The long-term effect of the Supreme Court Judgement in LaSalle on share price reactions with regard to Chapter 11 bankruptcy filings

(1)

Bachelor Thesis

The Long-term Effect of the Supreme Court Judgement in LaSalle on

Share Price Reactions with regard to Chapter 11 Bankruptcy Filings

University of Amsterdam (UVA)

Economics and Business, track Economics and Finance Kirsten Schreuder, 10380876

Supervisor : Dr. R. Perez Ribas Date: 28th June 2015

(2)

Abstract

The Supreme Court in LaSalle acknowledged an exception to the “Absolute Priority Rule” for “new value contributions” on condition of a “market test”. This “market test” increased the probability for equity holders to lose their equity in the Chapter 11 procedures. This study examined the long-term effect of the “market test” on share price reactions with regard to Chapter 11 filings. An event study was conducted for each Chapter 11 filing to obtain the share price reactions. This study found that the share price reactions for the period before and after LaSalle were significantly different and were stronger for the period after LaSalle (p-value <0.0001). Furthermore, firms were in a worse financial condition after LaSalle, which suggests that firms tried to avoid bankruptcy but still ended up in Chapter 11. However, the effect of the “market test” on share price reactions, determined by a regression, was not statistical significant. This implies that the “market test” might have influenced firms, but that the asset price channel does not capture the effect of the “market test”. The results of this empirical research suggests that future research on the “market test” should use a channel which is more directly linked to “absolute priority deviations”.

(3)

This document is written by Kirsten D. Schreuder who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(4)

1

1 Introduction

The American bankruptcy law is designed to give individuals and entities a second chance if they did not succeed in realising their full potential and therefore could not meet their obligations. Firms can obtain this second chance by successfully reorganizing under Chapter 11. Chapter 11 protects the debtor against aggressive creditors by providing them time to restructure and if needed the possibility to “cramdown’’ the reorganization when one or more classes of creditors reject the plan. The approval of a bankruptcy court and at least one class of impaired claims is required for a “cramdown”. The bankruptcy court tests if the plan does not “discriminate” and if it is “fair and equitable’’1_{. If the “cramdown” is accepted, then the}

reorganization plan will be exercised regardless of the rejection of a specific class of creditors. Equity holders often used the provided time to renegotiate or to “cramdown” a reorganization plan, which incorporates retention of (a part of) the equity in the reorganized firm (Adler, 2012).

On May 3 1999 the Supreme Court made an unexpected judgement concerning the retention of equity by old partners in exchange for a new value contribution. The judgement of the Supreme Court on the Bank of America v. 203 North LaSalle Partnership had a big impact on equity holders who sought to retain their relationship with the debtor. The Supreme Court acknowledged, in her judgement, an exception to the “Absolute Priority Rule” (APR). The APR entails that no junior claimant will be paid before all senior classes are paid in full. It implies that equity holders, who are residual claimants, cannot receive any stake in the reorganized firm based on their junior claim unless all creditors are paid in full. The Supreme Court decided that the “new value contribution” exception is only allowed, when equity holders contribute significant new value in exchange for a stake in the reorganized firm

(Giambona, Lopez-de-Silanes and Matta, 2014). Furthermore, the Supreme Court ordered a “market test” to prevent equity holders from abusing their exclusive right to propose a reorganization plan2 which mainly benefits the equity holders by offering a low new contribution. The “market test” was not further specified by the Supreme Court, but a common method used after the judgement is lifting the period of exclusivity either by an auction or by allowing others to submit a competing plan (Adler, 2012).

Economic literature on bankruptcy mainly examines the short term consequences of filing for bankruptcy and mostly focusses on Chapter 11. To my knowledge, only Giambona,

1_{See 11 U.S.C. §1129(b)(2) for further requirements for a “cramdown”.}

2

(5)

2

Lopez-de-Silanes and Matta (2014) examined the market performance around the Supreme Court Ruling. They expected to find that the “market test” would increase funding availability due to the increased asset verifiability. This expectation was supported by their findings as a significant positive cumulative average abnormal returns of 1.62% was found five days surrounding the judgement.

This research contributes to previous economic literature by examining the long term effect of the “market test” for the “new value contribution”, determined in the LaSalle case, on share price reactions. The US legislation has two main insolvency procedures for firms, namely Chapter 7 and Chapter 11. Chapter 7 contains the legislation to end the life of a firm by liquidating their assets. This research however will focus on Chapter 11, which entails the reorganization of a firm. Furthermore, this paper will examine how market reactions to Chapter 11 cases changes by comparing two periods: before and after the judgement of the Supreme Court on May 3 1999. The market reactions for each period will be analysed with an event study.

This research is also relevant for the legal literature. On December 4th 2014 the American Bankruptcy Institute (ABI) presented their Final Report and Recommendations on the Reform of Chapter 11. Although this was an advise, it sends an important message that practitioners are ready for an renewed bankruptcy law (Weijs & Wessels, 2015). This paper can contribute to the ongoing discussion concerning the desirability of the “market test”.

