The Dynamics within Merger Waves

Evidence from the Industry Merger Waves of the 1990s


Doctoral Thesis / Dissertation, 2008

163 Pages, Grade: Summa cum Laude


Excerpt

Table of Contents

List of Tables

List of Abbreviations

1 Introduction
1.1 Background and objective
1.2 Organization of the dissertation

2 Literature review
2.1 Motives of M&A transactions
2.1.1 Value-increasing motives
2.1.1.1 The productive efficiency hypothesis
2.1.1.2 The collusion hypothesis
2.1.1.3 The buying power hypothesis
2.1.2 Value-decreasing motives
2.1.2.1 The hubris hypothesis
2.1.2.2 Agency theories
2.1.2.3 Overvaluation theory
2.1.2.4 Competitive advantage theory
2.1.3 Summary
2.2 Aggregate and industry-specific merger waves
2.2.1 The time-series behaviour of M&A activity
2.2.2 Reasons of the wave pattern
2.2.2.1 Potential explanations of merger waves
2.2.2.1.1 Neoclassical explanations of merger waves
2.2.2.1.2 Behavioural explanations of merger waves
2.2.2.2 Neoclassical versus behavioural explanations of merger waves
2.2.3 Summary
2.3 The success of M&A transactions
2.3.1 The event study methodology
2.3.1.1 Outline of an event study
2.3.1.2 Estimation of abnormal returns
2.3.1.3 Hypotheses testing
2.3.1.3.1 Single hypotheses tests of means
2.3.1.3.2 Single hypotheses tests of medians
2.3.1.3.3 Tests of equality
2.3.2 Empirical results for the transaction success of the merging firms
2.3.3 Summary
2.4 The determinants of the transaction success
2.4.1 Relative size of the target
2.4.2 Absolute size of the bidder
2.4.3 Excess cash of the bidder
2.4.4 Financial leverage of the bidder
2.4.5 Diversification of the transaction
2.4.6 Method of payment
2.4.7 Valuation of the bidder
2.4.8 Organizational form of the target
2.4.9 Summary
2.5 Intra-industry effects of mergers and acquisitions
2.5.1 Sources of intra-industry effects
2.5.1.1 The influence of the acquisition motive on the returns to rivals
2.5.1.2 The acquisition probability hypothesis
2.5.2 Empirical results for the intra-industry effects of M&A transactions
2.5.3 Determinants of rival returns
2.5.3.1 Relatedness of the acquisition
2.5.3.2 Type of target
2.5.3.3 Method of financing
2.5.3.4 Size of the transaction and the bidder
2.5.3.5 Industry concentration
2.5.3.6 Summary
2.6 Summary

3 Abnormal returns within merger waves
3.1 Data
3.1.1 Data sources
3.1.2 Transaction data
3.1.3 Industry merger waves
3.1.3.1 Industry assignment
3.1.3.2 Wave identification process
3.1.3.3 Sample of deals within industry merger waves
3.2 Abnormal returns to bidders and rivals within merger waves
3.2.1 Methodology
3.2.2 Abnormal returns to bidders within different stages of merger waves
3.2.3 Abnormal returns to rivals within different stages of merger waves
3.3 Summary

4 The robustness of the wave effect
4.1 Bidding firm and deal characteristics in merger waves
4.1.1 Operationalization of deal and bidding firm characteristics
4.1.2 Comparison of bidding firm and deal characteristics
4.2 The robustness of the wave effect to bidding firms
4.2.1 The wave effect to bidders across subsamples of bids
4.2.2 Multivariate regressions of abnormal returns to bidders
4.3 The robustness of the wave effect to rivals
4.3.1 The wave effect to rivals across subsamples of bids
4.3.2 Multivariate regressions of abnormal returns to rivals
4.4 Summary

5 The explanation for the returns within merger waves
5.1 Transaction motives at different stages of merger waves
5.1.1 The changing importance of motives within merger waves
5.1.2 Conditions for an adequate research design
5.2 The competitive advantage theory and the wave effect
5.2.1 Course of investigation
5.2.2 Do only late wave bidders overpay?
5.2.2.1 Theoretical predictions of the competitive advantage theory
5.2.2.2 Empirical results
5.2.2.3 Summary
5.2.3 Why do only late wave bidders overpay?
5.2.3.1 Scarcity of assets and the valuation ratio
5.2.3.2 Methodological approach
5.2.3.3 Empirical results
5.2.3.4 Summary
5.3 Other merger-related theories and the wave effect
5.3.1 The hubris hypothesis and the wave effect to bidders
5.3.2 Agency-related problems at early and late wave stages
5.4 Summary

