Corporate Bankruptcies in Germany: Recovery Rates in Insolvency Plans [1 ed.] 9783896737342, 9783896447340

Citation preview

Corporate Bankruptcies in Germany – Recovery Rates in Insolvency Plans

Studienreihe der Stiftung Kreditwirtschaft an der Universität Hohenheim Herausgeber: Prof. Dr. Hans-Peter Burghof

Band 55

Barbara Flaig

Corporate Bankruptcies in Germany Recovery Rates in Insolvency Plans

Verlag Wissenschaft & Praxis

Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar.

D100 ISBN 978-3-89673-734-2 © Verlag Wissenschaft & Praxis Dr. Brauner GmbH 2017 D-75447 Sternenfels, Nußbaumweg 6 Tel. +49 7045 930093 Fax +49 7045 930094 [email protected] www.verlagwp.de

Alle Rechte vorbehalten Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Druck und Bindung: Esser printSolutions GmbH, Bretten

TABLE OF CONTENTS

5

Table of Contents Tables ........................................................................................................................ 7 Figures ....................................................................................................................... 9 Abbreviations .......................................................................................................... 11 I

Introduction ...................................................................................................... 13

II

Efficiency of Insolvency Laws ........................................................................ 15 II.1

Ex-post Efficiency ..................................................................................... 15

II.1.a

Incentives of Stakeholders in Bankruptcy ........................................ 15

II.1.b

Solving Ex-post Inefficiencies ......................................................... 17

II.2

Ex-ante Efficiency ..................................................................................... 22

III Overview of the German and the US Insolvency Law ................................... 29 III.1 German Procedure ..................................................................................... 29 III.2 US Procedure............................................................................................. 33 III.3 Legal Differences between Germany and the US ..................................... 34 IV Empirical Evidence for Corporate Bankruptcies ............................................ 37 IV.1 Extent of Recovery Rates .......................................................................... 37 IV.2 Influencing Factors on Recovery Rates .................................................... 40 IV.2.a

Firm Recovery Rates ........................................................................ 40

IV.2.a.1 Overview ....................................................................................... 40 IV.2.a.2 Firm Characteristics....................................................................... 42 IV.2.a.3 Procedural Characteristics ............................................................. 50 IV.2.a.4 Macroeconomic Variables ............................................................. 54 IV.2.b

Secured and Unsecured Recovery Rates .......................................... 56

IV.2.c

Bank Recovery Rates ....................................................................... 65

IV.2.d

Summary ........................................................................................... 74

IV.3 Evidence for Germany .............................................................................. 75

6

TABLE OF CONTENTS

V Analysis of German Reorganization Procedures ............................................ 79 V.1

Research Design ........................................................................................ 79

V.1.a

Data Acquisition ............................................................................... 79

V.1.b

Data Cleansing and Definition of Variables..................................... 80

V.1.b.1 Firm-specific, Procedural and Macroeconomic Variables ............ 81 V.1.b.2 Recovery Rates .............................................................................. 85 V.2

Data Analysis ............................................................................................ 87

V.2.a

Descriptive Statistics ........................................................................ 87

V.2.a.1 Firm Characteristics....................................................................... 87 V.2.a.2 Procedural Characteristics ........................................................... 101 V.2.a.3 Distribution of Recovery Rates ................................................... 105 V.2.b

Influencing Factors on Firm Recovery Rates ................................. 109

V.2.b.1 Hypotheses................................................................................... 109 V.2.b.2 Methodology................................................................................ 111 V.2.b.3 Correlation Analyses ................................................................... 115 V.2.b.4 OLS Regressions ......................................................................... 117 VI Conclusion...................................................................................................... 125 References ............................................................................................................. 129

TABLES

7

Tables Table 1: Comparison of Insolvency Procedures in Germany and the US .............. 35 Table 2: Overview of Empirical Studies (Firm, Secured and Unsecured Recovery Rates) ........................................................................................ 38 Table 3: Overview of Empirical Studies (Bank Recovery Rate)............................ 40 Table 4: Influencing Factors on Recovery Rates .................................................... 41 Table 5: Firm Recovery Rates and Firm Characteristics (1) .................................. 48 Table 6: Firm Recovery Rates and Firm Characteristics (2) .................................. 49 Table 7: Firm Recovery Rates and Procedural Characteristics .............................. 53 Table 8: Firm Recovery Rates and Macroeconomic Variables .............................. 55 Table 9: Secured and Unsecured Recovery Rates - Firm Characteristics (1) ........ 59 Table 10: Secured and Unsecured Recovery Rates - Firm Characteristics (2) ...... 60 Table 11: Secured and Unsecured Recovery Rates - Procedural Characteristics .. 62 Table 12: Secured and Unsecured Recovery Rates - Macroeconomic Variables .. 64 Table 13: Bank Recovery Rates and Firm Characteristics ..................................... 66 Table 14: Bank Recovery Rates and Procedural Characteristics ........................... 69 Table 15: Bank Recovery Rates and Macroeconomic Variables ........................... 71 Table 16: Bank Recovery Rates and Terms of Credit as well as Business Connection .............................................................................................. 73 Table 17: Firm-specific, Procedural and Macroeconomic Variables ..................... 82 Table 18: Definition of Recovery Rates ................................................................. 87 Table 19: Summary Statistics ................................................................................. 88 Table 20: Comparison Total Debt (Filing vs. Termination) ................................... 89 Table 21: Expenses Insolvency Procedure ............................................................. 95 Table 22: Number of Creditors ............................................................................. 103 Table 23: Debtor-in-Possession Management ...................................................... 104 Table 24: Petitioner and Initiator Insolvency Plan ............................................... 105 Table 25: Distribution of Recovery Rates ............................................................ 106 Table 26: Distribution of Recovery Rates (Liquidation vs. Reorganization) ....... 108

8

TABLES

Table 27: Hypotheses and Control Variables ....................................................... 111 Table 28: Correlation Analysis between Firm RR and assumed Influencing Factors................................................................................................... 115 Table 29: Test for Significance and Point-Biserial Correlation ........................... 116 Table 30: Regression Results - Firm Recovery Rate ............................................ 118 Table 31: Influencing Factors on Firm RR in Continuations and Liquidations .......................................................................................... 121

FIGURES

9

Figures Figure 1: Typical Insolvency Procedure in Germany ............................................. 29 Figure 2: Difference Total Debt.............................................................................. 90 Figure 3: Coverage Rate ......................................................................................... 92 Figure 4: Differences Firm Recovery Rate - Coverage Rate.................................. 93 Figure 5: Liquidity Gap .......................................................................................... 94 Figure 6: Industries (Whole Sample) ...................................................................... 97 Figure 7: Industries (Liquidation vs. Continuation) ............................................... 98 Figure 8: Reasons for Bankruptcy .......................................................................... 99 Figure 9: Firm Age in Liquidation vs. Continuation .............................................. 99 Figure 10: Year of Insolvency .............................................................................. 100 Figure 11: Location of Insolvency Court.............................................................. 101

ABBREVIATIONS

Abbreviations APR

Absolute Priority Rule

APRD

Absolute Priority Rule Deviation

CES

Center for Economic Studies

CRSP

Center for Research in Security Prices

EAD

Exposure at Default

ESUG

Gesetz zur weiteren Erleichterung der Sanierung von Unternehmen

GBP

Great British Pound

GDP

Gross Domestic Product

InsO

Insolvenzordnung

ISIC

International Standard Industrial Classification of all Economic Activities

NACE

Nomenclature Statistique des Activités Économiques dans la Communauté Européenne

OLS

Ordinary Least Squares

RR

Recovery Rate

UK

United Kingdom

US

United States

U.S.C.

United States Code

USD

US Dollar

11

INTRODUCTION

I

13

Introduction

One of the major topics in finance literature is the prediction of bank recovery rates for distressed companies. Due to requirements according to credit risk assessments as well as the calculation of loan interest rates, banks have to determine recovery rates as accurately as possible.1 Thus, studies about recovery rates are mainly driven by banks and analyze corporate reorganizations before a formal procedure is initiated. Since 1999, German insolvency legislation provides two different mechanisms within one formal procedure. Besides liquidation, reorganization by means of an insolvency plan (called Insolvenzplanverfahren) is possible. Investigations about corporate reorganizations within a formal procedure and corresponding creditor recovery rates are rare. This is due to a lack of data as well as the fact that only a few cases have been passed since 1999. Official statistics do not reveal the quantity of approved insolvency plans for corporate bankruptcies in Germany. Nevertheless, other statistics indicate that the proportion of insolvency plans in corporate bankruptcies was never higher than 2% between 1999 and 2012 respectively.2 In comparison to Chapter 11, which is the corresponding reorganization chapter in US bankruptcy law, the German reorganization procedure has been relatively uncommon until now. The main objective of the amendments to the German insolvency statute (called Gesetz zur weiteren Erleichterung der Sanierung von Unternehmen – ESUG), which were gradually made effective in 2012 and 2013, is to strengthen continuation within the formal procedure. This may increase ex-post efficiency, which suggests that efficient firms shall be continued and inefficient firms shall be liquidated. According to the supplements to German insolvency legislation, a convergence to Chapter 11 can be observed. Indeed, insolvency legislation does not only affect distressed companies. Even creditors and shareholders of viable firms consider insolvency rules while calculating expected values. Consequently, different incentives are created by insolvency rules which may, for example, impact corporate finance before bankruptcy. Literature refers to these difficulties under the term “ex-ante efficiency” of insolvency legislation.

1

See Grunert and Weber (2009), p. 505.

2

See Kanzlei Schultze & Braun (2013).

14

INTRODUCTION

As a first step, theoretical literature about efficiency of insolvency legislation is discussed in chapter II. Based on this analysis, bankruptcy procedures of the US and Germany are presented and the two insolvency laws are appraised according to ex-post and ex-ante efficiency (chapter III). Subsequently, empirical literature concerning the distribution of creditor recovery rates as well as influencing factors on the extent of recovery rates is analyzed (chapter IV). As mentioned above, studies about insolvency plans in Germany are rare.3 The few existing studies concentrate predominantly on descriptive analyses about the economic situation of insolvent companies, or use combined data sets which include reorganizations and liquidations. The repayments to creditors in reorganization procedures as well as influencing factors on the extent of the recovery rate are not investigated in detail. Against this background, the empirical study in chapter V concentrates on the economic situation of insolvent companies at the time of filing as well as specific aspects of ex-post efficiency in Germany: The extent of different creditor recovery rates as well as influencing factors on the recovery rate for all creditors (so called firm recovery rate) in formal reorganization procedures. In addition, the data set is separated into two subsamples. The first subsample includes firms which continue after the formal procedure and the second one involves firms which are liquidated after the acceptance of an insolvency plan. Again, influencing factors on firm recovery rates are presented. Chapter VI summarizes the main results of the study.

3

See for example Kranzusch and Icks (2009), Icks and Kranzusch (2010) as well as Blazy, Petey and Weill (2012).

EFFICIENCY OF INSOLVENCY LAWS

15

II Efficiency of Insolvency Laws II.1 Ex-post Efficiency II.1.a Incentives of Stakeholders in Bankruptcy In case of a business failure, stakeholders have to decide between liquidation and reorganization of the firm. The decision ought to be based on the expected value after liquidation compared to the expected value after reorganization. When the decision between liquidation and reorganization maximizes the expected value, the solution is considered to be ex-post efficient.4 Regarding investigations about bankruptcy resolutions, theory focuses on the different incentives stakeholders face. Consequently, all studies exclude a conflict of interest between old management and shareholders. The following considerations assume that managers and owners have congruent goals. One main problem is that stakeholders want to maximize the expected value of their own returns instead of the expected value of the entire firm. This can lead to ex-post inefficiencies. Jensen and Meckling (1976) show that equity holders of leveraged firms face incentives to overinvest.5 The reasoning behind this is that shareholders receive all residual earnings in case of success, while their risk is limited to the amount of equity. This effect is also known as “go for broke”.6 Conversely, debt holders tend to underinvest.7 As these incentive effects increase with the leverage ratio, they apply in particular to distressed firms. The deviation of the incentives of shareholders and creditors can impede the reorganization and liquidation proceedings. According to Jensen and Meckling (1976) and Burghof (1998), shareholders tend to reorganize firms that should be liquidated while creditors tend to liquidate firms that should be reorganized. The renegotiation power of equity- and debt holders is highly affected by insolvency laws. But insolvency laws do not only enforce stakeholders to assert themselves. Moreover, the legal framework can also create new incentives that can affect the distribution of the firm value between the different stakeholders.

4

See for example Hart (2006), p. 4, Aghion, Hart and Moore (1992), pp. 532-533 or White (1989), p. 129.

5

See Jensen and Meckling (1976), pp. 334-337.

6

See Harris and Raviv (1991), p. 301.

7

See Burghof (1998), pp. 505 et seqq.

16

EFFICIENCY OF INSOLVENCY LAWS

When firms are liquidated, the proceeds from the asset sales are distributed according to the absolute priority rule (APR). Accordingly, senior creditors have to be paid in full before junior creditors and subsequently shareholders receive any payments.8 Besides liquidation, several countries additionally provide a framework to reorganize the firm within a formal procedure. Insolvency laws define structured bargaining processes between shareholders and creditors to deal with the individual incentives to continue or shut down business operations. Particularly insolvency laws which allow for debt to equity swaps require the bargaining process to evaluate those assets that are not sold.9 The renegotiation results are included in the insolvency plan and stakeholders have to agree on the result by voting. Regarding reorganizations, shareholders may receive payments even though debt cannot completely be paid back to secured and unsecured creditors.10 This phenomenon is called absolute priority rule deviation (APRD). Literature provides different explanations for this deviation and identifies several influencing factors. Bebchuk and Chang (1992) investigate in their bargaining model how the firm value is distributed to creditors and shareholders within a reorganization procedure in the US (called Chapter 11).11 Therefore, they focus on the bargaining power of shareholders within Chapter 11 as well as some firm characteristics. According to their study, APRD are caused by the ability of shareholders to delay the resolution and thereby increase the costs of insolvency. Furthermore, shareholders can increase liquidation costs for instance by shifting from Chapter 11 to Chapter 7.12 Bebchuk and Chung find in a descriptive empirical analysis, that the APRD increases with the volatility of asset values, high insolvency costs, the duration of the reorganization, the length of the period equity holders are allowed to suggest a plan, the level of liquidation costs and with the proportion of assets to liabilities.13 Bergman and Callen (1995) also analyze influencing factors of the APRD in Chapter 11 and informal workouts in a bargaining model. They find that deviations from the APR depend on the decline of firm value in Chapter 11, firm value before default and Tobin´s Q.14

8

See Eberhart, Moore and Roenfeldt (1990), p. 1457 or Weiss (1990), p. 286.

9

See Braham and Steffen (2003), p.422.

10

See Franks and Torous (1989), Betker (1995) or Eberhart, Moore and Roenfeldt (1990).

11

See Bebchuk and Chang (1992), p. 254.

12

See Bebchuk and Chang (1992), p. 255.

13

See Bebchuk and Chang (1992), p. 256.

14

See Bergman and Callen (1995), pp. 11-13.

EFFICIENCY OF INSOLVENCY LAWS

17

In contrast, Baird and Bernstein (2006) doubt that shareholders do have sufficient bargaining power to deviate from the APR. They argue that, especially in large Chapter 11 cases, shareholders often have no private information which could be used strategically.15 They conclude that the APRD ensues from valuation problems within negotiations.16 Other influencing factors on the distribution of values are country-specific voting rights that are crucial for the final acceptance of the insolvency plan. Insolvency laws differ in the classes that are allowed to vote as well as in the requisite majority. Furthermore, the voting rights can be assigned  equally within each class.  according to the face value of claims or  a combination of both. These aspects are taken into consideration by Braham and Steffen (2003). They conclude that voting rights in different countries influence the adoption of a plan as well as the distribution of values between shareholders and creditors.17 II.1.b Solving Ex-post Inefficiencies Several theoretical investigations discuss how an ex-post efficient outcome can be achieved. Consequently, the studies included in this section can be divided into two different groups. A first group of studies concentrates on existing insolvency laws respectively single insolvency rules and describes their impact on ex-post efficiency. The second group of studies analyzes how insolvency laws and procedures should ideally be set up to ensure ex-post efficiency. Within the first group of studies that analyze existing legal frameworks, the results are contradictory. While some studies conclude that US bankruptcy law increases ex-post efficiency, other investigations find that Chapter 11 hinders an ex-post efficient outcome. One of the studies which support the first view is the one by Berkovitch and Israel (1998). They assume that inefficient investment decisions result from conflicts of 15

See Baird and Bernstein (2006), pp. 1937-1938.

16

See Baird and Bernstein (2006), p. 1942.

17

See Braham and Steffen (2003), p.440.

18

EFFICIENCY OF INSOLVENCY LAWS

interest between owner and creditors.18 Assuming information asymmetry, managers do not only face incentives to overinvest and “go for broke” (see Chapter II.1.a). They also tend to underinvest when the debt overload is so large that a project with a positive net present value would only increase the value of debt, not of equity. In case of underinvestment, an informal workout can solve the problem when debt holders are willing to deviate from the absolute priority rule and thereby let the owner participate on the profits of new projects. In contrast, Chapter 7 gives debt holders the possibility to liquidate firms according to the APR and thereby cope with overinvestment of managers. Since Chapter 7 gives the power to liquidate to debt holders, managers can file for Chapter 11 to protect themselves against liquidation. Once a firm is under Chapter 11, creditors cannot file for Chapter 7 and this increases the bargaining power of shareholders.19 This special combination of Chapter 7 and Chapter 11 gives the owners the incentive and the bargaining power to renegotiate in case of overinvestment.20 Baird (1991) also concludes that US bankruptcy law increases efficiency. The starting point is Chapter 7. He argues that in Chapter 7, the interests of all creditors are grouped. Creditors lose their right to seek repayment on their own. Individual rights are waived against the rights of the group.21 However, Chapter 7 still has to be filed by individual creditors. Baird argues that those creditors capable of detecting financial distress of their debtor early are often better off by seeking repayment individually.22 One cause could be that well informed creditors are often also collateralized. Baird concludes that creditors are very unlikely to trigger Chapter 7. Consequently, bankruptcy has to be filed by the managers. Chapter 11 therefore provides the legal framework. However, one main goal of bankruptcy law is to protect creditors and therefore relying on managers to file for bankruptcy would seem unreasonable.23 Baird analyzes whether Chapter 7 and Chapter 11 still provide incentives to managers to only file for bankruptcy when it is in the collective interest of firms’ creditors. He argues that the best date for filing for bankruptcy is when individual creditors start to seize their assets and managers cannot meet the due payments. As long as managers are unable to raise new money, the firm is insolvent and a bankruptcy procedure is collectively desirable.24 Otherwise, 18

See Berkovitch and Israel (1998), p. 2.

19

See Berkovitch and Israel (1998), pp. 2-3.

20

See Berkovitch and Israel (1998), p. 3.

21

See Baird (1991), p. 223.

22

See Baird (1991), p. 224.

23

See Baird (1991), p. 232.

24

See Baird (1991), p. 230.

EFFICIENCY OF INSOLVENCY LAWS

19

the firm will be liquidated and the manager will lose his job. For fear of losing his job, the manager has an incentive to file for bankruptcy. Therefore, managers act in the interest of all stakeholders. To prevent managers delaying the petition or entering into risky investments, insolvency law should also enable creditors to file for bankruptcy. According to these rules, Baird states that US bankruptcy law is appropriate for increasing ex-post efficiency. Brown (1989) also addresses consequences of insolvency rules on ex-post efficient outcomes. He concludes that especially the exclusive period for the debtor to initiate an insolvency plan, the possibility for a cram down25 and the impairment rule26 help to solve hold-out problems. These rules contribute to the fact that a firm will only be reorganized when the reorganization value is higher than the liquidation value. 27 Consequently, these rules can increase ex-post efficiency. Annabi, Breton and Francois (2011) also investigate the influence of the exclusive period to initiate an insolvency plan on the distribution to stakeholders and the expost efficient solution. In a simulation, they investigate how giving the initiating rights to different stakeholders influences the value of claims of each group and the probabilities of liquidation, cram down, adoption of the plan and of future default.28 Firstly, the results of the simulation confirm the first mover advantage. Each group can maximize its outcome when it is given the right to initiate the plan.29 Furthermore, they find that the trade-off between not violating priority rules and the successful continuation of distressed firms can be regulated by giving the initiating rights to different groups.30 When the first priority is on adherence to the absolute priority rule, junior creditors should initiate the insolvency plan because they need a higher level of protection against shareholders than senior creditors.31 However, when the first goal of the bankruptcy procedure is the reorganization of the company, shareholders should initiate the plan. Then, the distressed firm is deleveraged most and the likelihood of future default is lowest.32

25

As long as the plan is fair and equitable it even can be confirmed when groups of creditors vote against the plan.

26

Solely creditor groups which legal rights are being changed by the plan are allowed to vote over the insolvency plan.

27

See Brown (1989), p. 121.

28

See Annabi, Breton and Francois (2011), pp. 3 et seqq.

29

See Annabi, Breton and Francois (2011), p. 14.

30

See Annabi, Breton and Francois (2011), p. 17.

31

See Annabi, Breton and Francois (2011), p. 16.

32

See Annabi, Breton and Francois (2011), p. 24.

20

EFFICIENCY OF INSOLVENCY LAWS

Blazy and Chopard (2004) find that conflicting interests of stakeholders can hinder an optimal decision about the allocation of resources. By deviating from the APR, the payments to the stakeholders can be modified in a way that all relevant groups vote for the ex-post efficient solution. The deviations reflect the payments stakeholders of the firm have to pay or receive for their individual impact (negative or positive) on the value-maximizing decision.33 Hence, maximizing the firm value is the main goal of the APRD. Beyond that, Blazy and Chopard define further relevant criteria. These are financial neutrality (positive payments and negative payments cancel each other out; firm value stays constant), legal tolerability (legal involvements should be as minimal as possible; APRD should be minimal) and democratic desirability (choice of optimal economic outcome is collectively desirable).34 In their study, they analyze different legal frameworks against the background of these criteria. They conclude that Chapter 11 proceedings obtain a higher level of ex-post efficiency than German and British insolvency laws.35 They argue that the high bargaining power of creditors impedes APRD. Other studies focus on the information asymmetry between equity holders and creditors in game-theoretic models. If creditors cannot evaluate whether liquidation or continuation would maximize the value of the firm, equity holders of inefficient firms might try to mimic an efficient firm to force continuation and maximize their own returns.36 White (1994) concludes that the US bankruptcy law is inefficient when creditors cannot distinguish between efficient and inefficient firms. However, Mooradian (1994) shows how Chapter 11 can still enhance ex-post efficiency by incorporating APRD into his model. If equity holders create additional value by not overinvesting, they can be compensated within the APRD.37 In this model, ex-post efficiency is ultimately not increased by reducing information asymmetry, but by adjusting payments to the different stakeholders to incentivize the better informed equity holders to maximize the value of the firm. Summarizing the results, there is a broad consensus in literature that Chapter 11 ensures that efficient firms are saved. The question of whether inefficient firms can be filtered and liquidated is still under debate. Literature provides evidence that also firms that should be liquidated are reorganized in Chapter 11, which would lead to ex-post inefficiency.

33

See Blazy and Chopard (2004), p. 455.

34

See Blazy and Chopard (2004), p. 453.

35

See Blazy and Chopard (2004), pp. 462-464.

36

See Mooradian (1994), p. 1404 and White (1994), p 277.

37

See Mooradian (1994), p. 1422.

EFFICIENCY OF INSOLVENCY LAWS

21

The second group of studies discussed in this chapter investigates how an optimal insolvency law should be written or an existing law should be converted to achieve ex-post efficiency. The main critiques on the widespread negotiationbased insolvency laws are the high direct and indirect costs. Furthermore, negotiations can lead to ex-post inefficiency when one group of stakeholders has too much bargaining power. As discussed earlier in this chapter, the US bankruptcy law tends to reorganize firms that should be liquidated.38 In contrast, giving more bargaining power to creditors, as the German insolvency law does, can lead to liquidations of firms that should have been reorganized. White (1994) analyzes the American insolvency law that is based on two different chapters, Chapter 7 and Chapter 11. She shows in her model that equity holders and managers face incentives to reorganizes inefficient firms in chapter 11.39 This can lead to ex-post inefficiency when the less informed creditors cannot distinguish between efficient and inefficient firms. White also discusses the effects on ex-post efficiency if only a liquidation chapter were to exist. In this case, ex-post inefficiency would arise from efficient firms that would be liquidated. In analyzing this trade off, she concludes that the total costs are lower in a legal framework that favors reorganization.40 One strand of literature expects improved efficiency due to a market based insolvency legislation.41 Hansen and Thomas (1998) as well as Hotchkiss and Mooradian (2003) propose to combine negotiations with an auction approach.42 Hansen and Thomas find that auctions are always efficient within a Chapter 11 procedure when stakeholders fail to reach an agreement within a certain time. In contrast, Bebchuk (1988, 2000) relies on a solution with options. If the firm value is unknown, every stakeholder receives options according to group membership and the amount of the claim.43 Subsequent to the issuance of options, two possibilities arise. Either the stakeholders get a particular price for the option or they are allowed to buy shares in the firm. Once the new shareholders are known, they can make the decision between liquidation and reorganization, which maximizes firm value.44

38

See for example Bebchuk (2000), p.832, Bradley and Rosenzweig (1992) or White (1994), p. 291.

39

See White (1994), p. 291.

40

See White (1994), p. 293.

41

See Hansen and Thomas (1998), pp. 164-165 and Hotchkiss and Mooradian (2003), p. 556.

42

See for example Hansen and Thomas (1998), p. 161.

43

See Bebchuk (2000), pp. 835-836.

44

See Bebchuk (2000), p 837.

22

EFFICIENCY OF INSOLVENCY LAWS

Beyond legal changes, ex-post efficiency can be enhanced by adjustments of the capital structure to reduce moral hazard through debt-equity swaps or equity issuances. Subsequently, the equity ratio increases and stakeholders incentives to under- or overinvest decrease.

II.2 Ex-ante Efficiency Several studies investigate how bankruptcy rules affect healthy firms even before they become bankrupt.45 Most studies focus on the APRD. Managers and creditors of healthy firms might be influenced in relevant decisions if they know that insolvency might cause a deviation from the APR. Ex-ante efficiency can be reached if the bankruptcy laws do not defer financing and investment decisions of healthy firms. Literature provides heterogeneous theories for both efficient and inefficient effects of the APRD. The most important theories that explain negative effects of the APRD on ex-ante efficiency analyze credit rationing, strategic defaults and gambling for resurrection. These theories will be described subsequently. Longhofer and Carlstrom (1995) find that APRDs have negative effects on ex-ante efficiency. Within their model, they investigate creditors´ willingness to provide capital, facing possible deviations from the APR.46 The study concludes that creditors claim higher interest rates when they fear deviations from the APR. If the deviations are high, they can even cause credit rationing. In addition, they point out that APRDs reduce the incentives to manage firms in a profitable manner because the deviation limits the loss of shareholders in case of default.47 Besides these negative effects on ex-ante efficiency, Longhofer and Carlstrom also highlight positive effects on ex-post efficiency. Especially for small companies in which the owner is inextricably linked with the business, APRD can help to incentivize the owner to reorganize the company.48 Longhofer and Carlstrom conclude that optimal insolvency legislation should make it possible to agree upon APRD or APR adherence in private contracts. In small manager-owned companies, APRD should be allowed. In contrast, bankruptcy procedures in large public companies should

45

See for example Pindado, Rodrigues and de la Torre (2008), Longhofer and Carlstrom (1995), Picker and Bebchuk (1993), Berkovitch, Israel and Zender (1997) and others.