This paper has the following structure: section 2 is a short overview of the literature on the possible effects of the LaSalle judgement and contains my hypothesis, section 3 provides the source and the construction of the dataset, section 4 describes the event study methodology. Section 5 reports and discusses the results. Finally, section 6 gives a conclusion.

2 Literature Review

There is an ongoing debate in the economic and legal research world concerning the effects and desirableness of the “market test”. The study of Giambona et al (2014) found that the “market test” increased the verifiability of the assets and showed that a higher verifiability increased the probability of a Chapter 11 filing and debt capacity for low-verifiability firms. They also pointed out that the Supreme Court ruling increased the recovery rate of creditors in Chapter 11. In addition, Gilson (2012) concluded that Chapter 11 and the debt restructuring industry got more efficient and helped reviving the economy. He argues that Chapter 11 not

(6)

3

only reduced agency problems, but also provided more liquidity for the debtor by means of “debtor in position” (DIP) financing, §363 sales3

, possibilities for reduction of the tax burden and a process to cope with non-financial liabilities like labour union agreements.

Furthermore, Bharath, Panchapegesan and Werner (2013) showed that Chapter 11 has become more creditor friendly. They argued that the “absolute priority deviations” (APD) in favour of equity declined drastically with respect to the 1980’s. They found that for the periods: 1980-1990, 1991-2005 and 2000-2005, respectively, the APD was in favour of equity holders corresponding with the following percentages: 75%, 22% and 9%. They ascribe these effects to the increased power of creditors, which was a results of the increasing importance of innovations like DIP financing and Key Employee Retention Programs (KERP).

Guy (2012) however pleads that LaSalle could have delayed economic recovery from the great recession. He pointed out several trends, one of them being the increased use of a sale either by a §363 sale or liquidation. The reason behind this trend was that the judgement of the Supreme Court in LaSalle made debtors reluctant to file for Chapter 11. The “market test” makes retaining a stake in the firm for the original shareholder uncertain and therefore risky, because the shareholders could lose their equity in an auction. Also, successfully restructuring outside Chapter 11 is more difficult due to the fact that the threat of filing for Chapter 11 is less credible. Therefore, the equity holders prefer to walk away and sell their assets rather than running a risk (Guy, 2012). The requirements imposed on the “new value contribution” makes Chapter 11 even less attractive for small and medium firms, since it is not likely that their shareholders own valuable assets equal to the market value of their claim (Markell, 2000). The ABI commissioned study on the reform of Chapter 11 (2014) included possible solutions to these concerns in their recommendations. Some of these recommendations were: the elimination of the requirement of at least one accepting impaired class of claims, another restructuring procedure for small and medium-sized enterprises, and restrictions on the use of §363 sales.

Based on the literature review, I expect that the “market test” caused self-selection, which means that firms only file for Chapter 11 when they expect their chances of emerging and retaining (part of) the equity is high. Therefore, I expect to find relatively higher, but still negative, returns for firms that filed for Chapter 11 after the judgement compared to those who filed before the judgement. Furthermore, based on the findings in the literature review

3_{A definition of a §363 sale can be found in the article of Hurley (1984) at p.234: “An alternative to a plan of}

reorganization may be a simple sale of all of the debtor's assets outside a plan of reorganization, pursuant to section 363(b) of the Code. Section 363 permits the debtor or trustee to use, sell or lease the debtor's property, both in the ordinary course of business and other than in the ordinary course of business.”

(7)

4

small and medium sized companies would rather go for a sale or an out of court solution than file for Chapter 11 as the risk of losing the equity will probably be higher for these firms. Thus, I also expect to find that the firms filing for bankruptcy after the judgement will be larger in size.

3 Data

3.1 Source

This study analysed bankruptcies of businesses covered by the US legislation. The UCLA-LoPucki Bankruptcy Research Database (BRD)4 was used to collect Chapter 11 bankruptcy events and the Compustat database was used to collect accounting and stock data. BRD included firms that filed for bankruptcy with a minimal asset value of $100 million (in 1980’s dollars). Financial firms were excluded from the samples, due to the fact that they are treated differently under the US bankruptcy legislation5. Financial firms were classified as firms with a Standard Industrial Classification (SIC) code between 6000 and 6999.

The first step in the data preparation procedure was extracting the data from the BRD. BRD provided some firms with GVKEY 99999, as a result of Compustat not providing these GVKEYs, and these firms were excluded in this analysis. BRD identified bankruptcies with the variable: Chapter. The cases with a Chapter identifier equal to “7” or “no order of relief” were also excluded. Also, the cases with a filing date in January 1990 or later were kept and the provided GVKEYs were used to collect the data from Compustat.

