The Long-Term Success of Mergers and Acquisitions in the International Automotive Supply Industry

Jan-Peer Laabs The Long-Term Success of Mergers and Acquisitions in the International Automotive Supply Industry GABLE...

Author: Jan-Peer Laabs

58 downloads 753 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Jan-Peer Laabs The Long-Term Success of Mergers and Acquisitions in the International Automotive Supply Industry

GABLER EDITION WISSENSCHAFT EBS Forschung Schriftenreihe der European Business School (EBS) International University · Schloss Reichartshausen Herausgegeben von Univ.-Prof. Ansgar Richter, PhD

Band 71

Die European Business School (EBS) – gegründet im Jahr 1971 – ist Deutschlands älteste private Wissenschaftliche Hochschule für Betriebswirtschaftslehre im Universitätsrang. Dieser Vorreiterrolle fühlen sich ihre Professoren und Doktoranden in Forschung und Lehre verpflichtet. Mit der Schriftenreihe präsentiert die European Business School (EBS) ausgewählte Ergebnisse ihrer betriebs- und volkswirtschaftlichen Forschung.

Jan-Peer Laabs

The Long-Term Success of Mergers and Acquisitions in the International Automotive Supply Industry With a foreword by Prof. Dr. Dirk Schiereck

GABLER EDITION WISSENSCHAFT

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de.

Dissertation European Business School, International University Schloss Reichartshausen, Oestrich-Winkel, 2009 D1540

1st Edition 2009 All rights reserved © Gabler | GWV Fachverlage GmbH, Wiesbaden 2009 Editorial Office: Claudia Jeske / Britta Göhrisch-Radmacher Gabler is part of the specialist publishing group Springer Science+Business Media. www.gabler.de No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the copyright holder. Registered and/or industrial names, trade names, trade descriptions etc. cited in this publication are part of the law for trade-mark protection and may not be used free in any form or by any means even if this is not specifically marked. Cover design: Regine Zimmer, Dipl.-Designerin, Frankfurt/Main Printed on acid-free paper Printed in Germany ISBN 978-3-8349-1693-8

Foreword It is precisely the kind of highly variable success that Schaeffler KG was forced to endure during its acquisitions of FAG Kugelfischer and Continental, which provides a telling account of the myriad potential consequences of mergers and acquisitions in the automotive supply industry. The global competitive landscape for automotive suppliers is wholly unique, and not just because its consumers and key customers, the automotive manufacturers, have joined together in an ever-narrowing oligopoly over the last twenty years, nor because the current focus on suppliers is already helping restore an increasingly stable balance of power. Since automotive manufacturers serve international markets, the competitive environment for suppliers has also always been transnational. This becomes all the more relevant when we consider recent shifts in the automotive manufacturing value chain. Meanwhile, some suppliers are producing a higher value contribution during construction of a car than the actual manufacturer. In these situations, international M&A transactions seem especially attractive, and should, in actual fact, generate positive reactions on the capital markets. But is that really the case? Studies looking specifically into the long-term success of international M&A transactions have been a rare commodity until now, the current state of knowledge on the success of acquisitions in the automotive supplier industry is even more limited, and the transferability of evidence based on experiences in other sectors is highly questionable. There was, therefore, an urgent need to analyze the success factors for M&A in this industry This paper seizes on this considerable research gap, with great attention to detail and maximum possible accuracy. The primary objective was to use capital market data and financial accounting information to investigate the success of international M&A transactions in the automotive supply industry, and to identify the key factors in this success. Thus can we achieve an objective state of knowledge, which can be used as a

VI

Foreword

basis for deriving well-founded, recommended actions to be implemented in industry practice. Mr. Laabs is in the best possible position to fulfill his self-imposed objectives in terms of this dissertation. The paper contains a wealth of very interesting results, and is written in such a way that the reader will thoroughly enjoy studying this entire work from cover to cover. I wish this paper the extensive circulation it deserves.

Professor Dr. Dirk Schiereck

Preface Writing the following doctoral thesis has been an inspiring and shaping personal experience. However, the progress of my research would not have been as advanced, focused, and exciting without the support of many people whom I would like to acknowledge and thank with the following lines. They made a successful completion of my work possible. First, I would like thank my doctoral supervisor, Prof. Dr. Dirk Schiereck, for his academic guidance and encouragement in preparing this thesis. Besides providing valuable advice and instant feedback, he always managed to balance academic "judgment calls" with an extraordinarily constructive and collaborative working atmosphere. I appreciated the discussions and will remember our joint journey to an academic conference in Nashville as one of the highlights of these past two years. In addition, I would like to thank Prof. Dr. Ronald Gleich for readily agreeing to assume the role of secondary advisor to my thesis. His strategic perspective on the implications of my work represented an enriching addition to my finance-driven research approach. Second, I am thankful for the support of a number of friends and colleagues. First and foremost, I would like to acknowledge the exceptional help and support of my friend Nils Steiner. Besides being an inspirational discussion partner, he contributed significantly with his technical and modeling experience. Without his "hands-on" practical advice, some of my calculations might have taken considerably longer, or even, would not have found their way into the final version of this thesis. In addition, I would like to thank my colleagues and friends from the Frankfurt Office "Fellow Community" for the joint times we shared, both by discussing on or off topic or by playing numerous rounds of table soccer. All other friends I would like to thank for spurring my motivation by continuously asking for my work's progress. This book is the answer to these questions.

VIII

Preface

Last, but definitely not least, I would like to thank my family and, in particular, my parents. Their unconditional support and encouraging mindset gave me the freedom I needed to pursue my goals and the certainty of their helping advice and opinion when required. Finally, I am especially grateful to my long-time girl-friend Nadine and our daughter Lina-Marie. They not only managed to cope with the additional time constraints inherent in writing a doctoral thesis, but also seemed to outbalance any upcoming doubts on my side with an increased level of encouragement and support. Their support, patience, and love have been invaluable sources of inspiration and motivation to finish this thesis. I dedicate this book to them.

Jan-Peer Laabs

Content Overview List of Figures................................................................................................................XV List of Tables .............................................................................................................. XVII List of Abbreviations ................................................................................................... XIX 1

Introduction .............................................................................................................. 1

2

Research Foundations............................................................................................... 9

3

Study 1: Determinants of Capital Market Performance ......................................... 23

4

Study 2: Does Operating Performance Meet Market Expectations?...................... 93

5

Study 3: How a Good Bidder Becomes a Good Target – The Case of Continental AG Acquiring Siemens VDO ........................................................... 133

6

Conclusion............................................................................................................ 173

Appendix ...................................................................................................................... 181 References .................................................................................................................... 189

Table of Contents List of Figures................................................................................................................XV List of Tables .............................................................................................................. XVII List of Abbreviations ................................................................................................... XIX 1

Introduction .............................................................................................................. 1 1.1 Problem Definition and Objectives .................................................................. 1 1.2 Course of Analysis ........................................................................................... 5

2

Research Foundations............................................................................................... 9 2.1 The Automotive Supply Industry ..................................................................... 9 2.1.1 Definition and Scope ............................................................................ 9 2.1.2 Current Trends and Challenges .......................................................... 10 2.2 Mergers and Acquisitions............................................................................... 15 2.3 Measuring Success of Mergers and Acquisitions........................................... 16 2.3.1 Time Horizons .................................................................................... 16 2.3.2 Research Approaches ......................................................................... 19

3

Study 1: Determinants of Capital Market Performance ......................................... 23 3.1 Introduction .................................................................................................... 23 3.2 Literature Review and Hypotheses................................................................. 25 3.2.1 Related Literature ............................................................................... 25 3.2.2 Hypotheses ......................................................................................... 29 3.2.2.1 The Overall Effect ............................................................... 29 3.2.2.2 The Impact of Transaction Characteristics.......................... 29 3.2.2.3 The Impact of Acquirer Characteristics............................... 31 3.3 Data and Methodology ................................................................................... 33 3.3.1 Identifying Merging Companies......................................................... 33 3.3.2 Portfolio of Matching Firms ............................................................... 35 3.3.3 Econometric Strategy ......................................................................... 36 3.3.3.1 Short-term Methodology ..................................................... 36 3.3.3.2 Long-term Methodology...................................................... 40 3.4 Empirical Results............................................................................................ 43 3.4.1 The Overall Effect .............................................................................. 43 3.4.2 The Impact of Transaction Characteristics......................................... 47 3.4.2.1 Geographical Expansion...................................................... 47

XII

Table of Contents

3.4.2.2 Product Diversification........................................................ 52 3.4.2.3 Transaction Size .................................................................. 57 3.4.3 The Impact of Acquirer Characteristics.............................................. 61 3.4.3.1 Product Groups .................................................................... 61 3.4.3.2 Acquisition Experience........................................................ 67 3.5 Robustness Cheques and Cross-Sectional Regressions.................................. 72 3.5.1 Regression of Short-term CAARs ...................................................... 73 3.5.2 Regression of Long-term BHARs ...................................................... 83 3.6 Conclusion...................................................................................................... 90 4

Study 2: Does Operating Performance Meet Market Expectations?...................... 93 4.1 Introduction .................................................................................................... 93 4.2 Literature Review and Hypotheses................................................................. 95 4.2.1 Related Literature ............................................................................... 95 4.2.2 Hypotheses ......................................................................................... 98 4.2.2.1 The Overall Effect ............................................................... 98 4.2.2.2 The Impact of Transaction and Acquirer Characteristics.. 100 4.2.2.3 The Correlation of Accounting- and Event-Study Results 100 4.3 Data and Methodology ................................................................................. 101 4.3.1 Identifying Merging Companies....................................................... 101 4.3.2 Portfolio of Matching Firms ............................................................. 103 4.3.3 Econometric Strategy ....................................................................... 105 4.4 Empirical Results.......................................................................................... 108 4.4.1 The Overall Effect ............................................................................ 108 4.4.2 Determinants of Profitability............................................................ 113 4.5 The Correlation of Accounting- and Event-Study Results........................... 121 4.6 Conclusion.................................................................................................... 129

5

Study 3: How a Good Bidder Becomes a Good Target – The Case of Continental AG Acquiring Siemens VDO ........................................................... 133 5.1 Introduction .................................................................................................. 133 5.2 Literature Review and Industry Overview ................................................... 135 5.2.1 Related Literature ............................................................................. 135 5.2.2 Overview of the Automotive Supply Industry ................................. 139 5.3 Case Study Background ............................................................................... 142 5.3.1 The Transaction Partners .................................................................. 142 5.3.1.1 Continental AG.................................................................. 142 5.3.1.2 Siemens VDO Automotive................................................ 146 5.3.2 Transaction Motives ......................................................................... 149 5.3.3 The Acquisition Event ...................................................................... 153

