The Quarterly Review of Economics and Finance 45 (2005) 48–64

Contagion effects of the world’s largest bankruptcy: the case of WorldCom
Aigbe Akhigbea,∗ , Anna D. Martinb , Ann Marie Whytec a Department of Finance, College of Business Administration, University of Akron, Akron, OH 44325, USA b Department of Finance, Charles F. Dolan School of Business, Fairfield University, USA c Department of Finance, School of Business, University of Central Florida, USA Received 16 June 2003; received in revised form 23 December 2003; accepted 27 July 2004 Available online 26 November 2004

Abstract On July 19, 2002 WorldCom sought protection from its creditors when it filed for Chapter 11 bankruptcy, earning the distinction as the largest bankruptcy filing in U.S. history. The events surrounding this history-making occurrence provide an important opportunity to examine the repercussions for WorldCom’s stakeholders. We especially focus on the valuation effects of the WorldCom failure on exposed financial institutions for their important monitoring roles as institutional investors and creditors. Despite the heightened uncertainty facing investors during this period, we find that the market is remarkably efficient in distinguishing among the various types of stakeholders. In particular, institutional investors and creditors are largely unaffected by the events, which is expected based on the benefit of diversification. In contrast, large and key competitors are adversely affected by the events, which may be attributed to scrutiny of rivals that are perceived to be facing similar problems. Furthermore, for large and key competitors, these results indicate that contagion effects dominate competitive effects. © 2004 Published by Board of Trustees of the University of Illinois.
JEL classification: G33; G14 Keywords: WorldCom; Bankruptcy; Contagion; Institutional investors; Creditors; Competitors

∗

Corresponding author. Tel.: +1 330 972 6883; fax: +1 330 972 5970. E-mail address: aigbe@uakron.edu (A. Akhigbe).

1062-9769/$ – see front matter © 2004 Published by Board of Trustees of the University of Illinois. doi:10.1016/j.qref.2004.07.002

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

49

1. Introduction Few could have imagined that a company with the stature and the size of WorldCom could collapse so quickly. WorldCom had attained a market value of $180 billion, was the largest Internet carrier, and was the second largest long-distance carrier. Nonetheless, barely five months passed from the time any widespread news report indicated that trouble might be ahead to the time they filed for bankruptcy. On February 6, 2002, the New York Times published an article that focused on WorldCom’s aggressive accounting in reporting revenue. On July 19, 2002, WorldCom, with $103.9 billion in assets as of December 31, 2001, made history as the largest bankruptcy in U.S. history, surpassing Enron, which at the time of its filing had $63.1 billion in assets. WorldCom’s problems stemmed, in part, from its highly publicized admission that it had overstated profits by $3.8 billion. Long before the firm’s accounting irregularities came to light, however, WorldCom and most of its competitors in the telecommunications industry were being hurt by negative market forces.1 Among these negative forces were overcapacity in their networks, the slowing economy, which had reduced business demand for telecommunications services, and the ongoing price wars that had reduced consumer long-distance rates considerably. Indeed, WorldCom’s bankruptcy filing added to an already substantial list of telecommunications firms that had filed for bankruptcy protection in recent years, including Northpoint Communications Group and Global Crossing. Several telecommunications firms were also being probed by the Securities and Exchange Commission (SEC), including Qwest Communications International and Global Crossing, further contributing to a generally unstable environment in the industry. The existing literature provides evidence that bankruptcy filings have significant repercussions for both the bankrupt firm and a variety of associated stakeholders, including rival firms, client firms, and creditors. For example, Lang and Stulz (1995) and Ferris, Jayaraman, and Makhija (1997) show that bankruptcy announcements generate a dominant industry contagion effect; that is, the stock prices of competitors decline because the bankruptcy reveals adverse information about industry asset values and future prospects. Datta and Iskandar-Datta (1995) also show that stockholders and unsecured creditors are adversely affected by the filing, whereas secured debtholders are unaffected. These studies analyze the effects of a sample of bankruptcies on different classes of stakeholders. There is evidence, however, that in some cases the sheer magnitude and scope of the bankruptcy of a single firm is sufficiently significant to warrant separate examination.2 In such instances, the failure of a single firm provides a natural laboratory for understanding the ramifications of the event. For example in 1990, Laventhol and Horwath, then the seventh
1 Despite these negative forces, there were some positive trends in the industry. Some telecommunications firms had emerged from bankruptcy protection, including Covad Communications Group Inc. Another positive development was Berkshire Hathaway Inc.’s purchase of $100 million in convertible bonds of Level 3 Communications Inc. This was generally viewed as a positive signal in the industry. 2 Prior studies have also examined the contagion and competitive effect of events (other than bankruptcy) involving a single firm. For example, Johnston and Madura (2000) show that the Citigroup–Travelers merger in 1999 had significant implications for other banks, insurance companies and securities firms, and Akhigbe and Martin (2002) find that Microsoft’s acquisitions within the internet/online services segment had adverse valuation effects for internet/online services rivals.