Compustat Fundamentals Annually database was used to collect the accounting data from the first available fiscal year end before the filing. If the data concerning the one year prior to the filing was not available, then the observations were dropped from the sample. The Price-Close-Daily (PRCCD) variable was retrieved from the Compustat – Security Daily database. If a firm stopped trading before the filing, for example because it went private, then it was also excluded from the sample.

Additionally, the Equal-Weighted Returns (including distributions) were retrieved from CRSP Stock Market Indexes, and the Fama and French (1993) factors were retrieved from Fama-French Portfolios and Factors. See Appendix Table 1 for a description of the variables.

4_{For more information on the UCLA-LoPucki Bankruptcy Research Database see: http://lopucki.law.ucla.edu/.}

5_{At the seminar ‘Revision of Chapter 11 vs. Revision of the Dutch Bankruptcy Act’ (18 May 2015) was}

(8)

5 15 39 44 30 19 14 9 3 5 10 15 1 2 7 5 0 10 20 30 40 F re q u e n cy 1999 2001 2003 2005 2007 2009 2011 2013 Year

Figure 2: distribution of 218 filings over the period May 3 1999 until 2013

After LaSalle Chapter 11 filings 12 14 9 6 8 6 8 6 15 8 0 10 20 30 40 F re q u e n cy 1990 1992 1994 1996 1998 2000 Year

Figure 1: distribution of 92 filings over the period 1990 until 2 may 1999

Before LaSalle

Chapter 11 filings

3.2 Sample selection

This research focussed on the event of bankruptcy. Panel data from Chapter 11 cases were collected and divided into two samples. The first sample covered the period from January 1990 up to May 2 1999 and the second sample covered from May 3 1999 up to 2013. The judgement of the Supreme Court was unexpected and it is therefore assumed that the daily stock returns for the period January 1990 up to May 2 1999 will not reflect the judgement of the Supreme Court.

An event study was conducted for each Chapter 11 filing in the sample. Prior research on bankruptcies used announcement windows equal to [-1,+1] trading days, because firms can file for bankruptcy after the market closes (Dawkins, Bhattacharya & Bamber, 2007). Therefore, investors cannot always immediately react on bankruptcy news. In this study an announcement window of [-1,1] trading days was also applied. The event window was set to one month or 21 trading days. The reason behind the chosen width of this event window was so that it would not include more than one event as some firms file for bankruptcy more than once.

Moreover, not more than one event should be taken into account in an event window, due to the fact that a specification bias can arise due to cross-correlation (Khotari and Warner, 2006). Khotari and Warner mention that this bias increases the homogeneity of a sample and that this especially occurs in long event windows. To minimize the cross sample correlation the method used by Cox and Peterson (1994) was applied. This method entails keeping only the first observation, after sorting on trading day and firm names, when more than one event takes place on a specific trading day. Furthermore, only the first filing will be included per sample. This means that a firm can only appear once in each of the two samples, but can appear twice in the total sample.

The total sample consisted of 310 events, including 92 events before and 218 events after the LaSalle case. See Figure 1 and Figure 2 for the distribution of the events.

(9)

6 5.963 15.14 22.48 23.39 14.68 11.93 6.422 0 10 20 30 40 Pe rc e n t 0 1 2 3 4 5 6 7 8 9 SIC

Figure 4: industry distribution of 218 events over the period May 3 1999 until 2013

After LaSalle Industry distribution 8.696 15.22 17.39 15.22 36.96 3.261 3.261 0 10 20 30 40 Pe rc e n t 0 1 2 3 4 5 6 7 8 9 SIC

Figure 3: industry distribution of 92 events over the period 1990 until May 3 1999

Before LaSalle

Industry distribution

Also, the industry distribution is provided in Figure 3 and Figure 4. There are ten major divisions in the Standard Industrial Classification (first digit SIC). There were no firms from division nine (public administration) and division zero (agriculture, forestry and fishing) present in the sample. In addition, division six was excluded from the sample as it consisted of financial firms.

4 Methodology

First the daily returns were calculated for every firm starting with 20 trading days before the bankruptcy date. The number of trading days till the filing is represented by “t” and the filing date is “t” equal to zero. The following formula was used:

R_it = log(P_t) − log(P_t−1 ) (1)

After calculating the returns, the returns were adjusted for the stock market returns. Instead of using normal returns calculated with, for example, the Capital Assets Pricing Model (CAPM) the returns were adjusted for the daily Equal-Weighted Returns (EWRET) retrieved from the CRSP database. The reason for using EWRET was that the company betas change before bankruptcy (Coelho & Taffler, 2008). It is therefore standard practice to use the Equal-Weighted Return (Rm) to calculate abnormal returns (AR). Dawkins, Bhattacharya and

Bamber (2007) pointed out that firms that file for bankruptcy generally have a lower median assets value and therefore the Equal-Weighted Returns should be used instead of the Value-Weighted Returns from the market index. These adjustments were incorporated in formula 2.