46

See Longhofer and Carlstrom (1995), p. 24.

47

See Longhofer and Carlstrom (1995), p. 26.

48

See Longhofer and Carlstrom (1995), p. 25.

EFFICIENCY OF INSOLVENCY LAWS

23

stick to the APR because a change in management could help to improve the situation.49 Further negative effects can arise when expected deviations from the APR create incentives for strategic default. Renegotiating contracts with creditors in Chapter 11 could then increase shareholders' wealth in certain cases.50 A broad range of studies discusses how strategic default can be prevented by arrangements in credit contracts.51 Bebchuk (2002) also analyzes the consequences of APRDs on ex-ante efficiency. He discusses moral hazard and the negative impact of APRDs on decisions of the management regarding investments (overinvestment problem), the distribution of dividends and financing of the enterprise.52 Concerning investment decisions, Bebchuk concludes that owners of a debt-financed company are incentivized to prefer risky projects, irrespective of whether deviations from the APR are expected. However, the incentive to invest in risky projects increases under regimes that allow for APRDs.53 Furthermore, APRDs increase incentives to shareholders for inappropriately high withdrawals (like dividends or salaries) and unnecessary additional borrowings.54 The basic idea behind these theories is that creditors have to bear the costs of managerial failure while owners benefit from the advantages alone. Bebchuk only considers APRDs between creditors and shareholders. Consequences of APRDs between different creditor groups are not investigated.55 Pindado, Rodrigues and de la Torre (2008) empirically analyze how different insolvency codes affect investments of healthy firms by empirically comparing Tobin's q for firms in different countries. Therefore, they examine specific differences for other countries and firms.56 They find that investments are less attractive in countries with high insolvency costs.57 Furthermore, they conclude that underinvestment incentives increase with several procedural characteristics such as APRDs or automatic stay, for example, while overinvestment problems arise due

49

See Longhofer and Carlstrom (1995), p. 27.

50

See Vychodil (2010), p. 106.

51

See Bolton and Scharfstein (1996) or Bester (1994).

52

See Bebchuk (2002), p. 447.

53

See Bebchuk (2002), p. 452.

54

See Bebchuk (2002), p. 456.

55

For this purpose see Bebchuk and Fried (1996), pp. 863, 865 and 870.

56

See Pindado, Rodrigues and de la Torre (2008), p. 228.

57

See Pindado, Rodrigues and de la Torre (2008), p. 229.

24

EFFICIENCY OF INSOLVENCY LAWS

to insolvency rules like debtor-in-possession management.58 Especially the result that APRDs increase underinvestment is contradictory to most other studies. One cause might be that Pindado, Rodrigues and de la Torre do not solely analyze APRD between shareholders and creditors, but also between the different groups of creditors.59 Other studies find that APRD induces positive incentives with respect to investment decisions, information disclosure and monitoring. Picker and Bebchuk (1993) developed a human capital model which assumes that management can increase its own skills by investing in their human capital at the beginning of a project. By doing so, the firms payoff can be increased in both cases, if the firm is running well and in case of bankruptcy.60 However, the additional value created by investing in human capital is specific to the current management. Picker and Bebchuk state that under strict adherence of APR and when creditors control the firm in bankruptcy, the specific skills of the managers are not used to create an additional value anymore. For this reason, it is less attractive for managers to invest into human capital at an early stage compared to regimes that allow for APRDs.61 Because managers have to bear the investment costs in human capital, they also want to reap the profits. As investments in human capital increase the value of a firm, a new owner is willing to pay more in case of bankruptcy. Under absolute priority, the managers will invest in human capital only as long as the extra payments cover the costs. This can lead to an “underinvestment in human capital”.62 In a regime that allows deviations from the APR, the original owners benefit from the payments of new owners and the amount according to the APRD.63 Thus, the incentive to invest in human capital increases. Berkovitch, Israel and Zender (1997) also investigate how bankruptcy codes influence investments in human capital of healthy firms. They developed a model in which the manager can choose to increase his skills. Firstly he can choose to develop his firm-specific knowledge that would increase the value of the firm, irrespective of whether the firm goes bankrupt or not. Secondly, he can improve his general skills that would increase the salary he could earn outside the company.64 58

See Pindado, Rodrigues and de la Torre (2008), p. 236.

59

See Pindado, Rodrigues and de la Torre (2008), p. 230.

60

See Picker and Bebchuk (1993), pp. 13-14.

61

See Picker and Bebchuk (1993), pp. 14-16.

62

See Picker and Bebchuk (1993), p. 15.

63

See Picker and Bebchuk (1993), p. 5.

64

See Berkovitch, Israel and Zender (1997), p. 489.

EFFICIENCY OF INSOLVENCY LAWS

25

Berkovitch, Israel and Zender show that the legal structure of bankruptcies already influences this investment decision in human capital of the management ex-ante. To determine the first best solution, they determine how much the manager would invest in firm specific knowledge without debt.65 As soon as the firm is indebted, the investment in human capital declines because the additional value of human capital then accrues to the creditors. Allowing for deviations from the APR only can mitigate the underinvestment into firm-specific knowledge because even with APRDs, the additional value still has to be shared with the creditors in case of insolvency.66 Berkovitch, Israel and Zender conclude that within their model, an auction-based bankruptcy procedure can achieve both ex-ante and ex post efficiency.67 First of all, an auction-based procedure maximizes the firm value in bankruptcy. Secondly, the bargaining power of the manager increases in an auction system in line with the value he can add to the firm, which in turn increases his firm-specific knowledge. However they find that the auction needs specific restrictions to avoid unwanted incentives like strategic defaults.68 Gertner and Scharfstein (1991) analyze the different effects of bank debt and public debt. They conclude that inefficiency arises from difficulties to renegotiate public debt with numerous debt holders.69 As debt always does, public debt on the one hand sets incentives to overinvest because equity holders do not bear the whole risk but receive the residual earnings. On the other hand, public debt can lead to underinvestment when it has to be paid back at face value and new bank debt has to be funded to realize positive net present value projects. Then, the new bank debt has to bear the risk while public debt holders benefit because their claims can be paid back at face value.70 In their model they show, that efficiency can be increased by Chapter 11 under certain circumstances. Chapter 11 gives a lot of bargaining power to equity holders, which makes it difficult for the bank to liquidate the firm.71 So in many cases, liquidations are too expensive and the bank has only two alternatives left: continue the firm under Chapter 11 or provide new capital to the firm for additional investments, thereby preventing an official bankruptcy procedure.72 According to the model, the bank will always choose the more 65

See Berkovitch, Israel and Zender (1997), p. 490.

66

See Berkovitch, Israel and Zender (1997), p. 491.

67

See Berkovitch, Israel and Zender (1997), pp. 492-493.

68

See Berkovitch, Israel and Zender (1997), p. 492.

69

See Gertner and Scharfstein (1991),p. 1215.

70

See Gertner and Scharfstein (1991), p.1191.

71

See Gertner and Scharfstein (1991), p.1213.

72

See Gertner and Scharfstein (1991), p.1214.

26

EFFICIENCY OF INSOLVENCY LAWS

efficient option out of these two alternatives. Additional investments can therefore be increased by Chapter 11. However, in cases in which liquidations would be the best outcome, efficiency can be hindered by Chapter 11 because it makes liquidations less attractive to debt holders.73 Heinkel and Zechner (1993) developed a model to analyze how the debtors can use private information strategically. Information asymmetry does not only provide incentives for the well discussed strategic defaults. It also can incentivize owners to hide financial distress. This occurs when the owners have private information that future prospects of the firm are bad and they try to squeeze out the already distressed firm with inadequately high dividends.74 Short term debt can help to solve this inefficiency. If the short term debt cannot be paid back in time, bankruptcy will already be triggered at an early stage.75 However, the ability to pay back short term debt can only be used as a signal for the firms´ future prospects when cash flows are correlated over time.76 If the correlation is low, APRD can provide incentives for the owners to file for bankruptcy when the expected value of equity is higher in insolvency than by squeezing out the firm.77 Berkovitch and Israel (1999) investigate which bankruptcy codes are optimal for market-based and bank-based financing markets. They assume that creditors are less informed in market-based system than in a bank-based system. Therefore, managers can use their private information strategically more easily in market based systems.78 Berkovitch and Israel show that a deviation from the APR can create incentives to reveal private information and thus consider the APRD to be an “information rent”.79 Cornelli and Felli (1997) analyze how incentives to monitor are affected by insolvency laws. They argue that two chapters are therefore needed. One that allows violations of the APR between senior and junior creditors, and one that sticks to the APR.80 Cornelli and Felli conclude that in most cases, deviations from the APR increase incentives to monitor.81 73

See Gertner and Scharfstein (1991), p.1214.

74

See Heinkel and Zechner (1993), p. 532.

75

See Heinkel and Zechner (1993), p. 533.

76

See Heinkel and Zechner (1993), p. 533.

77

See Heinkel and Zechner (1993), pp. 550-552.

78

See Berkovitch and Israel (1999), p. 367.

79

See Berkovitch and Israel (1999), p. 370.

80

See Cornelli and Felli (1997), pp. 484-485.

81

See Cornelli and Felli (1997), p. 484.

EFFICIENCY OF INSOLVENCY LAWS

27

Bigus (2002) explores the impact of different insolvency rules, such as implementing a bankruptcy trigger, a debtor in possession rule and the right to file for bankruptcy by creditors. He investigates their effects on forming inefficient coalitions. Bigus states that coalitions between owners and junior creditors lead to risky investment activities, resulting in disadvantages for senior creditors. On the other hand coalitions between owners and secured creditors lead to investments which are less risky, thereby discriminating junior creditors.82 Bigus claims that formal insolvency procedures help to avoid inefficient coalitions and therefore are superior to private resolutions. Beyond, the profits for members of inefficient coalitions have to be minimized. A bankruptcy trigger like over-indebtedness in Germany is not able to prevent the formation of inefficient coalitions because the formulation is not precise enough to lead to early filings for bankruptcy.83 However, inefficient coalitions can be prevented by the possibility of creditors filing for bankruptcy. This option is limited for creditors in the US. In contrast, the debtor-in-possession rule and the possibility to deviate from APR inherent in Chapter 11 offer greater involvement of the owners and reduce their incentive to form an inefficient coalition. Summarizing the findings literature provides evidence of a trade-off between an ex-post and an ex-ante efficient solution. Several studies show that deviations from absolute priority can induce an ex-post efficient outcome. On the other hand, APRDs can cause ex-ante inefficiencies. Thus, the trade-off must be wellbalanced.84 Consequently, country specific factors like market- or bank based credit markets have to be taken into consideration. The following chapter will discuss the insolvency legislation of the US and Germany against this background.

82

See Bigus (2002), p. 110.

83

See Bigus (2002), pp. 124-126.

84

See for example Aghion, Hart and Moore (1992) or Berkovitch, Israel and Zender (1998).

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

29

III Overview of the German and the US Insolvency Law III.1 German Procedure The new German Insolvency Statute has been in force since 1999. Some important changes were made effective in 2012 and 2013. As the empirical analysis in chapter V uses data of corporate bankruptcies before the incorporation of the amendments, the next sequence will focus on the primary code of 1999. Nevertheless, the major amendments of 2012 and 2013 will be highlighted. The main objective of the German insolvency legislation is to maximize the payoff to creditors. In contrast to the US insolvency law, both liquidation and reorganization are incorporated into the same procedure. Figure 1 illustrates how insolvencies in Germany are typically carried out.

Figure 1: Typical Insolvency Procedure in Germany

A crucial requirement for filing for bankruptcy is the existence of one of three insolvency triggers. As defined in §§ 17-19 InsO, reasons to open insolvency pro-

30

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

ceedings can be illiquidity or imminent illiquidity as well as over-indebtedness for corporate entities. Both debtors and creditors are allowed to file for bankruptcy.85 Following the insolvency petition, courts usually appoint a preliminary administrator86 who writes an initial report to verify the correct filing. In addition, the court can immediately arrange for the company's assets to be frozen to protect creditors’ claims. The advice of the preliminary administrator is decisive for the opening or dismissal of insolvency proceedings by the court. In a first step, he validates whether bankruptcy is triggered and preliminary evaluates future prospects of the firm. Furthermore, the administrator estimates whether the insolvency costs could be covered.87 If the free assets of the firm along with voluntary extra payments do not cover the expected insolvency costs, the court can refuse the opening of insolvency proceedings or defer the costs. Otherwise, the insolvency procedure starts when bankruptcy is triggered and costs are covered. After the decision to open the procedure, the court appoints an administrator88 who takes control over the company´s assets89. Only if the debtor successfully applies for debtor-inposition management, he is allowed to manage the company under supervision of a trustee.90 Furthermore, debt holders are called to file their claims and inform the administrator about their collaterals.91 The court then determines dates for a report- and a verification meeting.92 These two meetings are usually combined at one date. In the verification meeting, the filed claims are validated.93 During the report meeting, the administrator presents his analysis of whether the business operations should be continued or liquidated.94 Not only expected returns and the value of assets95 but also the administration expenses96 of the insolvency procedure have to be taken into account. Based on the report, the creditors´ assembly decides whether the firm will be liquidated 85

See § 13 InsO.

86

See § 21 InsO.

87

See § 26 InsO.

88

See § 27 InsO in conjunction with § 56 InsO.

89

See § 80 InsO in conjunction with § 148 InsO.

90

See §270 InsO.

91

See § 28 InsO.

92

See § 29 InsO.

93

See § 176 InsO.

94

See § 156 InsO.

95

See § 35 InsO.

96

See § 53 InsO.

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

31

or reorganized. Furthermore, debt holders have the option of establishing a creditors´ committee97 which has the power to replace the administrator98. If the creditors decide to reorganize the company, an insolvency plan has to be worked out by the debtor or the administrator.99 Within the plan, deviation from the absolute priority rule is allowed, as long as creditors’ returns do not fall below their estimated payments in case of liquidation.100 Within the insolvency plan, debt holders are grouped in classes according to their seniority.101 After the submission of a plan, the court announces a date to vote for the plan.102 Usually, the voting date coincides with a further verification meeting to validate additional claims submitted after the initial verification meeting.103 In the run up to the voting procedure, the insolvency plan can still be modified in collaboration with the stakeholders. The adoption of the plan requires a twofold majority vote within each impaired class.104 Within each creditor class, at least fifty per cent of face-value of debt and also fifty per cent of the number of creditors have to accept the plan. To enhance the chances of acceptance, the obstruction by creditors is prohibited when their returns in reorganization are higher than in case of liquidation.105 Furthermore, a debtor´s consent is assumed as long as he does not object. The court has to confirm the accepted insolvency plan.106 After confirmation and discussion about the final report107, the court can close the proceedings108 and the debtor returns control over the assets of the company109. However, within the insolvency plan it is possible to determine the supervision of the plan for a longer period.110

97

See § 68 InsO.

98

See § 57 InsO.

99

See § 157 InsO and § 218 InsO.

100

See § 217 InsO.

101

See § 222 InsO

102

See § 235 InsO.

103

See § 177 InsO in conjunction with § 236 InsO.

104

See § 244 InsO and § 77 InsO.

105

See § 245 InsO.

106

See § 248 InsO.

107

See § 197 InsO.

108

See § 258 InsO.

109

See § 259 InsO.

110

See § 260 InsO.

32

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

As mentioned earlier in this chapter, some important amendments were made effective in 2012 and 2013 (called Gesetz zur weiteren Erleichterung der Sanierung von Unternehmen - ESUG). The main objective is to disburden the reorganization of companies within a formal procedure. In the context of this study, three major changes are important. Firstly, one goal was to strengthen the influence of creditors. Therefore, courts now are obliged to establish a preliminary creditors´ committee if the firm remains in business and two out of three requirements concerning total assets, turnover and number of employees are met.111 Furthermore, the preliminary creditors´ committee is allowed to propose a specific administrator or to define the requirements the administrator has to fulfill.112 Secondly, the insolvency procedure should be enhanced by facilitating corporate actions and changes in voting rights. Until the new act came into force, a modification of shareholder rights was difficult. Now it is possible to convert creditors´ claims into ownership rights as long as two conditions are met. The changes have to be written down in the insolvency plan and creditors have to agree to the conversion.113 The probability of realizing a debt to equity swap and thereby strengthening the equity base of the firm is improved respectively. The interference in the rights of shareholders is the reason why they now have to agree to the plan and therefore a group of shareholders has to be formed additionally.114 Previously, shareholders were not allowed to vote on the insolvency plan. The third goal was to enforce debtors to dispose over the firms´ assets and thereby manage the firm after filing (so called debtor-in-possession management). Prior to the amendments, courts frequently rejected the application for debtor-inpossession management. After the supplements, the court has to accept an application as long as the preliminary creditors´ committee supports the application.115 Furthermore, firms which are not illiquid and have positive prospects for the future have to be encouraged to solve their problems contemporary by keeping control over the firms´ assets and preparing an insolvency plan within three month after filing (so called Schutzschirmverfahren).116

111

See § 22a InsO.

112

See § 56a InsO.

113

See § 225a InsO.

114

See § 238a InsO.

115

See § 270 InsO.

116

See §§ 270a-c InsO.

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

33

III.2 US Procedure Under the Bankruptcy Act of the United States of America, insolvent corporations have the choice to file for Chapter 7 or for Chapter 11. Under Chapter 7, the assets of the firm are sold as a going concern or piecemeal.117 The proceeds are divided according to the seniority of all participants. Accordingly, distribution follows the absolute priority rule. The objective of a filing under Chapter 11 is to reorganize the firm and to continue operations as a going concern.118 As this is the counterpart to the German insolvency plan procedure forming the basis of this study, this section will focus on Chapter 11. In Chapter 11, the petition is made by the debtor (voluntary petition) or, under specific circumstances, by the creditors (forced petition).119 In contrast to the German law, there are no specific events which trigger a formal bankruptcy procedure. Thus, the corporation which files for bankruptcy may be solvent.120 Usually, the debtor retains control over the assets of the corporation (called debtor-inpossession management).121 After the filing, an automatic stay protects the debtor. For example, payments to creditors can be stopped and secured creditors are prohibited from liquidating their collateral.122 In addition, the debtor has the possibility to acquire new financial resources. Due to a super-priority status, the new creditors face a lower default risk than the old creditors.123 As mentioned before, the main objective according to Chapter 11 is to reorganize the firm. Therefore, a reorganization plan has to be developed. In the scope of the plan, claims of creditors with similar characteristics are grouped together.124 For each group, the plan regulates the reallocation of the financial claims. A plan can be prepared by the debtor or any other participant of the procedure. However the debtor has the exclusive right to propose a plan within the first 120 days after filing.125 Consequently, the bargaining power of the debtor is high because creditors are merely able to accept or reject the plan. The approval of the plan requires a 117

See Bebchuk (1988), p. 775.

118

See Hotchkiss, John, Mooradian and Thorburn (2008), p. 12.

119

See 11 U.S.C. §§ 301, 303.

120

See Davydenko and Franks (2008), p. 570.

121

See 11 U.S.C. § 363 (c).

122

See 11 U.S.C. § 362 and Hotchkiss, John, Mooradian and Thorburn (2008), p. 12.

123

See 11 U.S.C. § 364.

124

See 11 U.S.C. § 1122 (a).

125

See 11 U.S.C. § 1121 (b).

34

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

two-thirds majority on face value and a simple majority in number within each group.126 To avoid rejection by a few creditors, a cram down is possible. However a cram-down is only possible as long as every class is treated fairly and equitably. In this case, the court has to identify the going concern value within a special hearing. Thereby the verification is time-consuming and costly.127 To finalize the procedure the court has to confirm the plan.128

III.3 Legal Differences between Germany and the US The comparison between the American bankruptcy law and the insolvency statute before ESUG in Germany (see table 1) shows that Chapter 11 is rather debtorfriendly, focusing on continuation of the firm whereas the German rules emphasize creditor rights. Accordingly, in the US the debtor retains control over the assets after filing for bankruptcy, while in Germany an administrator is appointed and debtor-in-possession management is less likely. Other issues reflecting the debtor-friendliness of Chapter 11 are for example the rules concerning the filing for bankruptcy, the initiation of an insolvency plan and the voting rules. In Germany, creditors have a stronger influence over the procedure. This is caused by lower requirements for creditors to file for bankruptcy, the right to initiate an insolvency plan without any exclusive period for debtors and voting rights. As a result, the American bankruptcy law focuses on the continuation of firms and hence rather on ex-post efficient outcomes, whereas the German rules concentrate on creditor rights and thereby ex-ante efficiency. The German insolvency law interplays well with the highly developed credit market and facilitates lending conditions by accepting lower ex-post efficiency. However, relationship banking in Germany also contributes to increase ex-post efficiency. Banks anticipate future business in case of continuation and therefore face incentives to reorganize distressed firms. In the US, firms are reorganized more frequently. In contrast to the insolvency legislation of Germany, Chapter 11 focuses on the interest of debtors. This is in line with the less developed credit market. To counteract ex-ante inefficiencies, American creditors claim covenants in credit agreements more frequently.129 This illustrates that there are substantial interactions between the content of insolvency laws 126

See 11 U.S.C. § 1126 (c).

127

See Hotchkiss, John, Mooradian and Thorburn (2008), p. 14.

128

See 11 U.S.C §1129.

129

See for example Leuz, Deller and Stubenrath (1998), p. 112.

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

35

and credit markets. Consequently, the specifics of particular credit markets have to be taken into consideration when changes in insolvency laws are discussed.130

Table 1: Comparison of Insolvency Procedures in Germany and the US US

Germany Insolvency Statute new

Chapter 11

Insolvency Statute old

One chapter for reorganization  and liquidation. Reorganization  depends on the decision of the  creditors´assembly. Satisfaction of the creditors.

Like old law, but with the  possibility to accelerate the  process to reorganize.

Objective

Besides Chapter 7 for  liquidations, an independent  chapter for reorganization  procedures exists (Chapter 11). Reorganization

Owner rights

Interferences possible.

Interferences are difficult.

Interferences possible as long as  written down in the insolvency  plan.

Procedure

Satisfaction of the creditors. But  the new rules shall promote  continuation. Petition for bankruptcy Debtor or creditor. Debtor or creditor. Debtor or creditor. Trigger No insolvency test. One of three triggers have to be  One of three triggers have to be  verified: imminent illiquidity,  verified: imminent illiquidity,  illiquidity, over‐indebtedness. illiquidity, over‐indebtedness. Control rights Debtor‐in‐possession. Debtor  Court appoints an administrator  Like old law. Debtor‐in‐ or trustee in case of debtor‐in‐ retains control over the assets. possession management is  possession management.  encouraged. Creditors´ assembly votes on an  insolvency plan. Automatic stay For most creditor claims. Three month for secured,  Three month for secured,  unlimited for other creditors. unlimited for other creditors. Possible, but creditors have to  New financing Possible. New claims with priority  Possible, but creditors have to  agree. agree. over old claims. Initiate a plan Creditors and debtor, but exclusiv  Debtor or administrator on behalf  Debtor or administrator on behalf  period for debtor. of creditors. of creditors. Creditor classes and shareholders  Voting rules All classes must agree: secured,  Creditor classes vote; debtor´s  vote. Debtor´s consent assumed.  unsecured, debtor; 2/3 in value  consent assumed; simple  majority in value and number. Simple majority in value and  and 1/2 in numbers. number for creditor classes and  simple majority in value for  shareholder class.

The aim of the amendments to insolvency law in Germany is to enhance continuations of insolvent firms and thereby improve ex-post efficiency. Along with the new rules in Germany (after ESUG), a convergence with Chapter 11 occurs. As already described in chapter III.1, changes arise according to debtor-in-possession management, shareholder rights and voting rules. The amendments induce a more debtor-friendly law and improve ex-post efficiency. Although theoretically, creditor rights should be strengthened as well, the formation of a preliminary creditors´ 130

See Hart (2006), p. 3.

36

OVERVIEW OF THE GERMAN AND THE US INSOLVENCY LAW

committee before the court appoints an administrator seems to be complicated. Because of extensive conflicts between creditors which prevent negotiations outof-court, it seems to be unrealistic that a preliminary creditors´ committee proposes an administrator in time. Overall, the supplements seem to be suitable for enhancing the continuation of firms in financial distress. However, to date, there are no studies about survival rates before and after the law reform as well as about the impact on ex-ante efficiency. A convergence with the American bankruptcy rules might be deteriorating financing conditions for healthy firms. As a result, adjustments concerning the supply of capital are needed. For instance, the underdeveloped equity market in Germany should be promoted. Due to country-specific differences in insolvency legislation and credit markets, it becomes clear that results of studies about ex-post or ex-ante efficiency may diverge strongly. The following analysis deals with empirical studies concerning different insolvency legislations and procedures. This study therefore focuses on particular aspects affiliated with ex-post efficiency. Thus the next section presents empirical studies investigating the extent of different creditor recovery rates as well as influencing factors on these recovery rates.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

37

IV Empirical Evidence for Corporate Bankruptcies IV.1 Extent of Recovery Rates As mentioned earlier (see chapter II) the insolvency law regulates the division of value in corporate bankruptcies. The payments each creditor group receives in relation to their face value of claims is called recovery rate of the considered group.131 Until now, several empirical studies investigate recovery rates in different countries. Table 2 gives an overview over major studies that analyze recovery rates of the entire firm as well as recovery rates of secured and unsecured creditors. The results are grouped by reorganizations, liquidations and combined data sets. Furthermore, table 2 provides additional information according to the country and the time period as well as the size of the recognized data sample. Referring to recovery rates of the whole firm, reorganization via Chapter 11 offers the highest recovery rates between 51-73%. The lowest firm recovery rate can be observed for British corporations (approximately 20%). Thereby, average firm recovery rates of the illustrated studies range between 20% and 73%. Firm recovery rates in liquidations are considerably lower than those of reorganizations. Bris, Welch and Zhu (2006) find average recovery rates equal to 27% for liquidation procedures in the US.132 According to all examined studies about liquidations, average firm recovery rates range between 5% and 27%. Several studies investigate firm recovery rates by using combined data sets that include both reorganizations and liquidations. Some data sets additionally include informal workouts. Studies analyzing firm recovery rates in auction-based legislations, like in Sweden (35%) or the Netherlands (37%), find higher recovery rates. One possible explanation might be that there is a bidding competition in auctions which results in higher recovery rates. However, the comparability of the results based on combined data sets is limited as the proportion of continuations and liquidations varies for each data set.

131

See Thorburn (2000), p. 361.

132

See Bris, Welch and Zhu (2006), p. 1288.