50

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

largest public accounting firm in the U.S., filed for Chapter 11 bankruptcy. Baber, Kumar, and Verghese (1995) find that the bankruptcy had significant negative price implications for the firm’s clients for two reasons. First, they argue that investors rely on auditors to recover future investment losses. Thus, the failure invalidates this insurance function performed by the auditor. Second, the failure caused investors to reassess the quality of the firm’s audits, triggering a negative share price response for their clients. Other examples of major bankruptcies include the failure of Penn Square Bank in 1982 and Continental Illinois Corporation, a bank holding company, in 1984. In the case of Penn Square, the evidence generally indicates that other banks were adversely affected by the failure (Fraser and Richards, 1985; Karafiath and Glascock, 1989; Lamy and Thompson, 1986; Peavey and Hempel, 1988). However, the market’s reaction was not indiscriminate. In particular, banks with the greatest exposure to the failed bank, including upstream banks that had loan participations with Penn Square, were more adversely affected by the events surrounding the bank’s failure. In contrast, banks outside Penn Square’s economic region were largely unaffected by the events. In the case of Continental Illinois, Swary (1986) finds that the failure had a significant negative impact on other banks, particularly those with a relatively large amount of nonperforming assets. WorldCom’s bankruptcy is one that deserves special examination. Coming as it did on the heels of the Enron collapse, the firm’s demise occurred during a period of unprecedented investor awareness, anxiety, and uncertainty. Further, given that WorldCom’s bankruptcy is the largest in U.S. history, we expect this history-making event may have significant repercussions for the firm’s key stakeholders. The significance of WorldCom’s bankruptcy is further highlighted in a statement made by the chairman of the Federal Communications Commission, Michael Powell, shortly after the bankruptcy filing. Recognizing the integral role that WorldCom plays in the economy, the chairman issued a statement assuring the public that: . . . we do not believe this bankruptcy filing will lead to an immediate disruption of service to consumers or threaten the operation of WorldCom’s Internet backbone facilities. It is my understanding that WorldCom has obtained funding necessary to continue operations during the pendency of its bankruptcy proceeding . . . This Commission stands ready to intervene in bankruptcy proceedings as necessary to ensure that the bankruptcy court is aware of and considers our public interest concerns. The bankruptcy of WorldCom clearly has consequences for its shareholders, who watched their stocks fall from $6.97 on February 5, 2002, the day before the first negative news event regarding the firm came to light, to $0.83 on July 19, 2002 when the firm ultimately filed for bankruptcy. In this study, we examine how information released about WorldCom in the months prior to its bankruptcy filing affected institutional investors, creditors, and competitors. Despite the heightened uncertainty facing investors during this period, we find that the market is remarkably efficient in distinguishing among the various types of stakeholders. In particular, institutional investors and creditors are largely unaffected by the events leading to WorldCom’s failure. These results are consistent with the basic benefit of diversification; a single bankruptcy, even the world’s largest bankruptcy, should have no impact on well-diversified shareholders. In contrast, large and key competitors are

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64 Table 1 Events indicating concern over the future viability of WorldCom 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

51

February 6, 2002: Concerns over aggressive accounting practices March 11, 2002: Request for information from the SEC relating to accounting procedures and loans to officers April 3, 2002: 4% of overall work force to be eliminated April 22, 2002: Standard & Poor’s and Moody’s cut credit ratings April 30, 2002: Resignation of WorldCom CEO, Bernard Ebbers May 9, 2002: Standard & Poor’s and Moody’s cut credit rating to junk status May 21, 2002: Dividend payments and two tracking stocks to be eliminated June 4, 2002: Sales of assets and business units June 25, 2002: CFO fired after uncovering improper accounting of $3.8 billion in expenses over five quarters starting in 2001; 20% of overall work force to be eliminated; fraud charges filed by SEC July 1, 2002: Indication of further fraud regarding reversals of reserve accounts; lenders notify WorldCom that they could demand immediate repayment due to default July 19, 2002: Bankruptcy filing expected on the next business day

This table briefly describes 11 major event dates that indicate some degree of concern over the future viability of WorldCom. We obtain the event dates and descriptions from Lexis–Nexis news articles on events that led up to WorldCom’s bankruptcy filing.

adversely affected by the events, which may be attributed to scrutiny of those rivals that are perceived to be facing similar problems. Furthermore, contagion effects appear to dominate any effects from competitive repositioning, for large and key competitors. These results indicate that shareholders, analysts, and portfolio managers are considering how firms with indirect ties to a financially distressed company may be affected.

2. Analyzing the effects on key stakeholders News releases about the future viability of WorldCom began to be disseminated by the popular press on February 6, 2002, when there was an article published that focused on the aggressive accounting methods used by WorldCom in reporting revenue. Over the next few months, reports were released that indicated WorldCom may have engaged in a variety of fraudulent accounting tactics. A chronology of important events that calls into question the ongoing viability of WorldCom is provided in Table 1. These critical dates are depicted in a time series diagram of WorldCom’s stock price in Fig. 1. 2.1. Impact on institutional investors To put the extent of WorldCom’s institutional ownership in perspective, we note that as of December 31, 2001, there are 830 institutional investors that own 56.5% of WorldCom’s shares outstanding. By June 30, 2002, there are only 408 institutional investors that own 44.9% of their shares outstanding. Table 2 provides a partial list of WorldCom’s institutional investors. Specifically, the top 20 institutional owners and the number of shares as of December 31, 2001 are provided, along with the number of shares as of June 30, 2001 and an indication as to whether the largest institutional investors were making drastic changes in their WorldCom holdings.

52

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

Fig. 1. Time series of WorldCom stock price.

Table 2 Institutional investors with the largest holdings in WorldCom Institutional investor Wellington Management Co, LLP AXA Financial, Inc Barclays Bank PLC Allianz Dresdner Asset Mgmt AM Citigroup Inc State Str Corporation Vanguard Group JP Morgan Chase & Co Mellon Bank NA Oppenheimerfunds Inc Deutsche Bk Aktiengesellschaft Franklin Resources Inc State Farm Mut Automobile Ins College Retire Equities Invesco Capital Mgmt Inc Putnam Investment Management Lord, Abbett & Company Capital Research & Mgmt Co Primecap Management Company Merrill Lynch Inv Managers (NJ) Number of shares as of December 31, 2001 119,573,615 108,188,362 106,158,868 99,465,827 67,361,737 67,313,242 48,343,229 43,720,849 36,340,857 35,938,991 32,141,918 31,345,690 23,822,564 22,972,661 21,885,885 18,675,605 18,412,541 17,636,500 16,998,675 16,676,111 Number of shares as of June 30, 2002 58,470,090 358,688,259 59,451,051 154,839,807 31,768,349 27,143,370 15,097,670 8,169,460 8,895,508 64,600,221 15,345,315 11,259,811 0 18,502,355 0 30,638 5,612,598 31,500 39,371,805 0 Largest +/− in holdings − + − + − − − − − + − − − − − − − − + −

Source: Holdings are obtained from Thomson Financial’s Shareworld database.