(10)

7

After calculating every AR, the CARs were calculated using the standard event study method.

CAR_𝑖 = ∑ 𝐴𝑅_𝑖𝑡 (3)

The average of the CARs was used to obtain the cumulative average abnormal return (CAAR). In line with earlier bankruptcy studies, the CAARs were tested for significance using a parametric t-test and a nonparametric Wilcoxon rank-sum test (Rose-Green & Dawkins, 2002).

Additionally, a regression was conducted to assess the effect of the “market test” on the cumulative abnormal return from companies who filed for Chapter 11. The following model was used:

𝐶𝐴𝑅_𝑖𝑡 = 𝑏₀+ 𝑏₁𝑀𝑎𝑟𝑘𝑒𝑡𝑇𝑒𝑠𝑡 + 𝑏₂𝑍𝑠𝑐𝑜𝑟𝑒 + 𝑏₃𝐵𝑜𝑜𝑘𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒+ 𝑏₄ 𝐿𝑁(𝑀𝑉𝐸𝑄) +

𝑏5𝑆𝑀𝐵 + 𝑏6𝐻𝑀𝐿 + 𝑏7𝑈𝑀𝐷 + 𝑏8∑(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦) + 𝑌𝑒𝑎𝑟𝑖+ 𝑒𝑖𝑡 (4)

Table 2 gives a definition of the model variables used in Formula 4. The calculation and sources of these variables can be found in the Appendix Table 1.

Table 2: Definition of the model variables

Variables Description

MarketTest Dummy variable for fiscal year

1: 3 May 1999- 2013 0: 1990 – 2 May 1999

Zscore Z-score measured for the fiscal year end before the bankruptcy filing

Bookleverage Total debt divided by the book value of total assets

MVEQ Market value of the equity for the fiscal year end before the bankruptcy filing

SMB Small minus big

HML High minus low

UMD Up minus down

Industry Six dummy variables for the major divisions from the Standard Industrial Classification

(SIC)

(11)

8

MarketTest was the variable of interest and it was expected that the firms who filed for Chapter 11 after the judgement have relatively higher CARs compared to the firms who filed for bankruptcy before the judgement. Higher CARs after LaSalle could be the result of self-selection caused by the “market test”. Therefore, it was predicted that the coefficient for MarketTest was positive.

The coefficient on Zscore was also predicted to be positive. A higher Z-score means a lower probability of bankruptcy (Altman, 1968), thus the stock from a company with a lower probability of default is expected to be less risky and a higher price is appropriate.

Bookleverage was expected to have a negative coefficient, due to the fact that when a company has a high debt and few assets it has a greater probability of default. When such a company files for bankruptcy, shareholders would likely receive close to nothing due to their residual claimant status. An investment in such a company is risky and, as a result, a higher compensation is demanded. Therefore, investors pay low prices for the shares.

The same measurement as Rose-Green and Dawkins (2002) was used to measure the size of a company. This measurement is represented by the variable MVEQ. Rose-Green and Dawkins (2002) argued that larger firms are more likely to get attention of the media and therefore experience larger declines in stock prices. Likewise, I also expected a negative coefficient on the natural logarithm of MVEQ.

The variables SMB, HML, UMD were included in the model to correct for market returns. Recent economic literature indicated that UMD profits from momentum strategies occur more often for firms with a low credit rating compared to those with a high credit rating and that SMB and HML are closely related to distress risk (Garlappi & Yan, 2011).

Returns can differ between industries, thus if the composition of the sample changes it will influence the returns. Therefore, a dummy variable for each industry was included in the model as control variables. In addition to the industry control variables, a fixed effect with respect to years was added. Figure 1 and 2 indicate that the Chapter 11 filings were not equally divided between 1990 and 2013. To prevent overrepresentation of a specific year, the fixed effect was added to the model and this was represented by the variable: Yeari.

5 Results

The market reactions to Chapter 11 filings before and after the introduction of the “Market test” determined by the Supreme Court in LaSalle are examined in this section. As a short reminder, the hypothesis for this paper was that the “Market test” caused self-selection and as

(12)

9

a result caused a relatively higher, but still negative, CAAR for the period after the judgement compared to the period before the judgement. First, the univariate analysis was conducted to give a general overview of the market reactions for the periods before and after LaSalle. Furthermore, a univariate analysis was conducted to compare the two periods and to give a descriptive analysis. Afterwards, the hypothesis was tested with a multivariate analysis. The implications of the results of the univariate and multivariate analyses are also discussed in this section.

