The Spillover Effects of Underwriting Relationships on Firm Value
Hyoseok (David) Hwang* Rutgers University – Camden
This is a Preliminary Draft. Please do not cite, or circulate without permission.
JEL classification: G24, K22 Keywords: Equity Underwriting, Spillover Effect, Securities Class Action Lawsuits
Assistant Professor of Finance at Rutgers School of Business – Camden; Tel.: +1 856 225 6693; Fax: +1 856 225 6231; E-mail:
[email protected]; Rutgers School of Business at Camden, 227 Penn Street, Camden, NJ 08102 *
The author acknowledges partial financial support from the David Whitcomb Center for Research in Financial Services of Rutgers University.
ABSTRACT
I investigate how a firm’s underwriting relationship affects its value. Firms that rely on underwriters’ reputation for equity issuance could damage themselves when the reputation is contaminated. Using firms with class action lawsuits, I demonstrate that disclosed misconduct of an underwriting client could cause investors to distrust its underwriter’s due diligence and thus decrease value of the underwriter’s other clients as well – underwriting spillover effect. The spillover effect is more severe for relatively opaque, cheap, and well-performed firms. It also increases the propensity of a firm to be sued, suggesting that the spillover effect is driven partially by investors’ anticipation of lawsuits.
I.
Introduction The role of an underwriter is imperative when it comes to equity issuance because they help build a strategical offering process, manage investor relations and communications, and certify legitimacy of business and financial structures of their client firms. Firms, therefore, tend to hire an investment bank with good reputation for their financial milestone, with an expectation that reputational capital of the bank would smoothen the process of selling securities. Especially for young, small, and opaque firms, they prefer certification from reputable banks because those banks can lower the firms’ informational cost of capital due to high uncertainty (see, e.g., Carter and Manaster, 1990; Chemmanur and Fulghieri, 1994; Fang, 2005; Fernando, Gatchev, and Spindt, 2005). In this respect, the reputable underwriters would affect the value of issuing firms positively through a higher quality assurance. The underwriting relationship, however, is not always constructive. It could deteriorate the value of underwriting client firms as the trusted reputation of an underwriter is tarnished. For example, Beatty, Bunsis, and Hand (1998) find that an SEC investigation of an underwriter penalizes not only itself, but also its past clients.2 The legal enforcement changes the market’s perception of underwriters’ practices and thus, future cash flows of the underwriter’s clients. Stock return volatility of client firms is evident after the announcement of the SEC investigation, suggesting that the level of information asymmetry between underwriting clients and investors could increase due to the troubled underwriters. It, however, does not provide a comparative analysis between the clients of sanctioned underwriters and those of non-sanctioned underwriters, but emphasize on differences in characteristics of underwriters. This paper, thus, focus on the dark side of underwriting relationships for client firms and investigate whether the tainted reputation of an underwriter leads to an unexpected loss of value of the client firms.
2
In their study, the authors document indirect penalties of an SEC investigation on underwriters. Direct monetary penalties are rather small, which are viewed not effective as a legal deterrent for market abuse by underwriters. However, indirect penalties such as reputational damage hand the underwriters a significant economic loss.
We employ securities class action lawsuits as the event where investors may suspect underwriters’ misrepresentation for their clients. Securities class action is a lawsuit filed by investors, typically shareholders, who bought and sold a company’s securities during a specific time span– called class period – and suffered from economic loss due to violations of the securities laws (e.g. financial misrepresentation, stock price manipulation, insider trading allegations, violations of Generally Accepted Accounting Principles (GAAP), bond/equity issuance related, etc.). Firms with securities class actions experience a dramatic decrease in stock price at the time of public discovery on their fraudulent activities – class period end – because of the nature of the lawsuits where stock price is artificially inflated. Furthermore, these negative market reactions could spread over firms that investors may suspect their practices in place. In this paper, we provide an explanation about the spillover effects based on underwriting relationships. – underwriting spillover hypothesis. In our analysis, the underwriting spillover hypothesis suggests that negative market reactions spill over into the firms hiring the same lead underwriters for their equity offerings as the accused firms do. The underlying mechanism is that if an underwriter failed in the due process on a client, it may do so in processing the other clients. The sample is constructed to combine firms employing the lead underwriters (SUW firms) for accused firms and firms with different underwriters (DUW firms) from the accused firms in public offerings, which occurred over the past 3 years of public recognition on the accused firms’ misconduct. Consistent with the underwriting spillover effect hypothesis, we find that SUW firms suffer from a loss of between 0.36% and 0.67% at the end of the class period. 3 The results still hold after controlling for firm characteristics, accusation types, and industry and year fixed effects. Also, the spillover effects of underwriting relationship are pronounced when the economic damages of the corresponding accused firms are larger, and the client firms are more opaque, cheaper, smaller, and better-performed in the past. Our findings imply that firms with greater uncertainty and inflated stock price fall off more.
3
The losses are calculated as 2-day cumulative abnormal returns (0, +1), using market model, market-adjusted model, and Fama-French Three factor model. Considering the SUW firms’ average market cap, the economic losses are around $5 to $10 million.
Filing a lawsuit against a firm may result in multiple suits for firms running similar businesses because investors could suspect that those business practices of the accused firms are widely applied within industries – called industry spillover. Recent studies document that the negative market reaction to the accused firms ripples through firms in the same industry (Grande and Lewis, 2009) or rival firms in competitive industries (Goldman, Peyer, and Stefanescu, 2012). To control for the industry spillover effects, we classified the sample into two groups, same industry firms and different industry firms, matched with the corresponding accused firms.4 The evidence strongly supports the underwriting spillover hypothesis. We also test the industry competition effect that rival firms in less competitive industries benefit from the event, while in competitive industries, negative market reactions prevail. Our findings still hold after considering industry competition. Finally, we investigate the underlying mechanism of the underwriting spillover effects. The spillover effects could be motivated by changes in market perception about the lead underwriter’s due diligence, raising informational cost of capital for its clients, thereby increasing volatility of stock price. Additionally, investors anticipate that the other clients could be the next litigation targets due to the underwriter’s deficiency in screening and monitoring. Malatesta and Thompson (1985) document that stock price reactions reflect partial anticipation by investors on corporate events such acquisitions.5 Therefore, we examine whether the underwriting spillover effects increase the odds of firms being sued. Firms engaging with the same underwriter as the accused firms do are more likely to be sued than firms with different underwriters within 5 years after the public disclosure of wrongdoings.6 Furthermore, we show that the same underwriter-firms are more likely to switch to reputable underwriters for future offerings.
4
We use the Fama-French 48 industry classifications. Other corporate events that could be highly anticipated by investors and thus reflected in price include debt offerings (Chaplinsky and Hansen, 1993), lawsuits (Gande and Lewis, 2009), and bankruptcy (Lang and Stulz, 1992). 6 Considering relevance of potential lawsuits with current ones, we limit the potential lawsuit period to 5 years after the current event. 5
Firms that lost their confidence in the tarnished underwriters leave for new, reputable banks, which would alleviate the cost related to information asymmetry because investors receive better-quality assurance.7 Whether banking relationships create value for client firms is extensively studied. Equity underwriters help firms raise capital by lowering informational cost (Chemmanur and Fulghieri, 1994), monitoring (Hansen and Torregrosa, 1992), and building a base of institutional investors (Benveniste and Spindt, 1989; Benveniste and Wilhelm, 1990; Cornelli and Goldreich, 2001; Ritter and Welch, 2002; Gao and Ritter, 2010). In addition, the underwriters could stabilize the price of stock for their clients after initial public offering (Aggarwal, 2000). However, Fernando, May, and Megginson (2012) document that firms could be adversely affected by investment banking relationships as well. Using the collapse of Lehman Brothers in September 2008, they provide evidence that the financial meltdown of the bank has affected its clients over and above the negative impact on equity market in general. This paper adds to literature in this regard with a broader setting in which more than 250 underwriters and 2636 unique firms are examined without possible biases, raised when focusing on a specific event with a single underwriter during a certain period. Secondly, this paper sheds light on the spillover effects. The spillover effects are widely studied in different contexts, including industry spillover (Lang and Stulz, 1992; Bittlingmayer and Hazlett, 2000; Song and Walkling, 2000; Gande and Lewis, 2009; Goldman, Peyer, and Stefanescu, 2012), technology spillover (Bernstein and Nadiri, 1989; Bloom, Schankerman, and Reenen, 2013), and international spillover (Keller, 1998; Coe and Helpman, 1995; Park, 2004, Berrospide, Black, and Keeton, 2016). In contrast to those spillover effects, underwriting spillover is centered on the relationship between underwriters and firms, which is adversely affected to client firms as underwriters’ reputational capital is damaged (Beatty, Bunsis, Hand, 1998; Fernando, May, and Megginson, 2012).