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

53

Since shareholders of WorldCom’s institutional investors own equity positions in a firm that has equity positions in a variety of companies, theoretically, shareholders of institutional investors should be protected against the negative performance of one firm contained in a well-diversified portfolio, even in the extreme case where the negative performance leads to bankruptcy. Nonetheless, given the sheer magnitude and scope of the WorldCom bankruptcy along with an environment of heightened investor anxiety, shareholders of the institutional investors may suffer significant losses. We suspect that institutions making substantial changes to their exposure to WorldCom over the period of concern do not go unnoticed. In particular, the market may penalize those institutional investors that substantially increase their relative exposure to WorldCom. 2.2. Impact on creditors A particular class of firms that may be more vulnerable to a bankruptcy shock is the group of creditor firms. Since creditors are concerned with the probability of default, and recognizing the fact that a bankruptcy announcement increases this probability, shareholders of the affected creditor may become concerned. Therefore, we analyze the share price reaction of these creditors. Since creditors are higher in the priority structure in terms of their claim on the firm’s assets, the adverse effects of an announcement may be mitigated. However, Datta and Iskandar-Datta (1995) note that the information content of bankruptcy announcements is not necessarily the same for all types of debtholders. In other words, all creditors are not created equally. They find that secured creditors are not significantly affected by the bankruptcy announcement (in terms of the change in bond prices), whereas other classes of creditors, such as holders of convertible debt and unsecured debt, experience significant losses. Based on the findings of Datta and Iskandar-Datta (1995), the extent of the mitigating effect for creditors is likely to be a function of whether they are secured or unsecured. Further, the reaction may depend on the extent of creditor diversification (Winton, 1999). Winton finds that an institution’s credit risk depends on monitoring mechanisms that are in place as well as the extent of diversification. Thus, the impact of the events leading to the WorldCom bankruptcy on creditors is an empirical question. In WorldCom’s case, 35 creditors had a significant amount of exposure as shown in Table 3. The average credit exposure is approximately $1.5 billion per firm, and exposures range from just over $100 million to nearly $20.5 billion. The expected impact on these creditors is unclear. Creditors that maintain lending portfolios that are well diversified should not experience adverse valuation effects. However, it is possible that creditors suffer negative valuation effects, to the extent that they are not well diversified. For example, adverse affects may accrue to creditors with substantial exposure to WorldCom relative to their overall credit portfolio, and to creditors that specialize in lending to the telecommunications industry, to the extent that further bankruptcies or default in the industry is expected. 2.3. Impact on competitors Lang and Stulz (1995) and Ferris et al. (1997) show that bankruptcy announcements generate a dominant contagion effect; that is, the stock price of competitors declines because the bankruptcy reveals adverse information about industry asset values and future prospects.

54

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

Table 3 Creditors’ exposure to WorldCom Creditor JP Morgan Chase Mellon Financial Citigroup Bank of New York Bear Stearns Morgan Stanley Dean Witter State Street Goldman Sachs Deutsche Bank Suntrust Banks ABN Amro Holding Wilmington Trust Northern Trust UBS Wells Fargo Bank of America Firstar Credit Suisse First Boston Lehman Brothers FUNB-Phil Main West LB Merrill Lynch Fleet Boston Financial Mizuho Holdings BNP Paribas Intesabci Bank of Tokyo–Mitsubishi Verizon Communications UMB Financial Bank One Bank of Nova Scotia Credit Lyonnais Bayerische Landesbank Royal Bank of Scotland Lloyds TSB Bank Amount of exposure ($ millions) 20,473.4 6,700.2 5,273.1 2,722.9 2,720.0 2,028.4 2,020.0 1,520.0 1,439.6 1,220.0 956.3 750.0 649.9 369.2 352.1 333.7 298.3 272.6 270.3 186.5 171.6 155.5 150.3 150.3 150.3 150.3 140.2 121.2 112.3 100.2 100.2 100.2 100.2 100.2 100.2

They also provide evidence that bankruptcy announcements have competitive effects; that is, some competitors actually gain, including competitors in concentrated industries, which are more likely to benefit from the potential for increased demand and market share arising from the demise of a competitor. Nonetheless, Ferris et al. (1997) conclude that the contagion effects clearly dominate the competitive effects. Further, they find that even small firm bankruptcies have a dominant contagion effect for smaller sized competitors. Given the evidence in past research and the facts known about the telecommunications industry in general at the time of the bankruptcy filing, we hypothesize that the events leading to WorldCom’s bankruptcy may have a dominant contagion effect for competitors. As noted earlier, overcapacity and other adverse trends had created a generally unstable environment in the industry. Several other telecommunications firms had previously filed for bankruptcy,

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64 Table 4 Competitors of WorldCom Competitors Panel A: Largest firms in SIC 4813 Allegiance Telecom Inc AT&T Corp BCE Inc Bellsouth Corp Broadwing Inc Compania Anonima Nac Tel de Vene Centurytel Inc Citizens Communications Co Commonwealth Tele Entprs Inc Compania de Telecomm de Chile SA Deutsche Telekom AG Global Crossing Ltd Panel B: Key competitors AT&T General Communications Global Crossing