5.1 Univariate analysis

To provide a general overview of the market reactions to Chapter 11 filings in the periods 1990 until May 2 1999 and May 3 1999 until 2013, the evolution of the CAARs are presented in Figure 5. On May 3 1999 the Supreme Court demanded the “Market Test” for new value contributions. The CAARs are the cross sectional average of the CARs per day in the event window. Figure 5 shows the negative CAARs for the period before as well as after the judgement in LaSalle. Previous bankruptcy studies also found negative CARs before the Chapter 11 filings. Furthermore, Figure 5 indicates a relative larger decline in CAARs for the period after LaSalle. This decline is especially visible on the day of filing (day zero) as the CAAR on the filing day covering the period before LaSalle equals -0.6198 (-61.98%) and after LaSalle equals -0.9143 (-91.43%).

The second analysis addressed the question whether the CAARs significantly differed from each other. First of all, the CAARs across the event window were tested with a t-test on their significance. Afterwards the CAARs were compared and checked on their significance by

(13)

10

conducting another t-test and a Wilcoxon rank-sum test. The results of this analysis are presented in Table 3.

Contrary to my expectations, the results in Table 3 show that the firms who filed for Chapter 11 after the judgement have a more negative CAAR than the ones before LaSalle. Both the t-test and the Wilcoxon rank-sum test pointed out that the CAARs differed significantly, with a difference equal to 0.085 (both p-values <0.0001). A possible explanation for this finding is that firms who file for bankruptcy after the judgement tried to avoid bankruptcy, but still ended up filing for Chapter 11 bankruptcy. It is likely that these firms entered bankruptcy in a worse condition. Entering Chapter 11 in this condition will negatively influence the expected chance of emerging from Chapter 11 and thus increase the chances of losing the equity. This results in a relatively larger price drop as shareholders try to sell their shares.

Bharath et al (2013) had a similar line of thought, which corresponded with my explanation. They proved that the power of creditors increased in the Chapter 11 process as a result of the financial innovations, the limitations on exclusivity and the changes in bankruptcy law practice in 1984 and 19886. This resulted in managers trying to avoid Chapter 11 for as long as possible as they anticipated their reduced bargaining power. Consequently, firms became more insolvent, which reduced the credibility of the threat to the delay the process in Chapter 11 even further, and therefore reduced the chance of an “absolute priority deviation”.

Table 3: Market-Adjusted Cumulative Average Abnormal Returns for firms who filed for Chapter 11

Notes:

** Significant at α ≤ 0.01 (two-tailed test).

CAAR = cumulative average market-adjusted abnormal return (fractions).

The final analysis, a specific descriptive analysis, was conducted to examine if the companies that filed for Chapter 11 after LaSalle were indeed in a worse condition. The results are displayed in Table 4. Both samples had Z-score below 1.81, which shows that they were

6_{Bharath et al (2013) refer to the “Bankruptcy Amendments and Federal Judgeship Act” (1984) and the Third}

Circuit judgement on “First American Bank of New York v. Century Glove”(1988).

Trading Before LaSalle After LaSalle Mean t-test Wilcoxon

Day CAAR(n=9

6)

t-value CAAR(n=223) t-value Difference t-value z-value

(14)

11

bankrupt (Altman, 1968). The significantly lower Z-score, calculated for the end of the fiscal year prior to the filing, supports the expectation that firms after LaSalle were in a worse condition. The firms filing for Chapter 11 after LaSalle also had a significant higher Bookleverage, which indicates that the proportion debt to assets was higher. This is in line with Bharath et al (2013), who also found a higher level of indebtedness. The market value of the equity (MVEQ) is presented to examine whether the sizes of the firms differed between the two sample periods. The results in Table 4 show that the MVEQ was higher at the end of the fiscal year prior to the filing for the sample that filed after LaSalle, however this difference was not significant.

Table 4: Descriptive analysis

Note: ** Significant at α ≤ 0.01 (two-tailed test).

5.2 Multivariate analysis

The goal of the multivariate analysis was to quantify the effect of the MarketTest and it was conducted by means of four regressions. The results of these regressions are presented in the Appendix in Table 5. The first regression analysis only controls for the following variables: Zscore, Bookleverage and HML, SMB and UMD. In addition to these factors the second regression also controls for time effects. The third regression controls for industry effects and the fourth regression controls for time and the industry effect.

First the results with respect to the main variable of interest MarketTest are discussed. In regression 1 a significant, at an alpha of five percent, coefficient of -0.1 was found for the MarketTest variable. This means that companies that filed after LaSalle compared to the ones who filed before LaSalle had, on average, a 10% lower cumulative abnormal return in the month prior to the Chapter 11 filing. In short, regression 1 indicates that the market reacted stronger in the period after the judgement in LaSalle.