7
Krigman, Shaw, and Womack (2001) find that firms move up to reputable underwrites and buy influential analyst coverage from the new ones.
Thirdly, we document investors’ ability to anticipate future corporate events. The extant literature has studied various events, including acquisitions (Schipper and Thompson, 1983; Malatesta and Thompson, 1985), debt offerings (Chaplinsky and Hansen, 1993), and bankruptcy (Lang and Stulz, 1992). We consider corporate litigations as the events that investors could anticipate and capitalize beforehand. Finally, we add empirical evidence that firms switch to new underwriters for better-quality assurance, consistent with Krigman, Shaw, and Womack (2001) that firms graduate to higher reputation underwriters and acquire influential analyst coverage. The rest of the paper is organized as follows: In section 1, we discuss underwriting relationships and the mechanism of underwriting spillover. Section 2 describes our data and sample. In section 3 and 4, we present our empirical results. Section 5 concludes.
II.
Underwriting Spillover We discuss the importance of underwriters for the firms issuing securities, and then deliberate the mechanism of underwriting spillover. Focusing on the quality assurance role of underwriters, we review how important is the reputation of underwriters for clients. Next, we summarize the information spillover literature to develop the underwriting spillover hypothesis that companies employing the same underwriters of the accused firms experience negative market reactions to stock price as investors change their perception regarding the underwriters’ due diligence.
2.1. Reputation on Quality Assurance Information asymmetry provides a market for intermediaries selling quality assurance (see, e.g., Akerlof, 1970, Leland and Pyle, 1977, Campbell and Kracaw, 1980; Booth and Smith, 1986). For their successful business in certificating one’s quality, investment banks specialized in selling securities are required to build reputational capital by repeatedly engaging in the equity offering market (Beatty and Ritter,
1986; Chemmanur and Fulghieri, 1994). These reputation-building activities are critical to the underwriter in several reasons, such as a high premium (Stoughton, Wong, and Zechner, 2001; Fang, 2005), high-quality clients (Titman and Trueman, 1986), and investor network or book-building (Sherman and Titman, 2002).8 The issuing firm hiring a reputable bank could raise capital smoothly due to a lower informational cost. Especially, the reputable investment bank can offer the services with a better quality, and attract more investors who, otherwise, are reluctant to get into a non-visible, young, or small firm with a relatively high uncertainty in its business. The extensive empirical literature has been studying the relationship between the reputation of investment banks and the values of the IPO securities.9 The relationship, however, is not always constructive. One of the downside risks arises because of the agency problem between financial intermediaries and investors. Yet the intermediaries understand that lost investor confidence would lead to loss of future business.10 A few papers has documented reputationrelated costs for client firms. Beatty, Bunsis, and Hand (1998) show indirect penalties – reputational damages – for underwriters brought under a SEC investigation. Underwriters targeted by the SEC experience large declines in IPO market share and have their IPO clients suffer tougher regulatory scrutiny. The authors attribute these economic losses to an unexpected drop in the value of the underwriter’s assurance-based reputation capital. Fernando, May, and Megginson (2012) consider the collapse of Lehman Brothers to measure the value of investment banking relationships. Analyzing firms that have received underwriting, advisory, analyst, and market-making services from Lehman, they find only equity underwriting clients bear a substantial loss – around -5% – around the sudden fall of Lehman. Finally, Gopalan, Nanda, and Yerramilli (2011) examine the reputation-related costs in the loan syndication market. A large-scale bankruptcy among a lead bank’s borrowers adversely affects the bank’s reputation.
8
Stoughton, Wong, and Zechner (2001) find that high-quality firms are willing to pay for information acquisition in the stock market, since they benefit through better quality reputation in the product market. 9 See, Logue (1973), Beatty and Ritter (1986), Carter and Manaster (1990), and Tinic (1988) 10 See, Klein and Leffler (1981), Kreps and Wilson (1982), Rogerson (1983), Allen (1984), Chemmanur and Fulghieri (1994), and Pichler and Wilhelm (2001)
Focusing on equity-underwriting relationships, suggested by literature, this study examines whether the tainted reputation of an underwriter leads to an unexpected loss of value of its client firm – underwriting spillover effects. With class action lawsuits, we attempt to expand our investigation regarding underwriting spillover effects in a broader setting. The class action lawsuit helps us identify an event period during which an accused firm commits to a misconduct – class period. At the end of the class period when the firm’s wrongdoing is disclosed, the stock price of the firm plummets. This negative market perception on an underwriting client transfers to the other underwriting clients if they employ a same underwriter within the past 3 years.
2.2. Mechanism of Underwriting Spillover Spillover effects in general could occur through various mechanisms within or across firms, industries and even markets by technical innovations, policy changes, relationships or networks. 11 The spillover channels could affect firms, markets, economies either positively (positive market reactions) or negatively (negative market reactions). 12 Looking at the dark side of the spillovers, for example, a bankruptcy filing announcement of a firm leads to negative market reactions on other firms in the industry – bankruptcy spillover (Lang and Stulz, 1992). Antitrust enforcement constrained Microsoft’s dominance in operating systems and applications market, which inflicted capital losses on the entire computer sector – policy spillover (Bittlingmayer and Hazlett, 2000). A technical innovation also could transform an entire industry or a country as well (Coe and Helpman, 1995; Bloom, Schankerman, and Reenen, 2013). These information and knowledge transfers take place various ways.
11
See, Bernstein and Nadiri (1989), Lang and Stulz (1992), Keller (1998), Bittlingmayer and Hazlett (2000), Song and Walkling (2000), and Park (2004) 12 The positive side of the spillover includes improved productivity, cost-cutting, and efficient operation through technology innovation spillover and increased investment opportunities by deregulation through policy spillover (Bloom, Schankerman, and Reenen, 2013; Huang, 2008).
The banking relationship is one of the spillover mechanisms where information could spread across different firms or markets connected to a specific bank. Berrospide, Black, and Keeton (2016) examine the mortgage lending of banks operating in multiple cities and states. They find that local economic shocks can be transmitted to other regions through banks’ internal capital markets. A loan supply shock in other markets could cause the multimarket bank with limited funds to re-allocate money in all operating markets. Another example is the bankruptcy of Lehman Brothers in 2008. The bank’s failure shocked the market and led to a huge economic loss to its equity-underwriting clients.13 Those firms with stronger and broader security underwriting relationships with Lehman were more badly affected, supporting the underwriting spillover hypothesis. In this study, we attempt to explain the underlying mechanism of the underwriting spillovers. The spillover occurs in a combination of different reasons. First, investors distrust the underwriter of the accused firm in deficiency of due process, diminishing the bank’s reputational capital. Then, the downgraded market expectation on the bank’s quality assurance passes onto its past clients because of an increase in uncertainty. Therefore, negative market reactions are particularly strong to young, small, and depressed firms since they tend to rely more on reputation of underwriters for equity issuance. As a result, firms hiring the underwriters of the accused firms are more likely to suffer a loss from the disclosure of misconducts. Secondly, investors could anticipate a potential lawsuit targeting the other underwriting clients and capitalize possible damages prior to the lawsuit. Malatesta and Thompson (1985) find that market reactions indicate investors’ expectations on future corporate events. Especially, Grande and Lewis (2009) study industry spillover effect that a firm’s lawsuit filing in an industry signals to investors that suits against other firms in the same industry are following and the stock prices of related firms are adjusted downward accordingly. We, thus, hypothesize that the probability that firms with the same underwriters of the accused
13
In the paper, Fernando, May, and Megginson (2012) examine the negative impact of the event on the other business clients, such as lending, analyst, advisory, and market-making services. However, other client groups were not adversely affected.
firms are sued is higher than firms with underwriters different from those of the corresponding accused firms are. Finally, we also investigate how likely the value-loss clients replace their tarnished underwriters with new, high-quality underwriters in the future offerings. Issuing firms would internally discuss about their relationship with underwriters and switch to another underwriter if a reputational capital of the financial intermediary is damaged. Krigman, Shaw, and Womack (2001) document that firms graduate to higher reputation underwriters for better quality assurance as well as influential analyst coverage. Our third hypothesis, therefore, is that firms with the accused firms’ underwriters are more likely to switch to betterquality underwriters in subsequent offerings.