38

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

Table 2: Overview of Empirical Studies (Firm, Secured and Unsecured Recovery Rates)133 Authors & Procedure

Country

Period

Sample Size

Firm  Recovery  Rate in %

Secured  Unsecured  Recovery  Recovery  Rate in % Rate in %

Reorganization Baird/Bris/Zhu (2007) US Bris/Welch/Zhu (2006) US Franks/Torous (1994) US Tashijan/Lease/McConnell (1996) US Carapeto (2007) US Ivashina/Smith/Iverson/Strömberg (2011) US Armour/Hsu/Walters (2006) UK Armour/Hsu/Walters (2012) UK Blazy/Chopard/Nigam (2013) UK Blazy/Chopard/Nigam (2013) France Blazy/Chopard/Fimayer/Guigou (2011) France Sundgren (1998) Finland Germany Kranzusch/Icks (2009) Icks/Kranzusch (2010) Germany Thorburn (2000) Sweden

1995‐2001 1995‐2001 1985‐1990 1986‐1993 1989‐1998 1998‐2009 2003‐2004 2003‐2004 1998‐2005 1993‐2005 1989‐2005 1993‐1994 2002‐2008 2002‐2008 1988‐1991

139 225 37 41 57 136 195 195 199 164 171 63 148 5 142

69.4 50.9 72.9 ‐ 51 21 20.2 20.5 45.8 69.3 42.8 46.9 14.4 39

92.1 90.2 80.1 99.3 93 ‐ 61 ‐ 38.7 51.9 68.5 ‐ ‐ ‐ 77

52.3 51.6 28.9 64 43.7 ‐ 0.006 ‐ 3.5 38.2 67.9 ‐ ‐ ‐ 2

US UK UK UK UK France France Germany

1995‐2001 2001‐2003 2001‐2003 1992‐1999 1998‐2005 1993‐2005 1989‐2005 2002‐2008

61 153 109 42 100 100 575 5620

27.4 21 20.5 ‐ 8.6 19.6 14.2 5,4

51.4 55 ‐ 61.7 16.2 40.3 19.5 ‐

1.1 0.002 ‐ ‐ 7.7 3.1 5 ‐

UK France France Finland Sweden Germany Netherland

1993‐2005 1989‐2005 1993‐2005 1982‐1992 1988‐1991 1999‐2005 1983‐2000

520 942 264 83 210 126 137/114

13.27 14.7 20.67 35.1 35 21.46 37

29.2 20.1 35.28 ‐ 69 76.71 ‐

6.36 5.5 5.82 ‐ 2 10.1 ‐

Liquidations Bris/Welch/Zhu (2006) Armour/Hsu/Walters (2006) Armour/Hsu/Walters (2012) Citron/Wright/Ball/Rippington (2003) Blazy/Chopard/Nigam (2013) Blazy/Chopard/Nigam (2013) Blazy/Chopard/Fimayer/Guigou (2011) Kranzusch/Icks (2009) Combined Data Sets Blazy/Petey/Weill (2012) Blazy/Chopard/Fimayer/Guigou (2011) Blazy/Petey/Weill (2012) Ravid/Sundgren (1998) Thorburn (2000) Blazy/Petey/Weill (2012) Couwenberg/de Jong (2008)

Focusing on secured creditors, studies find average recovery rates between 39% and 99% in case of continuation and between 16% and 62% for liquidations. As 133

Own illustration according to Couwenberg and de Jong (2008), pp. 110-111.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

39

can be expected, the results of studies using combined data sets lie in between with average recovery rates ranging from 20% up to 77%. Thus, the results vary a lot. One study which addresses these differences is the one by Blazy, Chopard and Nigam (2013). Depending on the country and the procedure, they show that the differences are significant. The secured recovery rate for reorganized firms is higher in France (52%) than in the UK (39%).134 In case of liquidations, secured recovery rates in France average at 40% and at 16% for the UK. US studies which concentrate on Chapter 11 cases reveal highest recovery rates for secured claims (over 80% on average). Average recovery rates for unsecured creditors are lowest in liquidations and combined data sets at a level of 10% and less. However, unsecured average recovery rates in reorganizations across all countries are frequently much higher. The study of Blazy, Chopard, Fimayer and Guigou (2011) finds an average recovery rate of 68% for France. The study by Blazy, Chopard and Nigam (2013) also investigates the repayments to unsecured creditors. As for secured creditors, they show that the extent of recovery rates for unsecured creditors depends on the country and the procedure.135 For Germany, solely one study investigates the recovery rate of different creditor groups. Blazy, Petey and Weill (2012) find in their sample that all groups of creditors on average receive 21% of their claims, whereas the recovery rates for secured and unsecured creditors average out at 77% and 10% respectively.136 Accordingly, they examine a combined data set which includes reorganizations and liquidations. The study by Kranzusch and Icks (2009) analyzes recovery rates for creditors in reorganizations (47%) and liquidations (5%).137 Within a further study, Icks and Kranzusch (2010) find an average recovery rate of 14% for five reorganization cases in Germany.138 Unfortunately, both studies by Kranzusch and Icks are limited by the data set. Firstly, they only have aggregated information on a firm level. Thus, they are not able to investigate group-specific recovery rates. The lack of group specific data might lead to a systematic bias. Secured creditors often receive payments as early as during the insolvency procedure. As the data set used in their study does not provide information about these payments, the average firm recovery rate might be underestimated.

134

See Blazy, Chopard and Nigam (2013), p. 1943.

135

See Blazy, Chopard and Nigam (2013), p. 1950.

136

See Blazy, Petey and Weill (2012), p. 24.

137

See Kranzusch and Icks (2009), p. 27 and 19.

138

See Icks and Kranzusch (2010), p.90.

40

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

Selected studies concerning the recovery rate of bank creditors are summarized in table 3. Average bank recovery rates range between 47% and 86%, with highest values for the US and the Netherlands. German studies find average recovery rates between 57% and 72%. However, a comparison of the results is difficult as many studies do not differentiate between reorganizations and liquidations as well as formal procedures and out-of court restructurings. This assumption is confirmed by Grunert and Weber (2009) for the German market. They find that bank recovery rates are significantly higher in case of continuation.139

Table 3: Overview of Empirical Studies (Bank Recovery Rate)140 Authors

Country

Period

Khieu/Mullineaux/Yi (2012) Altman/Jha (2003) Acharya/Bharath/Srinivasan (2007) Franks/Torous (1994) Schuermann (2004) Mora (2013) Franks/Sussman (2005) Davydenko/Franks (2008) Davydenko/Franks (2008) Couwenberg/de Jong (2008) Ingermann/Hesse/Bélorgey/Pfingsten (2013) Grunert/Weber (2009) Davydenko/Franks (2008)

US US US US US US UK UK France Netherland Germany Germany Germany

1987‐2007 1996‐2002 1982‐1999 1985‐1990 1970‐2003 1970‐2008 1997‐1998 1984‐2003 1984‐2003 1983‐2000 2006‐2011 1992‐2003 1984‐2003

Sample Size

Bank  Recovery  Rate in %

1364 262 358 37 151 655 130 645 460 139 131 120 198

84.14 69 81.12 86.4 63.1 67.1 74‐77 69 47 80 57 72.45 59

IV.2 Influencing Factors on Recovery Rates IV.2.a Firm Recovery Rates IV.2.a.1 Overview For Europe and the US, several studies investigate influencing factors on different recovery rates. Many studies solely calculate the recovery rates without analyzing influencing variables. Tables 2 and 3 mainly focus on more profound studies that 139

See Grunert and Weber (2009), p. 511.

140

Own illustration according to Grunert and Weber (2009), p. 507.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

41

typically use regression analysis to identify relevant factors. The dependent variables are the recovery rate of the firm, the recovery rate of secured or unsecured creditors as well as the recovery rate of bank creditors. According to this classification, chapter IV.2 will concentrate on identifying influencing factors on the extent of these recovery rates. Therefore, the influencing factors can be divided into three major groups: firm-specific characteristics, procedural characteristics and the macroeconomic surroundings. Accordingly, the first group includes variables concerning the economic situation of the firms at the time of the bankruptcy filing, while the second group of variables addresses the insolvency procedure itself. The third group of variables controls for the surrounding conditions such as the economic cycle of the country or industry. Table 4 summarizes potentially relevant factors for each of the three major groups. Table 4: Influencing Factors on Recovery Rates Firm Characteristics Firm Size Asset Structure Debt Structure Collateralization No. of Creditors (Secured/Unsecured/Bank) Over‐Indebtedness Liquidity Firm Age

Legal Form Location/Country/State Industry Reason for Default Equity owned by Management Operational Performance pre Filing Prior Default

Procedural Characteristics Petitioner Survival after Bankruptcy/Procedure Pre‐packaged Bankruptcy Continued Operations after Filing Time in Bankruptcy Asset Sales Macroeconomic Variables Economic Performance (National/Regional) Industry Distress Unemployment Rate Interbank Interest Rate 

Intervenion by Courts Share Distribution to Management Commitee Expenses for the Procedure Cash used in the Reorganization Receiver/Administrator/Judge

42

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

Furthermore, studies investigating bank recovery rates additionally incorporate factors reflecting the relationship between bank and customer or details of credit contracts. Therefore, bank recovery rates will be discussed separately in chapter IV.2.c. Before discussing the results of empirical studies dealing with firm recovery rates, a brief overview presents the aim of each included investigation. As mentioned earlier, the objective of all studies is to investigate ex-post efficiency within different insolvency legislations. Couwenberg and de Jong (2008) and Thorburn (2000) focus on questions regarding liquidation-based insolvency law in the Netherlands and Sweden respectively. They compare their results with existing American studies. Both data samples include piecemeal auctions as well as going concern auctions and thereby complement literature on auction based bankruptcies.141 In contrast, the study by Franks and Torous (1994) deals with corporations filing for bankruptcy under Chapter 11. They compare their results with firms using a distressed exchange to solve financial distress. Thus, Franks and Torous (1994) only shed light on continuations. Bris, Welch and Zhu (2006) compare Chapter 11 cases and Chapter 7 cases. Ivashina, Smith, Iverson and Strömberg (2011) also examine Chapter 11 cases and study the effect of creditor concentration on recovery rate. Sundgren (1998) investigates another aspect. He examines the effect of the amendment to insolvency law in Finland on recovery rates. Analogously, Armour, Hsu and Walters (2012) investigate recovery rates before and after a reform of UK bankruptcy legislation. Blazy, Chopard and Nigam (2013) test the effect of different insolvency rules on firm recovery rates in France and the UK while Blazy, Petey and Weill (2012) test significant differences between recovery rates in Germany, the UK and France. Blazy, Chopard, Fimayer and Guigou (2011) concentrate on influences affiliated with decisions and actions of a court. IV.2.a.2 Firm Characteristics One of the most analyzed firm-specific influencing factors on firm recovery rates is the size of the firm. Many authors use the amount of total assets or the number of employees as dependent variable. The underlying assumption is that even with intense monitoring, information asymmetries stay higher for large firms and that therefore difficulties in restructuring increase with firm size. Following this consideration, firm recovery rates should be lower for large firms. In empirical literature, the results are very diverse. For a sample of British firms Armour, Hsu and

141

See for example Bebchuk (1988), Hotchkiss and Mooradian (2003), Hansen and Thomas (1998) or Aghion, Hart and Moore (1992).

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

43

Walters (2012) confirm this hypothesis.142 However, the studies by Bris, Welch and Zhu (2006) for the U.S., Couwenberg and de Jong (2008) for the Netherlands and Thorburn (2000) for Sweden, which analyze a combined data sample of liquidations and reorganizations respectively, find no significant results.143 This result is in line with the studies of Sundgren (1998) for the Finnish and Blazy, Chopard, Fimayer and Guigou (2011) for the French market analyzing liquidations and reorganizations separately.144 Solely Blazy, Chopard and Nigam (2013), who concentrate on British and French firms, find a positive relationship between firm recovery rate and firm size.145 The authors themselves unfortunately do not discuss their contrary results. One possible explanation might be that the costs of restructurings increase under proportionally with firm size. Several studies examine variables which refer to the asset structure of a company. Couwenberg and de Jong (2008) find a positive relationship between the ratio of fixed assets to total assets and firm recovery rate.146 They assume that this ratio reflects the assets with substantial value that could be sold. Blazy, Chopard, Fimayer and Guigou (2011) confirm this result for their reorganization sample but reject it for their liquidation sample.147 Furthermore, they find that the quota of outstanding receivables has no significant effect on the extent of recovery rates. On the other hand, their reorganization sample shows a positive relationship between the share of inventory and the recovery rate of the firm. The study provides no further explanation for the differing results for the two subsamples. Blazy, Chopard and Nigam (2013) choose another proxy for asset structure: Cash assets to total assets. They find a significant and positive relation to the firm recovery rate.148 With the exception of Couwenberg and de Jong (2008), the studies provide no hypotheses for asset structure variables. The effects of highly firm- or industry-specific assets (like machinery, equipment, inventory, intangible assets and work in progress) are investigated for a Swedish sample of auctions by Thorburn (2000). She assumes that specific assets can be sold only at a substantial discount. According to this hypothesis, a high proportion

142

See Armour, Hsu and Walters (2012), p. 123.

143

See Bris, Welch and Zhu (2006), p. 1290, Couwenberg and de Jong (2008), p. 121, Thorburn (2000), p. 363.

144

See Sundgren (1998), p. 192 and Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

145

See Blazy, Chopard and Nigam (2013), p. 1945.

146

See Couwenberg and de Jong (2008), p. 121.

147

See Blazy, Chopard, Fimayer and Guigou (2011), p. 133.

148

See Blazy, Chopard and Nigam (2013), p. 1945.

44

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

of industry-specific assets should lead to lower recovery rates. However, Thorburn (2000) finds no significant evidence for this hypothesis.149 Different authors investigate the debt structure at the beginning of bankruptcy procedures. Ivashina, Smith, Iverson and Strömberg (2011) find that in case of a high concentration of debt on few creditors, measured by a Herfindahl-Hirschmann index, firm recovery rates decrease.150 One of the results of the study by Couwenberg and de Jong (2008) is that the proportion of bank debt to total debt has a significant positive impact on firm recovery rates.151 Blazy, Chopard, Fimayer and Guigou (2011) as well as Blazy, Petey and Weill (2012) examine claims with priority status. Accordingly, the latter show that the weight of senior creditors in total due amounts has a positive impact on firm recovery rates.152 The study by Blazy, Chopard, Fimayer and Guigou (2011) confirms the positive influence for the reorganization sample but neglects any influence for the liquidation sample.153 On the other hand, both studies investigate the impact of new money claims on recovery rates. Due to higher priority, Blazy, Chopard, Fimayer and Guigou (2011) find that the quota of new money claims has a positive impact on firm recovery rate. Again, this result is solely correct for their reorganization sample.154 The positive relationship between new money claims and recovery rate is confirmed by Blazy, Petey and Weill (2012).155 Among the firm-specific characteristics, collateralization is another frequentlyanalyzed factor. Literature argues that secured creditors approve credit only to the amount of the value of their collaterals. Accordingly, high collateralization, measured by secured debt to total debt, increases recovery rates of the firm. Bris, Welch and Zhu (2006), Thorburn (2000) and Blazy, Chopard and Nigam (2013) confirm the hypothesis while Blazy, Chopard, Fimayer and Guigou (2011) find no significant results.156 The number of secured and unsecured creditors is analyzed by Bris, Welch and Zhu (2006). The study reveals that the number of unsecured creditors has no im-

149

See Thorburn (2000), p. 363.

150

See Ivashina, Smith, Iverson and Strömberg (2011), p. 41.

151

See Couwenberg and de Jong (2008), p. 121.

152

See Blazy, Petey and Weill (2012), p. 31.

153

See Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

154

See Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

155

See Blazy, Petey and Weill (2012), p. 31.

156

See Bris, Welch and Zhu (2006), p. 1290, Thorburn (2000), p. 363, Blazy, Chopard and Nigam (2013), p. 1945 and Blazy, Chopard, Fimayer and Guigou (2011), p. 133.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

45

pact on recovery rates. However, there is a negative relationship between the number of secured creditors, which is interpreted as a measure for dispersion, and the firm recovery rate157 In addition Bris, Welch and Zhu (2006) try to filter the effects when banks are involved in secured or unsecured debt. Therefore, they incorporate dummy variables. However, the study shows no significant effects on recovery rates in this respect.158 Another important influencing factor investigated by several studies is overindebtedness at the beginning of bankruptcy proceedings. Usually, the assumption is that high indebtedness leads to lower recovery rates. Couwenberg and de Jong (2008), Bris, Welch and Zhu (2006) as well as Armour, Hsu and Walters (2012) analyze the ratio of total debt to total assets. Couwenberg and de Jong (2008) as well as Bris, Welch and Zhu (2006) find a negative relationship and thereby support the hypothesis.159 In contrast Armour, Hsu and Walters (2012) surprisingly state increasing recovery rates for a high level of indebtedness. However, their results are based on a low level of significance.160 Furthermore, the study does not provide a theory for a positive relationship. Blazy, Chopard, Fimayer and Guigou (2011), Sundgren (1998), Blazy, Chopard and Nigam (2013) as well as Blazy, Petey and Weill (2012) investigate the coverage rate (total assets/total debts). In analogy to the hypothesis for over-indebtedness, the relationship between coverage rate and recovery rate should be positive. This hypothesis is confirmed by all mentioned studies.161 The results are robust for all kinds of samples (reorganization samples, liquidation samples and combined data samples). Solely Ivashina, Smith, Iverson and Strömberg (2011) find no significant results.162 Liquidity of the firm is solely investigated by Couwenberg and de Jong (2008). They use the ratio current assets to current liabilities as a proxy for the potential of a firm to recover and assume that firms with high liquidity ratios have higher recovery rates. They empirically confirm their hypothesis.163 Studies investigating firm age do not find a significant influence. Neither the combined data sample investigated by Blazy, Chopard and Nigam (2013) nor the liq-

157

See Bris, Welch and Zhu (2006), p.1291.

158

See Bris, Welch and Zhu (2006), p.1291.

159

See Bris, Welch and Zhu (2006), p.1290, Couwenberg and de Jong (2008), p. 121.

160

See Armour, Hsu and Walters (2012), p. 123.

161

See Blazy, Chopard, Fimayer and Guigou (2011), p. 134, Sundgren (1998), p. 192 Blazy, Chopard and Nigam (2013), p. 1945 and Blazy, Petey and Weill (2012), p. 43.

162

See Ivashina, Smith, Iverson and Strömberg (2011), p. 41.

163

See Couwenberg and de Jong (2008), p. 121.

46

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

uidation based sample analyzed by Blazy, Chopard, Fimayer and Guigou (2011) provide evidence that there is an impact on firm recovery rates.164 Another factor that is frequently incorporated into investigations is the legal form of the company. For corporate entities, recovery rates should be lower due to limited liability of the shareholders. This hypothesis is confirmed by Balzy, Chopard and Nigam (2013), Blazy, Petey and Weill (2012) as well as Blazy, Chopard, Fimayer and Guigou (2011) for their sample of restructurings.165 Solely the liquidation based sample analyzed by Blazy, Chopard, Fimayer and Guigou (2011) reports higher recovery rates in case of a company with limited liability.166 Furthermore, Blazy, Petey and Weill (2012) explore country-specific differences between Germany, UK and France. Their results reveal that recovery rates in Germany are significantly higher.167 They explain these differences by the particular insolvency rules and procedures of each country. Thorburn (2000), Blazy, Chopard and Nigam (2013), Sundgren (1998), Franks and Torous (1994) and Blazy, Chopard, Fimayer and Guigou (2011) investigate whether recovery rates vary for different industries. However, none of the studies detects any significant effects.168 Blazy, Chopard and Nigam (2013), Blazy, Petey and Weill (2012) and Blazy, Chopard, Fimayer and Guigou (2011) focus on the reasons for insolvency. Blazy, Chopard and Nigam (2013) and Blazy, Petey and Weill (2012) incorporate dummy variables for different reasons like problems within the production, the management or the firms’ strategy. The former assume that production problems lead to lower recovery rates, while management mistakes have a positive impact on recovery rates.169 However, Blazy, Petey and Weill (2012) do not find a significant impact on recovery rates.170 Blazy, Chopard, Fimayer and Guigou (2011) confirm the negative impact of production problems.171 Instead of high recovery rates due to mistakes by the management they reveal external factors as a source for high

164

See Blazy, Chopard and Nigam (2013), p. 1945 and Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

165

See Blazy, Chopard and Nigam (2013), p. 1945 and Blazy, Petey and Weill (2012), p. 43.

166

See Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

167

See Blazy, Petey and Weill (2012), p. 43.

168

See Thorburn (2000), p. 363, Blazy, Chopard and Nigam (2013), p. 1945, Sundgren (1998), p. 192, Franks and Torous (1994), p. 361 and Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

169

See Blazy, Chopard and Nigam (2013), p. 1945.

170

See Blazy, Petey and Weill (2012), p. 43.

171

See Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

47

recovery rates. The results only hold for the reorganization sample of the study. The liquidation based sample provides no significant results. Other firm-specific factors are rarely analyzed. For example Bris, Welch and Zhu (2006) find no significant relationship between firm recovery rates and the proportion of shares held by management as well as a specific judicial district.172 Moreover, Franks and Torous (1994) find no significant impact on recovery rates by operative performance within the last two years before bankruptcy filing.173 The same finding is revealed by Thorburn (2000) who cannot find any relationship between recovery rate and cash flow performance.174 Table 5 and 6175 illustrate the findings mentioned in the previous chapter.

172

See Bris, Welch and Zhu (2006), p.1291 and 1292.

173

See Franks and Torous (1994), p. 361.

174

See Thorburn (2000), p. 363.

175

Based on Böttger, Guthoff, Heidorn (2008), p. 18 and supplemented by further studies.

Asset Structure Fixed assets

Firm Size

Influencing Factor Negative

Firm Recovery Rate

Blazy et al. (2011)

Blazy et al. (2012)

Thorburn (2000)

Blazy et al. (2011)

Blazy et al. (2011)

Bris et al. (2006)

Ivashina et al. (2011)

Bris et al. (2006), Couwenberg/de  Armour et al. (2012) Jong (2008), Thorburn (2000),  Sundgren (1998), Blazy et al.  (2011) 

Not Significant

Number and Type of Creditors No. secured creditors No. unsecured creditors Bris et al. (2006) Secured debt includes banks  Bris et al. (2006) (Dummy) Unsecured debt includes banks  Bris et al. (2006) (Dummy)

Collateral Secured debt to total debt

Cash assets Outstanding receivables Inventory Industry specific assets Debt Structure Firm  Creditor concentration Characteristics Bank debt Seniority

Category

Table 5: Firm Recovery Rates and Firm Characteristics (1)

Bris et al. (2006), Thorburn (2000)  and Blazy et al. (2013)

Couwenberg/de Jong (2008) Blazy et al. (2012),  Blazy/Petey/Weill (2012) 

Blazy et al. (2011)

Couwenberg/de Jong (2008),  Blazy et al. (2011) Blazy et al. (2013)

Blazy et al. (2013)

Positive

48 EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

Legal Form

Liquidity Firm Age

Coverage rate

Over‐indebtedness Debt over assets

Influencing Factor

External factors Equity Owned by Management Jurisdiction Performance pre Filing

Reason for Default Management Production

Firm  Country Characteristics Industry

Category

Blazy et al. (2011) Bris et al. (2006) Bris et al. (2006) Franks and Torous (1994),  Thorburn (2000) 

Thorburn (2000), Blazy et al.  (2013), Sundgren (1998),  Franks/Torous (1994), Blazy et al.  (2011) 

Balzy et al. (2013), Blazy et al.  (2011) 

Ivashina et al. (2011) 

Not Significant

Table 6: Firm Recovery Rates and Firm Characteristics (2)

Positive

Blazy et al. (2013), Blazy et al.  (2011) Blazy et al. (2011)

Blazy, Chopard and Nigam (2013)

Balzy et al. (2013),  Blazy et al. (2011) Blazy/Petey/Weill (2012), Blazy et  al. (2011) country specific

Couwenberg/de Jong (2008), Bris  Armour et al. (2012) et al. (2006) Blazy et al. (2011), Sundgren  (1998), Blazy et al. (2013),  Blazy/Petey/Weill (2012) Couwenberg/de Jong (2008)

Negative

Firm Recovery Rate

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

49

50

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

IV.2.a.3 Procedural Characteristics The second set of influencing factors involves variables concerning the insolvency procedure itself (see table 7). A first investigated variable concerns the petitioner who files for bankruptcy. The underlying idea is that a bankruptcy petition by the debtor leads to higher recovery rates than a petition by creditors. Thorburn (2000) concludes that a forced petition implies lower firm quality. Thorburn (2000) as well as Couwenberg and de Jong (2008) do not find significant results whereas Bris, Welch and Zhu (2006) confirm this relationship.176 The data samples of the studies by Couwenberg and de Jong (2008), Thorburn (2000) and Armour, Hsu and Walters (2012) compare recovery rates of liquidations and continuations by incorporating a dummy variable. In this respect, they assume that recovery rates are lower for liquidations. Couwenberg and de Jong (2008) and Thorburn (2000) confirm this idea while Armour, Hsu and Walters (2012) find no significant evidence.177 A similar variable is used by Bris, Welch and Zhu (2006). In their paper they investigate differences between Chapter 11 (reorganization) and Chapter 7 (liquidation) of the U.S. Bankruptcy Code. They find higher recovery rates for reorganizations under Chapter 11.178 Furthermore Sundgren (1998) and Armour, Hsu and Walters (2012) assume that a revision of insolvency statute can result in higher recovery rates.179 They confirm this argument when legislation on continuation changes, like in Finland or the UK respectively. However, the positive effect of reorganization on the extent of recovery rates is limited due to the influence on healthy firms. Strengthening continuation and thereby ex-post efficiency leads to interdependencies, for example with corporate finance decisions of creditors and thereby ex-ante efficiency (see chapter III.3). Furthermore Thorburn (2000) examines pre-packaged180 sales of the firm as a going concern. However she doesn’t find a significant impact on recovery rates.181 Another variable concerns the continuation of business operations directly after filing for bankruptcy. Couwenberg and de Jong (2008) and Armour, Hsu and Wal-

176

See Thorburn (2000), p. 363, Couwenberg and de Jong (2008), p. 120, Bris, Welch and Zhu (2006), p.1290.

177

See Couwenberg and de Jong (2008), p. 120, Thorburn (2000), p. 363 and Armour, Hsu and Walters (2012), p. 121.

178

See Bris, Welch and Zhu (2006), p.1290.

179

See Sundgren (1998), p. 192 and Armour, Hsu and Walters (2012), p. 122.

180

A pre-packed bankruptcy is a special procedure in which a sale or insolvency plan is agreed before the filing for bankruptcy.

181

See Thorburn (2000), p. 363.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

51

ters (2012) find that firms which continue after filing have higher recovery rates than firms which immediately cease business operations.182 As long as insolvency proceedings are ongoing, the management of the firm is governed by various restrictions that limit its capacity to act. Consequently, indirect insolvency costs increase over time. Long-term proceedings should therefore lower recovery rates. Blazy, Chopard and Nigam (2013), Couwenberg and de Jong (2008), Bris, Welch and Zhu (2006), Blazy, Petey and Weill (2012) and Armour, Hsu and Walters (2012) investigate the effect of the duration of bankruptcy proceedings.183 Collectively they choose the number of days or months between filing and closing of the insolvency proceedings as an independent variable. In contrast to all other studies Armour, Hsu and Walters (2012) detect a positive correlation between time in bankruptcy and recovery rate.184 In parts of their study, they confirm this relationship without giving any further explanation. All other studies mentioned above do not find significant results. Furthermore, several studies investigate how asset sales during the procedure affect recovery rates. Following the results of Shleifer und Vishny (1992), Franks and Torous (1994) examine the proportion of average annual asset sales to firm value after restructuring.185 They argue that assets are often sold below their fair value in times of distress. Thus, asset sales should lead to lower recovery rates. In addition Couwenberg and de Jong (2008) analyze the timing of asset sales, the period over which asset sales occur and the number of potential buyers. Their analysis of an auction-based sample doesn’t reveal significant findings.186 Thorburn (2000) also analyzes auctions and finds that in 54% of cases, the main shareholders repurchase the assets to continue business.187 She assumes that main shareholders possess private information and only would repurchase the assets of viable firms. The same argument can be applied to relationship banks that finance repurchases. In the empirical analysis, Thorburn (2000) does not find higher recovery rates for repurchases by main shareholders.188 However, she finds empirical evidence that relationship banks that finance a repurchase do have a positive effect on recovery rates. 182

See Couwenberg and de Jong (2008), p. 121, Armour, Hsu and Walters (2012), p. 123.