55

Level 3 Communications Inc Magyar Tavkozlesi Vallalat RT Nippon Telegraph & Telphone Corp Qwest Communications Intl Inc SBC Communications Inc Sprint Corp Surewest Communications Telefonica de Argentina SA Telefonica de Espana SA Telus Corp Time Warner Telecom Inc Verizon Communications Qwest Communication Sprint FON

but the magnitude of WorldCom’s filing may still result in strong industry contagion effects. However, given the allegations of accounting fraud surrounding the bankruptcy, significant competitive effects may dominate to the extent that other telecommunications companies are not perceived as engaging in fraudulent activities. In this instance as well, it is difficult to decipher the likely effect because the SEC was probing several other companies in the industry for fraud. In addition, given that WorldCom’s collapse came so soon after Enron’s collapse (which was largely attributable to accounting fraud), there may have been a perception among investors that the problems are systemic causing widespread negative effects. Another factor that may affect the reaction of rivals is that, given the fragility of the industry, the events leading to the collapse of a major player like WorldCom may signal increased likelihood of further industry consolidation. Under this scenario, smaller rivals are more likely to become acquisition targets, because they are easier and less costly to acquire, and the acquisition is less likely to be challenged on antitrust grounds than the acquisition of larger firms. Table 4 provides a partial list of WorldCom’s competitors. Panel A lists the largest firms who share the telecommunications industry SIC code 4813. Broadly speaking, these firms are considered to be rivals. There are a total of 96 publicly traded firms that are categorized as operating in SIC code 4813 according to Standard & Poor’s Compustat database. Panel B also lists a more focused set of rivals.3
3 We thank an anonymous reviewer for suggesting that a more focused set of competitors may reveal better results. We combine two sources to generate this potentially better set of competitors. A Merrill Lynch analyst report on January 4, 2002 specifically states that WorldCom’s major competitors are AT&T, Sprint FON, and to a lesser degree Qwest Communication. Industry comparison data in the NetAdvantage database by S&P indicates that WorldCom competitors as of May 31, 2001 are AT&T, General Communication, Global Crossing, and Sprint FON.

56

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

3. Data and methodology To gauge the fallout from information released about the future viability of WorldCom, we examine the response of portfolios returns for institutional investors, creditors, and industry rivals. We require that sample firms have daily stock returns available from CRSP between January 2, 2002 and December 31, 2002. This requirement provides us with a sample of 64 institutional investors, 22 creditors, and 96 competitors. We use the seemingly unrelated regression (SUR) technique developed by Johnston (1984). He shows that in the presence of contemporaneous correlation, the SUR methodology generates more efficient estimates. The first model we use to estimate the share price response is as follows:
11

Rp,t = β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + k=1 λkp Dkt + β4p δt Rm,t + ep,t (1)

where Rp,t is the return on the portfolio on day t; Rm,t + 1 the return on the market portfolio on day t + 1; Rm,t the return on the market portfolio on day t; Rm,t − 1 the return on the market portfolio on day t − 1; Dkt is 1 for event k, and 0 otherwise (Table 1 lists the date of each event); δt is 1 on the final event day (July 19, 2002) and all days following this event, and 0 otherwise; β0p the intercept term for the portfolio; β1p , β2p , β3p the lead, contemporaneous, and lagged market betas for the portfolio; λkp the coefficient measuring the abnormal return for event k for the portfolio; β4p the shift in systematic risk of the portfolio due to the bankruptcy filing; ep,t the disturbance term for the portfolio on day t. We estimate Eq. (1) using daily returns for the period from January 2, 2002 to October 9, 2002. The CRSP value-weighted NYSE/AMEX index is used to represent the market portfolio. Similar to the intervention analysis used in Box and Tiao (1975), Scholes and Williams (1977), Larcker, Gordon, and Pinches (1980), Saunders and Smirlock (1987) and Bhargava and Fraser (1998), we include the concurrent, as well as the lead and lagged market returns to control for nonsynchronous trading. We also include a dummy variable to account for a potential shift in systematic risk.4 Negative and significant coefficient estimates for λ1 to λ11 , which represent abnormal returns for each of the events, may be revealed for the portfolios of institutional investors and creditors. For rival firms, the expected sign on these coefficients is unclear as previously discussed. Positive and significant coefficient estimates for the risk shift parameters, β1 to β4 , indicate that the portfolio has greater market risk exposure in the period following WorldCom’s bankruptcy filing. The second model we use to estimate the abnormal share price movement is: Rp,t = β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + γcomb,p Dkt + β4p δt Rm,t + ep,t (2)
4 The advantage of the SUR approach over the ordinary least squares (OLS) method is that by adding this dummy variable it allows the returns and systematic risk following an event to be examined at the same time. We evaluate different risk shift dates and find qualitatively similar results to those reported in the paper. Furthermore, Durbin–Watson statistics do not indicate that first-order auto-correlation is a problem in the models.

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

57

where γ comb,p is 1 for each event k, and 0 otherwise; and all other variable were previously defined. The coefficient, γ comb , represents the average abnormal share price response to the set of important events related to the deterioration of WorldCom.

4. Results We evaluate the potential contagion effects related to the important events leading up to the bankruptcy of WorldCom that are displayed in Table 1. First, we analyze the impact of the information disseminated about WorldCom on their shareholders. These results are reported in Table 5. Second, we analyze whether portfolios of WorldCom stakeholder firms were affected. In Tables 6–8, we report the results for all the stakeholder firms in each category, and for several subgroups.5 4.1. Impact on WorldCom shareholders Panel A of Table 5 shows that the share price of WorldCom is significantly affected by only the most severe event day (July 1, 2002).6 Not until further fraud is uncovered and default occurs does the share price of WorldCom elicit a significantly negative reaction. It is likely that at this point in time that bankruptcy was unequivocal; this also explains why the actual bankruptcy filing did not generate a significant reaction. Panel B of Table 5 indicates that, on average, the events leading up to the bankruptcy have significant and negative valuation effects. 4.2. Impact on institutional shareholders In Table 2, we list WorldCom’s 20 largest institutional investors as a partial representation of institutional owners. Data for institutional ownership are from Thomson Financial’s Shareworld database that makes available quarterly institutional holdings based on the SEC’s requirement to file 13F forms. We identify all institutional owners with positions as of December 31, 2001, the end of the quarter immediately preceding the first release of information, and June 30, 2002, the end of the quarter immediately preceding their bankruptcy filing. In total, there are 893 institutional investors that owned WorldCom shares during one or both periods.
5 We evaluate equations (1) and (2) using the standard event window of t = −1 to +1. In doing so, we find that t = 0 captures the WorldCom reaction, whereas day +1 captures the reaction by stakeholders that are connected to WorldCom. Also, the intra-industry study of bankrupt firms by Lang and Stulz (1995) documents the most pronounced effects on day +1 among non-bankrupt firms during bankruptcy announcements, indicating the market needs time to assess the economic impact of the announcements. Thus, the results we report for WorldCom use t = 0 and the results we report for all the stakeholders use t = +1. 6 It should be noted that Eqs. (1) and (2) are adjusted by omitting the risk shift factor, since WorldCom’s stock was delisted on July 30, 2002. Furthermore, we examine their cumulative abnormal return surrounding the bankruptcy filing. The 3-day CAR [−1, +1] is 35.76%, statistically significant at the 1% level. This suggests that all price adjustments resulting from the publicized difficulties for WorldCom must have occurred prior to the official announcement of bankruptcy.