III.
Data Description and Sample Selection 3.1. Securities Class Action Lawsuits A securities class action lawsuit is a case brought pursuant to Federal Rule of Civil Procedure 23 on behalf of a group of persons who purchased the securities of a company during a specified period – the class period. The complaint generally contains allegations that the company and/or certain of its officers and directors violated one or more of the federal or state securities laws. A suit is filed as a class action because the relevant parties are too many to be resolved in separate cases. The Securities Class Action Clearinghouse (SCAC) provides detailed information on securities class action lawsuits, such as prosecution, defense, and settlement of federal class action securities fraud litigation.14 The database tracks securities class actions filed in federal courts since passage of the Private Securities Litigation Reform Act of 1995.15 For empirical analysis, we hand-collected key information on
14
Find more details here, http://securities.stanford.edu/ The Private Securities Litigation Reform Act of 1995 (PSLRA) implemented several substantive changes in the United States, affecting certain cases brought under the federal securities laws, including changes related to pleading, discovery, liability, class representation, and awards fees and expenses. Before the PSLRA, plaintiffs – usually 15
each lawsuit, including a lawsuit filing date, class period, case status, and allegations. Our sample only consists of public firms traded on New York Stock Exchange (NYSE), American Stock Exchange (AMEX), and NASDAQ from 1996 to 2013.
3.2. Equity Underwriting Relationships As literature suggests, we focus on equity underwriting relationships between issuing firms and investment banks. The SDC Platinum New Issues database offers details of issuance, such as issue date, issuing firm, proceeds, identification of underwriters, the role of underwriters. We exclude American depositary receipts, closed-end funds, units, real estate investment trusts, limited partnerships, financial firms, utilities and firms with an offering price less than $5. For each equity offering, we identify its lead underwriters, offer date and firm identification. Individual stock’s market information is obtained from CRSP, which includes return, trading volume and shares outstanding. We also obtain accounting figures in financial statements from COMPUSTAT. One of the important reasons for an issuing firm that hires top underwriters is to buy influential analyst coverage (Krigman, Shaw, and Womack, 2001). Therefore, we use I/B/E/S for brokerage house information, including the number of analyst following, the level of the consensus recommendation, as well as earnings announcement dates for firms.
3.3. Descriptive Statistics
shareholders – could file a lawsuit by simply showing their economic damages (stock price drops) and sought potential fraud during the litigation process. After the PSLRA, however, plaintiffs were required to provide evidence of fraud.
To investigate the spillover effects of underwriting relationships, we identify a sued firm’s lead underwriter and its underwriting clients around a class period ending date on which the accused firm’s misconduct is discovered.16 The time flow of a firm’s equity offering, and lawsuit is depicted in Figure 1.17
[Insert Figure 1 here]
We first select accused firms issuing equity within 3 years prior to the class period ending date (CPE) and identify firms hiring the same lead underwriter for equity offerings as the sued firms do over the past 3 years before the CPE. We assume those non-sued underwriting client firms went through similar due process to the accused firm by the same underwriter. For matching firm analysis, we also collect non-sued firms issuing equity, but with underwriters different from the accused firm over the same period. Based on literature, we control for industry and avoid industry spillover effects (Gande and Lewis, 2009; Goldman, Peyer, and Stefanescu, 2012). Figure 2 shows our sample classification.
[Insert Figure 2 here]
Table I provides the summary statistics for lawsuits from 1996 to 2013. Panel A reports the number of equity offerings with lawsuits. 730 of the total 14,477 equity offerings are qualified with our sample selection criteria in which a firm’s class period ending date is less than 3 years of its offering date. We restrict our sample firms with firm characteristics similar to the corresponding accused firm for each lawsuit and delete firms with earnings announcement two days before or after the CPE, which further drops the
Each lawsuit document provides three key dates – a class period starting date, a class period ending date, and a filing date. The class period starting date is the date on which an accused firm begins wrongdoings. The class period ending date is the date on which public discovers the misconducts. The filing date is the date a lawsuit filed against the firm. 17 For some companies (3,560 out of 15,728 in total observations), equity was offered after the CPS. 16
final number of equity offerings to 680. The sample firms are comprised of firms with the same lead underwriters (SUW) of the accused firms and firms with different lead underwriters (DUW) from the corresponding accused firms. The yearly total number of sample firms is 6,100, SUW firms (3,406) and DUW firms (2,694). 18 Panel B presents various accusation types for the sample lawsuits. Among 680 lawsuits, 672 suits are related to false or misleading statements, and 587 cases are alleged of artificial stock price inflation. In addition, 212 lawsuits charge investment banks as co-defendants. Panel C shows the number of days for relevant event periods. Class period is, on average (median), 346 (295) days. Moreover, it takes 125 (39) days for investors to file a lawsuit since the end of the class period.
[Insert Table 1 here]
Table 2 compare firm characteristics for our sample. Size is calculated as price times the number of shares outstanding, reported in millions. Book-to-market uses a firm’s common stock amount relative to its size. Leverage is long-term liabilities divided by total assets. Firm Performance is calculated as abnormal returns over the past six months of a quarter prior to the CPE, using the market-adjusted model.19 Trading Volume measures the three-month moving average of the monthly trading volume, reported in millions. Analyst Coverage is the average number of analysts following a sample firm, and Analyst Consensus is the mean value of analyst recommendations. Spread is calculated as closing ask price minus closing bid price and divide it by closing midpoint. Industry competition is measured using the Herfindahl-Hirschman index (HHI). Number of offerings is the total number of public offerings for a firm before the CPE. Offering period is the number of days between the latest public offering for a firm and the CPE. These variables are measured at the beginning of the quarter prior to the CPE.
18
Those are not the numbers of unique firms since there are firms involved in different lawsuits. Following Wharton Research Data Services (WRDS), the market-adjusted model uses abnormal returns defined in excess of CRSP value-weighted market return, assuming market beta of 1. 19
[Insert Table 2 here]
Looking at the differences between SUW firms and DUW firms, SUW firms are larger, more leveraged. There are significant differences in trading volume and analyst coverage. Significantly higher trading volume and analyst coverage for SUW firms than those of DUW firms suggest that SUW firms are more visible and liquid. For Spread, a proxy for asymmetric information, SUW firms are narrower than DUW firms. The product market competition for SUW firms are more severe as well.
IV.
Empirical Results 4.1. Market Reactions to Public Disclosure of Frauds This section investigates the spillover effects of underwriting relationships on firm value. We first calculate market reactions to the accused firms’ public disclosure on sample firms using the market model, the market-adjusted model, and Fama-French (1993) three-factor model.
(1)
𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡 − 𝑅𝑀𝑅𝐹𝑡
(2)
𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡 − 𝛼𝑖 − 𝛽1 𝑅𝑀𝑅𝐹𝑡
(3)
𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡 − 𝛼𝑖 − 𝛽1 𝑅𝑀𝑅𝐹𝑡 − 𝛽2 𝑆𝑀𝐵𝑡 − 𝛽3 𝐻𝑀𝐿𝑡
where Rit is the daily return of the firm i in excess of a risk-free return over day t; RMRFt is the excess return of CRSP value-weighted market return over risk-free return over day t; SMBt, and HMLt are returns on zeroinvestment factor-mimicking portfolios for size and book-to-market, respectively. The coefficients α and β are ordinary least squares (OLS) estimates based on each model over the 100-day estimation period with at least 70-day non-missing returns. Next, we compute cumulative abnormal returns (CARs) as follows;
(4)
𝜏2 𝐶𝐴𝑅𝑖 [𝜏1 , 𝜏2 ] = ∑𝑡=𝜏 𝐴𝑅𝑖𝑡 1
where day 0 is the day of public release on frauds (i.e. class period ending date).