183

See Blazy, Chopard and Nigam (2013), p. 1945, Couwenberg and de Jong (2008), p. 121, Bris, Welch and Zhu (2006), p. 1291, Blazy, Petey and Weill (2012), p. 43 and Armour, Hsu and Walters (2012), p. 123.

184

See Armour, Hsu and Walters (2012), p. 123.

185

See Franks and Torous (1994), p. 361.

186

See Couwenberg and de Jong (2008), p. 120.

187

See Thorburn (2000), p. 363.

188

See Thorburn (2000), p. 363.

52

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

Blazy, Chopard, Fimayer and Guigou (2011) mainly focus on interventions by courts during insolvency procedures in France. As they find that French courts tend to force continuation to preserve employment they question whether costs that reduce recovery rates are incurred. However, they do not find any significant results for this aspect.189 Several other factors are investigated in addition. Bris, Welch and Zhu (2006) investigate the impact of a committee for unsecured creditors as well as the extent of expense ratio on firm recovery rate.190 Franks and Torous (1994) assume that the proportion of cash repaid to creditors is positively correlated with recovery rates.191 None of the mentioned studies reveals significant results.

189

See Blazy, Chopard, Fimayer and Guigou (2011), p. 134.

190

See Bris, Welch and Zhu (2006), p.1291 and 1292.

191

See Franks and Torous (1994), p. 361.

Procedural  Charakteristics

Category

Asset Sales Asset sales to firm value Timing and duration of asset  sales Type of buyer Financing by the old bank Intervention by Courts Committee Expense Ratio Cash Used in the Reorganization

Prepack Time in Bankruptcy

Continued Operations

Reorganization vs. Liquidation  (Liquidation=1)

Petitioner (Debtor=1)

Influencing Factor

Blazy et al. (2011) Bris et al. (2006)  Bris et al. (2006)  Franks/Torous (1994)

Thorburn (2000)

Couwenberg/de Jong (2008)

Thorburn (2000) Balzy et al. (2013),  Couwenberg/de Jong (2008), Bris  et al. (2006), Blazy/Petey/Weill  (2012) 

Couwenberg/de Jong (2008),  Thorburn (2000)  Armour et al. (2012)

Not Significant

Firm Recovery Rate

Franks/Torous (1994)

Couwenberg/de Jong (2008),  Thorburn (2000), Bris et al. (2006),  Sundgren (1998), Armour et al.  (2012)

Negative

Table 7: Firm Recovery Rates and Procedural Characteristics

Thorburn (2000)

Armour et al. (2012) 

Couwenberg/de Jong (2008),  Armour et al. (2012) 

Bris et al. (2006)

Positive

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

53

54

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

IV.2.a.4 Macroeconomic Variables The third set of influencing factors concerns the macroeconomic surroundings of insolvencies (see table 8). Thorburn (2000) investigates whether filings for bankruptcy in Sweden during the crisis year 1991 have an effect on recovery rates. As a result, the year dummy indicates that recovery rates are significantly lower.192 She also examines industry distress. She argues that if a large proportion of firms within an industry are in distress, industry specific assets can only be sold at lower prices and recovery rates decline. However, the empirical analysis shows that industry distress has no impact on the extent of recovery rates. Franks and Torous (1994) examine the impact of economic performance in general by incorporating stock market returns prior to insolvency. To measure market returns, they use the performance of the equally weighted CRSP193 index. Finally, they find that the performance of the index is positively related to firm recovery rates.194 Blazy, Chopard and Nigam (2013) as well as Ivashina, Smith, Iverson and Strömberg (2011) measure macroeconomic performance by means of GDP growth rates. However, the annual rate of change does not have an impact on the proceeds for creditors.195

192

See Thorburn (2000), p. 363.

193

Center for Research in Security Prices (CRSP).

194

See Franks and Torous (1994), p. 361.

195

See Blazy, Chopard and Nigam (2013), p. 1945 and Ivashina, Smith, Iverson and Strömberg (2011), p. 41.

Macroeconomic  Variables

Category

Industry Distress

Economic Performance

Influencing Factor Negative

Firm Recovery Rate

Blazy et al. (2013), Ivashina et al.  Thorburn (2000), Franks/Torous  (2011) (1994) Thorburn (2000) 

Not Significant

Table 8: Firm Recovery Rates and Macroeconomic Variables

Positive

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

55

56

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

IV.2.b Secured and Unsecured Recovery Rates Some of the studies presented in the previous chapter also shed light on the recovery rates of different groups of creditors and analyze the particular influencing factors.196 The study by Citron, Wright, Ball and Rippington (2003) solely investigates secured recovery rates from management buy-outs in the UK. The role of creditors´ committees is analyzed by Harner and Marincic (2011). One part of their study investigates the impact of different categories of committees on unsecured recovery rates. A more descriptive analysis by Baird, Bris and Zhu (2007) concentrates on Chapter 11 cases and how the size of the firm affects unsecured recovery rates. Overall, it becomes obvious that the number of studies analyzing secured and unsecured recovery rates separately is limited. As a result, the investigations examine the same influencing factors on secured and unsecured recovery rates as on firm recovery rates. The empirical analyses are often based on the same hypothesizes as described in chapter IV.2.a. In the following, solely new theories and arguments will be described. Analogous to chapter IV.2.a, the results concerning firm-specific variables are presented in a first step (see table 9 and 10). Studies about the relationship between the size of the firm and secured or unsecured recovery rates come to different conclusions. While Bris, Welch and Zhu (2006) and Thorburn (2000) neglect any relationship between firm size and secured recovery rates, Blazy, Petey and Weill (2012) and Citron, Wright, Ball and Rippington (2003) do find significant results.197 However, Blazy, Petey and Weill (2012) report a positive and Citron, Wright, Ball and Rippington (2003) reveal a negative relation on secured recovery rate.198 One major difference between those studies that might cause the contradictory findings is that Bris, Welch and Zhu (2006) analyze solely Chapter 11 reorganizations, while all other studies use combined data sets. Concerning unsecured recovery rates, the results are diverse too. Bris, Welch and Zhu (2006) again do not find a significant influence, while Blazy, Petey and Weill (2012) as well as Harner and Marincic (2011) confirm a positive relationship.199 Baird, Bris and Zhu (2007) support this finding for a sample of Chapter 11 bankruptcies.200 Within

196

These studies are: Bris, Welch and Zhu (2006), Thorburn (2000), Blazy, Petey and Weill (2012), Ivashina, Smith, Iverson and Strömberg (2011).

197

See Bris, Welch and Zhu (2006), p.1293 and Thorburn (2000), p. 363.

198

See Blazy, Petey and Weill (2012), p. 45 and Citron, Wright, Ball and Rippington (2003), p. 156.

199

See Bris, Welch and Zhu (2006), p.1294, Blazy, Petey and Weill (2012), p. 45 and Harner and Marincic (2011), p. 787 et seqq.

200

See Baird, Bris and Zhu (2007), p. 24.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

57

their descriptive study, they conclude that Chapter 11 does not sufficiently protect unsecured creditors of small firms. Asset structure variables are only rarely investigated. Solely Thorburn (2000) analyzes the effect of industry-specific assets on secured recovery rates. In analogy to the results for firm recovery rates, Thorburn finds no significant relationship for her Swedish data sample.201 The influence of asset structure variables on unsecured recovery rates has not been investigated yet. The only influencing factor concerning debt structure is analyzed by Ivashina, Smith, Iverson and Strömberg (2011). They examine debt ownership concentration and its impact on secured and unsecured recovery rate. For secured and unsecured recovery rates, they find a significant and positive influence of debt ownership concentration.202 With regard to the relationship between recovery rates and collateralization solely Bris, Welch and Zhu (2006) use the proportion of secured debt to total debt as independent variable in their regression analysis. In contrast to the results for firm recovery rates above, they do not confirm an impact on secured or unsecured recovery rates.203 The number of secured creditors is analyzed by Citron, Wright, Ball and Rippington (2003) as well as Bris, Welch and Zhu (2006). Citron, Wright, Ball and Rippington (2003) reject an impact of the number of secured creditors on the secured recovery rate. In contrast, Bris, Welch and Zhu (2006) find that the number of secured creditors has a negative influence on the extent of secured and unsecured recovery rates.204 For the relationship between the number of unsecured creditors and secured or unsecured recovery rates, the study does not provide significant results.205 Furthermore Bris, Welch and Zhu (2006) find that unsecured recovery rates increase when there is unsecured debt financed by banks.206 Furthermore, there is no distinct relationship between recovery rates and overindebtedness. Ivashina, Smith, Iverson and Strömberg (2011) find that the coverage rate has no significant impact on secured recovery rates.207 Bris, Welch and Zhu (2006) find that over-indebtedness measured by a dummy (debt/assets>100%) 201

See Thorburn (2000), p. 363.

202

See Ivashina, Smith, Iverson and Strömberg (2011), p. 41.

203

See Bris, Welch and Zhu (2006), p. 1293 and 1294.

204

See Citron, Wright, Ball and Rippington (2003), p. 156 and Bris, Welch and Zhu (2006), p. 1293 and 1294.

205

See Bris, Welch and Zhu (2006), p. 1293 and 1294.

206

See Bris, Welch and Zhu (2006), p. 1294.

207

See Ivashina, Smith, Iverson and Strömberg (2011), p. 41.

58

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

leads to lower secured recovery rates.208 Both studies reveal the same results for unsecured recovery rates. Blazy, Petey and Weill (2012) analyze firm age and legal form and find no significant impact on unsecured recovery rates.209 However, they confirm that older firms have higher secured recovery rates. Secured creditors of corporate enterprises have lower recovery rates. Citron, Wright, Ball and Rippington (2003) find no significant relationship between firm age and repayments to secured creditors.210 A further factor concerns the location of filing within the United States to analyze state specific differences. Bris, Welch and Zhu (2006) confirm that secured recovery rates are significantly lower in Arizona than in New York.211 However no impact is found according to unsecured recovery rates. Furthermore, Thorburn (2000) and Citron, Wright, Ball and Rippington (2003) investigate whether secured recovery rates vary over different industries.212 Neither study finds any significant impact. Unsecured recovery rates have not yet been investigated in this respect yet. Analyzing the reason for default, Citron, Wright, Ball and Rippington (2003) and Blazy, Petey and Weill (2012) find consistent results. Both studies state that secured recovery rates are higher when the bankruptcy is caused by mismanagement.213 Blazy, Petey and Weill (2012) also reveal that secured recovery rates increase when bankruptcy mainly arises from financial reasons. Furthermore, they conclude that two other factors are relevant to unsecured recovery rates: Mistakes referring to the sales quantity and macroeconomic developments. Both reasons lead to significant lower recovery rates for unsecured creditors. Accordingly, the reason for bankruptcy allows statements about payments to secured and unsecured creditors.

208

See Bris, Welch and Zhu (2006), p. 1293.

209

See Blazy, Petey and Weill (2012), p. 45.

210

See Citron, Wright, Ball and Rippington (2003), p. 156.

211

See Bris, Welch and Zhu (2006), p. 1292.

212

See Thorburn (2000), p. 363 and Citron, Wright, Ball and Rippington (2003), p. 156.

213

See Citron, Wright, Ball and Rippington (2003), p. 156 and Blazy, Petey and Weill (2012), p. 45.

Firm Size

Influencing Factor

Bris et al. (2006),  Thorburn (2000)

Not Significant

Asset Structure Industry specific assets Thorburn (2000)  Debt Structure Creditor concentration Collateral Secured debt to total debt Bris et al. (2006) Number and Type of Creditors Firm  No. secured creditors Citron et al. (2003)  Characteristics No. unsecured creditors Bris et al. (2006)  Secured debt includes  Bris et al. (2006)  banks (Dummy) Unsecured debt includes  Bris et al. (2006)  banks (Dummy) Over‐Indebtedness Debt over assets Coverage rate Ivashina et al. (2011) Firm Age Citron et al. (2003) 

Category

Bris et al. (2006) 

Bris et al. (2006)

Citron et al. (2003) 

Negative

Secured Recovery Rate

Blazy/Petey/ Weill  (2012)

Ivashina et al. (2011)

Blazy/Petey/ Weill  (2012) 

Positive

Ivashina et al. (2011) Blazy/Petey/ Weill  (2012)

Bris et al. (2006)  Bris et al. (2006) 

Bris et al. (2006)

Bris et al. (2006) 

Not Significant

Table 9: Secured and Unsecured Recovery Rates - Firm Characteristics (1)

Bris et al. (2006) 

Bris et al. (2006)

Negative

Bris et al. (2006) 

Ivashina et al. (2011)

Blazy/Petey/ Weill  (2012), Harner/Marincic  (2011), Baird et al.  (2007)

Positive

Unsecured Recovery Rate

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

59

Firm  Characteristics

Category

Jurisdiction

Sales quantity

Corporate finance

External factors

Reason for Default Management

Industry

Legal Form

Influencing Factor

Thorburn (2000),  Citron et al. (2003) 

Not Significant

Bris et al. (2006) 

Blazy/Petey/ Weill  (2012)

Negative

Secured Recovery Rate

Blazy/Petey/ Weill  (2012)

Citron et al. (2003),  Blazy/Petey/ Weill  (2012)

Positive

Bris et al. (2006) 

Blazy/Petey/ Weill  (2012)

Not Significant

Table 10: Secured and Unsecured Recovery Rates - Firm Characteristics (2)

Blazy/Petey/ Weill  (2012)

Blazy/Petey/ Weill  (2012)

Negative

Positive

Unsecured Recovery Rate

60 EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

61

As mentioned earlier, a second set of variables which affect recovery rates concerns the insolvency procedure (see table 11). Several factors, like a forced petition by creditors (Thorburn (2000) as well as Bris, Welch and Zhu (2006)), a prepackaged bankruptcy (Thorburn (2000)), the duration of the procedure (Blazy, Petey and Weill (2012)) and Bris, Welch and Zhu (2006)), the existence of a committee for unsecured creditors (Bris, Welch and Zhu (2006)), the expense ratio (Bris, Welch and Zhu (2006)) as well as the type of receiver (Citron, Wright, Ball and Rippington (2003)) do not provide significant results for secured recovery rates.214 However, Thorburn (2000) and Citron, Wright, Ball and Rippington (2003) confirm that secured recovery rates are significantly lower for liquidations than for reorganizations.215 Analyzing auction based reorganizations, Thorburn (2000) finds a significant impact on secured recovery rates when the initial owner repurchases the firm or when the main bank finances the buyer of the firm. 216 Studies investigating procedural characteristics and the impact on unsecured recovery rates are rare. Bris, Welch and Zhu (2006) and Blazy, Petey and Weill (2012) do not find an effect of the time of the bankruptcy proceedings.217 Bris, Welch and Zhu (2006) find that the existence of a committee of unsecured creditors has no impact on unsecured recovery rates.218 However, Harner and Marincic (2011) contradict this result.219 They conclude that unsecured recovery rates increase when at least one committee exists. In contrast to their results mentioned for secured recovery rates, Bris, Welch and Zhu (2006) confirm that the petitioner and the expense ratio have a significant impact on unsecured recovery rates.220 Repayments to unsecured creditors are significantly lower when the bankruptcy is filed by creditors. The relationship between debtor expenses to pre-assets and unsecured recovery rates is significantly positive in their analysis. Furthermore Bris, Welch and Zhu (2006) confirm the thesis that Chapter 11 proceedings lead to higher recovery rates than Chapter 7 cases.221

214

See Thorburn (2000), p. 363, Bris, Welch and Zhu (2006), p. 1293, Blazy, Petey and Weill (2012), p. 45 and Citron, Wright, Ball and Rippington (2003), p. 156.

215

See Thorburn (2000), p. 363 and Citron, Wright, Ball and Rippington (2003), p. 156.

216

See Thorburn (2000), p. 363.

217

See Bris, Welch and Zhu (2006), p. 1294 and Blazy, Petey and Weill (2012), p. 45.

218

See Bris, Welch and Zhu (2006), p. 1294.

219

See Harner and Marincic (2011), p. 787 et seqq.

220

See Bris, Welch and Zhu (2006), p. 1294.

221

See Bris, Welch and Zhu (2006), p. 1297.

Procedural  Charakteristics

Category

Thorburn (2000),  Citron et al. (2003)  Blazy/Petey/Weill  (2012), Bris et al.  (2006)

Thorburn (2000), Bris  et al. (2006)

Not Significant

Expense Ratio Receiver/Administrator/  Judge

Bris et al. (2006) Citron et al. (2003)

Asset Sales Type of buyer Financing by the old bank Committee Bris et al. (2006)

Time in Bankruptcy

Survival after  Bankruptcy/Procedure  (Liquidation=1) Prepack

Petitioner (debtor=1)

Influencing Factor

Thorburn (2000),  Citron et al. (2003) 

Negative

Secured Recovery Rate

Thorburn (2000) Thorburn (2000)

Positive

Bris et al. (2006)

Blazy/Petey/Weill  (2012), Bris et al.  (2006)

Not Significant

Table 11: Secured and Unsecured Recovery Rates - Procedural Characteristics

Bris et al. (2006) 

Negative

Harner/Marincic  (2011)  Bris et al. (2006)

Bris et al. (2006) 

Positive

Unsecured Recovery Rate

62 EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

63

Referring to macroeconomic conditions (see table 12) Thorburn (2000) provides significant results. Analogous to her analysis of firm recovery rates, she also investigates how secured recovery rates are affected by general economic crises.222 Accordingly, she confirms the hypothesis that filing in the crisis year 1991 leads to lower secured recovery rates. However, the studies by Citron, Wright, Ball and Rippington (2003), Blazy, Petey and Weill (2012) and Ivashina, Smith, Iverson and Strömberg (2011) do not confirm any impact of economic performance on secured and unsecured recovery rates.223 Furthermore, Thorburn (2000) examines distress within a specific industry and its impact on secured recovery rates. The results for this investigation are not significant.224

222

See Thorburn (2000), p. 363.

223

See Citron, Wright, Ball and Rippington (2003), p. 156, Blazy, Petey and Weill (2012), p. 45 and Ivashina, Smith, Iverson and Strömberg (2011), p. 41.

224

See Thorburn (2000), p. 363.

Macroeconomic  Variables

Category

Industry Distress

Economic Performance

Influencing Factor Negative

Citron et al. (2003),  Thorburn (2000)  Blazy/Petey/ Weill  (2012), Ivashina et al.  (2011)  Thorburn (2000) 

Not Significant

Secured Recovery Rate Positive Blazy/Petey/ Weill  (2012), Ivashina et al.  (2011) 

Not Significant

Table 12: Secured and Unsecured Recovery Rates - Macroeconomic Variables

Negative

Positive

Unsecured Recovery Rate

64 EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

65

IV.2.c Bank Recovery Rates For banks, a precise prediction of recovery rates is crucial since they are incorporated into credit risk modeling, the calculation of equity requirements or risk adjusted interest rates. To reach this goal, banks have to investigate which influencing factors affect the extent of bank recovery rates (or loss given default). Since this research question is so crucial to banks, literature provides a multitude of studies covering this topic. This section will provide an overview of relevant empirical studies. These studies examine the same groups of influencing factors which are mentioned in chapter IV.2.a. Besides firm characteristics, procedural characteristics and macroeconomic variables, the studies about bank recovery rates examine variables concerning the terms of credit and variables affiliated with the business connection between bank and borrower. The reason why studies about bank recovery rates frequently have information about these two groups of variables is that data samples are directly provided by banks. Among the group of firm characteristics, firm size is frequently analyzed (see table 13). Referring to the amount of total assets Grunert and Weber (2009), Hurt and Felsovalyi (1998), Eales and Bosworth (1998) as well as Couwenberg and de Jong (2008) find evidence that bank recovery rates decrease when the absolute amount of total assets is high.225 However, Bartlett (2000), Franks, Servigny and Davydenko (2004) and Davydenko and Franks (2008) contradict any relationship.226 Asset and debt structure variables and their effect on the extent of bank recovery rates are rarely investigated. Solely Couwenberg and de Jong (2008) analyze the association between bank recovery rates and the proportion of fixed assets to total assets as well as bank debt to total debt respectively.227 However, both variables provide no significant results. A more detailed investigation is required. Once again, collateralization is studied in particular. Davydenko and Franks (2008), Grunert and Weber (2009) as well as Ingermann, Hesse, Bélorgey and Pfingsten (2013) state that the association between the proportion of collateral to the exposure at default and bank recovery rates is significant and positive.228

225

See Grunert and Weber (2009), p. 510, Hurt and Felsovalyi (1998), p. 6, Eales and Bosworth (1998), p. 59 and Couwenberg and de Jong (2008), p. 122.

226

See Bartlett (2000), p. 5, Franks, Servigny and Davydenko (2004), p. 93 and Davydenko and Franks (2008), p. 594.

227

See Couwenberg and de Jong (2008), p. 123.

228

See Davydenko and Franks (2008), p. 594, Grunert and Weber (2009), p. 510 and Ingermann, Hesse, Bélorgey and Pfingsten (2013), p. 32.

Firm  Characteristics

Category

Prior Default

Manufactoring

Firm Age Legal Form Country Industry Trading

Secured bank debt to bank  debt Type of security Over‐Indebtedness Debt over assets

Asset Structure Fixed assets Debt Structure Bank debt Collateral Collateral to EAD

Firm Size

Influencing Factor Negative

Bank Recovery Rate

Grunert/Weber (2009) 

Couwenberg/de Jong (2008)

Couwenberg/de Jong (2008)

Dermine/Neto de Carvalho  (2006), Davydenko/Franks (2008) Dermine/Neto de Carvalho (2006) 

country specific

Couwenberg/de Jong (2008),  Khieu et al. (2012) 

 type and country specific

Bartlett (2000), Franks et al.  Grunert/Weber (2009),  (2004), Davydenko/Franks (2008) Hurt/Felsovalyi (1998),  Eales/Bosworth (1998),  Couwenberg/de Jong (2008) 

Not Significant

Table 13: Bank Recovery Rates and Firm Characteristics

Khieu et al. (2012)

Dermine/Neto de Carvalho (2006) 

Davydenko/Franks (2008),  Grunert/Weber (2009),  Ingermann et al. (2013)  Couwenberg/de Jong (2008)

Positive

66 EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

67

Furthermore, several descriptive studies find a positive relation between collateralization and repayment to bank creditors.229 Another observed variable is the type of security. Khieu, Mullineaux and Yi (2012) detect that collaterals like inventory or/and accounts receivable lead to higher bank recovery rates.230 According to the study by Davydenko and Franks (2008), there is also an impact of the proportion of accounts receivable or real estate to EAD and bank recovery rates.231 In France, a high level of accounts receivable leads to higher bank recovery rates while in Germany and the UK, real estate securities are responsible for higher bank recovery rates. Furthermore, Couwenberg and de Jong (2008) examine the influence of secured banks and discover that a high level of secured bank debt to bank debt has a significant and positive effect on bank recovery rates.232 Unambiguous results can be found with respect to the level of over-indebtedness (debt over assets). Couwenberg and de Jong (2008) as well as Khieu, Mullineaux and Yi (2012) conclude that there is a significant and negative relation to bank recovery rates.233 Studies about firm age are rare. Dermine and Neto de Carvalho (2006) find that bank recovery rates are higher for old firms.234 Grunert and Weber (2009) investigate the relation of legal form and recovery rates. However their results do not show any significant relationship.235 Country-specific differences are analyzed by Davydenko and Franks (2008). They use a data sample including French, German and UK loans and conclude that in France there is a negative impact on bank recovery rates.236 Industry classification and the impact on recovery rates are analyzed by Araten, Jacobs Jr. and Varshney (2004), Dermine and Neto de Carvalho (2006), Davydenko and Franks (2008) and Gupton, Gates and Carty (2000). Significant negative effects are found by Dermine and Neto de Carvalho (2006) and Davydenko and Franks (2008) for trading industry and by Dermine and Neto de Carvalho (2006) 229

See Carty, Hamilton and Moss (1998), pp. 24-25, Gupton, Gates and Carty (2000), p. 80, Araten, Jacobs Jr. and Varshney (2004), p. 32, Franks, Servigny and Davydenko (2004), Dermine and Neto de Carvalho (2006), Van de Castle, Keisman and Yang (2000), p. 63, Khieu, Mullineaux and Yi (2012).

230

See Khieu, Mullineaux and Yi (2012), p. 928.

231

See Davydenko and Franks (2008), p. 594.

232

See Couwenberg and de Jong (2008), p. 123.

233

See Couwenberg and de Jong (2008), p. 123 and Khieu, Mullineaux and Yi (2012), p. 930.

234

See Dermine and Neto de Carvalho (2006), p. 1238.

235

See Grunert and Weber (2009), p. 511.

236

See Davydenko and Franks (2008), p. 593.

68

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

for manufacturing.237 Other studies like the one by Araten, Jacobs Jr. and Varshney (2004) and Gupton, Gates and Carty (2000) conclude that there is no significant relationship between industries and recovery rates.238 The last influencing factor among the group of firm characteristics concerns former insolvencies of a corporation. Khieu, Mullineaux and Yi (2012) analyze insolvent firms which previously have been bankrupt at least once. They provide evidence that a positive relationship on bank recovery rates exists.239 They conclude that a firm which has been in financial distress before is strictly monitored by creditors. Furthermore, agreements concerning bank lending are more restrictive. Procedural characteristics are rarely incorporated into studies about bank recovery rates (see table 14). Nevertheless Couwenberg and de Jong (2008) find out that a bankruptcy petition by the debtor leads to higher bank recovery rates.240 Grunert and Weber (2009) also discover a positive relationship between bank recovery rate and a continuation dummy indicating firms which continue business activities after filing.241 Furthermore, Couwenberg and de Jong (2008) find support for the hypothesis that liquidation in contrast to reorganization (survival after the procedure) leads to lower recovery rates.242 Khieu, Mullineaux and Yi (2012) provide evidence that bank creditors within a pre-packaged bankruptcy are better off than creditors of any other procedure.243 Couwenberg and de Jong (2008) do not verify an influence of the time in bankruptcy on bank recovery rates, while Ingermann, Hesse, Bélorgey and Pfingsten (2013) find a positive relationship and Khieu, Mullineaux and Yi (2012) conclude that long time in bankruptcy leads to lower recovery rates.244 Furthermore, Couwenberg and de Jong (2008) examine the number of potential buyers of the firm. Their results indicate that there is no significant relationship on the extent of bank recovery rates.245 237

See Dermine and Neto de Carvalho (2006), p. 1237 and Davydenko and Franks (2008), p. 595.