Total Sample

Before LaSalle After LaSalle Mean t-test Wilcoxon

mean Mean Mean Difference t-value z-value

Zscore -0.133 0.312 -0.321 0.632 8.329** 11.433** Bookleverage 0.460 0.411 0.480 -0.0698 -6.420** -5.794** MVEQ 126.928 88.365 143.202 -54.836 -0.827 0.018 SMB 0.0001 -0.0002 0.0002 -0.0004 -2.252** -3.778** HML 0.0001 -0.0003 0.0004 -0.0007 -3.492** -4.480** UMD 0.0004 0.0005 0.0004 .00008 0.2830 -1.851

(15)

12

The coefficient on MarketTest remained negative in regression 2, 3, and 4. But, in regression 2, controlled for time fixed effect, and in regression 4, controlled for industry and time fixed effect, the coefficients on MarketTest were not statistically significant. Thus, controlling for time fixed effects impacts the statistically significance of the coefficient on MarketTest, because regression 3, where only industry effects control variables were added, shows that the coefficient on MarketTest was equal to -0.097 (-9.7%) and still significant at an alpha of five percent.

These results could be interpreted in multiple ways. First of all, it could mean that the judgement in LaSalle had no effect on share price reactions. This implies that the foresight of a “market test” does not cause investors to react stronger to a Chapter 11 bankruptcy filing. A second interpretation, which contradicts the first interpretation, is that the “market test” did influence share price reactions, but that the inclusion of the fixed effects with respect to time in the model was incorrect. McKinnish (2000) found that fixed effect models on panel data possibly underestimate the effect of the variable of interest. She demonstrated that fixed effect models, which control for unobserved time-constant characteristics, can worsen measurement errors. To prevent hasty conclusions stating that no effect exists, she advises to be cautious with interpreting estimates of fixed effect models. Finally, the lack of significance could be caused due to the asset price channel possibly not reflecting the effects of the imposed “market test”. Bharath et al (2013) argued that the frequency of absolute priority deviations (APD) reduced drastically, and as a result, they found it unlikely that the expectation of APD was reflected in asset prices. They emphasized examining institutional factors that lead to APD instead of using APD as an assumption.

The last interpretation seems to be the most likely and could be an explanation for the fact that the regression models, presented in the Appendix Table 5, have relative low explanatory power (low R-square).

The control variables were also analysed. A significant negative coefficient for Zscore was found in all of the four regressions (α<1%). This finding was supported by Rose-Green and Dawkins (2002) who also found a significant negative Zscore. However, a negative coefficient for the Altman Z-score was unexpected as it implies that firms who had a better financial position at the end of the fiscal year prior to the filing, and therefore a higher Z-score, experienced a larger price decline. A negative coefficient on Zscore might hint that investors were more surprised of the bankruptcy filings from companies with a better Z-score. The coefficient from the natural log of the MVEQ variable was (as expected) negative, but the coefficients were only significant in regression 1 and 2 (α<5%). The other control

(16)

13

variables Bookleverage, SMB, HML, UMD were not significant in all of the regression models and therefore have no extra explanatory value.

6 Conclusion

On May 3 1999 the Supreme Court did an unexpected judgement in the case Bank of America v. 203 North LaSalle Partnership. The Supreme Court granted permission for a deviation to the “absolute priority rule”(APR) with respect to old equity holders who in exchange for a sufficient value contribution could retain their equity after the firm emerges from Chapter 11 bankruptcy. But, on the condition that a “market test” would be conducted. Deviations from the APR in favour of equity holders was not a rare occurrence, however the condition concerning the “market test” was unexpected. It implied that equity holders could lose their equity by means of an auction or a competing plan.

This research investigated the long term effects of the judgement in LaSalle on share price reactions. The hypothesis of this study was that the “market test” caused self-selection, which meant that the companies who filed for Chapter 11 estimated their chance of emerging high. Thus, the returns for companies who filed after LaSalle were expected to be higher, but still negative, compared to those who filed before LaSalle. To test the hypothesis an

univariate and multivariate analysis was conducted. In the univariate analysis the cumulative average abnormal returns and the sample characteristics were compared for the period before and after the judgement. The univariate analysis showed results contradicting the hypothesis. The cumulative average abnormal returns were lower after the LaSalle case. A possible explanation for this result was that firms tried to avoid bankruptcy, but still ended up in Chapter 11. The descriptive analyses confirmed that the firms who filed after LaSalle were in a poorer financial condition. Also, the size of the firms were not significantly different before and after LaSalle.

The multivariate analysis was conducted to quantify the effect of the “market test”. The results showed a negative effect equal to 9.7%, when controlled for industry effect. However, when controlling for fixed effects, with respect to time, the effect of the “market test” was not significant. This could be interpreted in several ways. The first possible interpretation assumes that the “market test” had no effect on share price reactions. The second possible interpretation could be that the “market test” did have an effect, but that the inclusion of the fixed effect with respect to time in the regression was incorrect. However, the most likely explanation is suggested by Bharath et al (2013) who argued that the asset prices

(17)

14

are unlikely to reflect the expectation of the “absolute priority deviation”, because the frequency of these events have rapidly declined. Thus, future research on the effect of “absolute priority devotions” should focus on other more direct channels rather than asset pricing.