58

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

Table 5 Abnormal returns and risk shift for WorldCom Parameter Panel A: Abnormal returns for each event β0 β1 β2 β3 λ1 λ2 λ3 λ4 λ5 λ6 λ7 λ8 λ9 λ10 λ11 Adjusted R2 F-value Panel B: Abnormal returns across all events β0 β1 β2 β3 γ comb Adjusted R2 F-value WorldCom 0.0029 −1.0902 1.4231 1.5598 −0.1692 −0.0589 0.0230 −0.1267 0.0477 −0.0085 −0.0706 −0.1031 −0.0464 −0.7921 −0.0693 0.2547 4.05*** 0.0032 −1.1852 1.8331 2.0462 −0.1178 0.1265 5.52*** 0.24 −1.14 2.02** 2.27** −2.59*** t-statistic 0.23 −1.05 1.65* 1.82* −1.29 −0.45 0.23 −0.96 0.36 −0.06 −0.54 −0.78 −0.35 −5.95*** −0.50

This table shows the coefficient estimates for WorldCom in response to the 11 events that indicated concern over its future viability. We obtain the estimates using OLS with daily returns from January 2, 2002 to December 31, 2002. We estimate the abnormal returns and risk shift in two ways: Rt = β0p + β1 Rm,t+1 + β2 Rm,t + β3 Rm,t−1 + 11 k=1 λk Dkt + et (Panel A) and Rt = β0 + β1 Rm,t+1 + β2 Rm,t + β3 Rm,t−1 + γcomb Dkt + et (Panel B), where Rt is the return on WorldCom on day t; Rm,t + 1 , Rm,t , and Rm,t − 1 are market returns on the day t + 1, t, and t − 1, respectively; Dkt is a dummy variable equal to 1 for event k and 0 otherwise; β0p is the intercept, β1p , β2p , and β3p are the lead, contemporaneous, and lagged market betas for WorldCom; λk is WorldCom’s abnormal response to event k; γ comb is WorldCom’s average abnormal response to all the 11 events; and et is the error term. We also report the R2 , adjusted R2 , and F-values. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

In Table 1, we report the number of shares owned on December 31, 2001, the number of shares owned on June 30, 2002, and provide an indicator of whether the institutional investor was in the top 5% of institutional investors with the greatest increase (+) or decrease (−) in the number of WorldCom shares between these two periods. Interestingly, each of the 20 largest institutional investors clearly expressed their opinions on the future prospects for WorldCom, by either substantially reducing or substantially increasing their exposure to WorldCom. The vast majority, 16 of the top 20 holders, significantly reduced their exposure to WorldCom during the period. Only 4 of the top 20 holders significantly increased their exposure to WorldCom.

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64 Table 6 Abnormal returns and risk shift for institutional investors Parameter All institutions t-statistic Large decrease in exposure 0.0005 0.0616 0.8394 −0.0393 0.0001 −0.0005 −0.0004 −0.0003 −0.0005 −0.0001 −0.0004 −0.0005 −0.0006 −0.0001 −0.0000 0.1803 16 0.8820 125.60*** 0.0005 0.0615 0.8386 −0.0392 −0.0007 0.1811 16 0.8834 379.76*** 1.31 2.75*** 23.03*** −1.78* −0.18 3.92*** t-statistic Large increase in exposure 0.0008 0.0078 0.9909 0.0375 −0.0023 −0.0020 −0.0021 −0.0021 −0.0020 −0.0022 −0.0019 −0.0020 −0.0018 −0.0022 −0.0023 0.2064 16 0.8187 76.25*** 0.0008 0.0079 0.9914 0.0375 −0.0021 0.2059 16 0.7500 145.78*** 0.45 0.24 18.24*** 1.14 −0.83 2.99***

59

t-statistic

Panel A: Abnormal returns for each event 0.0005 1.70* β0 −0.0024 −0.12 β1 β2 0.8024 24.14*** β3 −0.0082 −0.41 0.0001 0.08 λ1 λ2 −0.0002 −0.10 −0.0001 −0.08 λ3 λ4 −0.0000 −0.01 −0.0003 −0.15 λ5 λ6 −0.0000 −0.00 −0.0001 −0.08 λ7 λ8 −0.0002 −0.09 −0.0002 −0.11 λ9 0.0000 0.01 λ10 λ11 0.0001 0.08 0.1758 4.17*** β4 N Adjusted R2 F-value 64 0.8927 139.58***

1.30 2.75*** 22.90*** −1.77* 0.03 −0.27 −0.19 −0.14 −0.26 −0.05 −0.21 −0.24 −0.30 −0.06 −0.01 3.88***

1.44 0.23 17.94*** 1.12 −0.88 −0.76 −0.80 −0.78 −0.74 −0.82 −0.72 −0.77 −0.68 −0.83 −0.84 2.95***

Panel B: Abnormal returns across all events 0.0005 1.72* β0 −0.0026 −0.13 β1 β2 0.8019 24.39*** β3 −0.0082 −0.41 γ comb −0.0012 −0.91 0.1763 4.23*** β4 N Adjusted R2 F-value 64 0.8948 426.19***