[Insert Table 3 here] Table 3 reports CARs of the sample firms with various event windows. Panel A presents market model based CARs around the CPE. The accused firms suffer from huge negative market reactions to their frauds. They are down, on average, 8% on the disclosure and further down to around 23% a day later. 3day CARs produce about a negative 25%. Mean CARs calculated on market-adjusted model and FamaFrench Three Factor model generate similar results in Panel B and Panel C, respectively. Looking at SUW firms and DUW firms, 1-day CARs are consistently negative for SUW firms, while DUW firms have mixed results. 2-day CARs are negative for both SUW firms and DUW firms at 1% level significance, even though the amount of a loss is two to four times bigger for SUW firms. 3-day CARs negative and significant for SUW firms, while DUW firms have positive and significant CARs. Overall, the results show that SUW firms experience negative market reactions to the incidents. On average, firms hiring the same lead underwriter as accused firms lost about 0.36% - 0.67% (2-day CARs) depending on the estimation model. The damage amount of underwriting spillover to past clients (i.e., SUW firms) is statistically and economically significant, worth around $4 - $7 million. 20 They suggest that investors downgrade their perception about equity underwriting clients with the same underwriter of the accused firms in that they seem not to be free of potential malpractice. It is consistent with Beatty, Bunsis, and Hand (1998) that an SEC investigation of an underwriter imposes indirect penalties on the underwriter and its past clients, particularly IPO clients. The results from the univariate tests may mislead because we do not
20
The mean market cap of SUW firms is $1,080,000.
rule out alternative explanations. We, thus, control for various factors that possibly affect the value of client firms during the event period.
4.2. Underwriting Spillover Effects on Firm Value To test the spillover effect of underwriting relationships on firm value, we estimate cross-section regressions for CARs on underwriting relationships around the CPE. We conduct the analyses with a combined sample of SUW firms and DUW firms. Standard errors are heteroskedasticity-robust and corrected for clustering at the event (lawsuit) level. We estimate the following equation:
(5)
CARi = β0 + β1SUWi + X'iB + ɛi
where CAR is the 1-day, 2-day, or 3-day cumulative abnormal returns, and day 0 is the event date (i.e. CPE). The cumulative abnormal returns are calculated on market model, market-adjusted model, Fama-French Three-Factor model. SUW is a dummy variable that is equal to one if a firm employed the same lead underwriter of the corresponding accused firm and zero otherwise (i.e. DUW firms). X is a group of control variables for various market and firm-specific characteristics found to affect firm returns. The details of the variables are in Appendix A. Table 4 reports the determinants of market reactions to the misconduct of accused firms on underwriting clients. The primary interest of variable is SUW. The coefficients on SUW are -0.223, -0.533, and -0.705 for 1-day, 2-day, and 3-day CARs, respectively. 21 They are statistically and economically significant, implying that the firm with a lead underwriter of the corresponding accused firm could suffer a loss of 0.22% in value on the date of public recognition of the sued firm’s wrongdoing, 0.53% a day after the disclosure, or 0.71% around the CPE, relative to the firm hiring different underwriters. The results are
21
We also test CAR (0, +2), and CAR (-1, +2) and find no significant differences.
consistent with the underwriting spillover hypothesis, which predicts that equity-underwriting relationships could affect the value of client firms. Stock prices of its past clients deteriorate significantly due to the underwriter’s reputation down.
[Insert Table 4 here]
Investors may perceive a firm’s misconduct differently, depending on the type of accusations, economic damages, industry, and so on. Therefore, we measure and control for the severity of the event (i.e. disclosure of wrongdoing) with a 2-day cumulative abnormal return (0, +1) of accused firms, where day 0 is the announcement date.22 Damage is positively correlated with the dependent variables, indicating that a more severe damage (i.e. larger negative market reactions to the news) to accused firms leads to greater negative market reactions to firms with recent equity offerings. Firm performance and book-tomarket ratio are negative and significant at the 1% level. Investors may suspect highly performed firms of artificially inflated stock price. Spread is a proxy for a firm’s opaqueness. The negative coefficients on Spread imply that firms that are opaquer (i.e., higher spread) tend to suffer larger losses due to higher uncertainty. All three estimation models generate qualitatively similar results. Using only SUW firms, we run the same regressions to examine the firm characteristics that affect firm value in Panel B. The results show that the negative market reactions to SUW firms are positively correlated to the amount of damage to corresponding sued firms. Firms that performed well over the past 6 months suffer from a loss of value, suggesting the winning performance may not be driven by real operating performance, but artificially inflated. The underwriting spillover effects are severe for relatively more opaque, cheaper (high book-to-market), and smaller firms, implying that firms with high informational cost and depressed asset value tend to experience more declines in value. It is consistent with Fernando, May,
22
We also use different windows such as (0, 0), (-1, 0), and (-1, +1) for a measure of damage, and we find no significant differences.
and Megginson (2012) that those firms depend more on the reputation of their underwriters and thus, negative market reactions reflect the increased market’s perception of uncertainty.
4.3. Industry Spillover Now we examine whether our results are driven by the industry spillover effect. Gande and Lewis (2009) weigh on the spillover effects of fraud firms on industry. Investors suspect that firms in the same industry may run business with similar mal-practices. Therefore, shareholders partially anticipate potential lawsuits against other firms in the same industry and take part of these losses prior to a lawsuit filing date.
[Insert Table 5 here]
Table 5 reports the estimation results with specifications including industry classification variable. Same Industry, a dummy variable, indicates a firm belonging to the same industry as the sued firm.23 In Panel A, the coefficients on Same Industry are insignificant, suggesting that the disclosure of a firm’s wrongdoing does not affect the value of others in the same industry. In contrast, the underwriting spillover effects (SUW) still hold. Next, we test industry competition hypothesis, which predicts that a rival firm will benefit from the public disclosure of a firm’s fraud in the form of reduced competition from the sued firm.24 It suggests that in competitive industries, the industry spillover effect prevails, resulting in negative returns to rival firms. Therefore, we rank each industry from 1 (most competitive) to 10 (least competitive) based on HerfindahlHirschman index each year and label the top two quintiles as high competition industry. Panel B of Table 5 presents the estimations of key variables and their interaction terms. The coefficients on SUW are negative and significant, while both the high competition and the interaction terms are insignificant. The findings
23
We use Fama-French 48 industry classification. Goldman, Peyer, and Stefanescu (2012) find that the information spillover effect varies among different industries, depending on the level of competition. 24
are consistent with the underwriting spillover hypothesis, suggesting that the underwriting spillover effect dominates the industry spillover effect.
4.4. Different Underwriter Reputation Recovery Period Our empirical analysis adapts a restriction that underwriters recover their reputation in 2 years. In the sample data construction, the accused firms’ underwriters stay out of our sample for 2 years so that DUW firms have no underwriting relationships with the tarnished underwriters during the restriction period. We follow the work of Gopalan, Nanda, and Yerramilli (2011) that investigate the effect of poor performance on financial intermediary reputation by estimating the effect of large-scale bankruptcies among a lead arranger’s borrowers on its subsequent syndication activity. They find a significant reputation damage for lead arrangers of failed borrowers and it lasts for up to 2 years. Now we relax the restriction for recovery and re-test our regressions.
[Insert Table 6 here]
Table 6 reports regression results adopting 1-year recovery of underwriter reputation. Consistent with the underwriting spillover hypothesis, the coefficients on SUW are negative and significant. Interestingly, the same industry variables turn out significant on 2-day and 3-day CARs. It is possible that firms in the same industry with their corresponding accused firms are added as their underwriters are now qualified for the sample. The results, however, still support for the underwriting spillover hypothesis even after controlling for the possible industry spillover effects.
V.
Implications of Underwriting Spillovers 5.1. Propensity for Class Action
Now we turn our attention to whether the underwriting spillover effects manifest investors’ anticipation on further lawsuits for other clients of the trouble underwriter. Gande and Lewis (2009) document that shareholders could exploit lawsuits against other firms in the same industry by capitalizing losses prior to a lawsuit filing date. Therefore, we hypothesize that an underwriter’ lack of ability to monitor and scrutinize one client could apply to the other clients where investors suspect possible mismanagement and manipulations, thereby increasing the probability of class action lawsuits for them. Therefore, we estimate the following equation.
(6)
Prob. (Sued=1)i = β0 + β1SUWi + β2Same Industryi + X'iB + ɛi
where the dependent variable is a dummy variable, equal to one if a firm is sued within 5 years of fraud disclosure by its corresponding sued firm, and zero otherwise. Our variable of interest is SUW, indicating whether a firm hired the same lead underwriter of its corresponding accused firm for the most recent public offering. Same Industry is another indicator variable for firms that belong to the same industry as their respective accused firms. Other control variables remain the same as previous regressions. Excluding duplicates of future lawsuits leave us 15,523 observations.