A limitation of this research is that the bankruptcy law was changed in 2005, which might have influenced my results. “The Bankruptcy Abuse Prevention and Consumer Protect Act” (BAPCPA) took effect on October 17, 2005. This act makes it less easy to liquidate in Chapter 7 and the “automatic stay” in Chapter 11 is shorter if a company refiles within one year (Xu, 2013). Xu (2013) compared five years before and after the BAPCPA and found that Chapter 11 cases did not change in the degree of liquidity, but their operating performance and coordination did improve after the BAPCPA. In my total sample there are six firms who filed twice and none of them refiled within one year, therefore the shorter automatic stay probably has not influenced this research. The improved operating performance could be reflected as a higher Altman score. Yet, the results in Table 4 showed a significant low Z-score despite the possible upwards influence. The improved operating performance most likely has not influenced the results and would still be in line with my conclusions.

It could be interesting for future research to examine the effect of the “market test” on small and medium sized companies. Hereby, a comparison could be made between companies with different sizes. The bankruptcy data for this research was kindly provided by UCLA-LoPucki. The BRD database includes firms with a minimal asset value of 100 million dollars. Therefore, the sample could be enriched by collecting bankruptcy events from Compustat by following the approach of Corbae and D’Erasmo (2014). Although, Compustat only reports the delisting date and not the filing date, the filing dates could be found by consulting for example Factiva. However, it should be taken into account that the process of searching for the correct filing date per company is very time-consuming and labor-intensive.

So in short, significant results were found that confirmed stronger share price

reactions and the worse financial condition of the firms filing after LaSalle. But, the effect of the “market test” on share price reactions was not statistical significant. This implies that the “market test” might have influenced firms, but that the asset price channel does not reflect the effect of the “market test”. Thus, future research could be conducted on the difference in the effect of the “market test” on companies with different sizes by making the sample more diverse. Moreover, my results demonstrate that it is advisable to focus on a more direct channel with respect to the “absolute priority deviation”.

(18)

15

References

Adler, B. E. (2000). Emergence of Markets in Chapter 11: A Small Step on North LaSalle Street, The. Sup. Ct. Econ. Rev.,8, 1, 1-21.

Altman, E. I. (1968). Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The journal of finance, 23(4), 589-609.

American Bankruptcy Institute Commission to study the reform of chapter 11 (2014).

2012-2014 Final Report and Recommendations. Retrieved from:

https://abiworld.app.box.com/s/vvircv5xv83aavl4dp4h.

Bharath, S. T., Panchapagesan, V., & Werner, I. M. (2014). The changing nature of Chapter 11. IIM Bangalore Research Paper, (461).

Coelho, L., John, K., & Taffler, R. (2010). Bankrupt Firms: Who’s Buying? Working Paper. Retrieved from http://papers.ssrn.com/abstract=1572733

Corbae, D., & D’Erasmo, P. (2014). Reorganization or liquidation: Bankruptcy choice and

firm dynamics. Working paper, University of Wisconsin, 1-37.

Cox, D. R., & Peterson, D. R. (1994). Stock returns following large one-day declines: Evidence on short-term reversals and longer-term performance. Journal of Finance, 255-267.

Dawkins, M. C., Bhattacharya, N., & Bamber, L. S. (2007). Systematic share price fluctuations after bankruptcy filings and the investors who drive them. Journal of

Financial and Quantitative Analysis, 42(02), 399-419.

Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of financial economics, 33(1), 3-56.

Garlappi, L., & Yan, H. (2011). Financial Distress and the Cross‐section of Equity Returns. The journal of finance, 66(3), 789-822.

Giambona, E., Lopez-de-Silanes, F., & Matta, R. (2014). Improved Creditor Protection and Verifiability in the US, 1-39.

Gilson, S. (2012). Coming through in a crisis: How Chapter 11 and the debt restructuring industry are helping to revive the US economy. Journal of Applied Corporate

Finance, 24(4), 23-35.

Guy, B. (2012). Has LaSalle Decision Delayed Economic Recovery? Dearth of Chapter 11s Slows Capital Reallocation. The Journal of Corporate Renewal. Retrieved from: https://www.turnaround.org/Publications/Articles.aspx?objectID=13723

(19)

16

Hurley, J. J. (1984). Chapter 11 Alternative: Section 363 Sale of all of the Debtor's Assets Outside a Plan of Reorganization. Am. Bankr. LJ, 58, 233-252.

Kothari, S. P., & Warner, J. B. (2006). The econometrics of event studies. Available at SSRN 608601.

Markell, B. A. (1999). LaSalle and the Little Guy: Some Initial Musings on the Ultimate Impact of Bank of America, NT & SA v. 203 North LaSalle Street Partnership. Bankr.

Dev. J., 16, p. 335-360.