This table shows the coefficient estimates for portfolios of institutional investors in response to the 11 events that indicated concern over the future viability of WorldCom. We obtain the estimates using SUR and daily returns from January 2, 2002 to December 31, 2002. We estimate the abnormal returns and risk shift in two ways: Rp,t = β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + 11 λkp Dkt + β4p δt Rm,t + ep,t (Panel A) and Rp,t = k=1 β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + γcomb,p Dkt + β4p δt Rm,t + ep,t (Panel B), where Rp,t is the return on the portfolio of exposed firms on day t; Rm,t + 1 , Rm,t , and Rm,t − 1 are market returns on the day t + 1, t, and t − 1, respectively; Dkt is a dummy variable equal to 1 for event k and 0 otherwise; δt is equal to 1 on the final event day (July 19, 2002) and all subsequent days and 0 otherwise; β0p is the intercept, β1p , β2p , and β3p are the lead, contemporaneous, and lagged market betas for the portfolio; λkp is the portfolio’s abnormal response to event k; γ comb is the portfolio’s average abnormal response to all the 11 events; β4p is the portfolio’s shift in systematic risk; and ep,t is the error term. We also report the OLS R2 , adjusted R2 , and F-values. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

60

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

Table 7 Abnormal returns and risk shift for creditors Parameter All creditors t-statistic Least exposure 0.0007 0.1054 1.0833 −0.0395 −0.0027 −0.0028 −0.0028 −0.0027 −0.0026 −0.0027 −0.0028 −0.0028 −0.0029 −0.0027 −0.0032 0.3875 6 0.8406 88.88*** 0.0007 0.1055 1.0827 −0.0393 −0.0017 0.3882 6 0.8389 261.29*** 1.19 2.80*** 17.60*** −1.06 −0.60 4.98*** t-statistic 1.82* 2.81*** 17.67*** −1.06 −0.22 −0.38 −0.51 −0.39 −0.54 −0.38 −0.65 −0.59 −0.79 −0.42 −0.55 4.99*** Most exposure 0.0008 0.0560 1.2728 −0.0756 0.0016 0.0009 0.0005 0.0010 0.0005 0.0009 0.0002 0.0003 −0.0004 0.0008 0.0004 −0.0033 6 0.8417 89.64*** 0.0008 0.0561 1.2723 −0.0754 0.0006 −0.0028 6 0.8394 262.37*** 1.41 1.57 21.85*** −2.14** 0.23 −0.04 t-statistic 1.42 1.58 21.98*** −2.16** 0.48 0.29 0.14 0.32 0.15 0.29 0.05 0.09 −0.12 0.25 0.12 −0.05

Panel A: Abnormal returns for each event 0.0006 1.23 β0 0.0712 2.48*** β1 1.1463 24.40*** β2 β3 −0.0483 −1.70* λ1 0.0001 0.04 λ2 −0.0004 −0.15 −0.0007 −0.25 λ3 λ4 −0.0004 −0.14 −0.0008 −0.29 λ5 λ6 −0.0003 −0.13 −0.0010 −0.40 λ7 λ8 −0.0009 −0.33 −0.0014 −0.53 λ9 −0.0005 −0.18 λ10 λ11 −0.0008 −0.29 0.1782 2.99*** β4 N Adjusted R2 F-value 22 0.8883 133.48***

Panel B: Abnormal returns across all events β0 0.0006 1.23 β1 0.0713 2.47*** 1.1458 24.31*** β2 β3 −0.0482 2.64*** γ comb −0.0006 −0.29 0.1787 2.99*** β4 N Adjusted R2 F-value 22 0.8871 393.89***

This table shows the coefficient estimates for portfolios of creditors in response to the 11 events that indicated concern over the future viability of WorldCom. We obtain the estimates using SUR and daily returns from January 2, 2002 to December 31, 2002. We estimate the abnormal returns and risk shift in two ways: Rp,t = β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + 11 λkp Dkt + β4p δt Rm,t + ep,t (Panel A) and Rp,t = k=1 β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + γcomb,p Dkt + β4p δt Rm,t + ep,t (Panel B), where Rp,t is the return on the portfolio of exposed firms on day t; Rm,t + 1 , Rm,t , and Rm,t − 1 are market returns on the day t + 1, t, and t − 1, respectively; Dkt is a dummy variable equal to 1 for event k and 0 otherwise; δt is equal to 1 on the final event day (July 19, 2002) and all subsequent days and 0 otherwise; β0p is the intercept, β1p , β2p , and β3p are the lead, contemporaneous, and lagged market betas for the portfolio; λkp is the portfolio’s abnormal response to event k; γ comb is the portfolio’s average abnormal response to all the 11 events; β4p is the portfolio’s shift in systematic risk; and ep,t is the error term. We also report the OLS R2 , adjusted R2 , and F-values. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

As a reference point, the same analysis is conducted for the 20 largest Enron institutional investors as of June 30, 2001, the end of the quarter immediately preceding the first release of negative information on the future viability of Enron. In the precedent-setting case of Enron, only 10 of the top 20 institutional investors significantly decreased their exposure to Enron

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64 Table 8 Abnormal returns and risk shift for competitors Parameter SIC 4813 t-statistic Small rivals t-statistic Large rivals 0.0006 0.0516 1.1–700 −0.1062 −0.0098 −0.0134 −0.0111 −0.0090 −0.0129 −0.0129 −0.0121 −0.0122 −0.0260 −0.0133 −0.0117 −0.1736 24 0.5696 23.06*** 1.34 0.88 3.69*** −1.21 0.39 −0.29 0.0006 0.0523 1.1689 −0.1100 −0.0131 −0.1727 24 0.5333 58.14*** 0.59 0.80 10.95*** −1.70* −2.67*** −1.28 t-statistic Key rivals 0.0016 0.0439 1.8119 −0.1972 −0.0178 −0.0232 −0.0188 −0.0151 −0.0195 −0.0233 −0.0184 −0.0248 −0.0579 −0.0233 −0.0165 −0.7912 5 0.4231 13.22*** 0.0016 0.0457 1.8056 −0.2080 −0.0235 −0.7853 5 0.3313 25.77***