[Insert Table 7 here]
Table 7 presents the propensity for class action lawsuits to be filed against sample firms. We separate our sample into two groups for the analysis based on the filing dates and class period ending dates. Firms could be sued before their corresponding accused firms are litigated (filing dates), but after the fraudulent activities by the accused firms are recognized (class period ends). It is because the filing periods – the number of days between a class period ending date and a filing date – vary among different lawsuits, depending on various reasons, such as types of accusations, economic damage, and presence of institutional
lead plaintiff. First two columns show estimations for filing dates and last two columns for class action ending dates. Consistent with the underwriting spillover hypothesis, the coefficients on SUW are positive and statistically significant, implying that the propensity for the firms with the same lead underwriter to be sued increase by about 7% and 5%, for lawsuit filing dates and for class period ending dates respectively.25 Several control variables (firm characteristics) turns out to be related to the propensity for class action lawsuits. Firm size is positively related to the odds of firms to be sued. And the greater trading volume, analyst coverage, and analyst consensus, the higher the possibility of lawsuits. The results suggest that larger firms that typically are liquid, visible, highly recommend have deep pockets and could pay greater settlement amounts if they lose the cases.
5.2. Switching Underwriters Firms that suffer a big loss due to their business relationship with underwriters could switch them to better-quality ones. Krigman, Shaw, and Womack (2001) explain the replacement with a firm’s strategic decision to graduate to higher reputation underwriters and buy quality analyst coverage. Fernando, Gatchev, and Spindt (2005), however, find that issuers do not just switch to a higher ability underwriter for its SEO, but try to match themselves with underwriters based on characteristics at time of issuance. The underwriting spillover hypothesis suggests that firms with questionable underwriters are more likely to change to ones that are more reputable. Therefore, we investigate whether firms switch their troubled underwriters in subsequent offerings and thus, estimate the following equation.
(7)
Prob. (Switch=1)i = β0 + β1SUWi + β2Same Industryi + X'iB + ɛi
Table 8 presents the test results of switching underwriters. The dependent variable is a dummy variable equal to one if firms hire a new lead underwriter for its subsequent public equity offering after the
25
The marginal effects of SUW are calculated at the means.
CPE and zero otherwise. In column (1), the coefficient on SUW is positive and significant at 1% level, indicating that SUW firms are more likely to switch to another lead underwriter. Same industry or damage does not affect the decision to change a firm’s underwriter. Higher performed firms as well as higher rated firms tend to move to a new underwriter in their subsequent offerings. It is possible that the firms have earned their own reputation and so graduate to better underwriters for the future issuance. Also, consistent with Krigman, Shaw, and Womack (2001), lower analyst coverage firms are more like to switch to another underwriter with an expectation of getting influential analyst coverage.
[Insert Table 8 here]
We, thus, test whether these firms seek to hire better-able underwriters for their next equity offerings (Krigman, Shaw, and Womack, 2001; Fernando, Gatchev, and Spindt, 2005). Column (2) and (3) employ a reputation measure based on market shares in underwriting business. Top10 indicates that an underwriter belongs to a top-tier underwriting business based on market shares each year. Quality measures a size of each underwriter’s market share relative to entire market each year. The results are qualitatively similar to the previous column, supporting the explanations of firms to switch to higher-quality underwriters.
VI.
Conclusion This paper provides evidence on how a firm’s underwriting relationship affects its value. Given the role of underwriters in equity issuance, such as quality assurance, certification, and monitoring, I show that underwriters misrepresenting current clients could damage their past clients, consistent with the underwriting spillover hypothesis. The underwriting spillover effects are especially severe for firms that are opaquer, cheaper, smaller and better-performed in the past. The evidence is in line with prior literature on underwriter reputation (Carter and Manaster, 1990; Chemmanur and Fulghieri, 1994; Fang, 2005; Fernando, Gatchev, and Spindt, 2005; Fernando, May, and Megginson, 2012; Golubov, Petmezas, and
Travlos, 2012), supporting the argument that reputable banks can mitigate information asymmetry between investors and issuing firms. Our results remain strong after controlling for so-called industry spillover, which finds that investors’ view on a firm easily carry onto other firms in the same industry. This information transfer within the industry vary among industries. Goldman, Peyer, and Stefanescu (2012) find that rivals in less (more) competitive industries benefit (suffer) from the negative event such as litigation. We also test the industry competition hypothesis and find that the underwriting spillover effect dominates the competition effect. The underwriting spillover effects lead to two consequences for client firms. First, the odds of being sued for underwriting clients are higher than otherwise similar firms. The spillover effects, therefore, simply reflect investors’ anticipation of potential lawsuits. This is consistent with Grande and Lewis (2009) that market reactions indicate investors’ expectations on future corporate events and, therefore, the negative spillover effects could be a result of investors anticipation on potential lawsuits. Secondly, the value-loss clients are more likely to switch their underwriters for better quality in future offerings. Firms that issued equity with underwriters in question could suffer a loss due to raising informational costs that the underwriters are supposed to reduce. Therefore, underwriting spillover effects increase the odds of client firms seeking highly reputable underwriters for subsequent offerings.
References Aggarwal, R., 2000, Stabilization activities by underwriters after initial public offerings, Journal of Finance, 55, 1075-1103. Akerlof, G., 1970, The market for “Lemons”: quality uncertainty and the market mechanism, The Quarterly Journal of Economics, 84, 488-500. Allen, F., 1984, Reputation and product quality, RAND Journal of Economics, 15, 311-327. Beatty, R., Bunsis, H., Hand, J., 1998, The indirect economic penalties in SEC investigations of underwriters, Journal of Financial Economics, 50, 151-186. Beatty, R., Ritter, J., 1986, Investment banking, reputation, and the underpricing of initial public offerings, Journal of Financial Economics, 15, 213-232. Benveniste, L., Spindt, P., 1989, How investment bankers determine the offer price and allocation of new issues, Journal of Financial Economics, 24, 343-361. Benveniste, L., Wihelm, W., 1990, A comparative analysis of IPO proceeds under alternative regulatory environments, Journal of Financial Economics, 28, 173-207. Bernstein, J., Nadiri, I., 1989, Research and development and intra-industry spillovers: an empirical application of dynamic duality, Review of Economic Studies, 56, 249-269. Berrospide, J., Black, L., Keeton, W., 2016, The cross-market spillover of economic shocks through multimarket banks, Journal of Money, Credit and Banking, 48, 957-988. Bittlingmayer, G., Hazlett, T., 2000, DOS Kapital: has antitrust action against Microsoft created value in the computer industry?, 55, Journal of Financial Economics, 329-359. Bloom, N. Schankerman, M., Reenen, J., 2013, Identifying technology spillovers and product market rivalry, 81, Econometrica, 1347-1393. Booth, J., Smith, R., 1986, Capital raising, underwriting and the certification hypothesis, Journal of Financial Economics, 15, 261-281. Campbel, T., Kracaw, W., 1980, Information production, market signaling, and the theory of financial intermediation, Journal of Finance, 35, 863-882. Cater, R., Manaster, S., 1990, Initial public offerings and underwriter reputation, Journal of Finance, 45, 1045-1067. Chaplinsky, S., Hansen, R., 1993, Partial anticipation, the flow of information and the economic impact of corporate debt sales, Review of Financial Studies, 6, 709-732. Chemmanur, T. J., Fulghieri, P., 1994, Investment bank reputation, information production, and financial intermediation, Journal of Finance, 49, 57-79.