Table of Contents

XIII

5.4 Acquisition Performance .............................................................................. 156 5.4.1 The Capital Market Perspective ....................................................... 156 5.4.2 Performance Analysis....................................................................... 164 5.5 Discussion and Conclusion........................................................................... 170 6

Conclusion............................................................................................................ 173

Appendix ...................................................................................................................... 181 References .................................................................................................................... 189

List of Figures Figure 2.1: M&A Transactions in the Automotive Supply Industry.............................. 13 Figure 3.1: CAARs to Acquirers, Targets, and Combined Entities ............................... 45 Figure 5.1: Transaction Volume in the Automotive Supply Industry .......................... 140 Figure 5.2: Share Price Development of Continental AG and Siemens AG................ 158 Figure 5.3: Long-term Relative Share Price Development of Continental AG............ 161 Figure 5.4: Revenue Split of Continental AG across Divisions ................................... 169

List of Tables Table 3.1:

Overview of Hypotheses and Predicted Value Impact on Acquirers ........ 30

Table 3.2:

Overview of the Transaction Sample – Descriptive Statistics .................. 35

Table 3.3:

CAARs to Acquirers.................................................................................. 44

Table 3.4:

BHARs to Acquirers.................................................................................. 46

Table 3.5:

Abnormal Returns (FF3F) to Acquirers .................................................... 47

Table 3.6:

Acquirer CAARs – Differences by Geographical Scope .......................... 49

Table 3.7:

Acquirer CAARs – Diferences by Continental Scope............................... 50

Table 3.8:

Acquirer BHARs – Differences by Geographical Scope .......................... 51

Table 3.9:

Abnormal Returns to Acquirers – Differences by Geographical Scope.... 51

Table 3.10: Acquirer CAARs – Differences by Product Scope ................................... 54 Table 3.11: Acquirer BHARs– Differences by Product Scope .................................... 55 Table 3.12: Abnormal Returns to Acquirers – Differences by Product Scope............. 56 Table 3.13: Acquirer CAARs – Differences by Transaction Size................................ 58 Table 3.14: Acquirer BHARs – Differences by Transaction Size................................ 60 Table 3.15: Abnormal Returns to Acquirers – Differences by Transaction Size ......... 61 Table 3.16: Acquirer CAARs – Differences by Product Group ................................... 63 Table 3.17: Acquirer BHARs – Differences by Product Group ................................... 65 Table 3.18: Abnormal Returns to Acquirers – Differences by Product Group ............ 67 Table 3.19: Acquirer CAARs – Differences by Acquisition Experience ..................... 69 Table 3.20: Acquirer CAARs – Differences for Bidder Champions............................ 70 Table 3.21: Acquirer BHARs – Differences by Acquisition Experience ..................... 71 Table 3.22: Abnormal Returns to Acquirers – Differences by Acquirer Experience... 72 Table 3.23: Regression of Short-term CAARs to Acquirers ........................................ 77 Table 3.24: Step-wise Regression of Short-term CAARs to Acquirers ....................... 78 Table 3.25: Regression of Long-term BHARs to Acquirers ........................................ 86 Table 3.26: Step-wise Regression of 36-month BHARs to Acquirers ......................... 87 Table 4.1:

Overview of the Transaction Sample – Descriptive Statistics ................ 103

Table 4.2:

Average Performance Following M&A Transactions............................. 109

XVIII

List of Tables

Table 4.3:

Abnormal Performance Changes in the Automotive Supply Industry.... 111

Table 4.4:

Regression of Median Post-Merger Abnormal Performance .................. 112

Table 4.5:

Abnormal Performance Change – Differences by Product Scope .......... 115

Table 4.6:

Abnormal Performance Change – Differences by Geographical Scope . 116

Table 4.7:

Stepwise Regression of Abnormal Performance Changes ...................... 119

Table 4.8:

BHARs to Acquirers Applied in the Accounting Study.......................... 122

Table 4.9:

Regression of 36-month BHARs on Abnormal Performance Changes .. 123

Table 4.10: Relationship between Event- and Accounting-Study Results ................. 127 Table 5.1:

Milestones in the Continental/Siemens VDO Transaction...................... 154

Table 5.2:

Abnormal Announcement Returns to Continental AG and Siemens AG 159

Table 5.3:

Buy-and-Hold Abnormal Returns to Continental AG............................. 162

Table 5.4:

Abnormal Returns to Continental AG versus Different Peer Groups ..... 163

Table 5.5:

Quarterly Balance Sheet/Income Statement Items of Continental AG ... 165

Table 5.6:

Unadjusted Quarterly Performance Indicators of Continental AG ......... 167

Table 5.7:

Abnormal Quarterly Performance Indicators of Continental AG ........... 168

List of Abbreviations Adj.R2

Adjusted Coefficient of Determination

AG

Aktiengesellschaft (Stock Corporation)

Avg.

Average

BHAR

Buy-and-Hold Abnormal Return

BHR

Buy-and-Hold Return

bn

Billion

BVA

Book Value of Assets

CAAR

Cumulative Average Abnormal Return

CAGR

Compound Annual Growth Rate

CAR

Cumulative Abnormal Return

CF

Operating Cash Flow

COGS

Cost of Goods Sold

Conti.

Continental AG

Corp.

Corporation

CO2

Carbon Dioxide

DAX

Deutscher Aktien Index (German Stock Index)

DWS

Durbin-Watson Statistic

EBIT

Earnings Before Interest and Taxes

EBITDA

Earnings Before Interest, Taxes, Depreciation, and Amortization

EMH

Efficient Market Hypothesis

EUR

Euro

FDI

Foreign Direct Investment

FF3F

Fama-French-3-Factor

FTSE

Financial Times Stock Exchange

GHz

Gigahertz

GmbH

Gesellschaft mit beschränkter Haftung (Limited Liability Corporation)

XX

List of Abbreviations

GST

Generalized Sign Test

HML

High Minus Low

i.e.

that is

IFRS

International Financial Reporting Standard

IPO

Initial Public Offering

M&A

Mergers and Acquisitions

MDN

Median

NOPLAT

Net Operating Profit Less Adjusted Taxes

OEM

Original Equipment Manufacturer

OLS

Ordinary Least Squares

R&D

Research and Development

SDC

Securities Data Corporation

SG&A

Selling, General and Administrative Expense

SIC

Standard Industrial Classification

SMB

Small Minus Big

S&P

Standard & Poor's

S&P 500

Standard & Poor's 500 Index

US

United States

USD

United States Dollar

US-GAAP

United States – Generally Accepted Accounting Principles

VDO

Vereinigte Deutsche Tachometerwerke und OSA Apparate GmbH

vs.

versus

1

Introduction

1.1

Problem Definition and Objectives As a result of economic and structural changes, the automotive supply industry

has been facing significant consolidation activity over the last twenty years. The pressure to produce better equipped and less expensive automobiles created a growing trend towards specialization and internationalization among automotive suppliers. For many players, mergers and acquisitions (M&A) became a common strategic response to these trends inducing increasingly dominated product ranges and a truly global competition. Consequently, the number of existing suppliers has been continuously decreasing: In Europe, for example, the number of direct suppliers dropped from 10,000 in the early 1970s to 3,000 in 1995 and to an estimated 500 in the year 2000 (Sadler (1999)). Between 1991 and 1999, horizontal M&A transactions with more than USD 50 million transaction value steadily increased both in total numbers as well as in total volume. While the influence of this potential merger wave appeared to weaken after the year 2000, the multi-billion dollar transactions involving German Continental AG in 2007 and 2008 have recently revived consolidation discussions and potentially indicate the start of a next major consolidation wave. The unique competitive situation within the automotive supply industry promotes this on-going consolidation. By the early 21st century, automotive suppliers are facing increasing pressure from both sides of the automotive value chain. On the one side, the Original Equipment Manufacturers (OEMs) as the suppliers' main customers consolidate and, at the same time, become more and more sophisticated. While the globalization of OEMs forces suppliers to costly follow abroad, a change in the manufacturers' sourcing preferences additionally strains the suppliers' profit situation. Since OEMs now prefer to source full systems of components from a limited number of suppliers, they transfer an increasing degree of production and development tasks down the value chain (Sadler (1999), Von Corswant and Fredriksson (2002)). On the other side of

2

1 Introduction

the value chain, increasing raw material prices for crude oil, natural rubber as well as many metals compile a threat for the remaining supplier profits. Consequently, competition among suppliers remains fierce and is still growing. Where a few international players dominate the product range, strong barriers to entry exist. Where specialization does not create a competitive edge, the comparably high number of existing suppliers encourages a very strong rivalry, as apparent in the leather and tire industries (Aktas, de Bodt, and Derbaix (2004)). Due to these specific industry characteristics, previous research on the value creation potential of M&A among automotive suppliers identifies the industry as an outlier: capital markets generally seem to appreciate M&A as a valuable response strategy by granting significant positive abnormal announcement returns to acquirers. These positive announcement returns represent the industry-specific synergy and efficiency potentials underlying the transactions as perceived by investors and capital markets (Mentz and Schiereck (2008)). Although preceding M&A literature on the general announcement effect of M&A also concludes that M&A create value for the combined entities, positive announcement returns to acquiring companies are clearly an exception. Comparing a variety of different studies, Bruner (2002) finds announcement returns for acquirers to be essentially zero, while Loughran and Vijh (1997) even conclude them to be overall negative. Consequently, the outstanding competitive characteristics of the automotive supply industry appear to also translate into an outstanding value creation potential for suppliers engaging in takeover activities. However, given that announcement returns represent a short-term assessment of value creation building on investors' expectations about future performance, the question remains whether acquirers are in fact able to sustain their exceptional positive returns beyond a short-term perspective. Previous literature on the general long-term capital market performance of acquirers provides a rather negative outlook. The majority of studies including the early work of Mandelker (1974) and the methodological milestones of Franks, Harris and Titman (1991) and Loughran and Vijh (1997) point to-

1.1 Problem Definition and Objectives

3

wards significant value losses for acquirers over three to five years following an M&A announcement. In the light of these consistently negative findings, it becomes even more relevant to assess the long-term performance of acquirers in the automotive supply industry. If suppliers were able to sustain their exceptional positive announcement returns in the long-run, the industry could truly be regarded as a consistent positive outlier representing a unique investment opportunity for industry in- and outsiders. If, however, acquirers were not able to sustain their positive announcement returns, this result would reveal their inability to fulfill capital market expectations about the inherent synergy potentials. In the light of on-going consolidation, such an inability poses a challenge to the management of merging companies and calls for a comprehensive analysis of determinants for the observed value effect. Besides the negative outlook delivered by preceding literature, a number of theoretical considerations makes the assessment of long-term M&A success generally more challenging than determining short-term announcement returns. Firstly, the application of longer time horizons requires well-specified methodological approaches to overcome inherent statistical biases. As a result, studies assessing long-term performance usually apply either statistically advanced methodological approaches or a combination of different methodologies to arrive at valid conclusions (Barber and Lyon (1997)). Secondly, long-term success of M&A can be determined based on different data sources. Besides the capital market perspective as expressed in abnormal share returns, publicly available accounting information also frequently serves as the basis for analyzing the long-term post-merger performance of acquirers. While accounting studies focus exclusively on historic data, both the resulting general effect of M&A on acquirer performance as well as the correlation of results under the two approaches remain ambiguous (Healy, Palepu, and Ruback (1992), Fridolfsson and Stennek (2005)). Consequently, a combination of different approaches and data bases is required to generate a comprehensive view on the long-term success of M&A and to explore potential determinants behind long-term performance.