238

See Araten, Jacobs Jr. and Varshney (2004), pp. 28-29 and Gupton, Gates and Carty (2000), p. 78.

239

See Khieu, Mullineaux and Yi (2012), p. 929.

240

See Couwenberg and de Jong (2008), p. 123.

241

See Grunert and Weber (2009), p. 511.

242

See Couwenberg and de Jong (2008), p. 123.

243

See Khieu, Mullineaux and Yi (2012), p. 929.

244

See Couwenberg and de Jong (2008), p. 123, Ingermann, Hesse, Bélorgey and Pfingsten (2013), p. 16 and Khieu, Mullineaux and Yi (2012), p. 929.

245

See Couwenberg and de Jong (2008), p. 122.

Influencing Factor

Petitioner (debtor=1) Survival after  Bankruptcy/Procedure  (Liquidation=1) Procedural  Continued Operations Charakteristics Prepack Time in Bankruptcy Asset Sales No. of potential buyer

Category

Couwenberg/de Jong (2008)

Couwenberg/de Jong (2008)

Not Significant

Bank Recovery Rate

Khieu et al. (2012)

Couwenberg/de Jong (2008),  Ingermann et al. (2013)

Negative

Table 14: Bank Recovery Rates and Procedural Characteristics

Grunert/Weber (2009)  Khieu et al. (2012) Ingermann et al. (2013) 

Couwenberg/de Jong (2008)

Positive

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

69

70

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

In contrast to studies about firm recovery rates as well as secured and unsecured recovery rates, the group of macroeconomic variables plays an important role in studies about bank recovery rates (see table 15). However the influence of the GDP on recovery rates is not clear since literature provides ambiguous results. While Grunert and Weber (2009), Davydenko and Franks (2008) and Dermine and Neto de Carvalho (2006) find no significant relation at all,246 Khieu, Mullineaux and Yi (2012) provide evidence that there is a positive relationship, whereas Ingermann, Hesse, Bélorgey and Pfingsten (2013) report negative effects.247 Furthermore, Grunert and Weber (2009) examine the regional growth of GDP, loan loss provision and unemployment rate.248 Nevertheless, none of the mentioned variables show significant results. However Ingermann, Hesse, Bélorgey and Pfingsten (2013) discover a significant and positive relationship between unemployment rate and the bank recovery rate.249 The impact of the interbank interest rate is analyzed by Dermine and Neto de Carvalho (2006). However, the study finds no significant influence.250

246

See Grunert and Weber (2009), p. 511, Davydenko and Franks (2008), p. 594 and Dermine and Neto de Carvalho (2006), p. 1238.

247

See Khieu, Mullineaux and Yi (2012), p. 929 and Ingermann, Hesse, Bélorgey and Pfingsten (2013), p. 17.

248

See Grunert and Weber (2009), p. 511.

249

See Ingermann, Hesse, Bélorgey and Pfingsten (2013), p. 17.

250

See Dermine and Neto de Carvalho (2006), p. 1238.

Economic Performance

Influencing Factor

Macroeconomic  Variables Regional Economic Performance Interbank Interest Rate Unemployment Rate

Category Negative

Bank Recovery Rate

Grundert/Weber (2009),  Ingermann et al. (2013)  Davydenko/Franks (2008),  Dermine/Neto de Carvalho (2006)  Grunert/Weber (2009) Dermine/Neto de Carvalho (2006)  Grunert/Weber (2009)

Not Significant

Table 15: Bank Recovery Rates and Macroeconomic Variables

Ingermann et al. (2013) 

Khieu et al. (2012) 

Positive

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

71

72

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

A new group of variables, which is rarely analyzed by studies about firm recovery rates, refers to the terms of credit (see table 16). Potential influencing factors may be the amount and duration of credit, loan syndication or creditworthiness. Grunert and Weber (2009) as well as Khieu, Mullineaux and Yi (2012) neglect any relationship between the amount of credit and the repayments to bank creditors.251 However, Hurt and Felsovalyi (1998), Dermine and Neto de Carvalho (2006) and Ingermann, Hesse, Bélorgey and Pfingsten (2013) provide evidence that higher amounts of credit result in lower bank recovery rates.252 Furthermore, Khieu, Mullineaux and Yi (2012) investigate whether recovery rates are lower for long term loans. They confirm this relationship and declare that banks are able to react quickly in case of a short term loan.253 Grunert and Weber (2009) also analyze the risk premium as a proxy for creditworthiness and find a significant negative relationship.254 The study by Gupton, Gates and Carty (2000) confirms this result while Khieu, Mullineaux and Yi (2012) neglect any relationship.255 The duration and intensity of the business connection is analyzed by Grunert and Weber (2009) as well as Dermine and Neto de Carvalho (2006). The hypothesis implies that a close relationship between borrower and debtor improves monitoring and thereby increases bank recovery rates. The effect of relationship banking, for example measured by multiple loans or the length of relationship, on recovery rates is confirmed by Grunert and Weber (2009) while Dermine and Neto de Carvalho (2006) do not find any significant results (see also table 16).256

251

See Grunert and Weber (2009), p. 511 and Khieu, Mullineaux and Yi (2012), p. 928.

252

See Hurt and Felsovalyi (1998), p. 3, Dermine and Neto de Carvalho (2006), p. 1237 and Ingermann, Hesse, Bélorgey and Pfingsten (2013), p. 15 and 16.

253

See Khieu, Mullineaux and Yi (2012), p. 928.

254

See Grunert and Weber (2009), p. 510.

255

See Gupton, Gates and Carty (2000), p. 79 and Khieu, Mullineaux and Yi (2012), p. 930.

256

See Grunert and Weber (2009), p. 510 and Dermine and Neto de Carvalho (2006), p. 1238.

Business  Connection

Terms of Credit

Category

Intensity Length

Loan Type (Long Term Loan=1) Creditworthiness

Loan Size

Influencing Factor

Dermine/Neto de Carvalho (2006)

Khieu et al. (2012) 

Grunert/Weber (2009), Khieu et  al. (2012) 

Not Significant Hurt/Felsovalyi (1998 ),  Dermine/Neto de Carvalho  (2006), Ingermann et al. (2013) Khieu et al. (2012)  Grunert/Weber (2009), Gupton et  al. (2000)

Negative

Bank Recovery Rate

Table 16: Bank Recovery Rates and Terms of Credit as well as Business Connection

Grunert/Weber (2009)

Positive

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

73

74

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

IV.2.d Summary Various empirical studies analyze which factors influence recovery rates. However, some other factors that might be relevant are barely investigated. Studies which analyze the same factors frequently find contradictory results. Therefore, it is difficult to clearly identify and summarize effects on recovery rates for several variables. However, literature provides an indication as to which variables influence recovery rates. According to the firm recovery rate, it becomes obvious that within the group of firm-specific variables, collateralization and firm size are frequently investigated. Consequently, the results for firm size are contradictory while collateralization provides positive effects on firm recovery rates. Among the other firm-specific factors, variables referring to the debt structure, over-indebtedness, illiquidity, the legal form of the company and the reason for default show significant results. The analysis of the procedural characteristics indicates that repayments to creditors are significantly higher in reorganization than in liquidation. Other potential influencing factors referring to the person who files for bankruptcy or the establishment of a committee are rarely investigated. Further investigations are required to clearly identify the effects. Surprisingly, the time in bankruptcy seems to be irrelevant for the extent of recovery rates. Macroeconomic variables and their impact on firm recovery rates are also rarely examined. Only a few studies investigate the effect of economic performance and industry distress. It is therefore obvious that the results for economic performance depend on the proxy. While GDP growth rates indicate no relationship, other studies confirm the association between low economic performance and low recovery rates. The studies which provide results about influencing factors on the extent of secured and unsecured recovery rates show that firm size is again one of the most analyzed variables. However, the results are ambiguous again. Collateralization, which is one of the major influencing factors in studies about firm recovery rates, is rarely examined. Furthermore, the results are more or less not significant. All other associations are also analyzed by solely one or two studies. As a result, further research is required to verify previous findings. According to bank recovery rates, collateralization again shows significant and positive results. Although rarely considered, procedural characteristics show significant effects on bank recovery rates. Future work could incorporate further variables into their analysis. Furthermore, variables referring to the terms of credit provide significant results. For example, high amounts of credit and low creditworthiness have a negative impact on bank recovery rates.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

75

As a result, chapter IV.1 shows that there are large deviations between the recovery rates of different subsamples (firm, secured, unsecured, bank recovery rate). Furthermore, the analysis shows that recovery rates vary not only between different countries but also between reorganizations and liquidations. As illustrated in chapter IV.2, literature concludes that these differences are significant. Consequently, results are not easily applicable to other countries or procedures.

IV.3 Evidence for Germany Studies concerning corporate bankruptcies in Germany are rare. Furthermore only a few investigations consider recovery rates within formal reorganizations (so called “Insolvenzplanverfahren”). The reason for this might be that it is hard to acquire data and the poor data quality. One of the few investigations about recovery rates in German bankruptcy proceedings is the study from Kranzusch and Icks (2009). The authors also address the difficult access to data.257 Their data sample, collected out of insolvency statistics, includes recovery rates of firms filing for bankruptcy in North Rhine-Westphalia between 2002 and 2007 and finished the formal process before the end of 2008. However the used database lacks relevant information. Furthermore, the data points are not standardized for all firms. For instance, information delivered while the proceedings are ongoing is not assigned to the same phases of the proceedings properly. Thus, the suitability of the data set for more profound analyses is limited. Altogether, Kranzusch and Icks (2009) investigate 14,839 liquidations and 148 insolvency plans. For liquidations, they report a mean recovery rate for unsecured creditors of about 3.6%.258 Their descriptive analysis suggests that the size of the company, the legal form, the particular industry and the petitioner have an impact on the extend of recovery rates. A subsample, which only involves liquidations with positive recovery rates (n=5,620), reveals unsecured recovery rates of 5.4%. According to the analyzed 148 insolvency plans, the recovery rate amounts to 46.9%.259 Accordingly, the calculation of recovery rates can only be determined approximately. Insolvency statistics concerning corporate bankruptcies which continue after the formal procedure solely collect data about the outstanding claims 257

See Kranzusch and Icks (2009), p. 3.

258

Therefore, it is noticeable that creditors receive repayments in two thirds of the cases. In all other cases they receive nothing because of high insolvency costs or little amounts repayable.

259

Due to data availability it is not clear if expenses for the insolvency procedure are included. The authors anticipate a recovery rate of 23% if the expenses are not considered. For further information see Kranzusch and Icks (2009), p. 27.

76

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

waived. Therefore, in contrast to liquidation procedures, the expected pay-off to creditors has to be calculated. As a result Kranzusch and Icks (2009) find that recovery rates of creditors are substantially better in continuation than in liquidation. Another study by Icks and Kranzusch (2010) confirms the good prospects for creditors for insolvency plan procedures in contrast to liquidations. Based on court files, they find a mean recovery rate of about 14.4% for a sample of five insolvency plans.260 Another noticeable result is that there is a huge gap between expected recovery rate (also called coverage rate) amounting to 66.4% and realized recovery rate. They find indications within their data set that a high coverage rate might lead to a high recovery rate at the end of the proceedings. However, they do not analyze this relationship econometrically. Blazy, Petey and Weill (2012) investigate recovery rates in Germany, France and the UK. While firm recovery rates in Germany (21.5%) and France (20.7%) are nearly similar, they are lower in the UK (13.3%). Within their cross-country analysis they investigate court files of 126 German procedures.261 However, they use a combined data set and do not differentiate between reorganizations and liquidations for German cases. Besides the extent of firm recovery rate, they also present results for different groups of creditors, like senior, junior and new money claims. Accordingly, senior and junior claims approximately correspond to secured and unsecured claims.262 For combined procedures in Germany, they reveal recovery rates for senior creditors of 76.7% and for junior creditors of 10.1%.263 The recovery rates for the UK and France are much lower (see table 2). Next, they try to identify the country specific factors that cause the divergent recovery rates. They therefore incorporate country specific variables concerning the proceedings, the quality of assets and the structure of claims into different regression models.264 They conclude that the differences of recovery rates between these countries can be explained by these factors.265 There is also another study investigating recovery rates across different countries including Germany. Davydenko and Franks (2008) focus on recovery rates of banks in Germany, France and the UK. They find that bank recovery rates are affected by different levels of creditor protection, in particular bankruptcy laws.

260

See Icks and Kranzusch (2010), pp. 102-104.

261

See Blazy, Petey and Weill (2012), p. 22.

262

See Blazy, Petey and Weill (2012), p. 4.

263

See Blazy, Petey and Weill (2012), p. 41.

264

See Blazy, Petey and Weill (2012), p. 27.

265

See Blazy, Petey and Weill (2012), p. 38.

EMPIRICAL EVIDENCE FOR CORPORATE BANKRUPTCIES

77

Median recovery rates for formal bankruptcies are lowest in France (39%) and highest in the UK (82%) with Germany in between (61%).266 In addition, they investigate whether a lack of creditor protection gives rise to adjusted loan contracts at the beginning of the lending relationship. They find that banks claim for higher collaterals when expected recovery rates are low. However, the differences in recovery rates across countries still remain. The key factor for increasing recovery rate is the level of collateral. One subsample of their study especially explores bank recovery rates for corporations which reorganize within a formal procedure.267 They find that accounts receivable are positively related to bank recovery rates in France, whereas higher levels of real estate positively influences bank recovery rates in the U.K. and Germany. Out of five industry dummy variables, only trading industry has a significant influence that lowers the recovery rate of banks. Grunert and Weber (2009) also analyze the extent and influencing factors on bank recovery rates in Germany. They conclude that the amount of collaterals and the intensity of the relation between borrower and bank are factors which have a significantly positive impact on the recovery rates of banks.268 On the other hand, the size of the firm as well as the creditworthiness of the borrower have a negative impact on bank recovery rates. Grunert and Weber (2009) also investigate influencing factors on the probability for high recovery rates (higher than 99%) as well as low recovery rates (lower than 50%). Thus, high recovery rates are positively affected by high exposures at default while the possibility of low recovery rates decreases as long as the relationship between bank and borrower is close, the risk premium is low, the amount of collaterals is high and business operations continue within bankruptcy procedures.269

266

See Davydenko and Franks (2008), p. 582.

267

See Davydenko and Franks (2008), p. 595.

268

See Grunert and Weber (2009), p. 509.

269

See Grunert and Weber (2009), p. 512.

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

V

79

Analysis of German Reorganization Procedures

V.1 Research Design V.1.a Data Acquisition Analyzing firms that reorganize within formal bankruptcy proceedings in Germany faces several problems regarding the data set. Until now, there is no central data bank containing all the required information. Rudimentary information is provided by official statistics. However, the data is not standardized and the comparability between different firms is limited. The only way to collect a data set that allows a profound analysis of court-supervised reorganizations is analyzing the court files of insolvencies and extracting the relevant information manually for each case. In a first step, different associations of insolvency administrators where requested if they could provide cases for this analyzes. Unfortunately, they did not provide files. In a second step, the relevant files were requested from all 192 bankruptcy courts within Germany. However, due to two major problems many courts were unable to approve the inspection. Firstly, several courts declared that since 1999 no formal reorganization process was passed. Secondly, a retrospective distinction between liquidations and continuations was not possible due to archiving problems. Therefore, judges had to identify the relevant cases from memory. Overall, the data set comprises 106 corporate insolvencies, filing for bankruptcy between 1999 and March 2012.270 Furthermore, all considered cases were closed before March 2012. Consequently, the results of the study are not diluted by the amendments to the German Insolvency Statute called “Erleichterte Sanierung von Unternehmen Gesetz” (ESUG) which entered into force in 2012 and 2013. Out of the 106 insolvency cases, four are excluded because the files are not complete. Thus, the final data set contains 102 bankruptcies. This is approximately in line with the sample sizes reported by other studies analyzing reorganization procedures.271 Sample sizes range from 37 cases (Franks and Torous (1994)) up to 225 cases (Bris, Welch and Zhu (2006)). 270

For detailed information about the location in association with the number of cases see chapter V.2.a.1 figure 11.

271

See chapter IV.1 table 2.

80

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Out of the files, three groups of data are extracted. Financial figures are computed at the time of filing as a starting point of the insolvency procedure. To calculate recovery rates for the firms and the different groups of creditors, the financial figures also have to be extracted at the date of the termination of the procedure. Thirdly, procedural information is extracted. To analyze macroeconomic performance, two further variables are included. Information about the annual GDP growth rate is collected from the German Federal Statistical Office. Cyclical effects are measured by the Ifo Business Climate Index. The relevant figures on a monthly basis are collected from the Center for Economic Studies (CES). V.1.b Data Cleansing and Definition of Variables To statistically evaluate the collected data, it is necessary to make the data comparable. First of all, the case files have to be structured consistently. Therefore, it is beneficial for the files to include similar reports by insolvency administrators. For example, almost every court file includes a report which refers to the date of filing as well as the first creditor´s meeting respectively, an insolvency plan and a report at the time of closing. Thus, it is generally possible to conduct analyses which correspond to a specific report and include all collected insolvency cases. However, the reports are prepared at different stages of the proceedings. Furthermore, insolvency files often contain inconsistent information. As a result, the extraction of the relevant data and a comparison is difficult in practice. For example, within a few corporate bankruptcy cases the submission of an insolvency plan corresponds to the application to file for bankruptcy. However, at this early stage the amount of claims is not reliable because the insolvency administrator is not able to proof them beforehand. Creditors have to file their claims not before a certain date which is fixed by courts in the decision to initiate the proceedings. In this case, an analysis of insolvency plans has to consider at least one addition to the initial plan. Depending on the volume and complexity of the case file, it is very time consuming to find the decisive updated version. Furthermore, the administrators use different calculation methods and definitions. Consequently, the calculation of the relevant variables has to be standardized. An investigation like in chapter V.2, which concerns the economic situation of all included firms at the beginning of bankruptcy, has to consider this problem. With regard to the initial report different property items, liabilities and associated securities are affected. The subsequent section presents the variables which are included in the empirical analysis and required data cleansing.

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

81

V.1.b.1 Firm-specific, Procedural and Macroeconomic Variables According to the three groups of influencing factors introduced in chapter IV, the data collected for this study is classified into firm-specific, procedural and macroeconomic variables (illustrated in table 17). Firm-specific variables are primarily collected out of the initial report. In other respects, the information originates from other parts of the case file. Information about the size of the firm (value of TotalAssets and number of Employees) and the asset structure (FixedAssets und CashAssets) is obtained from the financial statement included in the initial report. Depending on future prospects, asset values are measured on the basis of goingconcern or liquidation values. Within the sample (75% of the cases), the assets are reported on the basis of liquidation values. Furthermore, information about the total liabilities (TotalDebt) and the proportion of bank debt to total debt (BankDebt), as well as for secured and unsecured bank creditors (SecBankDebt and SecOtherDebt) is collected. As mentioned earlier, problems referring to the amount of liabilities at the beginning of bankruptcy exist. Creditors have to file their claims after preparation of the initial report. Accordingly, the first report does not cover all claims and solely estimates expected values for the amount of liabilities. The ultimate value of debt is not known until the end of the proceedings. Another interesting factor refers to collateralization and is measured by the ratio secured debt to total debt (SecDebt). However, the collection of data is difficult. Administrators frequently abstain from revealing assets in financial statements when assets are fully collateralized. For example, if a bank finances a machine and is secured in full by the collateralization of the machine, neither the debt nor the machine appear in the balance sheet of the initial report. The value of these assets and corresponding liabilities have to be identified by studying the court files in detail. The ratio of secured debt is also collected for banks (SecBankDebt) and non-banks (SecOtherDebt) separately. Another problem refers to the amount of collaterals which are included in financial statements. Administrators use different methods to measure the value of collaterals. According to § 171 InsO secured creditors have to pay costs of determination (4% of asset value) and disposition (5% of asset value). These lump sum payments are not generally included in financial statements. Furthermore, there is a difference when the costs of determination and disposition are included and when they are not. As long as the value of collateral is higher than the revenue from the asset sale, the creditor receives the whole proceeds.

82

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Table 17: Firm-specific, Procedural and Macroeconomic Variables Influencing Factor

Variable

Definition

TotalAssets Employees FixedAssets CashAssets FreeAssets

Firm Characteristics Firm Size

Firm Age Legal Form Jurisdiction

TotalDebt BankDebt SecBankDebt UnSecBankDebt SecDebt SecBankDebt SecOtherDebt DebtAssets CoverageRate LiquidityGap CostInso CostObligInso TotalCosts RecoveryCosts NoCreditors SecCreditors UnsecCreditors Age LegalForm (0,1) Location (0,1)

Industry Classification Cause of Default

Industry (0,1) Reason (0,1)

Prior Default Workers´ Council Year of Insolvency

PriorDefault (0,1) WorkersCouncil (0,1) YearInso (0,1)

Book value of total assets Number of emloyees Ratio fixed over total assets Ratio cash over total Assets  Total assets minus the value of collateral corrected by the costs of  determination and disposition plus expected proceeds in insolvency Book value of total liabilities Ratio bank debt to total debt Ratio secured bank debt to total debt Ratio unsecured bank debt to total debt Secured debt to total debt Secured debt to total debt held by banks (see above) Secured debt to total debt held by non‐banks Total debt over total assets Total assets and assets in insolvency over total debt Percentage of debt due that is not covered by current assets Expected expenses according to §54 InsO Expected expenses according to §55 InsO Expected expenses according to §54 and §55 InsO Value of free assets minus the expenses according to § 54 InsO  Total number of creditors Number of secured creditors Number unsecured creditors Firm age in years Dummy variable =1 if limited liability Dummy for each location (Dresden, Erfurt, Landshut, Leipzig,  Mannheim, Neuruppin, Stuttgart, Zweibrücken) Dummies for different industries; classification: NACE Rev. 2 Dummies for different reasons like management, financial, strategic,  sales quantity, macroeconomic,private and other, yes=1 Dummy in case of a previous insolvency, yes=1 Dummy in case of the existence of a staff council, yes=1 Dummy for the year of filing for bankruptcy (1999‐2012), yes=1

ContInso (0,1) Survival (0,1) Committee (0,1) ManagementApplied (0,1) ManagementCanceled (0,1) ManagementAccepted (0,1) Petitioner (0,1) InitiatorPlan (0,1) TimeDays

Continuation after filing for bankruptcy=1 Liquidation=0 or Continuation =1 after the procedure Dummy=1 if a committee exists Dummy=1 if the debtor applies for debtor‐in‐possession management Dummy=1 if the debor cancels the application Dummy=1 if courts accept debtor‐in‐possession management  Dummy=1 if petitioner is the debtor Dummy=1 if initiator of the insolvency plan is the debtor Duration of the procedure (filing to closing in days)

GDP Ifo

GDP growth rate in the year of filing Ifo Business Climate Index (in the month of filing)

Asset Structure

Debt Structure

Collateral

Over‐Indebtedness Coverage Rate Liquidity Costs

Number of Creditors

Procedural Charakteristics Continued Operations Outcome Influence Creditors Influence Debtor

Filing/Forced Petition Initiator Insolvency Plan Time in Bankruptcy Macroeconomic Variables Economic Conditions

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

83

Due to the costs of determination and disposition the proceeds decrease. A new claim therefore ensues.272 If the administrator fails to reduce the value of collaterals by corresponding costs, this has to be corrected manually. Furthermore, financial statements provide information about over-indebtedness and illiquidity at the beginning of bankruptcy. Over-indebtedness is measured by total debt over total assets (DebtAssets). However, DebtAssets does not include profits generated during the procedure (Assets in Insolvency). The coverage rate includes these assets and thus can be considered as the expected firm recovery rate estimated by the administrator.







Besides over-indebtedness, there are two other events in German insolvency law that trigger a bankruptcy procedure: illiquidity and imminent illiquidity (see chapter III.1). Within the data sample, imminent illiquidity never initiates a formal procedure. Within the initial report, a liquidity statement gives information about the liquidity gap. The gap is calculated as the percentage of debt due that is not covered by current assets.





A corporation has to file for bankruptcy when the liquidity gap is higher than 10% and an improvement in the near future is unlikely.273 A liquidity gap below 10% suggests that the firm is liquid. Besides the extent of the liquidity gap, the amount of time for which a firm is unable to meet its payments is important. A filing for bankruptcy is not necessary if the firm is expected to become solvent within a three week period. According to this definition, only 2% of the corporations are liquid within the data sample. This shows that illiquidity is a frequent trigger for bankruptcy. 272

See Heeseler (2002), p. 926.

273

See BGH 24.5.2005, IX ZR 123/04.