McKinnish, T. G. (2000). Model sensitivity in panel data analysis: Some caveats about the interpretation of fixed effects and differences estimators. terra, 303, 492-6770.

Rose-Green, E., & Dawkins, M. (2002). Strategic bankruptcies and price reactions to bankruptcy filings. Journal of Business Finance & Accounting, 29(9‐10), 1319-1335. doi: 10.1111/1468-5957.00471

Weijs, D., Jakob, R., & Wessels, B. (2015). Proposed Recommendations for the Reform of Chapter 11 US Bankruptcy Code. Amsterdam Law School Research Paper, (2015-14). Xu, K. (2013). The Choice of Bankruptcy Reorganization: Chapter 11, Prepackaged and

(20)

17

Appendix

Table 1: description and source variables

Variables Description Source

GVKEY Company identifier Compustat (WRDS)

Chapter Identifier for the bankruptcy law chapter UCLA-LoPucki

Bankruptcy Research Database

PRCCD Price-Close-Daily Compustat (WRDS)

EWRET Equal-Weighted Returns (including distributions) CRSP (WRDS)

MarketTest Dummy variable for fiscal year, where 1 represents May 3 1999-

2013 and zero represents 1990 – 2 May 1999

Zscore _1.2𝑊𝐶𝐴𝑃 𝐴𝑇 + 1.4 𝑅𝐸 𝐴𝑇+ 3.3 𝐸𝐵𝐼𝑇 𝐴𝑇 + 0.6 𝑀𝑉𝐸𝑄 𝐿𝑇 + 0.999 𝑆𝐴𝐿𝐸𝑆 𝐴𝑇 Altman (1969)

WCAP Working Capital Compustat (WRDS)

RE Retained Earnings Compustat (WRDS)

EBIT Earnings Before Interest and Taxes Compustat (WRDS)

MVEQ MKVALT and if not available PRCC_F *CSHO Rose-Green & Dawkins

(2002)

MKVALT Market Value – Total -Fiscal Compustat (WRDS)

PRCC_F Price Close – Annual Fiscal Compustat (WRDS)

CSHO Common Shares Outstanding Compustat (WRDS)

SALES Sales Compustat (WRDS)

LT Liability -Total Compustat (WRDS)

AT Book Assets -Total Compustat (WRDS)

Bookleverage 𝐷𝐿𝐶 + 𝐷𝐿𝑇𝑇

𝐴𝑇

Giambona et al (2014)

DLC Debt in Current Liabilities Compustat (WRDS)

DLTT Long-Term Debt - Total Compustat (WRDS)

SMB Small minus big Fama-French Portfolios

and Factors (WRDS)

HML High minus low Fama-French Portfolios

and Factors (WRDS)

UMD Up minus down / Momentum Fama-French Portfolios

and Factors (WRDS)

Industry Major divisions in Standard Industrial Classification (SIC) UCLA-LoPucki

Bankruptcy Research Database

(21)

18

Table 5: results regression models on CARs from 310 companies who filed for Chapter 11

Notes:

* Significant at α ≤ 0.05 (two-tailed test).

** Significant at α ≤ 0.01 (two-tailed test).

In the parentheses the robust standard errors adjusted for clusters are presented.

(1) (2) (3) (4) Intercept -0.157** (0.042) -.233** (0.095) -0.098 (0.080) -0.016 (0.132) MarketTest -0.100* (0.040) -.008 (0.121) -0.097* (0.040) -0.016 (0.114) Zscore -0.021** (0.005) -0.020** (0.006) -0.023** (0.006) -0.024** (0.006) Bookleverage -0.004 (0.052) 0.017 (0.058) -0.007 (0.052) 0.018 (0.056) LNMVEQ -0.018* (0.009) -0.018* (0.009) -0.016 (0.009) -0.016 (0.009) SMB 0.929 (1.161) 0.232 (1.195) 0.771 (1.170) 0.172 ( 1.189) HML -0.853 (0.975) -0.289 (0.951) -0.626 (0.900) -0.165 (0.891) UMD -0.562 (0.833) -0.217 (0.845) -0.367 (0.824) -0.070 (0.836)

Industry (SIC): No No Yes Yes

2 - Light Manufacturing 0.021 (0.083) -0.002 (0.086) 3 - Heavy manufacturing -0.081 (0.080) -0.090 (0.088) 4 – Transportation & Utility -0.063 (0.890) -0.090 (0.097) 5 – Wholesale & Retail

Trade

-0.099 (0.081)

-0.091 (0.088) 7 – Personal & Business

Service -0.160 (0.123) -0.166 (0.128) 8 – Health, Educational service, Other -0.103 (0.147) -0.080 (0.147)

Year Fixed Effects No Yes No Yes

Model R2 0.0263 0.064 0.036 0.071

The long-term effect of the Supreme Court Judgement in LaSalle on share price reactions with regard to Chapter 11 bankruptcy filings