4

1 Introduction

In general, it becomes apparent that both the uniquely competitive situation within the industry as well as the theoretical considerations on the measurement of longterm M&A success significantly influence any approach assessing long-term performance within the automotive supply industry. Consequently, this industry represents a particularly relevant research object in order to assess the sustainability of positive abnormal announcement returns in the long-term and its underlying determinants. Overall, the following objectives serve as the basis for the upcoming empirical analysis: 1. The study empirically analyzes the long-term value creation potential of M&A in the automotive supply industry. In addition to arriving at an overall judgment on the long-term value creation potential, the question whether acquirers in the automotive supply industry are able to sustain their extraordinary positive short-term returns over a long-term time horizon is addressed in particular. Therefore, a combination of different methodological approaches including long-term event- and accounting-study methodologies is applied on a base sample of 230 M&A transactions within the global automotive supply industry between 1981 and 2007. 2. Based on the overall assessment of the value creation potential, determinant variables of long-term performance are analyzed for the magnitude and direction of their respective value impact. In this regard, variables identified to influence short- and long-term performance in preceding general finance literature are challenged against their effect within the specific competitive environment of the automotive supply industry. By identifying a comprehensive list of potential value drivers, this study does not only advance previous literature by an additional perspective on the effect of the various determinants, but also allows senior management of automotive suppliers to identify valuemaximizing strategies for future takeovers. 3. With regard to the described theoretical considerations, this study focuses on creating a comprehensive and differentiated perspective on the acquirers'

1.2 Course of Analysis

5

post-merger performance. For this purpose, the last objective of this study is to verify the derived effects between the different available methodologies. The long-term abnormal capital market returns are challenged against the abnormal performance of acquirers as published in their accounting statements. Thereby, a differentiation between genuine value impact and the subjective assessment of capital markets becomes feasible. A case study on a successful takeover from the industry provides an additional opportunity to verify the findings from the general event studies.

1.2

Course of Analysis This chapter defines and explains the overall objectives of this thesis and sets

the course of analysis for the following elaborations. Chapter 2 provides a number of research foundations including a definition of 'automotive supplier', 'mergers and acquisitions', and different research approaches. In addition, it provides a description of the structure and competitive environment within the automotive supply industry and of the different theoretical considerations in measuring long-term post-merger performance. Thereby, the chapter intends to create an understanding for the competitive pressures affecting automotive suppliers and the various methodological approaches available to determine long-term success of takeover activities. Chapters 3 to 5 each represent a self-contained empirical study addressing different research questions related to the overall topic. In sum, they compose the main empirical analyses of this thesis and are developed in subsequent order. Chapter 3 examines the short- and long-term value effects of horizontal mergers and acquisitions on the capital market performance of acquirers in the automotive supply industry. It answers to the questions whether acquirers in this industry are actually able to sustain their extraordinary short-term share returns in the long-run and what the potential determinants for the observed short- and long-term value effects are.

6

1 Introduction

For this purpose, the significant positive announcement returns determined by preceding research are at first updated and validated against a global sample of 230 takeover announcements between 1981 and 2007. Then, this chapter challenges the short-term announcement returns against the long-term capital market performance of the respective acquirers. By applying the Fama-French-3-Factor-model (FF3F-model) in calendar-time and the control-firm approach in event-time, a comprehensive perspective on the post-merger capital market performance is created revealing whether acquirers are able to sustain their positive short-term abnormal returns. After this, Chapter 3 analyzes a number of determining variables for both the short- and long-term capital market performance and tests for their significance using a cross-sectional regression model of the determined long-term Buy-and-Hold Abnormal Returns (BHARs). Chapter 4 addresses the research question whether the observed long-term capital market returns to acquirers are consistent with their post-merger operating performance. Besides creating an overall view on the accounting performance, this chapter also addresses the question how the results under both approaches correlate, i.e. to what extent the results under both approaches are consistent and could therefore serve as substitutes. For this purpose, the post-merger change in performance of acquirers is firstly determined based on a total of six performance indicators. Afterwards, the performance is normalized against a sample of non-merging peers from the same industry. As with the capital market performance in Chapter 3, Chapter 4 also analyzes the observed operating performance impact for determinant variables and tests for their significance with a cross-sectional regression model. In line with the primary objective of this thesis, it intends to develop a consistent perspective on the value creation potential of M&A across different data bases. In a later step, Chapter 4 addresses its second research question by determining a correlation between accounting- and event-study results in order to advance the methodological discussion on the substitutability of both methodologies. This section also differentiates results between different performance indicators and analyzes potential capital market premiums awarded for a correctly anticipated longterm profitability development.

1.2 Course of Analysis

7

Chapter 5 validates the derived empirical findings of Chapters 3 and 4 against a case study from the industry, namely the acquisition of Siemens VDO by Continental AG in July 2007. The main questions addressed include whether an industry leader of significant size can complete a consistently successful transaction across multiple analytical approaches and what characteristics contribute to Continental's present situation. In a first step, this chapter examines the motivation behind Continental's takeover of Siemens VDO and evaluates the motives against the present industry trends and transaction background. Afterwards, it assesses the short- and long-term post-merger wealth effects of this transaction on Continental's capital market and operating performance. Besides assessing the absolute value impact, the chapter focuses on comparing Continental's performance against a number of different peer benchmarks. By replicating the methodological approaches of the preceding chapters, this case study afterwards confirms the main performance drivers previously determined in the cross-sectional regressions. Overall, it provides an additional perspective on which M&A strategy potentially leads to long-term post-merger success within the automotive supply industry. Chapter 6 consolidates the main findings from the three empirical studies, identifies interrelations and determines key success factors for long-term success of M&A within the automotive supply industry. The study concludes with an outlook and presents potential areas for further academic research.

2

Research Foundations The following chapter provides a brief overview of the research foundations re-

lating to the three main components of this study's title, namely the automotive supply industry, mergers and acquisitions, and the measurement of M&A success. Besides providing fundamental definitions, this chapter also intends to define the scope of the following analyses and to create a basic understanding not only for the current situation within the automotive supply industry but also for the theoretical discussion around the measurement of long-term value creation through M&A transactions. Overall, it becomes apparent that the automotive supply industry provides a uniquely competitive environment in which automotive suppliers are increasingly seeking M&A activity as a response strategy. For determining M&A success, especially the consideration of a longer time horizon requires a differentiated use of various available methodologies.

2.1

The Automotive Supply Industry

2.1.1 Definition and Scope Although preceding literature comprises a number of different attempts to define the term 'automotive supplier,' an unambiguous definition has only recently been developed. According to Mentz (2006), an 'automotive supplier' is defined as any economic entity directly or indirectly delivering products and/or services to car producers, socalled OEMs, in order to be included in the production process of automobiles or eventually become part of the automobile itself. Since the focus of this thesis likewise lies in analyzing a comprehensive and global sample of M&A transactions in the automotive supply industry, adopting this definition for its purpose carries two major advantages. On the one hand, it extends the scope of the automotive supply industry to service companies enabling this study to create an even more comprehensive perspective on the specific long-term value creation potential within this industry. On the other hand, it

10

2 Research Foundations

confines the scope of the industry to genuine suppliers. Consequently, OEMs and the raw materials industries are excluded under the applied definition; both represent their respective ends of the observed automotive value chain. Based on the depth of the respective value added, the differentiation of automotive suppliers into different supplier tiers has become common and relates back to the pyramid-shaped supply chains developed mainly in the Japanese automotive and electronics industry. First-tier suppliers produce the highest value added by delivering subassembled units or modules (e.g. complete seats or transmissions) to OEMs. Second-tier suppliers produce less complex components usually delivered to first-tier suppliers for inclusion into their sub-assembled units. Third-tier suppliers produce standardized or basic products requiring a minimum of additional logistical services or planning (Von Corswant, Wynstra, and Wetzels (2003)). While this pyramid structure enables automotive producers to simplify their material flows, it also decreases the coordination efforts for producers towards their suppliers: first-tier suppliers coordinate their second-tier suppliers, second-tier suppliers the third-tier, and so on (Kamath and Liker (1994)). Consequently, manufacturers are able to reduce their direct relations to a few first-tier suppliers while the suppliers coordinate themselves down the supply chain (Von Corswant and Fredriksson (2002)). 2.1.2 Current Trends and Challenges This section outlines trends and challenges influencing automotive suppliers over the last twenty years. It aims to create an understanding for the competitive pressures promoting the continuous consolidation within the automotive supply industry. As the competitive environment for suppliers is by definition strongly related to their customers, the main set of trends relates to the customer-supplier relationship and the sourcing behavior of car producers. These trends include globalization, outsourcing, shorter product life cycles and an increasing degree of sophistication in the customersupplier relationship.