61

t-statistic

Panel A: Abnormal returns for each event β0 0.0008 0.94 0.0029 0.0217 0.42 0.1217 β1 0.9656 11.55*** 0.8346 β2 β3 −0.1050 −2.08** −0.1671 λ1 −0.0050 −1.16 0.0010 −0.0053 −1.25 0.0046 λ2 λ3 −0.0049 −1.15 0.0025 −0.0043 −1.01 0.0013 λ4 −0.0047 −1.09 0.0048 λ5 λ6 −0.0055 −1.29 0.0038 λ7 −0.0046 −1.08 0.0040 −0.0059 −1.38 0.0023 λ8 λ9 −0.0103 −2.39** 0.0122 λ10 −0.0055 −1.28 0.0043 λ11 −0.0043 −0.99 0.0040 −0.2130 −2.01** −0.0826 β4 N Adjusted R2 F-value 96 0.5276 19.61*** 24 0.0919 2.69***

1.33 0.87 3.66*** −1.21 0.08 0.40 0.22 0.12 0.42 0.33 0.35 0.20 1.06 0.37 0.34 −0.29

0.62 0.83 11.35*** −1.71* −1.46 −2.01** −1.67* −1.34 −1.94** −1.92* −1.81* −1.83* −3.89*** −1.98** −1.72* −1.33

0.87 0.37 9.19*** −1.66* −1.23 −1.63* −1.30 −1.04 −1.35 −1.63* −1.27 −1.72* −4.00*** −1.60 −1.12 −3.17***

Panel B: Abnormal returns across all events 0.0008 0.94 0.0029 β0 β1 0.0220 0.43 0.1211 β2 0.9646 11.62*** 0.8342 β3 −0.1066 −2.12** −0.1650 γ comb −0.0055 −1.43 0.0041 β4 −0.2120 −2.01** −0.0822 N Adjusted R2 F-value 96 0.5333 58.13*** 24 0.1074 7.01***

0.80 0.35 8.56*** −1.63* −2.42** −2.94***

This table shows the coefficient estimates for competitors of WorldCom in response to the 11 events indicating concern over the future viability of WorldCom. We obtain the estimates using SUR and daily returns from January 2, 2002 to December 31, 2002. We estimate the abnormal returns and risk shift in two ways: Rp,t = β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + 11 λkp Dkt + β4p δt Rm,t + ep,t (Panel A) and Rp,t = k=1 β0p + β1p Rm,t+1 + β2p Rm,t + β3p Rm,t−1 + γcomb,p Dkt + β4p δt Rm,t + ep,t (Panel B), where Rp,t is the return on the portfolio of exposed firms on day t; Rm,t + 1 , Rm,t , and Rm,t − 1 are market returns on the day t + 1, t, and t − 1, respectively; Dkt is a dummy variable equal to 1 for event k and 0 otherwise; δt is equal to 1 on the final event day (July 19, 2002) and all subsequent days and 0 otherwise; β0p is the intercept, β1p , β2p , and β3p are the lead, contemporaneous, and lagged market betas for the portfolio; λkp is the portfolio’s abnormal response to event k; γ comb is the portfolio’s average abnormal response to all the 11 events; β4p is the portfolio’s shift in systematic risk; and ep,t is the error term. We also report the OLS R2 , adjusted R2 , and F-values. ***, **, and * indicate significance at the 1%, 5%, and 10% level, respectively.

62

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

over their unraveling period. In comparison, 16 of the 20 largest WorldCom institutional investors decreased their positions in WorldCom, which is a statistically significant increase in the proportion of institutional investors reacting in this manner. This comparison indicates that institutional investors could have become more cautious and responsive in making and changing investments in the post-Enron environment. Table 6 shows that institutional shareholders are not significantly affected by the events leading up to WorldCom’s demise. Indeed, neither the reaction to each of the 11 individual events separately (Panel A) nor the average effect across all events (Panel B) is significant for the portfolio of all institutional investors. Furthermore, the reaction is invariant with respect to whether the institutions were increasing or decreasing their relative exposure to WorldCom. We separately examine a portfolio containing the quartile of institutional investors with the largest decrease in their relative exposure and a portfolio containing the quartile of those with the largest increase in their relative exposure.7 Not finding significant valuation effects for institutional investors is consistent with the idea that these institutions are well diversified, and are less vulnerable to shocks associated with a particular company, albeit even the world’s largest bankruptcy.8 Despite the heightened investor anxiety during the post-Enron period, the market appears to be efficient in distinguishing firms that are well diversified, and are, therefore, less vulnerable. The risk shift parameter, β4 , is positive and significant in Panel A and Panel B for all of the subsets. We surmise that this increased market exposure is most likely a result of general economic conditions. 4.3. Impact on creditors Table 7 reports the reaction of WorldCom creditors. The portfolio of all creditors is insignificant for each of the 11 event days (Panel A), as well as over the combined dates (Panel B). The quartile of creditors with the most relative exposure is separately examined from the quartile of creditors with the least relative exposure.9 Neither subset shows significant valuation effects. Thus, it appears that the creditors are also considered to be well diversified. Generally, the risk shift parameter, β4 , is positive and significant. Again, we believe it is likely that increased sensitivity to market conditions is due to general economic conditions. 4.4. Impact on competitors Table 8 reports the reaction of competitors. In Panel A, it can be seen that the reaction of all rivals that share SIC code 4813, is only significant for the most severe event date (July
7 Relative exposure is defined as the value of WorldCom holdings/market value of the institutional investor. We also examine the quartile of institutional investors with the largest relative WorldCom holdings separately from the quartile with the smallest relative holdings. These subsets also are not significantly affected by the WorldCom events. 8 We thank an anonymous reviewer for pointing out that the institutional ownership data do not allow us to ascertain how gradual or abrupt the position changes occurred, which may contribute to not detecting significant results. 9 Relative exposure is defined as the reported WorldCom credit exposure/market value of the creditor.