Coe, D., Helpman, E., 1995, International R&D spillovers, European Economic Review, 39, 859-887. Cornelli, F., Goldreich, D., 2001, Bookbuilding and strategic allocation, Journal of Finance, 56, 2337-2369. Fama, E., French, K., 1993, Common risk factors in the returns on stocks and bonds, Journal of Financial Economics, 33, 3-56. Fang, L., 2005, Investment bank reputation and the price and quality of underwriting services, Journal of Finance, 60, 2729-2761. Fernando, C., Gatchev, V., Spindt, P., 2005, Wanna dance? How firms and underwriters choose each other, Journal of Finance, 60, 2437-2469. Fernando, C., May, A., Megginson, W., 2012, The value of investment banking relationships: evidence from the collapse of Lehman Brothers, Journal of Finance, 67, 235-270. Gande, A., Lewis, C., 2009, Shareholder-initiated class action lawsuits: shareholder wealth effects and industry spillovers, Journal of Financial and Quantitative Analysis, 44, 823-850. Gao, X., Ritter, J., 2010, The marketing of seasoned equity offerings, Journal of Financial Economics, 97, 33-52. Goldman, E., Peyer, U., Stefanescu, I., 2012, Financial misrepresentation and its impact on rivals, Financial Management, 41, 915-945. Golubov, A., Petmezas, D., Travlos, N., 2012, When it pays to pay your investment banker: new evidence on the role of financial advisors in M&As, Journal of Finance, 67, 271-311. Gopalan, R., Nanda, V., Yerramilli, V., 2011, Does poor performance damage the reputation of financial intermediaries? Evidence from the loan syndication market, Journal of Finance, 66, 2083-2120. Hansen, R., Torregrosa, P., 1992, Underwriter compensation and corporate monitoring, Journal of Finance, 47, 1537-1555. Huang, R., 2008, Evaluating the real effect of bank branching deregulation: comparing contiguous counties across US state borders, Journal of Financial Economics, 87, 678-705. Keller, W., 1998, Are international R&D spillovers trade-related? Analyzing spillovers among randomly matched trade partners, European Economic Review, 42, 1469-1481. Klein, B., Leffler, K., 1981, The role of market forces in assuring contractual performance, Journal of Political Economy, 89, 615-641. Kreps, D., Wilson, R., 1982, Reputation and imperfect information, Journal of Economic Theory, 27, 253279. Krigman, L., Shaw, W., Womack, K., 2001, Why do firms switch underwriters?, Journal of Financial Economics, 60, 245-284.
Lang, L., Stulz, R., 1992, Contagion and competitive intra-industry effects of bankruptcy announcements: an empirical analysis, Journal of Financial Economics, 32, 45-60. Leland, H., Pyle, D., 1977, Informational asymmetries, financial structure, and financial intermediation, Journal of Finance, 32, 371-387. Logue, D., 1973, On the pricing of unseasoned equity issues: 1965-1969, Journal of Financial and Quantitative Analysis, 8, 91-103. Malatesta, P., Thompson, R., 1985, Partially anticipated events: a model of stock price reactions with an application to corporate acquisitions, Journal of Financial Economics, 14, 237-250. Park, J., 2004, International student flows and R&D spillovers, Economics Letters, 82, 315-320. Pichler, P., Wilhelm, W., 2001, A theory of the syndicate: form follows function, Journal of Finance, 56, 2237-2264. Ritter, J., Welch, I., 2002, A review of IPO activity, pricing, and allocations, Journal of Finance, 57, 17951828. Rogerson, W., 1983, Reputation and product quality, Bell journal of Economics, 14, 508-516. Sherman, A., Titman, S., 2002, Building the IPO order book: underpricing and participation limits with costly information, Journal of Financial Economics, 65, 3-29. Song, M., Walkling, R., 2000, Abnormal returns to rivals of acquisition targets: a test of the ‘acquisition probability hypothesis’, Journal of Financial Economics, 55, 143-171. Stoughton, N., Wong, K. P., Zechner, J., 2001, IPOs and product quality, Journal of Business, 74, 375-408. Tinic, S., 1988, Anatomy of initial public offerings of common stock, Journal of Finance, 43, 789-822. Titman, S., Trueman, B., 1986, Information quality and the valuation of new issues, Journal of Accounting and Economics, 8, 159-172.
Appendix A. Variable definitions Variable names
Definitions
SUW
An indicator variable which equals one if the firm employs the underwriter of the corresponding accused firm.
Same Industry
An indicator variable which equals one if the firm belongs to the industry of the corresponding accused firm.
Damage
The cumulative abnormal returns (0, +1), where day 0 is the class period ending date, calculated using the market-adjusted model.
Past Firm Performance
The cumulative abnormal returns over the past 6 months of a quarter prior to CPE, using the market-adjusted model.
Size
The natural log of the firm’s market capitalization (the price times the number of shares outstanding).
Book-to-Market
The natural log of the firm’s common stock amount relative to the size.
Leverage
The firm’s long-term liabilities divided by its total asset
Spread
The firm’s closing ask price minus closing bid price and divide it by closing midpoint.
Trading Volume
3 month moving average of monthly trading volume scaled by the firm’s market value.
Analyst Coverage
The average number of analyst following on the firm
Analyst Consensus
Mean value of analyst recommendations
High Competition
A rank variable for industry competition, based on Herfindahl-Hirschman index each year, and the top two quintiles are labeled as high competition industry.
Top 10
An underwriter’s market share is ranked within the highest 10 underwriters’ group each year
Quality
The underwriter’s market share relative to overall market in terms of equity issuance proceeds.
Figure 1. Time order of dates related to equity issuances and lawsuits
Equity Issuance
Class Period Start (CPS)
Class Period End (CPE)
Lawsuit Filing
Figure 2. Classification of sample firms
Same underwriter
Different underwriter
Same as sued firms
(1)
(2)
Different from sued firms
Industry
Underwriter
(3)
(4)
Table 1. Sample description for equity offerings and lawsuits This table provides the summary statistics for securities class action lawsuits from 1996 to 2013. Panel A presents the number of sued equity offerings. # Public Equity Offering is the number of equity offerings each year. # Sample Equity Offering is the number of equity offerings where an issuing firm’s class period ending date does not exceed 3 years of an offering date. # Sample firms is the number of relevant firms in our sample. SUW Firms are the firms hiring the same lead underwriter for their equity offerings as the accused firms. DUW Firms are those employing lead underwriters different from the accused firms in public offerings. Panel B reports the type of accusations. # Lawsuits is the number of sample lawsuits for a specific accusation. Panel C presents the number of days for the class period, from a class period starting date to a class period ending date, and the filing period, from a class period ending date to a lawsuit filing date. Panel A. Number of equity offerings with lawsuits # Sample Firms
Year
# Public Equity Offering
# Sample Equity Offering (Final)
# SUW Firms
# DUW Firms
Total
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
1,560 1,289 901 961 761 535 534 575 778 666 664 669 294 814 883 707 826 1,060
29 (29) 39 (39) 70 (67) 61 (59) 51 (51) 195 (164) 34 (31) 25 (25) 30 (30) 23 (22) 19 (17) 21 (21) 21 (14) 15 (14) 29 (29) 26 (26) 16 (14) 26 (23)
162 231 339 272 211 388 156 142 158 118 139 175 185 140 140 177 119 154
563 423 442 247 127 114 67 55 88 76 89 96 59 41 41 51 46 69
725 654 781 519 338 502 223 197 246 194 228 271 244 181 181 228 165 223
Total
14,477
730 (680)
3,406
2,694
6,100
Panel B. Lawsuit characteristics Accusations
# Lawsuits
False/Misleading Statement
672
Artificially Inflated Stock Price
587
IPO/SEO
321
M&A
75
Insider Trading
227
SEC 1933 Section 11
260
GAAP Violation
172
Investment Bank Co-Defendant
212
Panel C. Event periods # Days
# OBS
Mean
Median
Min.
Max.
Class Period
680
346
295
0
1,762
Filing Period
680
125
39
0
1,265
Table 2. Financial characteristics for sample firms This table present firms characteristics of sample firms. Sued firms are firms with class action securities lawsuits. SUW firms employ the sued firms’ lead underwriter for their equity issuance. DUW firms place their equity with underwriters other than the sued firms’ lead underwriters. Size is calculated as price times the number of shares outstanding, reported in millions. Book-to-market uses a firm’s common stock amount relative to its size. Leverage is long-term liabilities divided by total assets. Firm Performance is calculated as abnormal returns over the past six months of a quarter prior to the CPE, using the marketadjusted model. Trading Volume measures the three-month moving average of the monthly trading volume, reported in millions. Analyst Coverage is the average number of analysts following the sample firm, and Analyst Consensus is the mean value of analyst recommendations for the sample firm. Spread is the firm’s closing ask price minus closing bid price and divide it by closing midpoint. Industry competition is measured using the Herfindahl-Hirschman index. Number of offerings is the total number of public offerings for a firm before the end of the class period. Offering period is the number of days between the latest public offering for the sample firm and the class period ending date for the corresponding accused firm. The variables are measured at the beginning of the quarter prior to the class period ending date. (1) Sued Firms (N = 680)
(2) SUW Firms (N = 5,187)
(3) DUW Firms (N =10,541)
Mean
Median
STD
Mean
Median
STD
Mean
Median
STD
Size (million)
1,340
438
3,410
1,080
577
2,180
384
228
576
Book-to-Market
0.35
0.27
0.34
0.33
0.29
0.22
0.33
0.29
0.19
Leverage
0.14
0.01
0.22
0.16
0.06
0.20
0.13
0.05
0.17
Total Asset (million)
1,270
198
6,520
921
312
2,230
264
125
560
Firm Performance
-0.04
-0.04
0.10
-0.01
0.00
0.08
0.00
0.00
0.07
Trading Volume (million)
24.60
7.95
57.60
13.20
5.69
28.60
4.08
1.84
10.00
Analyst Coverage
8.13
7.00
5.77
7.82
7.00
4.89
4.65
4.00
3.11
Analyst Consensus
4.16
4.00
0.68
4.05
4.00
0.65
4.23
4.00
0.68
Spread
0.01
0.01
0.03
0.01
0.00
0.04
0.02
0.01
0.04
Industry Competition
0.05
0.04
0.03
0.05
0.04
0.06
0.06
0.04
0.06
Number of Offerings
1.38
1.00
0.70
1.37
1.00
0.69
1.20
1.00
0.50
Offering Period
430
392
251
491
442
285
620
644
307
Table 3. The stock market reactions to disclosure of corporate fraud Table3 presents cumulative abnormal returns (CARs) over event windows, where day 0 refers the class period ending date. Panel A, B, and C report CARs for sued firms, SUW firms (firms hiring the same lead underwriter as sued firms) , and DUW firms (firms hiring underwriters other than the sued firms’ lead underwriters), calculated using market model, market adjusted model, and Fama-French Three-Factor Model, respectively. t-statistics are reported. ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively.