2.1 The Automotive Supply Industry

11

Although recent literature finds that the transfer of production and product development activities by car producers into an increasing number of countries has recently slowed down, car producers are still enforcing globalization by other means (Von Corswant and Fredriksson (2002)). Main examples include the global use of common automotive platforms in different models as well as a growing interest of OEMs to source supplies from the same supplier on a world-wide basis (Sadler (1999)). In order to meet their customers' demands regarding just-in-time delivery as well as local regulative requirements including customs and in-country quotas, the pressure on automotive suppliers to create a costly global presence is growing. As a result, the average number of countries in which a sample of suppliers maintains production facilities increased from 5.8 in 1988 to 15.0 in 2003. Product development activities of the same group took place in 5.9 countries in 2003 as opposed to 2.6 in 1988 (Von Corswant and Fredriksson (2002)). Wherever financially affordable, it appears that automotive suppliers are following the internationalization of their customers either by geographical expansion or cross-border acquisitions (Abrenica (1998), Sadler (1999)). A second trend frequently addressed in academic literature is the growing degree of outsourcing (Mercer (1995), McIvor, Humphreys and McAleer (1998)). Automotive producers increasingly seek to outsource parts of their production facilities and to purchase full systems of components from their suppliers rather than individual parts (Sadler (1999)). Von Corswant and Fredriksson (2002) show that OEMs were expected to acquire more than 65 per cent of their total turnover in 2003 as purchased materials. While exposing them to the risk of losing control over the car as a whole, OEMs still benefit from outsourcing a significant degree of their production to a limited number of suppliers. As a result, they are able to focus their coordination efforts on a few relationships with first-tier suppliers (McIvor et al. (1998)). For the suppliers, however, coordination efforts likewise increase as first-tier suppliers also outsource a growing part of their activities. In the end, first-tier suppliers are increasingly required not only to manage their customer demands up-stream, but also a growing number of lower-tier suppliers down the value chain.

12


At the same time, product life-cycles within the automotive industry become shorter. As they converge towards a minimum product life-cycle length for the producers, the challenge for suppliers increases. On the one hand, suppliers are required to take over the product renewal responsibility and to upgrade their systems and components regularly. On the other hand, automotive producers are also increasingly transferring full product development tasks. While the producers' share of total product development resources averaged at around 70% in 1988, it dropped to approximately 60% ten years later (Von Corswant and Fredriksson (2002)). As a result, suppliers are acquiring valuable production expertise and product development capabilities in order to be capable of delivering innovative products at high frequency. The closer in the supply chain a supplier is situated to the automobile producer, the larger the degree of actual product development activity it shoulders (Von Corswant et al. (2003)). A last trend stems from a generally increasing degree of sophistication within the customer-supplier relationship. While reducing the number of relationships to direct suppliers, OEMs try to establish a collaborative partnership with a limited number of first-tier suppliers. As a result, the supplier is required not only to fulfill its customers' demands but also to actively engage in subcontracting, adhere to just-in-time logistics, and meet legal and warranty requirements. As the first-tier suppliers are expected to exhibit constant cost-reductions on a year-to-year basis, lower-tier suppliers are increasingly struggling to adhere to cost and contractual requirements (McIvor et al. (1998)). Besides these trends concerning the sourcing behavior of OEMs, the intensity of competition among suppliers also increases. As producers try to source complete systems from a limited number of first-tier suppliers, supply companies develop an increasing tendency towards specialization on particular products or segments. Some leading firms have even become inseparably connected with particular systems or technologies (Sadler (1999)). One relevant example for such a connection includes the speedometers of Siemens VDO, whose success story started with their prominent placement into the Volkswagen Beetle in 1939. Where this specialization strategy is successful, the level of competition is fairly low, due to the limited number of rivaling companies dominating

2.1 The Automotive Supply Industry

13

the product range (e.g., only a few players provide very complex products such as Xenon headlights). As a result, high barriers to entry exist almost ruling out new market entries into these product ranges. Where specialization is not able to create a competitive edge, the particularly high number of remaining suppliers creates a very strong competitive environment and fierce rivalry. This is the case, for example, in the leather and tire industries (Aktas et al. (2004)). Figure 2.1: M&A Transactions in the Automotive Supply Industry 25

3 2.5

20

2 15 1.5 10 1 5

Transaction Value (inflation adj., in USD billion)

2007

2005

2003

2001

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

0

0.5 0

Avg. Transaction Value

Source: Thomson One Banker, Bloomberg M&A database, own calculations

Lastly, automotive suppliers are also facing decreasing profit margins. While OEMs expect constant cost reductions decreasing the suppliers' sales volume (McIvor et al. (1998)), increasing prices for raw materials raise their costs of goods sold. Although the slowing down US economy and the weak US Dollar offset some of the price effects on suppliers, especially for European and Asian companies, the majority of raw materials reached record prices in 2007: For example, the average price of crude oil rose by 11%, the price of processed metals increased by 9% and of natural rubber by 10%. The majority of other raw materials including copper, steel, and nickel experienced similar

14


price increases with different volatilities (Continental (2007b)). In connection with the increasing pressure from customers and competitors, these costs for raw materials put a strong strain on the profit situation of automotive suppliers. As a result, many suppliers have realized losses or significant profit reductions over the first years of this century. Between 2000 and 2002, for example, many suppliers suffered from significant profit reductions of up to 50% (Fitzgerald (2002)). Given these challenging industry conditions, mergers and acquisitions represent a common response strategy for automotive suppliers in order to offset some of the challenges described. Although consolidation still remains far lower among automotive suppliers than among OEMs, consolidation activity has significantly increased among suppliers over the last two decades, especially in Europe, Northern America and Japan/South Korea. It is estimated that the number of direct suppliers in Europe, for example, dropped from 10,000 in the early 1970s to 3,000 in 1995 and to about 500 in the year 2000 (Sadler (1999)). Figure 2.1 emphasizes this consolidation activity and shows how a strong merger wave affected the automotive supply industry during the 1990s. Between 1991 and 1999, significant M&A transactions (with a transaction value of more than USD 50 million) steadily increased both in total number as well as in total inflation-adjusted transaction volume. A closer look at the average transaction values reveals that they peaked three times over the last 30 years: once during an early consolidation wave in the 1980s, once during the merger wave of the 1990s and once just recently with the USD 15.7 billion transaction of Continental AG. Consequently, it is likewise reasonable to assume that, instead of just one merger wave in the 1990s, consolidation in the automotive supply industry follows a continuing activity pattern with the 2007 Continental acquisition of VDO potentially representing the starting point of the next significant consolidation wave.

2.2 Mergers and Acquisitions

2.2

15

Mergers and Acquisitions Especially in US-American academic literature, 'mergers and acquisitions' have

developed into a widely-used collective term representing all corporate transactions in the course of which ownership and control of a firm are transferred from one hand to another (see Weston, Chung, and Siu (1990)). Jensen and Ruback (1983) describe the resulting market as a "market for corporate control", in which managers and management teams are actively competing for their right to control corporate resources. Alongside corporate expansions, these transactions in a broader sense also include corporate disposals as well as structural changes in the corporate control, corporate ownership or governance of a firm (Copeland and Weston (1988)). Corporate expansions are regarded as mergers and acquisitions in a narrower sense comprehending all transactions forming one economic unit from two or more previous ones (Weston et al. (1990)). As the purpose of this thesis is to determine long-term success of M&A as an attempt to realize synergy potentials in the automotive supply industry, this study focuses on expanding transactions (M&A in a narrower sense). Mergers and acquisition in a narrower sense can be distinguished based on the future legal status of the target and the degree of codetermination on the target side.1 In a merger, the management boards of both sides agree to combine their firms and jointly prepare a respective proposal to their shareholders. After the shareholders' agreement, one of the merging firms usually ceases to exist as a legal entity and becomes part of the other. As a special sub-form, a consolidation represents a merger in which both companies cease to exist and shareholders of both sides receive shares in a newly found corporate entity. By contrast, the legal status of the target usually remains unaltered in a corporate acquisition. Acquisitions are carried out either as share deals, in which the acquirer directly addresses the shareholders of the target to tender their shares, or asset deals, in which one firm acquires the assets of another. While a tender offer usually bypasses the management board of the target, no formal agreement of the shareholders is 1

The following description is based on Damodaran (2004).

16


needed. In an asset deal, however, the shareholders of the target have to formally approve the acquisition. Once the assets are transferred, the target firm can eventually be liquidated. For the purpose of this thesis, the described focus on expanding transactions is again restricted to only those transactions involving the transfer of majority control, i.e. where the acquirer gains at least 50% of the outstanding shares or private equity. These transactions represent the most comprehensive form of mergers and acquisitions and are expected to have the most profound impact measurable in form of a capital market reaction. For the remainder of this study, the terms 'mergers and acquisitions,' 'corporate acquisitions,' 'consolidation activity,' 'takeover transactions,' 'deals,' and 'corporate combinations' are used interchangeably and all refer to the described expanding transactions involving the transfer of majority control.

2.3

Measuring Success of Mergers and Acquisitions

2.3.1 Time Horizons One challenge in measuring the success of mergers and acquisitions lies in addressing success over different time horizons. While preceding literature distinguishes between short-term announcement returns and long-term abnormal returns, different statistical and methodological issues influence the observable performance under both time horizons. In addition, the past only produced common agreement on the measurement methodology assessing short-term announcement returns (Brown and Warner (1980)). The methodology to assess long-term abnormal returns is constantly refined and discussed in recent publications (see Mitchell and Stafford (2000)). Since this study focuses on the potentially more challenging measurement of long-term success, the following section briefly describes the main available methodologies for both time horizons and their inherent problems.

2.3 Measuring Success of Mergers and Acquisitions

17

While finding frequent application in the past, the event-study methodology usually serves as the means for measuring short-term M&A success. It determines to what extent the share price performance of a transaction partner is considered abnormal, i.e. different from what is expected (Brown and Warner (1980)). This approach builds on and, at the same time, challenges the Efficient Market Hypothesis (EMH), which assumes new information to be incorporated promptly and correctly into return information (Bruner (2002)). One common approach to determine expected returns lies in the two-factor capital asset pricing model, a market model estimating expected returns on the basis of a market portfolio and a risk premium (Black (1972)). Since this eventstudy methodology in connection with the market model is superior to a number of more sophisticated approaches (Brown and Warner (1980)) and robust against ignoring autocorrelation and variance changes in daily data (Brown and Warner (1985)), preceding literature establishes this approach as the predominant methodology for measuring short-term success. Besides methodological agreement, the determined value creation potential in short-term event studies has also converged towards a common opinion. Corporate control transactions in form of mergers or acquisitions are in general associated with significant wealth creation for the combined entity. However, this value creation is unevenly distributed between the shareholders of the target and the acquirer. While targets realize significant positive announcement returns of up to thirty per cent (Jensen and Ruback (1983)), the returns to acquirers are essentially zero (Bruner (2002)); Loughran and Vijh (1997) even conclude that acquirer returns are overall negative or, at most, insignificantly positive. Given the synergy potential in the automotive supply industry, this industry is clearly an outlier as acquirers are able to realize significant positive abnormal returns of +1.6% in the 11-day event period surrounding the announcement date (Mentz and Schiereck (2008)). By contrast, the measurement of long-term M&A success is subject to a number of theoretical pitfalls originating from analyzing longer time periods, the choice of vari-