84

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

In addition to over-indebtedness and illiquidity, the administrator examines whether the free assets of the company are sufficient to cover the expenses of the procedure according to § 54 InsO (RecoveryCosts). Otherwise, the procedure will be rejected due to a lack of assets. The amount of free assets (FreeAssets) is measured by the value of total assets minus the value of collateral corrected by the costs of determination and disposition. In case of continuation after filing for bankruptcy, the expected proceeds have to be added. The full expected expenses include the costs of the insolvency procedure according to § 54 InsO (CostInso) and other obligations according to § 55 InsO (CostObligInso). The former consist of court fees and expenses for the administrator or the members of a creditor´s committee. The latter arise due to the activity of the administrator or wage claims. Several other firm characteristics are collected. These are the number of creditors (NoCreditors), the number of secured (SecCreditors) and unsecured creditors (UnsecCreditors), the age of the firm (Age), the legal form (LegalForm), the venue (Location), a previous insolvency (PriorDefault), the existence of a staff council (WorkersCouncil) as well as the year of filing for bankruptcy (YearInso). The industry and the reason for bankruptcy are also investigated. The industry affiliation is classified by the “Statistical classification of economic activities in the European Community” (NACE Rev. 2). Developed by the European Union, NACE Rev. 2 is inspired by the “International Standard Industrial Classification of all Economic Activities” (ISIC Rev. 4) generated by the United Nations. For each branch of industry (Manufacturing; Electricity, Gas, Steam and Air Conditioning Supply; Construction, Wholesale and Retail Trade; Transporting and Storage; Accommodation and Food Service Activities; Financial and Insurance Activities; Real Estate Activities; Professional, Scientific and Technical Activities; Administrative and Support Service Activities; Education; Human Health and Social Work Activities; Other Services Activities; Other Industries) one dummy variable is established, taking the value one as long as the firm belongs to the specific industry. Different reasons for insolvency are also coded by a dummy variable. Each dummy indicates one of seven reasons (management, financial, strategic, sales quantity, macroeconomic, private or other) for bankruptcy. In many cases, different reasons cause the insolvency. In this context, solely the reason which is most prominent following the administrator is taken into account. Information about the proceedings originate from the initial report, the insolvency plan and the report at the end of bankruptcy. Furthermore, information is located in other parts of the case file like the application for debtor-in-possession management or restructuring subsidies. Two variables concern continuation or liquidation. Those are ContInso and Survival. The former dummy variable refers to con-

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

85

tinuation directly after filing while the latter implies continuation or liquidation at the end of the procedure. The influence of stakeholders on the proceedings is also analyzed. Creditors are able to affect the procedure by a creditors´ committee (Committee), while shareholders gain influence when courts accept debtor-in-possession management (ManagementAccepted). In addition to these two variables, stakeholder groups are able to affect the procedure depending on who files for bankruptcy (Petitioner) and who initiates the insolvency plan (InitiatorPlan). The duration of German insolvency proceedings is investigated by the variable TimeDays. Thereby the time in bankruptcy comprises the days between opening and closing decision. The omitted macroeconomic variables are the growth rate of the German gross domestic product (GDP) and the Ifo Business Climate Index (Ifo). GDP is calculated on an annual basis and assigned to the firms according to their individual year of filing. Information about the business climate is based on monthly data. V.1.b.2 Recovery Rates The distribution of recovery rates between different groups of creditors in German insolvency procedures is one of the major research questions of this study. In this investigation, the final recovery rate should be calculated on the basis of real payments to the creditors. Unfortunately, debt is often repaid long after the end of the court trial. From this, the payments to creditors cannot be extracted from the court files in most cases. Therefore, the expected recovery rate serves as a proxy for the final recovery rate in this study. The expected recovery rate can be analyzed at various points in time. First, indications about the extent of recovery rates are provided within the initial insolvency plan. Depending on the stage of the insolvency procedure, the parameters to measure the recovery rate are more or less uncertain. The required data like the amount repayable as well as the value of insolvency claims are based on preliminary estimations. The accuracy of the prediction increases according to later stages of the proceedings respectively later versions of the insolvency plan. The recovery rates calculated at the time of voting over the insolvency plan were examined by the administrator at least once. The present study therefore uses recovery rates at the date of voting, including previous payments to creditors within the procedure. Another difficulty of the study concerns the measurement of recovery rates. The author of the insolvency plan has to form different groups of creditors according to

86

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

their legal status.274 However, this classification does not allow for an immediate investigation of secured and unsecured recovery rates as well as firm recovery rates. The reason is that not every creditor claim is involved in the insolvency plan. Under German insolvency law, there is a distinction between two groups of secured creditors. One group of secured creditors is entitled to separate an object from the insolvency estate.275 Therefore, they do not have any claim in the insolvency procedure and are not included in the insolvency plan. The claims of the second group of secured creditors are included in the insolvency plan only if the plan affects their rights.276 This illustrates that several secured claims are not included in insolvency plans due to full satisfaction of claims or continuation of the contractual relationship after bankruptcy. This study includes these claims in the calculation of recovery rates. Another important aspect for the calculation of the secured recovery rate is that secured claims are grouped into at least one creditor class. When the value of the security is worth less than the liability, this creditor is represented in two different classes. He joins the group of secured creditors only with the amount covered by the security. According to § 222 InsO, the remaining claim is part of the group of non-lower-ranking creditors, which usually includes all unsecured creditors. This new classification of secured debt into unsecured debt distorts the calculation of originally secured claims. Therefore, it is necessary to revise the data. Furthermore, it is inappropriate to calculate secured, unsecured and firm recovery rates directly out of the insolvency plan at the time of voting. Rather, the entire case file has to be studied to identify all impaired claims, the realization of collaterals during the insolvency procedure as well as continued contracts, which can be seen as fully repaid. After the calculation of repayments and corresponding liabilities for secured and unsecured creditors, firm recovery rates can be computed. For the purpose of this study, subordinated insolvency claims are not taken into account. One reason is that these claims are waived according to § 225 InsO. Furthermore, creditors of subordinated claims frequently belong to the group of shareholders, their family members or firm management. As a result, the effect on recovery rates for ordinary creditors would be diluted. As a result the data sample enables the analysis of nine different recovery rates defined in table 18. As mentioned earlier, the firm recovery rate is calculated by the amount repayable to all creditors divided by corresponding debt claims. Fur-

274

See § 222 InsO.

275

See § 47 InsO.

276

See § 222 InsO.

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

87

thermore, several subsample recovery rates are measured like secured and unsecured recovery rates as well as bank and non-bank recovery rates. Table 18: Definition of Recovery Rates Variable

Definition

FirmRR BankRR OtherRR

Total pay‐out to creditors, scaled by total debts Total pay‐out to bank creditors, scaled by total debts of bank creditors Total pay‐out to other creditors, scaled by total debts of other creditors

SecRR SecBankRR SecOtherRR

Total pay‐out to secured creditors, scaled by total debts of secured creditors Total pay‐out to secured bank creditors, scaled by total debts of secured bank creditors Total pay‐out to secured other creditors, scaled by total debts of secured other creditors

UnsecRR Total pay‐out to unsecured creditors, scaled by total debts of unsecured creditors UnsecBankRR Total pay‐out to unsecured bank creditors, scaled by total debts of unsecured bank creditors UnsecOtherRR Total pay‐out to unsecured other creditors, scaled by total debts of unsecured other creditors

V.2 Data Analysis V.2.a Descriptive Statistics V.2.a.1 Firm Characteristics The initial report, prepared by the preliminary administrator, includes information about the economic situation at the moment of filing for bankruptcy as well as other firm characteristics. Table 19 summarizes information about firm characteristics for the whole sample. Furthermore, the initial sample is divided into two subsamples. One subsample includes firms which survive after the formal process. The other subsample involves companies which are liquidated. Different tests of significance like the t-test, the Wilcoxon-Mann-Whitney as well as the Mediantest compare the means of the subsamples and analyze whether there exist significant differences between these two subsamples.

92 92 92

84 92 92 100 85

TotalDebt SecDebt BankDebt

DebtAssets CoverageRate LiquidityGap Employees Age

4527% 36% 89% 13 13

2,740,172 19% 39%

721,572 45% 22%

377% 26% 97% 3 9

614,506 9% 35%

134,632 43% 9%

Total Mean Median

2047% 46% 85% 19 15

3,564,278 24% 40%

979,220 50% 22%

305% 39% 94% 7 9

765,229 21% 37%

301,074 52% 12%

Continuation Mean Median

Note: ***, ** and * denote significance at the 1%, 5% and 10% level.

92 84 84

Obs

TotalAssets FixedAssets CashAssets

Variables

Table 19: Summary Statistics

10524% 14% 96% 3 10

1,316,408 7% 36%

230,275 37% 24%

1147% 7% 100% 0 9

300,592 0% 25%

12,335 9% 0.2%

Liquidation Mean Median

1.70 ** ‐5.60 *** 3.32 *** ‐3.41 *** ‐1.16

‐0.78 ‐4.39 *** ‐0.49

‐1.08 ‐1.37 * 0.36

t‐test

3.86 *** ‐5.00 *** 4.27 *** ‐5.56 *** ‐1.11

‐2.46 ** ‐4.38 *** ‐0.80

‐4.57 *** ‐1.70 * ‐2.26 **

Wilcoxon

12.95 *** 19.40 *** 9.50 *** 27.56 *** 0.03

9.50 * 15.71 *** 0.78

15.71 *** 2.82 * 4.66 **

Median‐Test

88 ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

89

A close view on firm characteristics shows that the firms within the sample are small to medium-sized enterprises (see table 19). Total assets account for 722,000 € on average (median 135,000 €). In comparison to US studies, the firms in the sample are much smaller. However the studies by Armour, Hsu and Walters (2012) for the British market and Thorburn (2000) for Sweden examine similar firm sizes with total assets from about 3.3 million £ and 2.4 million US$ respectively.277 The number of employees also confirms that the included firms are rather small. Only 13 employees on average work for the insolvent companies, whereas the number of employees ranges between zero and 229. This explains why there are only two workers´ councils. The tests of significance show that there is a significant difference between the size of insolvent firms in liquidation and continuation. Consequently, firms which survive are significantly larger than those of liquidations. The assets of the firm include fixed assets, current assets and private assets. The observed firms show fixed assets of 45% and cash assets of 22% on average. The median for cash assets is only 9%. Again, the tests of significance show that there is a significant difference between liquidations and continuations for both asset structure variables. The first estimation of total debt at the moment of filing shows a huge difference between mean value (2.7 million €) and median (615,000 €). This indicates that a few firms have total debt values which are much higher than for most other companies. The difference between liquidations and continuations is also significant. In addition to the level of total debts at the moment of filing, table 20 summarizes the level of total debts at the end of the bankruptcy procedure (TotalDebtEnd).

Table 20: Comparison Total Debt (Filing vs. Termination) Variable TotalDebt TotalDebtEnd Difference %

277

Obs

Mean

Median

92 95 85 85

2,740,172 3,645,751 1,173,539 50%

614,506 559,040 25,559 15%

See Armour, Hsu and Walters (2012), p. 115 as well as Thorburn (2000), p. 345.

90

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Differences between these two values provide insights into the accuracy of estimation by the insolvency administrator. Table 20 shows that there is a difference between the expected and the final value of total debt. A Wilcoxon signed-rank test for related samples can be used to statistically prove that the difference between total debt at filing and at closing is statistically significant. The null hypothesis is rejected at the 1% level (z= -2.699, p= 0.0070). This indicates that estimations about the final recovery rate at the beginning of bankruptcy are difficult. Besides the expected recovery rate, other factors have to be taken into account. Figure 2 demonstrates the big gap between expected and final total debt values in relation to the expected debt values.

Figure 2: Difference Total Debt 200

Difference Total Debt (%)

150

100

50

0

-50 0

20

40 60 Number of Observation

80

100

The figure illustrates that most of the differences vary within an interval between 50% and +50%. However, distinctions over +50% occur frequently. A positive discrepancy shows that the amount of total debts increases during the procedure while a negative divergence indicates that the expectations have been too high. A further variable, illustrated in table 19, refers to collateralization. The proportion of secured debt to total debt (SecDebt) at the beginning of bankruptcy is 19% on average. Referring to continuation and liquidations, it is obvious that mean values

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

91

are significantly different at the 1% level. In continuations the level of collateralization (24%) is significantly higher than in liquidations (7%). The relationship between banks and insolvent companies is analyzed by the amount of bank debt to total debt (BankDebt). In about 17% of cases, firms have no bank debt. However, bank financing reaches 39% on average. The separation of the sample shows no significant differences for reorganizations compared to liquidations. All three tests of significance confirm this result. After calculating the value of assets and debts at the beginning of bankruptcy, it is possible to analyze the events which induce bankruptcy in Germany. Overindebtedness (DebtAssets) occurs in 95% of the included insolvency cases. Table 19 shows that in the median, liabilities are 3.7 times higher than the assets of the firm. Consequently, the mean value is strongly influenced by extreme values. As mentioned in chapter V.1.b.1 the coverage rate can be considered as the expected firm recovery rate at the time of filing for bankruptcy. For the entire sample, the coverage rate indicates that creditors will receive repayments amounting to 36% of their claims. The median is about 26%. This suggests that extreme values affect the data (see figure 3). For example, the bottom 10% of the distribution shows a coverage rate under 0.6%. Furthermore, the tests of significance show that the coverage rate in liquidations and reorganizations are significantly different. In case of reorganization, the coverage rate is about 46% on average, while in case of liquidation the coverage rate is just 14% on average. As a result, the coverage rate within reorganizations is significantly higher than in liquidations.

92

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Figure 3: Coverage Rate 2

Coverage Rate

1.5

1

.5

0 0

20

40 60 Number of Observation

80

100

Figure 4 illustrates a comparison between the expected recovery rate at the beginning of the procedure (CoverageRate) and the final firm recovery rate (FirmRR).

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

93

Figure 4: Differences Firm Recovery Rate - Coverage Rate

Difference Total Recovery Rate (in %)

50

8 0 -23 -50

-100

-150 0

20

40 60 Number of Observation

80

100

Thereby the differences are calculated according to: Consequently, positive differences arise when the firm recovery rate is higher than the coverage rate. One explanation is that the administrator's expectation about the final firm recovery rate is too pessimistic. Otherwise, a negative difference shows that the administrator was too optimistic about the final firm recovery rate. In 40% of the cases, the differences are positive while in 60% of the cases the differences are negative. Positive differences are about 8% (median 5%) and negative differences are about -23% on average (median -11%). Several reasons may affect the differences. For example, it is possible that the administrator fails to identify all relevant assets and liabilities within the financial statement at the beginning of bankruptcy. In addition, the initial assessment of asset values can be wrong. Another reason may be that the expectations about the proceeds after filing are inaccurate. Furthermore, recovery rates increase when the firm acquires new money (for example subsidies or loans). As long as the agreement to provide new money takes place after the preparation of the initial report, the coverage rate is calculated without it.

94

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Moreover, the sample provides evidence that over-indebtedness is not only a problem for corporate entities. Although over-indebtedness induces bankruptcy solely for corporate entities (see chapter III.1), it is also a problem for non-incorporated firms. Within the data sample, there are only a few cases which are illiquid but not over-indebted. On average, 89% of debt due is not covered by current assets (LiquidityGap). Thereby, in 73% of the cases the liquidity gap is at least 89%. Figure 5 shows that only two companies have a liquidity gap under 10%, which means that they are solvent (see chapter V.1.b.1). Furthermore, there is a significant difference between liquidations and reorganizations (see table 19). The liquidity gap is significantly higher in liquidations.

Figure 5: Liquidity Gap

Liquidity Gap

100

50

10 0

0

20

40 60 Number of Observation

80

100

Another aspect concerns the costs of the procedure. To verify whether the expenses of the procedure are covered, the amount of the free assets and the expenses according to § 54 InsO have to be calculated.

92 89 89 89

Obs

401,822 62,416 193,287 349,551

38,361 24,804 25,565 13,389

574,145 84,803 288,068 515,158

91,775 34,311 39,743 33,466

Continuation Mean Median

Note: ***, ** and * denote significance at the 1%, 5% and 10% level.

FreeAssets CostInsO TotalCosts RecoveryCosts

Variables

Total Mean Median

Table 21: Expenses Insolvency Procedure

88,289 21,308 34,369 66,981

7,450 5,461 6,933 658

Liquidation Mean Median ‐1.12 ‐1.63 * ‐0.88 ‐1.11

t‐test

‐5.10 *** ‐4.86 *** ‐4.72 *** ‐4.86 ***

Wilcoxon

27.93 *** 19.05 *** 27.61 *** 23.13 ***

Median‐Test

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

95

96

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

The median averages amount to 38,000EUR for free assets and 25,000EUR for expenses according to § 54 InsO (CostInso). This means that for the average firm, free assets are high enough to cover the expenses. However, it should be noted that the expenses are unknown in three cases. The variable RecoveryCosts implies that the value of free assets exceeds the expenses according to § 54 InsO (CostInso) by 13,000 € for the average firm. Due to the fact that the companies within the sample finished a formal procedure, it is obvious that the administrators concluded that the assets will cover the expenses. However, this does not mean that the assets exceed the costs in every single case and at any time. In 20% of the cases, the initial report shows that the expected costs are higher than the free assets. However, the application for bankruptcy was accepted. This is due to the possibility in German insolvency law that the expenses for the proceedings can be deferred. However, the amount of the deficit is only 2,500 € on average. Again, table 21 shows significant differences between liquidations and reorganizations. The reorganization sample indicates that there are significantly more free assets as well as expenses in comparison to the sample of liquidations. Besides the analysis of financial parameters, other firm characteristics can be examined. One collected variable examines previous insolvencies. Within the data set, eight companies have been bankrupt before (PriorDefault). Out of these, five companies are restructured at least a second time. The investigation concerning the legal form of the insolvent companies finds that the sample includes firms with limited liability in 26% of the cases. Other firms are often individual companies. This also shows that the included companies are rather small. A successful restructuring is achieved in 85% of all limited companies and in 58% of all other cases. Figure 6 illustrates that most of the observed firms belong to four different industries. These industries are: manufacturing (16%), wholesale and retail trade (15%), accommodation and food service activities (14%) as well as professional, scientific and technical activities (22%).

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

97

Figure 6: Industries (Whole Sample) % 20

15

10

5

0 Manufacturing

Real estate activities

Electricity, gas, steam and air conditioning supply

Administrative and support service activities

Construction

Professional, scientific and technical activities

Wholesale and retail trade

Education

Transporting and storage

Human health and social work activities

Accommodation and food service activities

Other services activities

Financial and insurance activities

Other industries

A further analysis shows that continuations are more frequent in the manufacturing industry as well as for professional, scientific and technical activities. This indicates that special-purpose machines are hard to sell and continuation is beneficial in these cases. However, wholesale and retail trade as well as accommodation and food service activities are liquidated more often (see figure 7). One explanation might be that accommodation and food service activities are frequently independent of one specific renter. In case of bankruptcy, the old renter can be replaced by a new one.

98

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Figure 7: Industries (Liquidation vs. Continuation) % 25

20

15

10

5

0

Liquidation Manufacturing Wholesale and retail trade

Continuation Accommodation and food service activities Professional, scientific and technical activities

Two reasons seem to be responsible for many bankruptcies. In 29% of the cases, management misconduct is the main reason for insolvency. The second reason is financial problems (30%). Figure 8 illustrates the minor importance of the other analyzed reasons.

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

99

Figure 8: Reasons for Bankruptcy % 30

20

10

0 Management Financial

Strategic Sales Quantity

Macroeconomic Private

Other

The firms examined are 13 years old on average (see table 19). A WilcoxonMann-Whitney test shows that there is no significant difference between the underlying distribution of the firm age for continuations and the firm age for liquidations (z = -1.110, p = 0.2670). Figure 9 illustrates that firms which continue are 15 years on average while liquidated firms are only ten years old on average. Figure 9: Firm Age in Liquidation vs. Continuation

Age of the Firm (Years)

15

10

5

0

Liquidation

Continuation

100

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

The frequency distribution in figure 10 illustrates how many filings occur in each year (YearInso). The figure shows that in the early years after the law reform in 1999, insolvency plans are rare. One reason is that the judges had to identify relevant cases from memory. It is possible that they changed jobs since 1999. Due to the fact that insolvency cases have to be finished to be included in the sample, the last two years are also of minor importance. The study integrates cases which are finished in 2012 but non which are filed in 2012. The highest proportion of filings, at almost 15%, is found in 2009. The German insolvency statistic shows this peak of corporate insolvencies too. In comparison to the previous year, the statistic reveals an increase of 11.6%.278 This development is not very surprising due to an annual growth rate of the GDP of -4% for the same period.279

Figure 10: Year of Insolvency % 15

10

5

0 19992000 2001 20022003 20042005 20062007 2008 20092010 2011

The regional focus of the investigated insolvency cases is illustrated in figure 11. Accordingly, a large proportion of the included cases originate out of the newlyformed German states. This is due to the approval to inspect the case files by in-

278

See Statistisches Bundesamt (2014a).

279

See Statistisches Bundesamt (2014b).

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

101

solvency courts and has nothing to do with the number of cases, including an insolvency plan in different German cities. To examine the relationship between two categorical variables like the location of the firm and a dummy variable that refers to continuation or liquidation after bankruptcy a Fisher´s exact test is used. The test shows that there is no statistically significant relationship between the location and the survival dummy (p = 0.373). Consequently, it cannot be said that liquidations or continuations occur significantly more frequently in specific cities.

Figure 11: Location of Insolvency Court % 40

30

20

10

0 Leipzig Mannheim

Neuruppin Zweibrücken

Landshut Erfurt

Dresden Stuttgart

V.2.a.2 Procedural Characteristics The survival dummy mentioned above belongs to the group of procedural characteristics. It investigates whether firms are able to continue or have to be liquidated after bankruptcy. Continuation is achieved in two thirds of the cases. All other firms are liquidated during the proceedings. It is noteworthy that immediately after filing, 76% of the sample firms continue their operations. This shows that the decision to restructure the firm does not automatically lead to continuation after bankruptcy.

102

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

On the basis of several variables, the influence of creditors on the bankruptcy procedure can be studied. A first variable concerns the number of creditors and another one the existence of a creditors´ committee (see table 22). On average, 50 creditors are involved, most unsecured. The number of secured creditors is only two on average. With regard to the number of creditors as a whole as well as the number of secured creditors and the number of unsecured creditors, significant differences between liquidations and reorganizations can be found. The number of each creditor group is significantly lower in liquidations.

94 93 92 101

Secured Creditors Unsecured Creditors Total Creditors Committee

2 47 50 6%

1 26 27 0

Total Mean Median 3 59 63

2 42 46

Continuation Mean Median

Note: ***, ** and * denote significance at the 1%, 5% and 10% level.

Obs

Variables

Table 22: Number of Creditors

1 25 26

0 21 21

Liquidation Mean Median ‐2.72 *** ‐3.40 *** ‐3.98 ***

t‐test

‐4.82 *** ‐3.42 *** ‐3.68 ***

Wilcoxon

19.39 *** 15.35 *** 16.03 ***

Median‐Test

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

103

104

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

A large number of creditors indicates the existence of conflicting interests. The proceedings are therefore affected by coordination problems. Therefore, the German bankruptcy law provides the possibility to form a creditors´ committee. However, a creditors´ committee is established in only 6% of cases. In each of those cases, continuation is achieved after the formal procedure. Furthermore, the establishment of a committee may depend on the number of creditors involved in the proceedings. The analysis shows that in firms in which a committee is established, the number of creditors is higher than the median. Besides the creditors, shareholders are also involved in the proceedings. A direct influence by shareholders respectively debtors is possible as long as the courts approve the application for debtor-in-possession management. Accordingly, three aspects promise interesting results. First of all, there are corporations which apply for debtor-in-possession management (Dummy ManagementApplied). Secondly, some of these corporations cancel the application (Dummy ManagementCanceled) and finally there are firms which are allowed to manage by themselves (ManagementAccepted). Table 23 shows that in 16% of the cases, the debtor applies for debtor-in-possession management, while in 25% of those cases the debtor cancels the application later on. Debtor-in-possession management is accepted in 50% of the cases which apply for it but only in 8% of the cases at all.

Table 23: Debtor-in-Possession Management Variable

Obs

Mean

Std. Dev.

Min

Max

Management Applied Management Canceled Management Accepted

98 16 102

16% 25% 8%

37% 45% 27%

0 0 0

1 1 1

In all cases in which debtor-in-possession management is accepted, reorganization is achieved. This implies that debtor-in-possession management has positive effects on the preservation of firms. As mentioned in chapter V.1.b.1, creditors and shareholders are allowed to file for bankruptcy or to initiate an insolvency plan. Depending on who files for bankruptcy and who initiates the insolvency plan, stakeholders are able to influence the procedure as well. Table 24 illustrates the results for the collected data sample.

105

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Table 24: Petitioner and Initiator Insolvency Plan Variable

Obs

Mean

Std. Dev.

Min

Max

Petitioner Initiator Plan

100 99

73% 37%

45% 49%

0 0

1 1

Accordingly, the debtor files for bankruptcy in 73% of the cases. The remaining cases are filed by a creditor, often the tax authority. In 68% of the cases in which the debtor filed for bankruptcy, a successful reorganization was achieved. In contrast, only 56% of the creditor filings achieve continuation after the formal process. However, there is no significant association between the petitioner and survival after bankruptcy (chi-square with one degree of freedom = 1.2371, p = 0.266). The insolvency plan is initiated by the debtor or his advisor in 37% of the cases. In all other cases, the insolvency administrator prepares the plan on behalf of the creditors. This suggests that the insolvency procedure is barely influenced by the debtor. Furthermore, in 68% of the cases an insolvency plan by a debtor results in continuation while a plan by the administrator leads to continuation in about 64% of the cases. The results of a Pearson´s Chi2 test shows that there is no significant relationship between the type of initiator and survival after bankruptcy (chi-square with one degree of freedom = 0.1342, p = 0.714). Another interesting result concerns the duration of the procedure. The time in bankruptcy within the whole sample is about 920 days on average (equals two and a half years). It can therefore be noted that on average, reorganization procedures (784 days) are faster than liquidations (1,207 days). However, the longest durations of bankruptcy can be found in reorganizations. The Wilcoxon-MannWhitney test shows that the null hypothesis can be rejected (z = 3.194, p= 0.0014). Therefore, there is a significant difference between the duration of procedures in reorganizations and liquidations. V.2.a.3 Distribution of Recovery Rates The firm recovery rate and the recovery rates for the different subsamples are summarized in table 25.

106

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Table 25: Distribution of Recovery Rates Variable

Obs

Mean

Median

Std. Dev.

Min

Max

FirmRR BankRR OtherRR

95 80 94

23.92% 41.74% 14.74%

19.62% 33.84% 9.71%

20.88% 35.14% 15.36%

0.18% 0.18% 0.18%

79.84% 100.00% 75.99%

SecRR SecBankRR SecOtherRR

69 56 36

65.56% 64.92% 65.67%

68.65% 68.50% 67.81%

29.71% 30.74% 35.59%

1.50% 1.50% 2.60%

100.00% 100.00% 100.00%

UnsecRR UnsecBankRR UnsecOtherRR

95 48 94

11.11% 9.56% 10.93%

7.50% 5.26% 7.71%

11.34% 15.80% 10.95%

0.18% 0.00% 0.18%

68.29% 100.00% 68.29%

The proportion of payments all creditors receive after the proceedings is about 24% on average. The results are similar to those of the studies by Thorburn (2000) and Armour, Hsu and Walters (2012) which analyze equal-sized firms. Thorburn (2000) reveals a slightly higher firm recovery rate of about 35% for Sweden, while Armour, Hsu and Walters (2012) report repayments of about 20-32 % for British reorganizations. 280 A more differentiated analysis of the firm recovery rate shows that banks recover 42% of their defaulted claims while the proportion for other creditors is only 15%. Related studies find higher bank recovery rates (57-72%, see chapter IV.1). These differences might not only be caused by country-specific factors. In contrast to other investigations, this study does not include informal cases in the calculation of bank recovery rates. The analysis of repayments to secured creditors shows recovery rates amount to 66%. This rate is comparable to the 69% revealed by Thorburn (2000). Unfortunately Armour, Hsu and Walters (2012) do not differentiate recovery rates for secured and unsecured creditors. However, within a working paper version Armour, Hsu and Walters (2006) identify a secured recovery rate of 61%.281 For the German sample the secured recovery rates for banks (65%) and other creditors (66%) are very similar. The unsecured recovery rates reported by Thorburn (2000) and

280

See Thorburn (2000), p. 360 and Armour, Hsu and Walters (2012), p. 116.

281

See Thorburn (2000), p. 360 and Armour, Hsu and Walters (2006), p. 20.

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

107

Armour, Hsu and Walters (2006) range between 0% and 2%.282 Therefore, the unsecured recovery rate in Germany is slightly higher (11%). Recovery rates up to 100% are found for secured creditors and banks, while unsecured creditors (with the exception of unsecured bank creditors) do not achieve recovery rates higher than 68%. Therefore, it can be assumed that there is a relationship between the extent of recovery rates and collateralization or relationship banking respectively. The tests of significance show differences between recovery rates in continuation and liquidation (see table 26). For the secured recovery rate and the two subsamples concerning bank creditors and other creditors, no significant difference can be found. Consequently, for the extent of the secured recovery rate, whether or not the firm continues or is liquidated after bankruptcy is irrelevant. However, all other recovery rates show significant differences according to liquidation and continuation. In other words, without being secured the survival after bankruptcy is decisive for the extent of recovery rates. Thus higher repayments are achieved in continuation.

282

See Thorburn (2000), p. 362 and Armour, Hsu and Walters (2006), p. 20.