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

63

1, 2002). Panel B of Table 8 does not indicate that, on average, the events leading up to the bankruptcy have significant and negative valuation effects on the portfolio of rivals that share SIC code 4813. Since pooling all rivals may obscure important effects, we consider the impact on small rivals, large rivals, and key rivals separately. Across both panels, the quartile of rivals that contains the smallest rivals, based on market value, do not show significant effects. However, the largest quartile of rivals and the key rivals exhibit a dominant contagion (negative) effect for most of the 11 events leading up to WorldCom’s failure. The average effect across all events (Panel B) is significantly negative for the larger rivals and key rivals, as well. An F-test shows the responses between the largest and smallest rivals are significantly different at the 0.01 level. This difference in reaction may be explained by several factors. First, given WorldCom’s size, the events leading to their bankruptcy sent strong signals to large rivals that they are not immune to scrutiny and/or bankruptcy. If there had been a perception in the industry that some firms were too big to fail, WorldCom’s failure eliminated this misconception and exposed their vulnerability. Further, given the overcapacity in the industry, many analysts were predicting consolidation. Under this scenario, smaller rivals are more likely to become acquisition targets than larger rivals, a fact that would accrue benefits to smaller rivals. The risk shift parameter, β4 , is negative and significant in Panel A and Panel B for the portfolio of rivals that share SIC code 4813 and for the portfolio of key rivals. This result indicates that competitors experienced a significant decrease in their sensitivity to market conditions in the post-bankruptcy period. This finding is plausible to the extent that the telecommunications industry had inflated market betas during the base period leading to the collapse of WorldCom. Inflated betas could be expected for the telecommunications industry that had overcapacity and a generally unstable environment in the base period.

5. Summary On July 19, 2002 WorldCom sought protection from its creditors when it filed for Chapter 11 bankruptcy, earning the distinction as the largest bankruptcy filing in U.S. history. The events surrounding this history-making occurrence provide an important opportunity to examine the repercussions for WorldCom’s stakeholders. We examine the impact of the events on the firm’s institutional investors, creditors, and competitors. Despite the heightened uncertainty facing investors during this period, we find that markets are remarkably efficient in distinguishing among the various types of stakeholders. In particular, institutional investors and creditors are largely unaffected by the events. These results are consistent with the basic benefit of diversification; a single bankruptcy, even the world’s largest bankruptcy, should have no impact on well-diversified shareholders. In contrast, large and key competitors are adversely affected by the events leading to the demise of WorldCom. The reactions may be attributed to scrutiny of rivals that are perceived to be facing similar problems. Thus, contagion effects appear to dominate any competitive repositioning for large and key competitors. This study indicates that shareholders, analysts, and portfolio managers are considering how firms with indirect ties to a financially distressed company may be affected.

64

A. Akhigbe et al. / The Quarterly Review of Economics and Finance 45 (2005) 48–64

References
Akhigbe, A., & Martin, A. D. (2002). Do Microsoft acquisitions benefit the computer industry? Financial Analysts Journal, 58, 19–27. Baber, W. R., Kumar, K. R., & Verghese, T. (1995). Client security price reactions to the Laventhol and Horwath bankruptcy. Journal of Accounting Research, 33, 385–395. Bhargava, R., & Fraser, D. R. (1998). On the wealth and risk effects of commercial bank expansion into securities underwriting: an analysis of Section 20 subsidiaries. Journal of Banking and Finance, 22, 447–465. Box, G. E. P., & Tiao, G. C. (1975). Intervention analysis with applications to economic and environmental problems. Journal of the American Statistical Association, 70, 70–79. Datta, S., & Iskandar-Datta, M. E. (1995). The information content of bankruptcy filing on security holders of the bankrupt firm: an empirical investigation. Journal of Banking and Finance, 19, 903–919. Ferris, S. P., Jayaraman, N., & Makhija, A. K. (1997). The response of competitors to announcements of bankruptcy: an empirical examination of contagion and competitive effects. Journal of Corporate Finance, 3, 367–395. Fraser, D. R., & Richards, R. M. (1985). The Penn Square bank failure and the inefficient market. Journal of Portfolio Management, 11, 34–36. Johnston, J. (1984). Econometric methods. New York: McGraw Hill. Johnston, J. J., & Madura, J. (2000). Valuing the potential transformation of banks into financial service conglomerates: evidence from the Citigroup–Travelers merger. Financial Review, 42, 17–36. Karafiath, I., & Glascock, J. (1989). Intra-industry effects of a regulatory shift: capital market evidence from Penn Square. Financial Review, 24, 123–134. Lamy, Robert E., & Thompson, G. R. (1986). Penn Square, problem loans, and insolvency risk. Journal of Financial Research, 9, 103–111. Lang, L. H. P., & Stulz, R. M. (1995). Contagion and competitive intra-industry effects of bankruptcy announcements: an empirical analysis. Journal of Financial Economics, 32, 45–60. Larcker, D. F., Gordon, L. A., & Pinches, G. E. (1980). Testing for market efficiency: a comparison of the cumulative average residual methodology and intervention analysis. Journal of Financial and Quantitative Analysis, 15, 267–287. Peavey, J. W., & Hempel, G. H. (1988). The Penn Square bank failure: Effect on commercial bank security returns—a note. Journal of Banking and Finance, 12, 141–150. Saunders, A., & Smirlock, M. (1987). Intra- and inter-industry effects of bank securities market activities: The case of discount brokerage. Journal of Financial and Quantitative Analysis, 22, 467–482. Scholes, M., & Williams, J. (1977). Estimating betas from nonsynchronous data. Journal of Financial Economics, 5, 309–327. Swary, I. (1986). Stock market reaction to regulatory action in the Continental Illinois crisis. Journal of Business, 59, 451–473. Winton, A. (1999). Don’t put all your eggs in one basket? Diversification and specialization in lending. Working paper, University of Minnesota.