18


ous performance benchmarks, and the applied statistics. As abnormal performance requires measurement against a benchmark, a longer time period automatically raises concerns about a potential new listing and rebalancing bias within the benchmark (Barber and Lyon (1997)). If abnormal performance is determined against a benchmark portfolio of non-merging firms, the results may be skewed, auto-correlated or exposed to heteroscedasticity. In the past, research focused on applying a combination of event-time and calendar-time approaches in order to overcome these various biases and problems. Buy-and-hold Abnormal Returns (BHARs) represent the most commonly-used methodology in determining long-term performance in event-time (see, for example, Ritter (1991), Barber and Lyon (1997), and Lyon, Barber, and Tsai (1999)). The BHARs are determined by normalizing the investment experience of an investor in an acquiring firm against the investment in a non-merging benchmark. The benchmarks are commonly determined by a character-based matching procedure based on market capitalization and market-to-book ratios (Lyon et al. (1999)). By carefully matching acquiring firms with reference portfolios or control-firms and by applying advanced test statistics on a sufficiently large sample, this method can overcome its inherent statistical pitfalls such as the new listing bias, the rebalancing bias and skewness (Lyon et al., (1999)). However, it is still exposed to significant cross-correlation, especially due to the application of not randomly selected matching samples. The Fama-French-3-Factor-model (FF3F) as a common representative of the calendar-time approaches eliminates the cross-correlation problem (Fama and French (1993)), while at the same time being exposed to other potential pitfalls such as heteroscedasticity. It determines abnormal returns of acquiring companies by regressing a time series of acquirers' excess returns (return less risk-free rate) on the time series of market excess returns, the time series of the difference in returns of small and big companies, and the time series of the difference in returns of companies with high and low marketto-book values (Fama and French (1993)). All in all, it becomes apparent that a comprehensive assessment of the long-term capital market performance of acquirers in the


19

automotive supply industry requires a differentiated and sophisticated methodological approach in order to arrive at valid conclusions about the general value-creationpotential within the industry. 2.3.2 Research Approaches Besides considering different time horizons in assessing the success of mergers and acquisitions, the variety of available research approaches also contributes to the complexity of long-term M&A analysis. Bruner (2002) describes four main research approaches for determining a perspective on M&A success and profitability: event studies, accounting studies, executive surveys, and clinical studies. The author also concludes that a comprehensive assessment of M&A success must look for patterns of confirmation across approaches and studies. Consequently, this study makes use of a variety of different approaches as it claims to provide a comprehensive view on the situation within the automotive supply industry. The following section provides an overview of the most frequently used research approaches within the field. Event studies represent the most commonly used research approach that dominates the field since the 1970s (Bruner (2002)). Event studies build on share return information and examine the abnormal returns to shareholders around or after the announcement of a transaction. As pointed out in the previous section, event studies can either have a short-term or a long-term focus. Abnormal returns are determined by comparing the returns of an acquirer or target to a benchmark return, usually the expected returns as determined by the capital asset pricing model in short-term event studies or the benchmark returns of a matching firm in long-term event studies. Event studies are generally forward-looking and build on the efficient market hypothesis: it is assumed that share prices represent the present value of expected future cash flows to shareholders (Bruner (2002)). However, given the subjectivity in assessing future cash flows, event studies are also exposed to the risk that some information may not be correctly incorporated into

20


the share price (Eberhart, Maxwell, and Siddique (2004)). In addition, short-term eventstudy results are sensitive to the event and estimation periods chosen; while abnormal returns are usually mitigated in larger event-windows, larger estimation periods are likely contaminated by other confounding events (Rhoades (1994)). The results obtained in long-term event studies, however, are often subject to the respective long-term methodology and the benchmark chosen. As pointed out in the previous section, the later requires a comprehensive methodological approach to overcome inherent statistical biases. Accounting studies represent the second research approach frequently applied in assessing value creation through M&A. These studies examine reported financial results and answer the question whether acquirers outperformed their non-merging peers. While focusing on performance indicators or general balance sheet structure, accounting studies are therefore primarily backward-looking and build on accredited published accounting statements. While published data carries credibility to the reader, a general comparability of published data across different years and reporting standards does not always apply (Bruner (2002)). To overcome the shortcomings inherent in the two broad research approaches, preceding literature commonly applies case or clinical studies as a third complementary approach. By focusing on a single transaction, they facilitate a more detailed analysis of outstanding transaction phenomena that usually exceed the scope of large-sample event or accounting studies (Kaplan, Mitchell, and Wruck (1997)). Case studies generally follow a structured approach inducing new insights from a detailed qualitative description of a real transaction and an evaluation of its performance along a number of pre-defined criteria that can be either quantitative or qualitative in nature (Eisenhardt (1989)). As a last research approach, Bruner (2002) presents executives surveys in which generalizations from a limited number of executive questionnaires are drawn. Although the respective studies are generally based on standardized questionnaires, a similar sampling bias applies as present in executive interviews (Kaplan et al. (1997)). As mainly


21

acquirers in successful transactions are willing to generously provide information, either in interviews or in questionnaires, this study refrains from applying this last approach. Instead, the following chapters will focus on providing a comprehensive overview of the long-term value creation through M&A in the automotive supply industry using a comprehensive mix of event-study, accounting-study, and case-study methodologies.

3

Study 1: Determinants of Capital Market Performance2

3.1

Introduction Over the last decades, wealth creation through mergers and acquisitions has

been extensively discussed in empirical finance research. With a variety of different approaches and foci, authors focusing on short-term announcement effects unambiguously conclude that mergers and acquisitions seem to create value. However, as this short-term value creation potential is mostly attributed to the shareholders of the target companies (Bradley, Desai, and Kim (1988), p.31), a closer look at the returns to acquiring firms reveals a different pattern: While short-term announcement returns for acquiring companies average at around zero (Bruner (2002)), long-term post-merger returns even indicate significant value losses on the acquirer side.3 In the light of ongoing merger activity and consolidation, these negative reactions pose a challenge to the management of merging companies and call for a comprehensive list of determinants for the direction and magnitude of the experienced value effect. Towards the turn of the last century, many industries including the automotive supply industry were facing increasing merger activity. The pressure to produce better equipped and less expensive automobiles created a growing trend towards specialization and internationalization of the industry. While some product ranges such as braking components are now dominated by very few players, acquisitions and foreign direct investments also led to geographical expansion of players across borders and across continents (Sadler (1999)). In the light of these specific industry characteristics, previous research in the automotive supply industry shows that acquiring companies realize significant positive short-term returns as an expression of the global synergy and efficiency potential underlying the transactions (Mentz and Schiereck (2008)). 2

3

An excerpt of this chapter has already been published in the Journal of Economics and Finance; see Laabs and Schiereck (2009). See, for example, Agrawal, Jaffe, and Mandelker (1992), who report a significant value loss of about 10% to acquiring companies over a five-year period following merger completion.

24

3 Study 1: Determinants of Capital Market Performance

However, the question, whether acquirers are able to sustain these exceptional positive returns beyond a short-term announcement window, remains open. Due to various methodological difficulties associated with studies on long-term abnormal performance, evidence from other industries remains scarce or narrows its focus on either one of the time horizons (short- vs. long-term) or on a single deal characteristic (method of payment, cross-border). In contrast to previous research, this study determines the shortand long-term performance of acquiring firms in the automotive supply industry, it uses a combination of Buy-and-Hold Abnormal Returns and expected returns based on the Fama-French-3-Factor-model to determine statistically reliable indications of long-term performance, and it analyzes a comprehensive list of transaction and acquirer characteristics for their respective impact on the short- and long-term wealth effect. The objective of this study is twofold: After updating previously published announcement effects on acquirers in the automotive supply industry, the short-term perspective is expanded by long-term abnormal returns based on event-time (BHARs) and calendar-time (FF3F). Then the observed return patterns are examined to detect and categorize determinant variables. The main focus lies in determining an effect based on the differences in geography, product range, transaction size, and bidder experience. To support the findings, a regression analysis is used to determine correlations and to test for statistical significance. The remainder of this study is organized as follows: Section 3.2 provides a brief overview of the relevant literature and outlines the derived hypotheses for the short- and long-term value creation effects on acquiring companies in the automotive supply industry. Section 3.3 presents the applied methodology as well as the sample selection procedure. The following Section 3.4 contains the empirical results including the overall short-term announcement effects and the long-term performance as well as the analyses concerning different value drivers. Section 3.5 complements these results with a crosssectional analysis before Section 3.6 summarizes the findings and concludes.

3.2 Literature Review and Hypotheses

3.2

25

Literature Review and Hypotheses

3.2.1 Related Literature A significant number of studies have examined the short-term announcement effect and the long-term capital market performance of acquiring firms. For the short-term announcement effect, previous M&A research has found that corporate control transactions in form of mergers or acquisitions are in general associated with significant wealth creation. Jensen and Ruback (1983) summarize early findings from the 1950s to 1970s with positive returns to bidders and targets: On average, they find positive cumulative abnormal returns (CARs) of +29.09% to target shareholders and +3.81% to bidder shareholders, both within the month of a successful merger or tender offer announcement. However, research has also shown that positive returns to acquirers are decreasing over time: Bradley et al. (1988) describe how the average abnormal return for bidders falls from a significant +4.1% in the 1960s, to +1.3% in the 1970s, and to a significant 2.9% in the 1980s. More recent studies support these findings and demonstrate a more negative effect for acquiring companies: using a sample of 4,256 merger events between 1973 and 1998, Andrade, Mitchell and Stafford (2001) find a negative but insignificant -0.7% return for acquiring companies within a three-day event-window surrounding the announcement date. Bruner (2002) provides a comprehensive overview of 44 studies investigating abnormal returns to acquiring companies and summarizes abnormal returns to be essentially zero; Loughran and Vijh (1997) conclude that acquirer returns are overall negative or, at most, insignificantly positive. In the light of this conclusion, previous findings from the automotive supply industry clearly identify the industry as an outlier. Using a sample of 201 M&A transactions in the automotive supply industry between 1981 and 2004, Mentz (2006) finds a significant positive abnormal return to acquiring companies of +1.6% in the eleven-day event period surrounding the announcement date. This result stands in clear contrast not

