structure, we find significant evidence that the effect of corporate conviction is. 31.

(37) positively associated with the spread of firm’s bank debt. This means that if our application of time adjustment does justice for the time lag for initial release, then the increase in the cost of debt caused by corporate misconduct is an ex post cost borne by the criminal firm. Moreover, when concerning interaction terms with conviction effect, we find that firms that are fined or convicted of fraud-related crimes tend to face higher loan spread than those aren’t after conviction; while at the same time, foreign-based convicted firms tend to face lower spread than those domestic-based after conviction. To sum up, this paper makes two contributions. First, it contributes to the loan. 政治大 conviction. Second, this paper is the first to link data of Loan Pricing Corporation’s 立 contracting literature by examining the relationship between loan spread and corporate. Dealscan and the database of Corporate Convictions. Through the collaboration of these. ‧ 國. 學. two distinctive yet potentially relevant database, we expect this paper to be the spark. ‧. for many studies of related issues to come.. n. er. io. sit. y. Nat. al. Ch. engchi. 32. i n U. v.

(38) Table 1 (5-year observation window) Summary Statistics (Full Sample) This table presents the summary statistics results of loan contract terms for convicted corporations both before and after corporate convictions (sample firms). Number of observations (N), mean, and standard deviation (STD) of debt contract terms are reported for loans in the full sample. The starting year of facilities in our sample ranges from 1998 to 2010. Variable. Mean. Median. Std Dev. Min.. Max.. N. Loan Spread(basis point). 196.64. 150. 167.01. 13. 650. 949. Loan Maturity(months). 43.16. 60. 23.99. 5. 87. 949. Loan Size(millions). 696. 290. 1099. 5. 6000. 949. number of covenants. 1.76. 2. 1.60. 0. 4. 949. Performance Pricing(dummy). 0.43. 0. 1. 949. 0. 20. 949. Number of Security. 0 0.5 治政 0.68 0 1.27 大. 立 13.04. Number of Lenders Lead Bank Share. 23.58. 9. 11.19. 2. 57. 862. 12.5. 26.95. 0. 100. 939. ‧. ‧ 國. 學. n. al. er. io. sit. y. Nat. Table 2 (5-year observation window) Summary Statistics (Pre/Post-Conviction Loans) Table 2 presents summary statistics of loan contract terms for sample firms. Number of observations (N), mean, and standard deviation (STD) of debt contract terms are reported for loans initiated before corporate conviction and loans initiated after corporate conviction. The means of the differences between the variables before conviction and after conviction are also reported. Significance at the 10%, 5% and 1% level is indicated by *, **, ***, respectively.. Ch. engchi. Pre-Conviction N. i n U. v. Post-Conviction Mean Std Dev. Difference. N. Mean. Std Dev. Mean. Loan Spread(basis point). 508 165.15 138.32. Loan Spread(basis point). 441. 232.91. 188.64. 67.76***. Loan Maturity(months). 508. 41.37. 25.70. Loan Maturity(months). 441. 45.22. 21.70. 3.85**. Loan Size(millions). 508. 682. 1022. Loan Size(millions). 441. 712. 1182. 29. number of covenants. 508. 1.53. 1.57. number of covenants. 441. 2.03. 1.60. 0.50***. Performance Pricing(dummy). 508. 0.49. 0.50. Performance Pricing(dummy). 441. 0.37. 0.48. -0.12***. Number of Security. 508. 0.69. 1.53. Number of Security. 441. 0.66. 0.88. -0.03. Number of Lenders. 438. 16.11. 12.12. Number of Lenders. 424. 9.87. 9.12. -6.24***. Lead Bank Share. 501. 22.28. 27.17. Lead Bank Share. 438. 25.06. 26.65. 2.77. 33.