Sued Firms (N = 680) Event Window
Mean CAR (%)
SUW Firms (N = 5,187)
DUW Firms (N =10,541)
t-stat
Mean CAR (%)
t-stat
Mean CAR (%)
t-stat
Panel A: Market Model (0, 0)
-8.08***
-13.84
-0.08
-1.24
-0.03
-0.70
(0, +1)
-23.14***
-27.32
-0.48***
-5.33
-0.25***
-3.96
(-1, +1)
-24.94***
-26.94
-0.21*
-1.80
0.04
0.51
Panel B: Market Adjusted Model (0, 0)
-8.46***
-14.62
-0.24***
-3.62
-0.01
-0.26
(0, +1)
-23.78***
-28.48
-0.67***
-7.37
-0.17***
-2.67
(-1, +1)
-25.10***
-26.74
-0.13
-1.04
0.22***
2.78
0.05
1.21
Panel C: Fama-French Three-Factor Model (0, 0)
-8.09***
-13.88
-0.01
-0.04
(0, +1)
-23.11***
-27.52
-0.36***
-3.98
-0.14***
-2.23
(-1, +1)
-24.66***
-26.55
-0.07
-0.57
0.27***
3.43
Table 4. The spillover effects of underwriting relationships on firm value This table presents the determinants of the spillover effects of underwriting relationships on firm value. The dependent variables are cumulative abnormal returns (CARs), calculated using the Fama-French three factor model. SUW firms employ the sued firms’ lead underwriter for their equity issuance. Damage is measured as the sued firm’s 2-day CARs (0, +1) around the class period ending date (CPE), using the market-adjusted model. Firm Performance is calculated as abnormal returns over the past six months of a quarter prior to the CPE, using the market-adjusted model. Size is calculated as the natural log of the firm’s market capitalization. Book-to-market uses the natural log of the firm’s common stock amount relative to the size. Leverage is long-term liabilities divided by total assets. Spread is the firm’s closing ask price minus closing bid price and divide it by closing midpoint. Trading Volume measures the three-month moving average of monthly trading volume scaled by the firm’s market value. Analyst Coverage is the average number of analysts following a sample firm, and Analyst Consensus is the mean value of analyst recommendations. t-statistics are reported. ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively. Panel A. Fama-French Three Factor Model Cumulative Abnormal Returns SUW Damage Firm Performance Size Book-to-Market Leverage Spread Trading Volume Analyst Coverage Analyst Consensus Constant
CAR (0, 0) -0.223**
CAR (0, +1) -0.533***
CAR (-1, +1) -0.705***
(-2.41)
(-3.98)
(-3.99)
0.705***
0.756**
2.524***
(3.12)
(2.35)
(5.12)
-4.116***
-6.312***
-11.417***
(-4.97)
(-5.47)
(-8.12)
-0.071
-0.040
0.021
(-1.27)
(-0.48)
(0.21)
-0.832***
-2.062***
-3.047***
(-3.15)
(-4.84)
(-5.20)
0.551***
1.079***
0.450
(2.88)
(3.90)
(1.20)
1.927
-0.716
-1.661
(1.05)
(-0.29)
(-0.47)
0.789***
0.620*
0.480
(2.98)
(1.76)
(1.09)
0.009
0.036*
0.007
(0.56)
(1.65)
(0.27)
0.087
0.005
-0.116
(1.53)
(0.07)
(-1.13)
1.009
-0.486
-0.515
(0.81)
(-0.23)
(-0.22)
Clustering
Lawsuit
Lawsuit
Lawsuit
Industry fixed
Yes
Yes
Yes
Year fixed Observations
Yes 15,728
Yes 15,728
Yes 15,728
R-squared
0.016
0.015
0.037
Panel B. SUW Firms Only Cumulative Abnormal Returns Damage Firm Performance Size Book-to-Market Leverage Spread Trading Volume Analyst Coverage Analyst Consensus
CAR (0, 0) 0.719*
CAR (0, +1) 1.109**
CAR (-1, +1) 2.753***
(1.87)
(1.98)
(3.59)
-4.422***
-4.511**
-8.284***
(-3.66)
(-2.52)
(-3.76)
0.180*
0.171**
0.338**
(1.73)
(1.99)
(2.01)
-0.889**
-1.878***
-3.739***
(-2.18)
(-3.09)
(-4.52)
0.454
1.856***
0.981
(1.24)
(3.66)
(1.63)
-3.335**
-6.928***
-11.225**
(-2.55)
(-3.37)
(-2.18)
0.410
-0.161
0.286
(1.27)
(-0.37)
(0.45)
-0.006
0.027
-0.027
(-0.24)
(0.88)
(-0.76)
-0.141
-0.330**
-0.449**
(-1.29)
(-2.04)
(-2.09)
-2.799
-5.315*
-4.208
(-1.26)
(-1.69)
(-1.15)
Clustering
Lawsuit
Lawsuit
Lawsuit
Industry fixed
Yes
Yes
Yes
Year fixed Observations
Yes 5,187
Yes 5,187
Yes 5,187
R-squared
0.017
0.025
0.029
Constant
Table 5. Effects of Industry Spillover on Firm Value This table presents the determinants of the spillover effects of underwriting relationships on firm value. The dependent variables are cumulative abnormal returns (CARs), where day 0 is the class period ending date, calculated using the Fama-French three factor model. SUW is a dummy variable, equal to one if a firm employs the sued firms’ lead underwriter for their equity issuance and zero otherwise. Same Industry is an indicator variable, equal to one is a firm belongs to the same industry (Fama-French 48 industry classification) as the sued firm and zero otherwise. Industry competition is measured using the Herfindahl-Hirschman index (HHI) and ranked into 10 deciles. The top two ranked industries are labeled as High Competition. The other control variables remain the same, not reported. t-statistics are reported. ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively. Panel A. Industry Spillover Cumulative Abnormal Returns SUW Same Industry
CAR (0, 0)
CAR (0, +1)
CAR (-1, +1)
-0.232**
-0.515***
-0.705***
(-2.40)
(-3.72)
(-3.93)
-0.068
-0.202
-0.448
(-0.36)
(-0.77)
(-1.19)
0.071
-0.101
0.055
(0.25)
(-0.25)
(0.10)
Clustering
Lawsuit
Lawsuit
Lawsuit
Industry fixed
Yes
Yes
Yes
Year fixed
Yes
Yes
Yes
Observations
15,728
15,728
15,728
R-squared
0.016
0.015
0.037
SUW*Same Industry
Panel B. Industry Competition Cumulative Abnormal Returns CAR (0, 0)
CAR (0, +1)
CAR (-1, +1)
-0.338*
-1.214**
-1.319**
(-1.84)
(-2.38)
(-2.07)
-0.266
-0.785
-0.324
(-0.47)
(-1.05)
(-0.33)
0.521
1.156
0.946
(0.97)
(1.55)
(0.91)
Clustering
Lawsuit
Lawsuit
Lawsuit
Industry fixed
Yes
Yes
Yes
SUW High Competition SUW*High Competition
Year fixed
Yes
Yes
Yes
Observations
1,812
1,812
1,812
R-squared
0.031
0.052
0.059
Table 6. Underwriter Reputation Recovery This table presents the determinants of the underwriting spillover effects on firm value with limiting an underwriter’ reputation recovery period to one-year. The dependent variables are cumulative abnormal returns (CARs), calculated using the Fama-French three factor model. SUW firms employ the sued firms’ lead underwriter for their equity issuance. Same Industry is an indicator variable, equal to one is a firm belongs to the same industry (Fama-French 48 industry classification) as the sued firm and zero otherwise. Damage is measured as the sued firm’s 2-day CARs (0, +1) around the class period ending date (CPE), using the market-adjusted model. Firm Performance is calculated as abnormal returns over the past six months of a quarter prior to the CPE, using the market-adjusted model. Size is calculated as the natural log of the firm’s market capitalization. Book-to-market uses the natural log of the firm’s common stock amount relative to the size. Leverage is long-term liabilities divided by total assets. Spread is the firm’s closing ask price minus closing bid price and divide it by closing midpoint. Trading Volume measures the three-month moving average of monthly trading volume scaled by the firm’s market value. Analyst Coverage is the average number of analysts following a sample firm, and Analyst Consensus is the mean value of analyst recommendations. t-statistics are reported. ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively. Cumulative Abnormal Returns SUW Same Industry SUW*Same Industry Damage Firm Performance Size Book-to-Market Leverage Spread Trading Volume Analyst Coverage Analyst Consensus Constant Clustering