26


only to the general results presented above but also to other event studies with industry focus.4 The author argues this finding to be the result of industry-specific synergy potentials perceived by capital markets: for the automotive supply industry, mergers and acquisitions seem to be a feasible measure to realize synergy- and efficiency-gains. Mentz (2006) also analyzes a number of underlying variables influencing the positive returns to bidders and concludes that bidder returns for automotive suppliers are positively influenced by a number of characteristics: an acquirer located in Northern America (vs. Asia and Europe), a transaction date between 1993 and 1998, a national transaction (vs. cross-border), deal financing without stocks and fully in cash, a public target, an acquirer of above average size, and an acquirer which has not completed another deal in the three years preceding the regarded transaction. These findings are discussed in more detail in the following section in order to derive appropriate hypotheses for the results of this study. The long-term post-acquisition performance of acquiring firms has been analyzed since the 1970s and can be characterized by three major research phases.5 Phase 1 contains the earlier work of the 1970s and 1980s, in which the analysis of long-term performance is mostly treated as a side-note to short-term event studies. Among the first to analyze long-term performance was Mandelker (1974): Based on a merger sample of 241 mergers, the author determines insignificant negative long-term abnormal returns of -1.4% over a 40-month period. By using the standard market model normally applied in short-term studies, Malatesta (1983) finds a significant -7.6% for 256 mergers in a 12month period following the merger event. Since the appearance of significant long-term abnormal returns contradicts the commonly-held Efficient Market Hypothesis (EMH), 4

5

Other event studies focusing on single industries mainly report negative abnormal returns to acquiring companies, for example -0.6% for electric utilities (Berry (2000)), -0.3% for telecommunications (Akdo÷u (2009)), and -0.1% for banks (Beitel, Schiereck, and Wahrenburg (2004)). Agrawal and Jaffe (2000) originally distinguish two major phases: Phase 1 represents the earlier work of the 1970s and 1980s and phase 2 the later, more advanced methodological research of the 1990s. However, since more recent studies apply additionally advanced methodologies, a third phase starting at the end of the 1990s is added to distinguish between the more recent work on methodologies and the methodological foundations of phase 2.


27

the interest in long-term return behavior increased in the 1980s. Other studies following in this period include Barnes (1984), Bradley and Jarrell (1988), and Franks and Harris (1989). Agrawal and Jaffe (2000) provide a comprehensive overview of these various early studies. Franks et al. (1991) represent the turning point in the analysis of long-term postmerger performance and the start of phase 2. Their study was the first to be exclusively devoted to the long-term performance of acquiring firms. By introducing sophisticated benchmarks and by combining calendar- and event-time approaches, the authors clearly set themselves apart from the simple statistical approaches employed during phase 1. Nevertheless, based on their calendar-time approach, they do not find evidence for abnormal returns to acquirers and conclude that previous findings are the result of misspecification in the appropriate benchmarks. About a year later, Agrawal et al. (1992) find that the non-existence of abnormal returns is specific to the time period examined by Franks et al. (1991): Overall, Agrawal et al. find a significant negative abnormal return of 10% in the five years following a merger. Similar results are presented by Loderer and Martin (1992), Kennedy and Limmack (1996), Gregory (1997), and Rau and Vermaelen (1998): All find negative abnormal returns between -1% and -18% within two to five years after the transactions. In general, the studies of phase 2 can be characterized by an advanced methodology to arrive at abnormal returns; most of the studies employ character-basedmatching approaches based on size, risk and market-to-book ratios or expected-return models such as the ten-factor-model employed by Franks et al. (1991). Although these models are generally more reliable and sophisticated than the ones employed in phase 1, they still fail to address some of the problems that usually afflict long-run abnormal performance. For example, Barber and Lyon (1997) find that test statistics calculated on the basis of reference portfolios are subject to the rebalancing and new listing biases; they also document positive skewness in CARs and in BHARs, a finding that inhibits any inference on the basis of a normality assumption. Mitchell and Stafford (2000) ar-

28


gue that the bootstrapping methodology used to infer from skewed abnormal returns is inappropriate. Since acquisitions are often clustered by industry, the underlying assumption of event independence is breached. Starting in the late 1990s, these methodological challenges are explicitly addressed by the advanced BHAR-approaches applied in the studies of phase 3. The first study associated with phase 3 is Loughran and Vijh (1997): Using a sample 947 transactions between 1970 and 1989, the authors find negative BHARs of -6.5% within a fiveyear time period following the transaction. They also document a difference between merger transactions and acquisitions: While mergers lead to negative returns of -15.9%, acquisitions yield positive returns of +43%. Likewise they find that stock-financed transactions yield negative returns while cash-financed transactions provide positive long-term abnormal returns to acquirers. Overall, Loughran and Vijh were the first to apply the advanced BHAR methodology in which they determine abnormal returns of acquirers on the basis of control-firms matched by market value and market-to-bookratio. Mitchell and Stafford (2000) calculate BHARs for a sample of 2,068 US transactions between 1961 and 1993 and find insignificant negative returns of -1% (equalweighted) and significant abnormal returns of -3.8% (value-weighted) over a three-year period. Since additional calculations using a 3-Factor model only yield significant returns with equal-weights, they conclude that their BHAR-results might be overstated. Bouwman, Fuller, and Nain (2009) show that significant long-term returns depend on the valuation of the market. While acquirers in highly-valued markets experience significant positive abnormal short-term returns and negative long-term returns, acquirers in lowly-valued markets experience short-term insignificant positive reactions and longterm insignificant positive returns. However, the majority of the more recent studies in this field point to negative abnormal returns. Significant abnormal returns are mostly derived using BHAR-approaches while calendar-time approaches yield mixed results.


29

Other examples of negative long-term returns are provided by Black, Carnes, and Jandik. (2001), Sinha (2004), and Gregory and Matatko (2004). 3.2.2

Hypotheses

3.2.2.1 The Overall Effect Previous research reports significant positive short-term announcement effects to acquirers in the automotive supply industry and thereby identifies the industry as a strong outlier compared to other industries (Mentz and Schiereck (2008)). Since this study is based on transactions within the same industry, it is also expected to determine positive short-term announcement returns. From a long-term perspective, the majority of preceding studies points to negative abnormal returns, even if the particular industry benefits from regulated and protected profit margins such as the telecommunications industry (Ferris and Park (2002)). Therefore, it is hypothesized that acquirers in the automotive supply industry realize negative long-term abnormal returns as well. Given the positive short-term return behavior, the long-term effect is assumed to be slightly negative and insignificant. H3.1: Acquirers in the automotives supply industry experience positive announcement effects, which carry over as an insignificant long-term underperformance. 3.2.2.2 The Impact of Transaction Characteristics The following two sections contain a number of hypotheses concerning the impact of transaction and acquirer characteristics on both the short-term announcement effect as well as the post-merger capital market performance of acquirers. Since one focus of this study lies on detecting long-term performance in the automotive supply industry and since long-term performance is often argued to be a function of long-term post-merger management and synergy realization, this section focuses on variables assumed to directly influence the available synergy potential, namely geographic expan-

30


sion, product diversification, transaction size, product sectors, and bidding experience. Other variables such as timing, public status of the target, and acquirer continent are also analyzed but omitted from the following univariate analysis. Table 3.1 provides an overview of the derived hypotheses and the assumed incremental effects. Table 3.1: Overview of Hypotheses and Predicted Value Impact on Acquirers Predicted Value Impact On Acquiring Companies Overall Effect Incremental Effect of Transaction Characteristics Geographical Expansion Cross-border deal Cross-continental deal Diversification Transaction Size Incremental Effect of Acquirer Characteristics Product Sector Engine Drive Components Electrical Components Acquisition Experience Multi Acquirer

Hypothesis 3.1

ShortTerm Positive (significant)

LongTerm Negative (insignificant)

None Negative Negative Positive

Negative Negative Negative Positive

Negative Negative Negative

Negative Negative Negative

Negative

Positive

3.2

3.3 3.4 3.5

3.6

A large number of studies focus on the effect of internationalization on shortterm value creation for acquirers, but yield mixed results. While some conclude that geographical expansion in form of cross-border transactions yield positive returns for acquirers (Goergen and Renneboog (2004)), others argue that these effects are characteristic to certain industries and countries (Dewenter (1995), Kiymaz and Mukherjee (2000)). For the automotive supply industry, Mentz and Schiereck (2008) do not find significant short-term return differences between national and cross-border transactions for acquirers while transcontinental deals have a negative influence. The literature on long-run performance is less ambiguous and unanimously points to long-run negative returns to acquiring companies in cross-border transactions due to the more challenging post-merger integration and imperfect information (Conn, Cosh, Guest, and Hughes (2005), Aw and Chatterjee (2004), and Black et al. (2001)).


31

H3.2: Cross-border transactions yield no significant impact on short-term announcement returns, but will significantly decrease long-term performance. In the past, activity diversification has normally been followed by a negative impact on long-term performance (Agrawal et al. (1992), Ferris and Park (2002)). These observed return patterns can be argued to be the result of a more difficult synergy realization process in diversifying mergers and acquisitions. Given the complexity and competition patterns in the automotive supply industry, where various suppliers produce a wide range of products from high-tech electrical components to tires, it is assumed that synergy- and efficiency-gains can be realized more easily in focusing transactions. Diversification into other product lines will have a negative impact on returns. H3.3: Product diversification negatively impacts short- and long-term returns to acquiring companies. The effect of transaction size on value creation for acquiring companies is regarded less extensively in prior research. Ferris and Park (2002) argue that larger transactions are more likely to result in economies of scale in research and production facilities and present corresponding supportive results for a long-term study in the telecommunications industry. As the automotive supply industry can primarily be regarded as a production industry, it is assumed that it shows similar reactions as previously determined for the telecommunications industry: The larger the transaction, the better the economies of scale and the stronger the positive influence on returns. H3.4: Larger transactions positively influence short- and long-term returns. 3.2.2.3 The Impact of Acquirer Characteristics The automotive supply industry contains a number of different product segments. Competitiveness within each segment strongly depends on the individual prod-

32


uct: As few suppliers deliver very complex products, for example Xenon-headlights, competition in this particular segment can be regarded as low. In other segments such as the tire or leather supply segment, competition is significantly higher due to a higher number of producers (Aktas et al. (2004)). Furthermore, some industry segments including suppliers of tires, forgings, and bearings suffer from over-capacity which in turn increases the degree of rivalry experienced within these segments (Carr (1993)). Therefore, it is inferred that the effects of M&A activity on acquirers within the automotive supply industry exhibit different magnitudes depending on the observed product sector of operations. Acquirers in product segments that are exposed to stronger competition and overcapacity are assumed to realize synergy gains more efficiently than other acquirers. Therefore, they are expected to experience a more positive impact on short- and long-term returns. H3.5: Takeovers involving acquirers from the exterior, interior or tires product segments yield positively influenced short- and long-term returns. The last acquirer characteristic to be analyzed here is the acquisition experience of the acquirer. Haleblian and Finkelstein (1999) determine a positive correlation between the squared number of completed transactions of an acquirer and the magnitude of the cumulative abnormal returns experienced in short-term event-windows. They argue that acquirers gain target-integrating experience that they can leverage in additional transactions. Fuller, Netter, and Stegemoller (2002), however, provide evidence that acquirers with more than five transactions within a three-year period before the observed transaction yield significant negative returns. For the automotive supply industry, acquirers without any bidding experience within the preceding three years of a transaction experience significantly higher positive announcement returns than experienced acquirers. Whether this observation can be attributed to management hubris or other explanations remains unanswered (Mentz (2006)).