95 80 94

69 56 36

95 48 94

FirmRR BankRR OtherRR

SecRR SecBankRR SecOtherRR

UnsecRR UnsecBankRR UnsecOtherRR

11.11% 9.56% 10.93%

65.56% 64.92% 65.67%

23.92% 41.74% 14.74%

7.50% 5.26% 7.71%

68.65% 68.50% 67.81%

19.62% 33.84% 9.71%

Total Mean Median

13.85% 12.32% 13.63%

65.81% 66.10% 64.84%

30.69% 52.71% 17.49%

9.86% 7.38% 9.90%

68.50% 68.65% 65.52%

24.41% 48.06% 11.29%

Continuation Mean Median

Note: ***, ** and * denote significance at the 1%, 5% and 10% level.

Obs

Variables

6.20% 6.81% 6.17%

64.65% 60.06% 67.84%

11.78% 20.20% 9.89%

3.46% 2.99% 3.46%

69.62% 62.82% 78.31%

9.51% 10.87% 5.37%

Liquidation Mean Median

Table 26: Distribution of Recovery Rates (Liquidation vs. Reorganization)

‐3.77 *** ‐1.21 ‐3.74 ***

‐0.13 ‐0.58 0.22

‐5.55 *** ‐4.72 *** ‐2.36 **

t‐test

‐3.95 *** ‐1.19 ‐4.05 ***

0.07 ‐0.43 0.11

‐4.53 *** ‐4.14 *** ‐3.18 ***

Wilcoxon

14.26 *** 3.00 * 14.93 ***

0.13 0.11 0.55

17.67 *** 12.58 *** 4.61 **

Median‐Test

108 ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

109

V.2.b Influencing Factors on Firm Recovery Rates V.2.b.1 Hypotheses Besides the examination of the economic situation at the time of filing (see chapter V.2.a.1 and V.2.a.2) and the extent of different creditor recovery rates (see chapter V.2.a.3), another goal of the empirical analysis is to investigate factors influencing firm recovery rates in Germany. As mentioned earlier, literature about insolvency plans in Germany is rare. Especially the extent of firm recovery rates and factors influencing firm recovery rates have not yet been analyzed into detail for formal restructurings. As mentioned in chapter IV.1, the firm recovery rate is defined as the amount repaid to all creditors divided by the value of creditors´ claims. This study focuses on four aspects affiliated with the German market and insolvency legislation. Based on the four aspects, seven hypotheses about potential influencing factors on recovery rates are developed which are analyzed in this chapter. The first aspect concerns collateralization283. Due to the possibility of selling collateralized assets in insolvency, the related hypothesis states that repayments to creditors increase when the amount of secured debt to total debt is high (H1). To obtain further information, the effect of secured debt held by banks as well as other creditors is analyzed separately. The second aspect deals with relationship banking. Due to the bank-based German financial system, this analysis is of special interest. The influence of banks can be analyzed by the proportion of bank debt to total debt. The corresponding hypothesis postulates that a high proportion of bank debt signals a close relationship and results in higher firm recovery rates (H2). The influence of bank creditors is analyzed for secured bank debt and unsecured bank debt separately. The third aspect is about triggering events which are specific to the insolvency legislation in Germany. As mentioned earlier, two (out of three) different events which initiate a formal procedure can be identified within the data sample: Overindebtedness and illiquidity. The third hypothesis states that a high debt overload measured by total debts over total assets results in lower firm recovery rates (H3). One further potential factor influencing the firm recovery rate is called liquidity gap. This describes the proportion of debt due that cannot be paid out of current assets. The related hypothesis postulates that a large liquidity gap leads to lower recovery rates (H4).

283

For definition of variables see chapter V.1.b.

110

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

The last aspect concerns stakeholders´ influence on the procedure. The first corresponding hypothesis concentrates on creditors´ power to influence the insolvency proceedings. Therefore, an interaction term is created that combines the number of creditors with the establishment of a creditors´ committee. The underlying assumption is that firm recovery rates decrease as long as many creditors with conflicting interests are involved and no committee exists due to coordination problems among creditors (H5). Another potential influencing factor is debtors´ influence on the proceedings. Therefore, the impact of debtor-in-possession management on firm recovery rates is investigated. One might argue that specific knowledge of the old management increases the value of the firm. On the other hand, specific knowledge of the management increases dependencies and thereby the renegotiation power of equity holders. For this reason, the hypothesis is that the recovery rate to all creditors decreases in case of debtor-in possession management (H6). The last potential influencing factor refers to the petitioner of bankruptcy. The assumption is that when a creditor files for bankruptcy, the firm recovery rate decreases. The underlying reason is that creditors only file when the management has delayed filing and the firm's situation has deteriorated (H7). Besides these potential influencing factors on firm recovery rates, several control variables (see table 27) which show significant results in other studies are tested.284 To reduce the influence of outliers some variables are transformed by the logarithm. All included variables are part of the three variable groups concerning firm-specific, procedural or macroeconomic variables. Table 27 gives an overview over all included factors which potentially influence the extent of firm recovery rates:

284

See for example Blazy, Chopard and Nigam (2013), Blazy, Petey and Weill (2012) or Thorburn (2000).

111

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Table 27: Hypotheses and Control Variables Potential Influencing Factors Hypotheses H1 Collateral H2 Relationship Banking H3 Over‐Indebtedness H4 Illiquidity H5 Creditors´ Power H6 Debtors´ Power (Dummy) H7 Petitioner (Dummy) Control Variables Economic Outcome (Dummy) Firm Size Legal Form (Dummy) Reason (Dummy) Economic Condition

Variables

SecDebt: SecBankDebt and SecOtherDebt BankDebt: SecBankDebt and UnsecBankDebt DebtAssets LiquidityGap NoCreditors x Committee ManagementAccepted Petitioner (Petition by Debtor)

Hypotheses

+ + ‐ ‐ ‐ ‐ +

EcoOutcome lnTotalAssets LegalForm ReasonManagement lnGDP lnIfo

V.2.b.2 Methodology To analyze the supposed relationships between the firm recovery rate and the mentioned variables, different techniques are used. In a first step, the relationships between two variables are tested without examining further influencing factors. Therefore, a correlation analysis investigates the relationship between the firm recovery rate and a single influencing factor. In a second step, a multiple regression analysis investigates the assumed causal relationship between the firm recovery rate as dependent variable and several independent variables. Depending on several assumptions like the level of measurement (nominal, ordinal, interval or ratio) different correlation coefficients can be calculated to measure the strength and direction between two variables. Within the scope of this study, the potential influencing factors are interval or nominal scale variables. The firm recovery rate is also interval. To identify the strength and direction of dependence between the firm recovery rate and other interval scale variables, the Pearson´s product-moment correlation coefficient is used. This coefficient is commonly used in literature to investigate the dependency between two interval scaled variables.285 To meet the assumptions for Pearson´s correlation coefficient, the relationship has to be linear and the two

285

See for example Grunert and Weber (2009).

112

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

variables have to be normally distributed. However, the correlation coefficient makes no presumption about the causal relationship. The coefficient is calculated by the covariance of the variables ( ) standardized by the two standard deviations ( , ):286 ∑ ∑

̅ ̅ ²∑

²

Whereas ,

denote the two interval variables

= Sample size. Pearson´s coefficient will be between 1

1.

Accordingly, a correlation of 1 (-1) indicates a perfect positive (negative) linear relationship while a correlation of zero indicates no linear relationship. The pvalue indicates weather the relationship is statistically significant (two-tailed test). Due to outliers identified in chapter V.2.a.1 as well as a rather small sample size, Spearman´s rho (nonparametric test) is also calculated. In contrast to Pearson´s coefficient, variable values are transformed into ranks. As long as no identical values exist the correlation coefficient by Spearman is computed by287:

1

6∗∑

² 1

Where is the difference between ranks of variable Sample size.

286

See for example Toutenburg and Heumann (2009), p. 131.

287

See Schulze (2003), p.130.

and variable

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

113

The interpretation of Spearman´s rho is analog to Pearson´s coefficient, whereas a monotonic relationship is assumed. The Pearson coefficient as well as Spearman´s rho are appropriate only for identifying the relationship between interval scaled variables. The correlation between the firm recovery rate and the dichotomous variables is measured by a pointbiserial correlation288: 1 Whereas is an interval variable, , are the mean values of the interval variable for group 1 and 2 of the dichotomous variable Y respectively, ,

are the number of observations in group 1 and 2, is the total sample size, and ∑

² is the standard deviation.

A point-biserial correlation is a special case of Pearson’s correlation. The results also range between -1 and 1 indicating a perfect negative respectively positive association between the dummy variable and the firm recovery rate. As mentioned earlier, the causal relationship between the firm recovery rate (as dependent variable) and several independent variables is investigated by several ordinary least squares (OLS) regression models. After considering several assumptions like linearity, random sampling, zero conditional mean, multicollinearity and homoskedasticity (also called Gauss-Markov assumptions), the OLS method seems to be an appropriate estimator.289 Furthermore, several empirical studies

288

See Dormann (2013), p. 100.

289

See for example Wooldridge (2009), pp. 103-104.

114

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

which investigate influencing factors on recovery rates use OLS regression analysis.290 The first regression model can be specified by the following equation, whereas defines the estimated coefficient and denotes the error term:

(1)







The initial model concentrates on the influence of firm characteristics. Besides variables concerning hypotheses H1-H4, different control variables are included (also illustrated in table 27). The second model investigates the power of stakeholders to affect the procedure and thereby the firm recovery rate. The associated hypotheses are H6 and H7. To examine the effect of continuation vs. liquidation, the dummy variable Survival is integrated within the second model:

(2)

The third model combines both groups of variables analyzed in model (1) and (2) and the fourth model examines macroeconomic variables, like the annual growth rate of the GDP and the business climate index. The models are specified by:

(3)









(4) 290

See for example Thorburn (2000), Blazy, Chopard and Nigam (2013), Blazy, Chopard, Fimayer and Guigou (2011), Couwenberg and de Jong (2008) or Bris, Welch and Zhu (2006).

115

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES







The last model which analyzes influencing factors on firm recovery rates solely includes variables with a significant impact on firm recovery rates:

(5)



Furthermore, two other models examine whether there are different results for continuations and liquidations respectively. Therefore, the data set is divided into two subsamples and potential influencing factors are tested on the basis of model (4). V.2.b.3 Correlation Analyses Before investigating the hypotheses by a multiple regression analysis, the dependencies between the potential influencing factors and the firm recovery rate are tested by correlation analysis as well as different tests of significance. As mentioned in chapter V.2.b.2 the relationship between two variables on interval scale are examined by the Pearson correlation coefficient. Table 28 illustrates the initial results:

Table 28: Correlation Analysis between Firm RR and assumed Influencing Factors Variables H1&H2 H1 H2 H3 H4 H5

Secured Bank Debt Secured Other Debt Unsecured Bank Debt Debt over Assets Liquidity Gap No. Creditors x Committee

Obs 85 85 85 77 85 92

Pearson´s correlation  coefficient 0.5570 *** 0.3197 *** ‐0.1394 ‐0.2070 * ‐0.1555 0.1059

Note: ***, ** and * denote significance at the 1%, 5% and 10% level.

Spearman´s correlation  coefficient 0.5861 *** 0.4288 *** ‐0.0789 ‐0.7186 *** ‐0.4205 *** 0.2657 **

116

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

The results show that there is a linear relationship between firm recovery rate and collateralization as well as between firm recovery rate and over-indebtedness, measured by debt over assets. Thus, there is initial evidence that hypotheses H1 and H3 might be confirmed. Both variables concerning the quota of secured debt to total debt are positively related to the firm recovery rate. The value for the Pearson correlation coefficient for secured debt held by banks is 0.557. The correlation coefficient concerning the relationship between firm recovery rate and secured debt held by non-bank institutions is slightly lower (0.3197). Both coefficients are significant at the 1% level. Hypotheses H3 assumes that high debt over asset ratios are related to low firm recovery rate. Pearson´s correlation coefficient confirms this assumption with a value of -0.2070. The coefficient is statistically significant at the 10% level. For all other Pearson coefficients, illustrated in table 28, a significant linear relationship cannot be observed. The Spearman´s correlation coefficient confirms the results for hypotheses H1 and H3. Furthermore the findings show that the coefficients for hypotheses 4 and 5 are significant. However, the relationship between the firm recovery rate and the interaction term, which reflects the influence of creditors, seems to be positive instead of negative. Hypotheses H6 and H7 concern the relationship between the extent of the firm recovery rate and two different dummy variables. Testing for point-biserial correlation shows that there is no statistically significant relationship between the firm recovery rate and (accepted or rejected) debtor-in-possession management. In addition, the coefficient between the firm recovery rate and the petitioner (debtor or creditor filing) is also not significant. The results are reported in table 29:

Table 29: Test for Significance and Point-Biserial Correlation Variables H6 H7

Management Accepted Petitioner

Obs

PB Correlation

95 93

0.03 0.14

Note: ***, ** and * denote significance at the 1%, 5% and 10% level.

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

117

For several reasons, the results of bivariate correlation analyses are restricted and have to be expanded by a multiple regression analysis. The reasons concern different third variable effects, like mediation, confounding or suppression.291 The consideration behind is that an additional variable has an impact on the relationship between a dependent and an independent variable. For example, the mediator effect implies that there is a connection between an independent variable and a dependent variable through a mediator variable. In other words, both variables have a combined effect on the dependent variable. In contrast to a confounder variable, a mediator variable implies a causal relationship. Both effects describe a reduction of the relationship between the dependent and the independent variable. On the other hand, the suppression effect gives an explanation for an increased extent of relationship between two variables. Including a third variable into the regression improves the prediction. Bivariate analyses are not able to examine such effects. V.2.b.4 OLS Regressions The combined influence of all potential influencing factors on the firm recovery rate is tested by a multiple regression analysis. As illustrated in table 30, five different regression models are tested. The first model (1) includes several firm characteristics. As a result, only collateralization has a significant impact on the firm recovery rate. A high level of collateralization results in a higher recovery rate, regardless of which type of secured creditor (bank vs. non-bank creditors) is involved. Just like in several other empirical studies, the first hypothesis is confirmed.292

291

See for example McKinnon, Krull and Lockwood (2000), pp. 173-175.

292

See Bris, Welch and Zhu (2006), Thorburn (2000) or Blazy, Chopard and Nigam (2013).

+ + + +

91 0.2140 0.1775

77 0.4489 0.3841

0.060 * 0.427 0.349 0.046 **

0.000 *** 0.002 *** 0.568 0.725 0.253 0.104 0.029 ** 0.031 **

76 0.5636 0.4805

0.12441 0.500

0.08433 0.058 *

0.19160 0.329

‐0.00015 0.746 ‐0.06751 0.362 0.05763 0.207

0.70842 0.58160 ‐0.04619 ‐0.00003 ‐0.11117 ‐0.00084 ‐0.15432 ‐0.09682

(3)

0.19771 0.000 *** 0.08233 0.01171 0.04649 0.08481

0.000 *** 0.002 *** 0.884 0.613 0.333

(2)

0.00101 0.943 0.00801 0.872 0.05443 0.209

0.65817 0.60331 ‐0.01210 ‐0.00005 ‐0.09932

(1)

0.071 * 0.413 0.351 0.049 ** 0.945 0.794 0.542

0.000 *** 0.002 *** 0.533 0.735 0.293 0.110 0.030 ** 0.033 **

76 0.5641 0.4641

0.08146 0.01250 0.04833 0.08547 0.04761 ‐0.34230 0.11624

0.70690 0.57709 ‐0.05508 ‐0.00003 ‐0.10590 ‐0.00089 ‐0.15740 ‐0.09756

(4)

0.08 *

85 0.5522 0.5178

0.07203 0.055 *

0.06396

0.10622 0.004 ***

‐0.09844 0.105 * ‐0.06223 0.098 *

0.69255 0.000 *** 0.71232 0.000 ***

(5)

Note: The table shows the hypotheses and regression results (OLS regressions) for the firm recovery rate. All influencing factors are defined in table 17. ***, ** and * denote significance at the 1%, 5% and 10% level.

Observations R² Adjusted R²

Survival lnTotalAssets LegalForm ReasonManagement lnGDP lnIfo Constant

SecBankDebt SecOtherDebt UnsecBankDebt DebtAssets LiquidityGap NoCreditorsXCommittee ManagementAccepted Petitioner

Hypothesen

Dependent variable: Firm Recovery Rate

Table 30: Regression Results - Firm Recovery Rate

118 ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

119

The second hypothesis concerns relationship banking which plays a significant role in the German banking system. Nevertheless, the impact on firm recovery rates is ambiguous. To confirm the hypothesis, both variables, SecBankDebt and UnsecBankDebt, should show a positive and significant effect. This is not the case. A potential explanation might be that the sample solely includes formal insolvency cases. However, it can be assumed that relationship banks can identify financial distress earlier than other creditors due to closer monitoring. Then banks try to conduct reorganization in an informal procedure in a first step to save the costs of insolvency. Those firms that still have to file for bankruptcy frequently failed to reorganize together with their banks earlier. As a result, relationship banking no longer plays a significant role at the stage of the procedure considered in this study. Literature about relationship banking and the influence on firm recovery rates is rare. Couwenberg and de Jong (2008) analyze the proportion of bank debt to total debt and find a significant and positive impact on firm recovery rates. The more differentiated analysis presented in this study cannot confirm the assumed relationship. The next two hypotheses refer to over-indebtedness and illiquidity. Neither variable shows a significant impact on firm recovery rate. One possible explanation for the result concerning debt-overload might be that a substantial amount can be obtained within the procedure. The value of assets at the beginning of bankruptcy thereby underestimates the payable amount at the end of bankruptcy. In contrast to existing studies, which also assume a negative relationship, the third hypothesis cannot be confirmed for Germany.293 In terms of illiquidity, measured by the variable LiquidityGap, further investigations are required. For example, it seems to be possible that illiquidity may have an influence on the extent of recovery rates when financial problems are the reason for insolvency. In literature, only Couwenberg and de Jong (2008) include a proxy for illiquidity in their analysis. In contrast to this study about German cases, Couwenberg and de Jong confirm a significant impact on the firm recovery rate. All other factors included in the first model are control variables referring to firm characteristics. So far, none of the factors show a significant impact on the firm recovery rate. To test a further set of variables, a second model (2) is established which includes solely procedural characteristics. Accordingly, hypotheses 5, 6 and 7 can be tested without the influence of firm-specific variables. In isolation, none of the three variables has a significant impact. Initially, this indicates that neither creditors nor shareholders are able to influence the extent of the recovery rate. The influence of 293

See for example Couwenberg and de Jong (2008) or Bris, Welch and Zhu (2006).

120

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

different stakeholders on the procedure has rarely been investigated in literature so far. None of the existing studies examines a combination of different variables concerning the power of stakeholders. Only the control variable (Survival) shows significant results. As in several empirical studies, firm recovery rates are higher in continuation than in liquidation.294 The third model (3) combines firm-specific and procedural variables. According to the increased adjusted R², the third model has the highest descriptive power. The results for hypotheses 1, 2, 3, 4 and 5 are robust. In contrast to the second model, hypotheses 6 and 7 can be confirmed. Accordingly, in case of debtor-in-possession management (H6), the firm recovery rate is lower than in firms managed by an administrator. This result confirms the assumption that despite good firm quality signaled by debtor-in-possession management, creditors receive less. One reason might be that shareholders have the power to strengthen their own position at the expense of creditors. According to the continuation of (viable) firms, which is also the aim of the amendments to German insolvency law, this might be appropriate. However, legislation has to protect repayments to creditors too. The influence of debtor-in-possession management on the extent of firm recovery rates has not been investigated in literature until now. The main reason is that in other countries like the US, debtor-in-possession management is the normal case and not the exception like in Germany. Hypothesis 7 concerns the petitioner of bankruptcy. In contrast to the hypothesis and the results by Bris, Welch and Zhu (2006), this analysis finds that a bankruptcy petition by a debtor results in lower recovery rates. An explanation for this result might be that debtors are more likely to file when their renegotiation power is high. In this case, the APRD would be higher and recovery rates would therefore decline. Strategic defaults are thoroughly investigated for the US market. However, due to the more creditor-friendly regime in Germany, strategic defaults are less likely in this sample. Among the control variables, two proxies show significant results. Like in model (2), firms which continue after the procedure show higher firm recovery rates. A second variable indicates that when the reason for bankruptcy is a failure of the management (in contrast to all other possible reasons), firm recovery rates increase. One explanation might be that it is much more difficult to solve other problems like a loss of receivables or a sales problem. To some extent, literature confirms this relationship.295

294

See Couwenberg and de Jong (2008) or Thorburn (2000).

295

See Blazy, Chopard and Nigam (2013).

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

121

Model (4) repeats the third regression but controls for two different macroeconomic variables. The results of the third regression are robust. As a result, hypotheses 1, 6 and 7, which refer to collateralization as well as the influence of debtors and creditors, can be confirmed. Furthermore, continuation after the procedure and the reason for insolvency are decisive for the extent of firm recovery rates. Neither the annual growth rate of the GDP nor the business climate index affect the extent of recovery rates. Model (5) includes only significant variables. The findings show that the former results are robust. Based on the results presented above, another analysis investigates the influencing factors on firm recovery rates in continuations and liquidations respectively. Therefore, the sample is separated into two subsamples. Overall, 52 continuations and 24 liquidations are examined.

Table 31: Influencing Factors on Firm RR in Continuations and Liquidations Dependent variable: Firm Recovery Rate Hypothesen

Continuation

Liquidation

SecBankDebt SecOtherDebt UnsecBankDebt DebtAssets LiquidityGap NoCreditorsXCommittee ManagementAccepted Petitioner

+ + + +

0.867901 0.664479 ‐0.025097 ‐0.000281 ‐0.093266 ‐0.001646 ‐0.195224 ‐0.155760

0.000 *** 0.004 *** 0.861 0.264 0.45 0.038 ** 0.029 ** 0.027 **

0.485693 0.251726 ‐0.074066 0.000004 ‐0.423489 0.000105

lnTotalAssets LegalForm ReasonManagement lnGDP lnIfo Constant

+ -

0.020542 0.094902 0.180497 0.297902 ‐0.881183 0.098584

0.369 0.168 0.014 ** 0.77 0.631 0.73

0.023063 ‐0.037640 0.014592 ‐0.560373 ‐0.565661 0.352077

Observations R² Adjusted R²

52 0.5459 0.3905

0.001 ** 0.548 0.197 0.925 0.187 0.845

‐0.084496 0.025 ** 0.069 * 0.360 0.718 0.240 0.687 0.298

24 0.8964 0.7833

Note: The table shows the hypotheses and regression results (OLS regressions) for the firm recovery rate for continuations and liquidations separately. All influencing factors are defined in table 17. ***, ** and * denote significance at the 1%, 5% and 10% level.

122

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

Based on firms that continue after the proceedings, the results of the entire sample can be confirmed (see table 31). Again, the quota of collateral (hypothesis 1) and the influence of debtors and creditors (hypotheses 6 and 7) on the extent of recovery rates show significant results. Besides hypotheses 1, 6 and 7, a further factor can be confirmed. The interaction term, which refers to the number of creditors and the establishment of a committee, is negatively correlated to the extent of the firm recovery rate. Therefore, it becomes clear that when no committee is established, high numbers of creditors lead to lower recovery rates. As a result, hypothesis 5 is confirmed for firms which continue after the formal proceedings. Unfortunately the data sample does not include firms which file for bankruptcy after the amendments to German insolvency law in 2012/2013. One aim of the revision is to strengthen the position of creditors by supporting the establishment of a creditors´ committee. The negative effect of high numbers of creditors on the recovery rate might, therefore, diminish under the new legislation. In line with the results of the whole sample, the continuation sample shows a positive and significant impact of a management failure on the extent of the recovery rate. In case of the liquidation sample, the results are slightly different (see table 31). The relationship between the firm recovery rate and secured bank debt is still significant. Otherwise, the proportion of secured debt financed by non banks does not generate higher recovery rates. In other words: in case of liquidations it is favorable for all creditors when a high proportion of secured debt is held by banks. Nevertheless, the other factor concerning relationship banking, UnsecBankDebt, remains not significant. Accordingly, the first and the second hypotheses cannot be confirmed without restrictions. As for the whole sample and the continuation sample, hypotheses three and four are not significant. In contrast to the continuation sample, the interaction term is not significant. This shows that establishing a committee in case of a high number of creditors has no effect in liquidations. One possible explanation might be that creditors do not expect high payments in liquidations. Therefore, they do not try to influence the proceedings. Among the procedural characteristics, the person who files for bankruptcy is a relevant influencing factor. Again, for liquidations the petitioner is still significant for the level of firm recovery rates. In addition, one control variable shows a significant result. The size of the firm, measured by the amount of total assets, is statistically significant. The only study investigating the size of the firm for liquidations is the one by Blazy, Chopard, Fimayer and Guigou (2011). In contrast to their result, which shows no significant relationship, the coefficient is positive for Germany. This indicates that creditors of large firms receive higher recovery rates. A possible reason might be that firm

ANALYSIS OF GERMAN REORGANIZATION PROCEDURES

123

size is a proxy for the amount repayable. As long as the amount repayable is high, the recovery rate increases. All other variables do not have a significant impact on firm recovery rate. In contrast to the other subsamples the reason for insolvency does not have any effect.

CONCLUSION

125

VI Conclusion Up to now, empirical research by formal and informal reorganizations has been limited mainly by the availability and quality of data. Studies concerning informal workouts are mainly driven by banks which provide their private data sets to researchers. Data sets of banks are typically well-structured. Additionally, informal workouts by banks are quite common so that the number of cases is high enough for a profound analysis. In contrast, formal reorganizations in Germany are still rare and the quality and availability of data is poor. For this reason, empirical studies in this field are more superficial in comparison to informal workouts and formal reorganizations in the US. This study analyzes a broad data set of formal reorganizations in Germany that includes both firm and procedural characteristics to provide a more comprehensive investigation. Starting with the theoretical economics of bankruptcies, literature provides two major criteria for insolvency law, ex-post and ex-ante efficiency. Ex-post efficiency claims that insolvency law should ensure that viable firms are continued and dilapidated firms should be put into liquidation. Therefore insolvency law needs to deal with different incentives of stakeholders who want to maximize their own return instead of the expected value of the entire firm. Studies analyzing ex-post efficiency focus mainly on US bankruptcy rules. However, the results are contradictory. Ex-ante efficiency postulates that insolvency laws should not negatively affect financing or investment decisions of healthy firms. Again, most existing studies analyze US insolvency law, focusing on the effects of deviations from the absolute priority rule. Literature provides theoretical evidence for both, positive and negative effects of the APRD on corporate finance decisions, risk taking and other investigated aspects. However, to a certain degree ex-ante and ex-post efficiency seem to contradict each other, which leads to a trade-off. Debtor-oriented insolvency laws such as Chapter 11 increase ex-post efficiency but often set negative incentives for the management of healthy firms. The German insolvency law that focuses on creditor protection, disciplines mangers of healthy firms but at the same time can lead to the liquidation of viable firms in bankruptcy. To provide a deeper insight, this study discusses the main differences between US and German bankruptcy law. Nevertheless different country specific factors, like the design of credit markets, can help to counteract undesirable effects. A change in insolvency legislation has to consider these country specificities.