CAR (0, 0)
CAR (0, +1)
CAR (-1, +1)
-0.211**
-0.625***
-0.473***
(-2.21)
(-4.53)
(-2.82)
-0.185
-0.701**
-0.894**
(-0.99)
(-2.51)
(-2.30)
0.196
0.521
0.770
(0.69)
(1.35)
(1.47)
0.633***
0.769**
2.178***
(2.85)
(2.54)
(4.94)
-4.346***
-6.625***
-11.095***
(-5.42)
(-6.26)
(-8.37)
0.006
0.068
0.141
(0.11)
(0.84)
(1.38)
-0.553*
-1.706***
-3.326***
(-1.95)
(-3.73)
(-4.46)
-0.001
0.704**
0.523
(-0.01)
(2.45)
(1.41)
2.612
-0.710
-2.483
(1.46)
(-0.30)
(-0.74)
0.745***
0.537*
0.318
(2.91)
(1.65)
(0.76)
-0.006
0.008
-0.029
(-0.43)
(0.41)
(-1.14)
-0.005
-0.225**
-0.488***
(-0.07)
(-2.38)
(-4.07)
-0.572
-0.625
0.242
(-0.50)
(-0.36)
(0.11)
Lawsuit
Lawsuit
Lawsuit
Industry fixed
Yes
Yes
Yes
Year fixed
Yes
Yes
Yes
Observations
16,508
16,508
16,508
R-squared
0.019
0.019
0.033
Table 7. Propensity for Class Action Lawsuits This table presents the determinants of the probability of a firm to be sued. The dependent variable is equal to one if a firm is sued, and zero otherwise. SUW firms employ the sued firms’ lead underwriter for their equity issuance. Same Industry is an indicator variable, equal to one is a firm belongs to the same industry (Fama-French 48 industry classification) as the sued firm and zero otherwise. Damage is measured as the sued firm’s 2-day CARs (0, +1) around the class period ending date (CPE), using the market-adjusted model. Firm Performance is calculated as abnormal returns over the past six months of a quarter prior to the CPE, using the market-adjusted model. Size is calculated as the natural log of the firm’s market capitalization. Book-to-market uses the natural log of the firm’s common stock amount relative to the size. Leverage is long-term liabilities divided by total assets. Spread is the firm’s closing ask price minus closing bid price and divide it by closing midpoint. Trading Volume measures the three-month moving average of monthly trading volume scaled by the firm’s market value. Analyst Coverage is the average number of analysts following a sample firm, and Analyst Consensus is the mean value of analyst recommendations. t-statistics are reported. ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively. The Odds of Firms to Be Sued Lawsuit Filing Date SUW
(1)
(2)
(3)
(4)
0.695***
0.655***
0.507***
0.483***
(11.50)
(10.07)
(8.10)
(7.20)
Same Industry SUW*Same Industry Damage Firm Performance Size Book-to-Market Leverage Spread Trading Volume Analyst Coverage Analyst Consensus Constant Observations
Class Period Ending Date
-0.136
-0.100
(-1.16)
(-0.84)
0.242
0.154
(1.64)
(1.00)
0.119
0.117
0.002
0.001
(0.85)
(0.84)
(0.02)
(0.01)
-1.237***
-1.221***
-0.175
-0.166
(-3.29)
(-3.24)
(-0.45)
(-0.42)
0.163***
0.165***
0.242***
0.243***
(4.33)
(4.36)
(6.27)
(6.29)
-0.012
-0.012
-0.199
-0.199
(-0.07)
(-0.07)
(-1.05)
(-1.04)
-0.268
-0.267
-0.290
-0.289
(-1.55)
(-1.55)
(-1.64)
(-1.64)
-3.372***
-3.378***
-2.629**
-2.635**
(-3.16)
(-3.17)
(-2.36)
(-2.36)
0.325***
0.331***
0.571***
0.574***
(2.93)
(2.98)
(5.23)
(5.26)
0.039***
0.039***
0.035***
0.035***
(5.02)
(4.95)
(4.43)
(4.39)
0.133***
0.132***
0.142***
0.141***
(3.02)
(3.00)
(3.12)
(3.11)
-19.601
-19.612
-20.963
-20.977
(-0.04)
(-0.03)
(-0.04)
(-0.04)
15,523
15,523
15,523
15,523
Pseudo R-squared
0.10
0.10
0.09
0.09
Table 8. Switching Underwriters This table presents the determinants of the probability of a firm to switch underwriters. The dependent variable is equal to one if a firm switches its underwriter, and zero otherwise. SUW firms employ the sued firms’ lead underwriter for their equity issuance. Same Industry is an indicator variable, equal to one is a firm belongs to the same industry (Fama-French 48 industry classification) as the sued firm and zero otherwise. Damage is measured as the sued firm’s 2-day CARs (0, +1) around the class period ending date (CPE), using the market-adjusted model. Firm Performance is calculated as abnormal returns over the past six months of a quarter prior to the CPE, using the market-adjusted model. Size is calculated as the natural log of the firm’s market capitalization. Book-to-market uses the natural log of the firm’s common stock amount relative to the size. Leverage is long-term liabilities divided by total assets. Spread is the firm’s closing ask price minus closing bid price and divide it by closing midpoint. Trading Volume measures the three-month moving average of monthly trading volume scaled by the firm’s market value. Analyst Coverage is the average number of analysts following a sample firm, and Analyst Consensus is the mean value of analyst recommendations. t-statistics are reported. ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively.
SUW Same Industry SUW*Same Industry Damage Firm Performance Size Book-to-Market Leverage Spread Trading Volume Analyst Coverage Analyst Consensus Constant Observations
(1) Switching Underwriters
(2) Top 10 Underwriters
(3) Quality Underwriters
0.315***
0.346***
0.517***
(6.41)
(6.27)
(5.27)
-0.068
-0.109
-0.204
(-0.69)
(-0.95)
(-1.37)
-0.031
-0.036
0.130
(-0.23)
(-0.23)
(0.54)
0.052
-0.027
0.074
(0.60)
(-0.26)
(0.46)
1.338***
0.239
-0.033
(4.76)
(0.77)
(-0.06)
0.055**
0.385***
0.706***
(2.07)
(12.48)
(14.27)
-0.201
-0.011
-0.110
(-1.57)
(-0.07)
(-0.50)
0.724***
0.795***
0.937***
(5.11)
(4.82)
(3.54)
-1.193**
-0.701
0.498
(-2.07)
(-1.15)
(0.57)
-0.343***
-0.351***
-0.219
(-3.12)
(-2.93)
(-1.37)
-0.055***
-0.071***
-0.110***
(-8.89)
(-10.19)
(-10.90)
0.150***
0.074**
0.023
(4.46)
(2.06)
(0.44)
-4.548***
-22.504***
-13.011***
(-4.00)
(-40.05)
(-12.95)
15,728
15,728
15,728
(Pseudo) R-squared
0.0811
0.0850
0.075