3.3 Data and Methodology

33

Previous findings from the long-term perspective support the argument led by Haleblian and Finkelstein: Antoniou and Zhao (2004) analyze a sample of 179 transactions by UK acquirers between 1991 and 1998 and find a significant positive influence of multi-bidders on post-merger capital market returns. While multi-bidders outperform the original sample average, single-bidders underperform with respect to the original sample performance. Therefore, it is assumed that the short-run outperformance of single-bidders in the automotive supply industry as described by Mentz (2006) holds true for the announcement effects, but fails to persist in the long-run. Over three years, the integration experience of multi-bidders will yield more positive returns and outperform the single-bidders. H3.6: Multi-bidders experience lower short-term announcement effects, but outperform single-bidders in the long-run.

3.3

Data and Methodology

3.3.1

Identifying Merging Companies The sample of mergers and acquisitions for the empirical event study is drawn

from the Securities Data Company (SDC)/Thomson One Banker Deals database and the Bloomberg M&A database. It includes all takeover events announced between January 1st, 1981, and September 1st, 2007. The total number of M&A deals is reduced to yield only those transactions meeting the following criteria.

At the time of the takeover, the target and acquirer company both possess active operations in the automotive supply industry.

The acquirer intends to purchase 50% or more of the outstanding shares or of the private equity, for publicly traded and privately held targets respectively.

The total transaction value accumulates to at least USD 50 million.

34


The acquiring company is located in one of the following geographic regions: Europe, North and South America, and Asia. In addition, all targets and acquirers are double-checked by a press research us-

ing the Factiva database to ensure that all transactions are horizontal and that announcement dates are correct as provided by the databases. All non-horizontal deals as well as deals involving financial investment companies are excluded. The described selection criteria result in a final takeover sample of 230 events in the automotive supply industry between the years 1981 and 2007. Table 3.2 provides an overview of the frequency distribution over time and reveals a strong concentration of events between the years 1995 and 2000. Since other recent research provides evidence of significant merger clustering within certain industries during the 1990s (see Mulherin and Boone (2000), p.123), and since Andrade and Stafford (2004) find that fifty percent of all mergers within an industry occur within a five-year time period, the distribution over time is assumed to be representative and valid. The average transaction value ranges from USD 63.0 million to USD 2,819.0 million, whereas the later is mainly driven by the USD 15.8 billion acquisition of Siemens VDO by Continental AG in 2007. This transaction likewise represents the largest transaction volume in the sample. In order to determine short- and long-term abnormal returns, daily and monthly adjusted stock prices for all public targets and acquirers are downloaded from the Thomson Datastream database.6

6

To reflect the influence of dividend payments as well as share issuances or repurchases on return data, the adjusted stock prices denoted by data type “RI” were selected.


35

Table 3.2: Overview of the Transaction Sample – Descriptive Statistics7 Year

Transactions

(%)

Avg.Trans.Value (USD million)

1981 1984 1985 1986 1987 1988 1989 1990 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Total

1 3 1 8 2 6 9 5 4 4 5 12 19 26 27 30 11 13 5 8 9 9 7 6 230

0.4 1.3 0.4 3.5 0.9 2.6 3.9 2.2 1.7 1.7 2.2 5.2 8.3 11.3 11.7 13.0 4.8 5.7 2.2 3.5 3.9 3.9 3.0 2.6 100.0

63.0 99.2 597.8 323.7 375.0 813.8 351.5 200.6 188.1 438.3 161.5 265.4 383.1 285.0 680.0 620.0 338.0 316.2 249.9 174.2 280.6 219.1 553.4 2819.0 208.18

Acquirer Region - Number of Transactions Americas Europe Asia Other 1 3 1 6 0 4 4 1 2 4 5 8 16 17 19 16 5 7 3 3 4 4 3 2 138

2 1 1 4 4 2

3 2 9 8 10 5 4 1 1 4 2 1 2 66

1 1 1

1 1

3 1 2 1 3 1 3 3 2 22

1

1

4

3.3.2 Portfolio of Matching Firms Both methods for determining long-term abnormal returns as employed in this study require a set of comparable, non-merging firms from the same industry. Since there is no broad global index available for the automotive supply industry which comforts the purpose of the following analyses, this study determines a portfolio of matching firms from the constituents' lists of broader country-specific indices. This portfolio

7

For the remainder of the document, Asian-Pacific deals outside the automotive triad (North America, Europe, Japan/South Korea) will be presented as "Other Regions" (namely India and Australia).

36


is then used to create an artificial industry index that serves as an input for the FF3Fapproach. The first step in constructing the portfolio of matching firms includes screening the constituents’ lists underlying the country-specific industry indices for companies in the automotive supply industry. Downloading all available constituents’ lists of indices in the automotive parts industry as supplied and constructed by the Thomson Datastream database results in 42 lists of international automotive suppliers.8 In a second step, all 42 constituents' lists are aggregated and compared to the respective lists from prior time periods. In a third step, the initial set of 109 companies derived from the Datastream database is cross-referenced to a list of the top 100 OEMs in the automotive supply sector as published on a yearly basis by the Automotive News (see Automotive News (2003)). Another 21 publicly traded automotive supply companies are added to yield a total number of 130 matching firms. The final portfolio contains 49 Asian, 46 American, and 35 European publicly traded companies. For each firm, daily and monthly stock returns as well as monthly market values and yearly market-to-book ratios are downloaded from the Thomson Datastream database. 3.3.3 Econometric Strategy 3.3.3.1 Short-term Methodology Short-term announcement returns are assessed using the event-study methodology in connection with the standard market model as described by Brown and Warner (1985). Although the choice of the applied model in a short-term context does not significantly influence derived results (see Brown and Warner (1980)), applying the standard model ensures direct comparability to the results from other industries: Both, Mentz (2006) for the automotive supply industry and the majority of other industry-

8

Constituents’ Lists are denominated by the industry-specific MNEMONIC code “LAUPRT” plus a two-digit country-specific suffix. For example, a list of U.S. automotive suppliers constituting the automotive parts industrial index in the US can be derived using the code “LAUPRTUS.”


37

related event studies make use of this approach (see Berry (2000) and Akdo÷u (2009)). Returns under the market model are determined on the basis of Equation (3.1).

Rit = α i + β i ⋅ Rmt + ε it

(3.1)

Rit and Rmt represent the returns in period t of security i and of the market portfolio respectively. İit stands for the zero-mean disturbance term, which is commonly referred to as the abnormal return. On the basis of this model, abnormal returns are determined as described by Equation (3.2).

ARit = Rit − (α i + β i ⋅ Rmt )

(3.2)

As the return of the market portfolio within the model generally refers to a market index associated with the given securities over time, local indices are determined for each country represented in the takeover sample (see Coutts, Mills, and Roberts (1994)). For example, the S&P 500 Composite return data applies for U.S. American companies, return data of the FTSE-All Shares index for companies located in Great Britain, and the DAX 30 index for German companies within the sample. Using different indices for each represented country accounts for regional differences in industry-returns and country-specific risk profiles. The market models are estimated by using Ordinary Least Squares (OLS) regression over a 200-trading-day period starting at trading day t = -250 relative to the earliest announcement date of the M&A event. On the basis of these estimated market model parameters, abnormal returns for all target and acquiring companies are derived.9 Cumulative abnormal returns are calculated as defined by Equation 9

In order to develop a perspective on total shareholder impact, the combined effect on an artificially combined entity is derived on the basis of market-value weighted returns: AR combined

;t

=

AR Acq , t * MV

Acq , t

+ AR Tar , t * MV Tar , t

MV

Acq , t

+ MV Tar , t

38


(3.3). The longest event-window is 41 days: t=[-20;+20], with t = 0 being the announcement date of the transaction.

m2

CAREventWindow ( m1, m 2) =

(3.3)

¦ AR

it

t = m1

1 n ¦ CARP,i n i =1 t= σ (CARP ) / n

(3.4)

To test for statistical significance of the short-term abnormal returns, this study employs three test statistics. The first represents a simple parametric t-test as described by Equation (3.4). The sample average CAR is divided by the standard deviation across the individual company returns over the square root of the number of observations in the sample. This test statistic follows the student's t-distribution for a sample or subsample below 30 observations and a normal distribution when regarding a larger sample (see Barber and Lyon (1996), p. 373f.). The cross-sectional test as proposed by Boehmer, Musumeci, and Poulsen (1991) accounts for a potential event-induced increase in standard deviation by combining variance information from the event and the estimation period. Equation (3.5) describes the corresponding test statistic and Equation (3.5a) the respective standardization procedure to standardize abnormal returns with the standard deviation of returns in the estimation period.10

(3.5)

10

t=

1 N

N

¦ SAR

i ,t

i =1

N SAR § 1 i ,t ¨ SAR − ¦ ¦ i ,t N ( N − 1) ¨ N i =1 ©

· ¸¸ ¹

2

For a more detailed description see Boehmer et al. (1991), as well as Mikkelson and Partch (1988).


(3.5a)

SARi ,t =

39

CARi ,e1 −e2 SDE ,i ,t

;with

SDE ,i ,t = SDi ,t

§ · ¨ ¸ 2 − ( R R ) 1 ¨ ¸ M ,t M 1+ + ¨ T ¸ T 2 ¨¨ ¦ ( RM , j − RM ) ¸¸ © j =1 ¹

With: SARi,t = Standardized abnormal return for company i on day t SDE,i,t = Standard error of the event period for company i on day t SDi,t = Standard error of the estimation period for company i

T = Number of days in the estimation period RM,t = Market return on day t of the event period RM,j = Market return on day j of the estimation period RM = Average return of the index in the event period Since prior research provides some evidence that non-parametric test statistics can be more powerful than parametric t-statistics (Barber and Lyon (1996), p.360), the non-parametric Generalised Sign Test (GST) as proposed by Cowan, Nandkumar, and Singh (1990) completes the test statistics applied in the short-term study. The GST tests whether the percentage of positive observations (p) within a sample is greater than the expected percentage of positive observations (p') (Equation (3.6)). The expected percentage of positive observations is derived on the basis of the abnormal returns in the estimation period (Equation (3.7)).11

(3.6)

11

z=

N ∗ p − E ( N ∗ p' ) N ∗ p'∗(1 − p' )

; with:

In addition, differences in means of two subsamples are tested for statistical significance with the methodology applied by Baradwaj, Dubofsky, and Fraser (1992) and Siems (1996).