126

CONCLUSION

Literature provides evidence that the extent of recovery rates for different creditor groups is associated with the choice of a specific insolvency procedure (e.g. continuation or liquidation) as well as the considered country. A closer examination of influencing factors on group specific recovery rates shows that several firm characteristics, procedural characteristics, macroeconomic variables as well as variables concerning the terms of credit or the business connection between bank and borrower have a significant impact. However, the same analyzed influencing factors frequently show contradictory results for different data sets. Therefore, the results are not applicable to different countries and procedures. For Germany, recovery rates for different creditor groups are barely investigated, especially within formal procedures. This study provides further insights into the extent of recovery rates for different creditor groups as well as influencing factors on firm recovery rates for firms which pass an insolvency plan in Germany. Based on 102 case files, the firm recovery rate for continuations in Germany equals 24% on average. A closer view on secured and unsecured creditors shows that secured creditors receive 66% on average while unsecured creditors receive 11% on average. Influencing factors on the firm recovery rate in Germany are tested in this study by different correlation coefficients and multiple regression analyses. The empirical investigation shows that firm recovery rate is positively related to collateralization. Furthermore, the impact of the petitioner for bankruptcy as well as accepted debtor-in-possession management on firm recovery rates is significant. Firm recovery rates decrease as long as debtors instead of creditors file for bankruptcy and as long as the debtor instead of an administrator is allowed to manage the firm after default. In addition, two control variables show significant results. Accordingly, firm recovery rates are higher for firms which continue after the filing for bankruptcy as well as in case of a failure of management instead of all other possible insolvency reasons. Two further regression analyses investigate influencing factors on firm recovery rates for firms which continue after the procedure in contrast to firms which are liquidated after bankruptcy. The results for continuations are the same as for the whole data set. Only one further variable shows significant results. It therefore becomes clear that high numbers of creditors lead to lower firm recovery rates, as long as no creditor committee is formed. In case of the liquidation subsample, the findings are slightly different. However, firm recovery rates still increase with the proportion of secured bank debt and decrease in case of a bankruptcy petition by the debtor. In contrast to the whole data sample, creditors of firms which end up in liquidation receive higher recovery rates in case of a large firm, measured by the amount of total assets.

CONCLUSION

127

Overall the study promotes the understanding of formal insolvency procedures in Germany especially about the firm recovery rate and its influencing factors. However, the results are strongly dependent on current legal rules. The data set contains insolvency cases before the law reform in 2012. Due to the new rules, which are more debtor friendly than the analyzed legislation in this study, further investigations are required. However, up to now the data set of completed bankruptcy procedures after the amendments is still too small.

REFERENCES

129

References Acharya, V. V.; Bharath, S. T. and Srinivasan, A. (2007): Does Industry-Wide Distress Affect Defaulted Firms? Evidence from Creditor Recoveries, in: Journal of Financial Economics, Vol. 85, pp. 787–821. Aghion, P.; Hart, O. and Moore, J. (1992): The Economics of Bankruptcy Reform, in: Journal of Law, Economics, & Organization, Vol. 8 (3), pp. 523-546. Altman, E. and Jha, S. (2003): Market Size and Investment Performance of Defaulted Bonds and Bank Loans: 1987-2002, Working Paper, NYU Stern Department of Finance Working Paper Series, pp. 1-46. Annabi, A.; Breton, M. and François, P. (2011): Game Theoretic Analysis of Negotiations under Bankruptcy. Available at: http://ssrn.com/abstract=1856255, pp. 1-27. Araten, M.; Jacobs Jr., M. and Varshney, P. (2004): Measuring LGD on Commercial Loans: An 18 Year Internal Study, in: The RMA Journal, Vol. 86 (8), pp. 96-103. Armour, J.; Hsu, A. and Walters, A. (2006): The Costs and Benefits of Secured Creditor Control in Bankruptcy: Evidence from the UK, Working Paper No. 332, Centre for Business Research, University Of Cambridge, pp. 1-36. Armour, J.; Hsu, A. and Walters, A. (2012): The Costs and Benefits of Secured Creditor Control in Bankruptcy: Evidence from the UK, in: Review of Law and Economics, Vol. 8 (1), pp. 102-135. Baird, D. G. (1991): The Initiation Problem in Bankruptcy, in: International Review of Law and Economics, Vol. 11, pp. 223-232. Baird, D. G. and Bernstein, D. S. (2006): Absolute Priority, Valuation Uncertainty, and the Reorganization Bargain, in: The Yale Law Journal, Vol. 115, pp. 1930-1970. Baird, D.; Bris, A. and Zhu, N. (2007): The Dynamics of Large and Small Chapter 11 Cases: An Empirical Study, in: American Law & Economics Association Annual Meetings, No. 2, pp. 1-36. Bartlett, F. (2000): Secured Loan Recovery Study: The UK Experience, Europe Loan Products, Fitch IBCA. Bebchuk, L. A. (1988): A New Approach to Corporate Reorganizations, in: Harvard Law Review, Vol. 101, pp. 775-804.

130

REFERENCES

Bebchuk, L. A. (2000): Using Options to Divide Value in Corporate Bankruptcy, in: European Economic Review, Vol. 44, pp. 829-843. Bebchuk, L. A. (2002): Ex Ante Costs of Violating Absolute Priority in Bankruptcy, in: The Journal of Finance, Vol. 57 (1), pp. 445-460. Bebchuk, L. A. and Chang, H. F. (1992): Bargaining and the Division of Value in Corporate Reorganization, in: Journal of Law, Economics, & Organization, Vol. 8, (2), pp. 253-279. Bebchuk, L. A. and Fried, J. M. (1996): The Uneasy Case for the Priority of Secured Claims in Bankruptcy, in: The Yale Law Journal, Vol. 105, pp. 857-934. Bergman, Y. Z. and Callen, J. L. (1995): Rational Deviations from Absolute Priority Rules, in: International Review of Financial Analysis, Vol. 4 (1), pp. 1-18. Berkovitch, E. and Israel, R. (1998): The Bankruptcy Decision and Debt Contract Renegotiations, in: European Finance Review, Vol. 2, pp. 1–27. Berkovitch, E. and Israel, R. (1999): Optimal Bankruptcy Laws Across Different Economic Systems, in: The Review of Financial Studies, Vol. 12 (2), pp. 347-377. Berkovitch, E.; Israel, R. and Zender, J. F. (1997): Optimal Bankruptcy Law and Firm-Specific Investments, in: European Economic Review Vol. 41, pp. 487-497. Berkovitch, E.; Israel, R. and Zender, J. F. (1998): The Design of Bankruptcy Law: A Case for Management Bias in Bankruptcy Reorganizations, in: The Journal of Financial and Quantitative Analysis, Vol. 33 (4), pp. 441-464. Bester, H. (1994): The Role of Collateral in a Model of Debt Renegotiation, in: Journal of Money, Credit and Banking, Vol. 26 (1), pp. 72-86. Betker, B. L. (1995): Management's Incentives, Equity's Bargaining Power, and Deviations from Absolute Priority in Chapter 11 Bankruptcies, in: The Journal of Business, Vol. 68, (2), pp. 161-183. Bigus, J. (2002): Bankruptcy Law, Asset Substitution Problem, and Creditor Conflicts, in: International Review of Law and Economics, Vol. 22, pp. 109-132. Blazy, R. and Chopard, B. (2004): Ex Post Efficiency of Bankruptcy Procedures: A General Normative Framework, in: International Review of Law and Economics, Vol. 24, pp. 447-471. Blazy, R.; Chopard, B.; Fimayer, A. and Guigou, J.-D. (2011): Employment Preservation vs. Creditors’ Repayment under Bankruptcy Law: The French

REFERENCES

131

Dilemma, in: International Review of Law and Economics, Vol. 31, pp. 126-141. Blazy, R.; Chopard, B. and Nigam, N. (2013): Building Legal Indexes to Explain Recovery Rates: An Analysis of the French and English Bankruptcy Codes, in: Journal of Banking & Finance, Vol. 37 (6), pp. 1936-1959. Blazy, R.; Petey, J. and Weill, L. (2012): Can Bankruptcy Codes Create Value? Evidence from Creditors’ Recoveries in France, Germany, and the United Kingdom, Working Paper, pp. 1-48. Available at: http://ssrn.com/abstract=2145849. Böttger, M.; Guthoff, A. and Heidorn, T. (2008): Loss Given Default Modelle zur Schätzung von Recovery Rates, Working Paper No. 96, Frankfurt School – Working Paper Series, pp. 1-49. Bolton, P. and Scharfstein, D. S. (1996): Optimal Debt Structure and the Number of Creditors, in: The Journal of Political Economy, Vol. 104 (1), pp. 1-25. Bradley, M. and Rosenzweig, M. (1992): The Untenable Case for Chapter 11, in: The Yale Law Journal, Vol. 101, pp. 1043-1095. Braham, M. and Steffen, F. (2003): Voting Rules in Insolvency Law: A SimpleGame Theoretic Approach, in: International Review of Law and Economics Vol. 22, 421–442. Bris, A.; Welch, I. and Zhu, N. (2006): The Costs of Bankruptcy: Chapter 7 Liquidation versus Chapter 11 Reorganization, in: The Journal of Finance, Vol. 61 (3), pp. 1253–1303. Brown, D.T. (1989): Claimholder Incentive Conflicts in Reorganization: The Role of Bankruptcy Law, in: The Review of Financial Studies, Vol. 2 (1), pp.109-123. Burghof, H-P. (1998): Die ökonomische Funktion von Nennwertaktien, in: Die Betriebswirtschaft, Vol. 58 (4), pp. 501-515. Carapeto, M. (2007): An Examination of Multiple Plans in Chapter 11 Reorganizations, in: Journal of Financial Stability, Vol. 3, pp. 279–293. Carty, L. V.; Hamilton, D. T. and Moss, A. (1999): Bankrupt Bank Recovery Rates, in: The Journal of Lending and Credit Risk Management, June 1999, pp. 20-26. Citron, D.; Wright, M.; Ball, R. and Rippington, F. (2003): Secured Creditor Recovery Rates from Management Buy-outs in Distress, in: European Financial Management, Vol. 9 (2), pp. 141–161.

132

REFERENCES

Cornelli, F. and Felli, L. (1997): Ex-Ante Efficiency of Bankruptcy Procedures, in: European Economic Review, Vol. 41, pp. 475-485. Couwenberg, O. and de Jong, A. (2008): Costs and Recovery Rates in The Dutch Liquidation-Based Bankruptcy System, in: European Journal of Law and Economics, Vol. 26, pp. 105-127. Davydenko, S. A. and Franks, J. R. (2008): Do Bankruptcy Codes Matter? A Study of Defaults in France, Germany and the UK, in: The Journal of Finance, Vol. 63 (2), pp. 565-608. Dermine, J. and Neto de Carvahlho, C. (2006): Bank Loan Losses-GivenDefault: A Case Study, in: Journal of Banking & Finance, Vol. 30, pp. 1219–1243. Dormann, C. (2013): Parametrische Statistik: Verteilungen, Maximum Likelihood und GLM in R, Springer Verlag, Berlin/Heidelberg. Eales, R. and Bosworth, E. (1998): Severity of Loss in the Event of Default in Small Business and Larger Consumer Loans, in: The Journal of Lending and Credit Risk Management, May 1998, pp. 58–65. Eberhart, A. C.; Moore, W. T. and Roenfeldt, R. L. (1990): Security Pricing and Deviations from the Absolute Priority Rule in Bankruptcy Proceedings, in: The Journal of Finance, Vol. 45, (5), pp. 1457-1469. Franks, J.; de Servigny, A. and Davydenko, S. (2004): A Comparative Analysis of the Recovery Process and Recovery Rates for Private Companies in the UK, France, and Germany, in: Standard and Poor‘s Risk Solutions Report, May 2004, pp. 1-93. Franks, J. and Sussman, O. (2005): Financial Distress and Bank Restructuring of Small to Medium Size UK Companies, in: Review of Finance, Vol. 9, pp. 65–96. Franks, J. R. and Torous, W. N. (1989): An Empirical Investigation of U.S. Firms in Reorganization, in: The Journal of Finance, Vol. 44 (3), pp. 747-769. Franks, J. R. and Torous, W. N. (1994): A Comparison of Financial Recontracting in Distressed Exchanges and Chapter 11 Reorganizations, in: Journal of Financial Economics, Vol. 35 (3), pp. 349–370. Gertner, R. and Scharfstein, D. (1991): A Theory of Workouts and the Effects of Reorganization Law, in: The Journal of Finance, Vol. 46 (4), pp. 1189-1222.

REFERENCES

133

Grunert, J. and Weber, M. (2009): Recovery Rates of Commercial Lending: Empirical Evidence for German Companies, in: Journal of Banking & Finance, Vol. 33, pp. 505-513. Gupton, G. M.; Gates, D. and Carty, L. V. (2000): Bank-Loan Loss Given Default, Moody’s Investors Service, Global Credit Research, November 2000, pp. 69-92. Hansen, R. G. and Thomas, R. S. (1998): Auctions in Bankruptcy: Theoretical Analysis and Practical Guidance, in: International Review of Law and Economics, Vol. 18, pp. 159 –185. Harner, M. M. and Marincic, J. (2011): Committee Capture? An Empirical Analysis of the Role of Creditors´ Committees in Business Reorganization, in: Vanderbilt Law Review, Vol. 64 (3), pp. 749-810. Harris, M. and Raviv, A. (1991): The Theory of Capital Structure, in: The Journal of Finance, Vol. 46 (1), pp. 297-355. Hart, O. (2006): Different Approaches to Bankruptcy. CESifo DICE Report, Ifo Institute for Economic Research, Vol. 4 (1), pp. 3–8. Heeseler, M. (2002): Kompensation der Kostenbeiträge gem. §§ 170, 171 InsO durch den Verwertungserlös, in: ZInsO, Vol. 19, pp. 924-926. Heinkel, R. and Zechner, J. (1993): Financial Distress and Optimal Capital Structure Adjustments, in: Journal of Economics & Management Strategy, Vol. 2 (4), pp. 531-565. Hotchkiss, E. S., John, K., Mooradian, R. M. and Thorburn, K. S. (2008): Bankruptcy and the Resolution of Financial Distress, pp. 1-89. Available at: http://ssrn.com/abstract=1086942. Hotchkiss, E. S. and Mooradian, R. M. (2003): Auctions in Bankruptcy, in: Journal of Corporate Finance, Vol. 9, pp. 555-574. Hurt, L. and Felsovalyi, A. (1998): Measuring Loss on Latin American Defaulted Bank Loans: A 27-Year Study of 27 Countries, pp. 1-8. Available at: http://ssrn.com/abstract=128672. Icks, A. and Kranzusch, P. (2010): Sanierungen in Insolvenzverfahren - übertragende Sanierungen und insolvenzplanbasierte Eigensanierungen in NRW, IfM-Materialien, Nr. 195, pp. 1-145. Ingermann, P.-H.; Hesse, F.; Bélorgey, C. and Pfingsten. A. (2013): Explaining the Recovery Rate for Retail and Commercial Customers in Germany with a Particular Focus on the Collateral, Working Paper, pp. 1-38.

134

REFERENCES

Ivashina, V.; Smith, D. C.; Iverson, B. and Strömberg, P. (2011): The Ownership and Trading of Debt Claims in Chapter 11 Restructurings, Working Paper, pp. 1-44. Jensen, M. C. and Meckling, W. H. (1976): Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, in: Journal of Financial Economics, Vol. 3 (4), pp. 305-360. Kanzlei Schultze & Braun (2013): Insolvenzplanindex 1999-2013. Available at: www.schubra.de/de/veroeffentlichungen/insolvenzstatistiken/Insolvenzplanin dex1999bis2013.pdf. Khieu, H. D.; Mullineaux, D. J. and Yi, H.-C. (2012): The Determinants of Bank Loan Recovery Rates, in: Journal of Banking and Finance, Vol. 36, pp. 923–933. Kranzusch, P. and Icks, A. (2009): Die Quoten der Insolvenzgläubiger in Regel und Insolvenzplanverfahren - Ergebnisse von Insolvenzverfahren nach der Insolvenzrechtsreform, IfM-Materialien, Nr. 186, pp. 1-50. Leuz, C.; Deller, D. and Stubenrath, M. (1998): An International Comparison of Accounting-Based Payout Restrictions in the United States, United Kingdom and Germany, in: Accounting and Business Research, Vol. 28 (2), 1998, pp. 111-129. Longhofer, S. D. and Carlstrom, C. T. (1995): Absolute Priority Rule Violations in Bankruptcy, in: Economic Review Federal Reserve Bank Cleveland, pp. 21-30. McKinnon, D.P.; Krull, J. L. and Lockwood. C.M. (2000): Equivalence of the Mediation, Confounding and Suppression Effect, in: Prevention Science, Vol. 1 (4), pp. 173-181. Mooradian, R. M. (1994): The Effect of Bankruptcy Protection on Investment: Chapter 11 as a Screening Device, in: The Journal of Finance, Vol. 49 (4), pp. 1403-1430. Mora, N. (2013): Creditor Recovery: The Macroeconomic Dependence of Industry Equilibrium, Working Paper, The Federal Reserve Bank of Kansas City Research Working Papers RWP 13-06, pp. 1-44. Picker, R. C. and Bebchuk, L. A. (1993): Bankruptcy Rules, Managerial Entrenchment, and Firm-Specific Human Capital, Working Paper No. 16, Program in Law and Economics, pp. 1-22.

REFERENCES

135

Pindado, J.; Rodrigues, L. and de la Torre, C. (2008): How do Insolvency Codes Affect a Firm’s Investment?, in: International Review of Law and Economics, Vol. 28, pp. 227–238. Ravid, S. A. and Sundgren, S. (1998): The Comparative Efficiency of Small-Firm Bankruptcies: A Study of the US and Finnish Bankruptcy Codes, in: Financial Management, Vol. 27 (4), pp. 28-40. Schuermann, T. (2004): What Do We Know About Loss Given Default?, in: Wharton Financial Institutions Center, Working Paper No. 04-01, pp. 1-29. Available at: http://ssrn.com/abstract=525702. Schulze, P.M. (2003): Beschreibende Statistik, Oldenbourg, München. Shleifer, A. and Vishny, R. W. (1992): Liquidation Values and Debt Capacity: A Market Equilibrium Approach, in: Journal of Finance, Vol. 47, pp. 1343-1366. Statistisches Bundesamt (2014a): Insolvenzen; Available at: www.destatis.de/ DE/ZahlenFakten/Indikatoren/LangeReihen/Insolvenzen/lrins01.html;jsessio nid=7751B6AE41429C7937EF4C8A0FDB1766.cae4?cms_gtp=152398_list %253D2&https=1. Statistisches Bundesamt (2014b): Volkswirtschaftliche Gesamtrechnung. Available at: www.destatis.de/DE/ZahlenFakten/Indikatoren/LangeReihen/ VolkswirtschaftlicheGesamtrechnungen/lrvgr02.html?cms_gtp=152422_list%253D2&https=1. Sundgren, S. (1998): Does a Reorganization Law Improve the Efficiency of the Insolvency Law? The Finnish Experience, in: European Journal of Law and Economics, Vol. 6, pp. 177–198. Tashjian, E.; Lease, R. C. and McConnell, J. J. (1996): An Empirical Analysis of Prepackaged Bankruptcies, in: Journal of Financial Economics, Vol. 40, pp.135-162. Thorburn, K. S. (2000): Bankruptcy Auctions: Costs, Debt Recovery, and Firm Survival, in: Journal of Financial Economics, Vol. 58 (3), pp. 337–368. Toutenburg, H. and Heumann, C. (2009): Deskriptive Statistik: Eine Einführung in Methoden und Anwendungen mit R und SPSS, 7. Auflage, Springer Verlag, Berlin, Heidelberg. Van de Castle, K.; Keisman, D. and Yang, R. (2000): Suddenly Structure Mattered: Insights into Recoveries of Defaulted Debt, in: Standard & Poor’s Credit Week, May Issue (2000), pp. 61-68.

136

REFERENCES

Vychodil, O. (2010): Essays on Corporate Bankruptcy, Center for Economics Research and Graduate Education, Charles University Prague. Available at: www.cerge-ei.cz/pdf/dissertations/2010-vychodil.pdf, pp. 1-123. Weiss, L. A. (1990): Bankruptcy Resolution. Direct Costs and Violation of Priority of Claims, in: Journal of Financial Economics, Vol. 27, pp. 285-314. White, M. J. (1989): The Corporate Bankruptcy Decision, in: The Journal of Economic Perspectives, Vol. 3 (2), pp. 129-151. White, M. J. (1994): Corporate Bankruptcy as a Filtering Device: Chapter 11 Reorganizations and Out-of-Court Debt Restructurings, in: Journal of Law, Economics, & Organization, Vol. 10 (2), pp. 268-295. Wooldridge, J. R. (2009): Introductory Econometrics. A Modern Approach, South Western, Cengage Learning.

Studienreihe der Stiftung Kreditwirtschaft an der Universität Hohenheim Bände 1 - 11 sind nicht mehr lieferbar. Band 12: Axel Tibor Kümmel: Bewertung von Kreditinstituten nach dem Shareholder Value Ansatz, 1994; 2. Aufl.; 1995. Band 13: Petra Schmidt: Insider Trading. Maßnahmen zur Vermeidung bei US-Banken; 1995. Band 14: Alexander Grupp: Börseneintritt und Börsenaustritt. Individuelle und institutionelle Interessen; 1995. Band 15: Heinrich Kerstien: Budgetierung in Kreditinstituten. Operative Ergebnisplanung auf der Basis entscheidungsorientierter Kalkulationsverfahren; 1995. Band 16: Ulrich Gärtner: Die Kalkulation des Zinspositionserfolgs in Kreditinstituten; 1996. Band 17: Ute Münstermann: Märkte für Risikokapital im Spannungsfeld von Organisationsfreiheit und Staatsaufsicht; 1996. Band 18: Ulrike Müller: Going Public im Geschäftsfeld der Banken. Marktbetrachtungen, bankbezogene Anforderungen und Erfolgswirkungen; 1997. Band 19: Daniel Reith: Innergenossenschaftlicher Wettbewerb im Bankensektor; 1997. Band 20: Steffen Hörter: Shareholder Value-orientiertes Bank-Controlling; 1998. Band 21: Philip von Boehm-Bezing: Eigenkapital für nicht börsennotierte Unternehmen durch Finanzintermediäre. Wirtschaftliche Bedeutung und institutionelle Rahmenbedingungen; 1998. Band 22: Niko J. Kleinmann: Die Ausgestaltung der Ad-hoc-Publizität nach § 15 WpHG. Notwendigkeit einer segmentspezifischen Deregulierung; 1998. Band 23: Elke Ebert: Startfinanzierung durch Kreditinstitute. Situationsanalyse und Lösungsansätze; 1998. Band 24: Heinz O. Steinhübel: Die private Computerbörse für mittelständische Unternehmen. Ökonomische Notwendigkeit und rechtliche Zulässigkeit; 1998. Band 25: Reiner Dietrich: Integrierte Kreditprüfung. Die Integration der computergestützten Kreditprüfung in die Gesamtbanksteuerung; 1998. Band 26: Stefan Topp: Die Pre-Fusionsphase von Kreditinstituten. Eine Untersuchung der Entscheidungsprozesse und ihrer Strukturen; 1999. Band 27: Bettina Korn: Vorstandsvergütung mit Aktienoptionen. Sicherung der Anreizkompatibilität als gesellschaftsrechtliche Gestaltungsaufgabe; 2000. Band 28: Armin Lindtner: Asset Backed Securities – Ein Cash flow-Modell; 2001. Band 29: Carsten Lausberg: Das Immobilienmarktrisiko deutscher Banken; 2001. Band 30: Patrik Pohl: Risikobasierte Kapitalanforderungen als Instrument einer marktorientierten Bankenaufsicht – unter besonderer Berücksichtigung der bankaufsichtlichen Behandlung des Kreditrisikos; 2001. Band 31: Joh. Heinr. von Stein/Friedrich Trautwein: Ausbildungscontrolling an Universitäten. Grundlagen, Implementierung und Perspektiven; 2002. Band 32: Gaby Kienzler, Christiane Winz: Ausbildungsqualität bei Bankkaufleuten – aus der Sicht von Auszubildenden und Ausbildern, 2002. Band 33: Joh. Heinr. von Stein, Holger G. Köckritz, Friedrich Trautwein (Hrsg.): E-Banking im Privatkundengeschäft. Eine Analyse strategischer Handlungsfelder, 2002.

Band 34: Antje Erndt, Steffen Metzner: Moderne Instrumente des Immobiliencontrollings. DCFBewertung und Kennzahlensysteme im Immobiliencontrolling, 2002. Band 35: Sven A. Röckle: Schadensdatenbanken als Instrument zur Quantifizierung von Operational Risk in Kreditinstituten, 2002. Band 36: Frank Kutschera: Kommunales Debt Management als Bankdienstleistung, 2003. Band 37: Niklas Lach: Marktinformation durch Bankrechnungslegung im Dienste der Bankenaufsicht, 2003. Band 38: Wigbert Böhm: Investor Relations der Emittenten von Unternehmensanleihen: Notwendigkeit, Nutzen und Konzeption einer gläubigerorientierten Informationspolitik, 2004. Band 39: Andreas Russ: Kapitalmarktorientiertes Kreditrisikomanagement in der prozessbezogenen Kreditorganisation, 2004. Band 40: Tim Arndt: Manager of Managers – Verträge. Outsourcing im Rahmen individueller Finanzportfolioverwaltung von Kredit- und Finanzdienstleistungsinstituten, 2004 Band 41: Manuela A. E. Schäfer: Prozessgetriebene multiperspektivische Unternehmenssteuerung: Beispielhafte Betrachtung anhand der deutschen Bausparkassen, 2004. Band 42: Friedrich Trautwein: Berufliche Handlungskompetenz als Studienziel: Bedeutung, Einflussfaktoren und Förderungsmöglichkeiten beim betriebswirtschaftlichen Studium an Universitäten unter besonderer Berücksichtigung der Bankwirtschaft, 2004. Band 43: Ekkehardt Anton Bauer: Theorie der staatlichen Venture Capital-Politik. Begründungsansätze, Wirkungen und Effizienz der staatlichen Subventionierung von Venture Capital, 2006. Band 44: Ralf Kürten: Regionale Finanzplätze in Deutschland, 2006. Band 45: Tatiana Glaser: Privatimmobilienfinanzierung in Russland und Möglichkeiten der Übertragung des deutschen Bausparsystems auf die Russische Föderation anhand des Beispiels von Sankt Petersburg, 2006. Band 46: Elisabeth Doris Markel: Qualitative Bankenaufsicht. Auswirkungen auf die Bankunternehmungsführung, 2010. Band 47: Matthias Johannsen: Stock Price Reaction to Earnings Information, 2010. Band 48: Susanna Holzschneider: Valuation and Underpricing of Initial Public Offerings, 2011. Band 49: Arne Breuer: An Empirical Analysis of Order Dynamics in a High-Frequency Trading Environment, 2013. Band 50: Dirk Sturz: Stock Dividends in Germany. An Empirical Analysis, 2015. Band 51: Sebastian Schroff: Investor Behavior in the Market for Bank-issued Structured Products, 2015. Band 52: Jan Müller: Optimal Economic Capital Allocation in Banking on the Basis of Decision Rights, 2015. Band 53: Helena Kleinert: The International Diversification Puzzle: Home Bias in Countries’ Investment Portfolios, 2016. Band 54: Frederic Dachs: Die Reform der Bankenrestrukturierung. Nachgelagerte und präventive Maßnahmen und ihre Wirksamkeit, 2017. Band 55: Barbara Flaig: Corporate Bankruptcies in Germany. Recovery Rates in Insolvency Plans, 2017.