40


N = Number of companies E = "Expectation Operator" P = Percentage of positive observations p' = Expected percentage of positive observations

(3.7)

p' =

1 n 1 E200 ¦ ¦ S i ,t ; with: n i =1 200 t = E1

Si,t = 1 for AR i,t >0 and AR i,t = 0 otherwise.

3.3.3.2 Long-term Methodology

To address the various statistical problems associated with the different methods of determining long-term capital market performance, a combination of the two most accepted methodologies is applied. BHARs represent the most commonly-used methodology in determining long-term returns in event-time (see, for example, Ritter (1991), Barber and Lyon (1997), and Lyon et al. (1999)). However, it is exposed to significant cross-correlation, especially due to the application of not randomly selected matching samples. The FF3F as a common representative of the calendar-time approaches eliminates the cross-correlation problem (Fama and French (1993)), while at the same time being exposed to other potential pitfalls such as heteroscedasticity. The BHARs are determined using a character-based matching approach, which individually matches non-merging counter-parts to the acquiring firms on the basis of market values and market-to-book ratios.12 The matching procedure follows the approach proposed by Lyon et al. (1999):

For each public acquiring firm in the takeover sample, the relevant size (market value) and market-to-book ratio is determined. Market values are determined as

12

Market values and market-to-book-ratios have been frequently used in prior research and appear to be the dominant matching criteria in the field. Examples include Lyon et al. (1999), Mitchell and Stafford (2000), Rau and Vermaelen (1998), and Loughran and Vijh (1997).


41

the last quote in the last June preceding the transaction, market-to-book ratios are derived for the last completed fiscal year before the transaction.13

Likewise, for each year, market values at the end of June and market-to-book ratios are downloaded for the full list of matching firms. For each acquisition, the total list of potential matches is then reduced to yield only those firms for which both descriptive data fields as well as monthly return data are available.

In a second step, the list of potential matches is reduced to those companies with a market value in the range of 70% to 130% of the acquirer's market value.

Finally, from the list of companies with a market value between 70% and 130% of the acquirer's market value, the one with the smallest absolute difference in market-to-book ratio is selected as the control-firm for the analysis. Based on this procedure, control-firms for each transaction and each acquiring

company between 1980 and 2004 are determined. These control-firms represent the benchmark in determining abnormal returns to acquirers. Equation (3.8) shows how the abnormal returns (BHARs) are derived as the difference between the Buy-and-Hold Return (BHR) of an investor in the acquiring company and the BHR of an investor in the control-firm. BHARs are determined for a period of 36 months following the takeover announcement. The average BHARs for the total sample as well as for the respective subsamples are calculated as an equal-weighted average and as a value-weighted average where the respective market value at the end of the preceding June serves as the weight. Statistical significance is tested using standard t-statistic.

13

All information is downloaded from the Thomson Datastream database.

42


s +T

(3.8)

s +T

BHARi ,T = ∏ (1 + Ri ,t ) −∏ (1 + Rcontrolfirm,t ) ; with: t =s

t =s

BHARi ,T = BHAR of company i over T months Ri ,t = Monthly return of company i in month t Rcontrolfirm ,t = Monthly return of the control-firm in month t

The second model employed relates to the most commonly used calendar-time approach originally developed by Fama and French (1993). The "Fama-French-3Factor-model" (FF3F) determines abnormal returns of acquiring companies by regressing a time series of acquirers' excess returns (return less risk-free rate) on the time series of market excess returns, the time series of the differences in returns of small and big companies (SMB), and the time series of differences in returns of companies with high and low (HML) market-to-book values (see Equation (3.9)).

(3.9)

R p ,t − R f ,t = a p − b p ( RM ,t − R f ,t ) + s p SMB + h p HML + e p ,t

The acquirers' return Rp,t is determined as the average return of a continuously changing acquirer portfolio: Each acquirer is part of the acquirer portfolio for 36 months after the announcement of its respective acquisition, the portfolio in month t therefore contains all acquirers which were active in the preceding 36-month period. The average return across the portfolio is calculated with equal-weights as well as value-weights and yields two time series of acquirer returns in calendar-time. The one-to-three-year US treasury rate serves as an approximation for the risk-free-rate. The market return in the automotive supply industry RM,t is calculated as the artificial index return of the matching portfolio as described in Section 3.3.2. SMB represents the monthly differences in returns between small and big companies in the matching portfolio; HML determines

3.4 Empirical Results

43

the return differences between companies form the matching portfolio with high and with low market-to-book ratios. The time series of SMB and HML are determined using the procedure proposed by Fama and French (1993): In a first step, all matching companies are ranked based on their market value at the end of June of each year. Using the median as a separator, two portfolios are created, one with the small (S) and one with the big (B) companies. Likewise, the sample is ranked by market-to-book ratios and divided into three portfolios with the 30% highest (H), the 30% lowest (L), and the remaining 40% middle (M) market-to-book ratio companies. The derived portfolios are combined to yield a total of six portfolios (S/L, S/M, S/H, B/L, B/M, and B/H). The return differences between small and big companies are calculated as the difference between the simple averages of the three portfolios containing big companies (B/L, B/M, and B/H) and the three portfolios containing small companies (S/L, S/M, and S/H). For the individual portfolio returns, value-weighted averages are applied. The time series of HML are determined in a similar way, the monthly return differences are represented by the difference in simple means of the two portfolios with high market-to-book ratios and the two portfolios with low market-to-book ratios. The factors ap, bp, sp, hp, and ep,t are the results of the regression analysis; the intercept ap represents the average monthly abnormal return to the observed acquirers over the 36 months following an acquisition. Statistical significance of these factors is determined using standard t-statistic.

3.4

Empirical Results

3.4.1 The Overall Effect

Table 3.3 reports the short-term announcement effect of mergers and acquisitions on the total sample of acquiring firms in the automotive supply industry. Upon the immediate announcement of a transaction, acquirers earn a highly significant 0.98% in the [-1,0] event-window. In the 31 and 41 days around the announcement day, the value

44


gain increases to approximately 2%. The majority of returns are highly significant at the 1%-level. On the one hand, this finding confirms prior research in this industry and, on the other hand, the exceptional position of the industry: Unlike results from the majority of other industries, positive short-term returns to acquirers represent the capital market's perception of extraordinary synergy potentials in the automotive supply industry. As a result, Figure 3.1 shows that all participating deal parties gain from this perception and realize positive abnormal returns upon the announcement of a deal.14 Table 3.3: CAARs to Acquirers Acquirers (n=206) Event-

t-Test

z-Test

Generalized Sign Test

Window CAAR t-value p-value z-value p-value P z-Value p-value [-20,20] 2.23% 1.94 0.03 ** 2.28 0.01 ** 52% 1.76 0.04 ** [-20,10] 2.03% 1.99 0.02 ** 2.37

The Long-Term Success of Mergers and Acquisitions in the International Automotive Supply Industry

Mergers and Acquisitions in the Machinery Industry

Mergers and acquisitions basics

Mergers and acquisitions

Mergers and Acquisitions

Acquisitions and Mergers

Mergers and Acquisitions

Advances in Mergers and Acquisitions, Volume 4 (Advances in Mergers and Acquisitions) (Advances in Mergers and Acquisitions)

The Management of Mergers and Acquisitions

Mergers, acquisitions, and corporate restructurings

Copper in the Automotive Industry

Advances in Mergers and Acquisitions, Volume 5

Plastics in the Automotive Industry

Mergers and Acquisitions (Express Exec)

Mergers, Acquisitions and Corporate Restructuring

German Mergers & Acquisitions in the USA

Advances in Mergers and Acquisitions, Volume 7 (Studies in Mergers and Acquisitions)

Advances in Mergers and Acquisitions, Volume 3 (Advances in Mergers and Acquisitions)

Intellectual Property Assets in Mergers and Acquisitions

The Social Side of Mergers and Acquisitions: Cooperation relationships after mergers and acquisitions

Mergers & Acquisitions

Advances in Mergers and Acquisitions, Volume 2 (Advances in Mergers and Acquisitions)

Value Creation in Mergers, Acquisitions, and Alliances

Mergers, Acquisitions and Corporate Restructuring

Mergers & Acquisitions

Advances in Mergers and Acquisitions, Volume 6 (Advances in Mergers and Acquisitions)

Advances in Mergers and Acquisitions, Volume 8

Intellectual Property Assets in Mergers and Acquisitions

Big Deal: Mergers and Acquisitions in the Digital Age

Big Deal: Mergers and Acquisitions in the Digital Age

The Management of International Acquisitions

The Long-Term Success of Mergers and Acquisitions in the International Automotive Supply Industry

Mergers and Acquisitions in the Machinery Industry

Mergers and acquisitions basics

Mergers and acquisitions

Mergers and Acquisitions

Acquisitions and Mergers

Mergers and Acquisitions

Advances in Mergers and Acquisitions, Volume 4 (Advances in Mergers and Acquisitions) (Advances in Mergers and Acquisitions)

The Management of Mergers and Acquisitions

Mergers, acquisitions, and corporate restructurings

Copper in the Automotive Industry

Advances in Mergers and Acquisitions, Volume 5

Plastics in the Automotive Industry

Mergers and Acquisitions (Express Exec)

Mergers, Acquisitions and Corporate Restructuring

German Mergers & Acquisitions in the USA

Advances in Mergers and Acquisitions, Volume 7 (Studies in Mergers and Acquisitions)

Advances in Mergers and Acquisitions, Volume 3 (Advances in Mergers and Acquisitions)

Intellectual Property Assets in Mergers and Acquisitions

The Social Side of Mergers and Acquisitions: Cooperation relationships after mergers and acquisitions

Mergers & Acquisitions

Advances in Mergers and Acquisitions, Volume 2 (Advances in Mergers and Acquisitions)

Value Creation in Mergers, Acquisitions, and Alliances

Mergers, Acquisitions and Corporate Restructuring

Mergers & Acquisitions

Advances in Mergers and Acquisitions, Volume 6 (Advances in Mergers and Acquisitions)

Advances in Mergers and Acquisitions, Volume 8

Intellectual Property Assets in Mergers and Acquisitions

Big Deal: Mergers and Acquisitions in the Digital Age

Big Deal: Mergers and Acquisitions in the Digital Age

The Management of International Acquisitions

Recommend Documents