6 Conclusion

Appendix

References

List of Tables

Table 2.1: Transaction motives and the gains to bidders
Table 2.2: The neoclassical and behavioural hypothesis of merger waves
Table 2.3: Summary of empirical results regarding the gains to bidding firms
Table 2.4: Determinants of bidder returns
Table 2.5: Transaction motives and the gains to rivals
Table 2.6: Determinants of the gains to rival firms

Table 3.1: Aggregate transaction value and number of deals sorted by announcement year
Table 3.2: The Fama and French (1997) industry classification scheme
Table 3.3: Industry merger waves
Table 3.4: CAR to bidders at different stages of merger waves
Table 3.5: CAR to rivals at different stages of merger waves

Table 4.1: Operationalization and sources of deal and bidder characteristics
Table 4.2: Descriptive statistics of deal characteristics
Table 4.3: Descriptive statistics of bidding firm characteristics
Table 4.4: CAR to bidders in merger waves by deal characteristics
Table 4.5: CAR to bidders in merger waves by bidder characteristics
Table 4.6: Regression results for CAR to bidders for method of payment, type of target, and focus based subsamples
Table 4.7: Regression results for CAR to bidders for cash holding, leverage, and size based subsamples
Table 4.8: Regression results for CAR to bidders for Tobin’s q of the bidder and relative size of the target based subsamples
Table 4.9: Regression results for CAR to bidders for method of payment, type of target, and focus based subsamples (alternative activity variable)
Table 4.10: Regression results for CAR to bidders for cash holding, leverage, and size based subsamples (alternative activity variable)
Table 4.11: Regression results for CAR to bidders for Tobin’s q and relative size based subsamples (alternative activity variable)
Table 4.12: CAR to rivals in merger waves across subsamples of bids
Table 4.13: Regression results for CAR to rivals for method of payment, type of target, and focus based subsamples
Table 4.14: Regression results for CAR to rivals for deal value and relative size based subsamples
Table 4.15: Regression results for CAR to rivals for method of payment, type of target, and focus based subsamples (alternative activity variable)
Table 4.16: Regression results for CAR to rivals for deal value and relative size based subsamples (alternative activity variable)

Table 5.1: CAR to rivals in different stages of merger waves for value-increasing and value-decreasing transactions
Table 5.2: CAR to most related rivals in different stages of merger waves for value- increasing and value-decreasing transactions
Table 5.3: Regression results for CAR to rivals for value-decreasing and value- increasing bids
Table 5.4: Summary of the results for the relationship between the return to bidders and rivals at the announcement of value-decreasing bids
Table 5.5: Summary of the results for the relationship between the return to bidders and rivals at the announcement of value-increasing bids
Table 5.6: Target valuation ratios at different stages of merger waves
Table 5.7: Regression results for CAR to rivals at early and late wave bids for publicly traded targets
Table 5.8: Regression results for CAR to bidders at early and late wave bids for publicly traded targets
Table 5.9: Premiums paid to targets at different stages of merger waves
Table 5.10: Summary for the empirical investigation of the target valuation at different stages of merger waves
Table 5.11: Regression results for CAR to frequent bidders at late wave stages
Table 5.12: Regression results for CAR to bidders at the announcement of early and late wave bids separately

List of Abbreviations

illustration not visible in this excerpt

1 Introduction

1.1 Background and objective

The M&A activity of the last century was characterized by five major waves, where most of the M&A transactions took place. These waves are characterized by a sudden heavy increase in both the number of transactions and the total dollar deal volume. However, not all sectors of the economy experience such a rise in M&A activity during aggregate merger waves, but within these waves mergers strongly cluster by industry. Thus, aggregate merger waves can be seen as a clustering of several industry-specific merger waves. Until recently, the financial research did not know the drivers of these industry-specific merger waves. Mitchell and Mulherin (1996) show that the causes of these industry-specific merger waves are economic, regulatory, or technological shocks. Harford (2005) shows that these exogenous shocks only propagate waves, if efficient capital liquidity is present, and that this macro-level component causes industry merger waves to cluster in time to create aggregate merger waves. Overall, the causes of industry-specific merger waves are examined extensively by the financial literature. However, so far no study has highlighted the importance of the special characteristics of these waves. This lack of evidence is surprising as the unique research environment offered by industry-specific merger waves may help to shed light on some major questions within the research field of mergers and acquisitions.

In general, several studies show that cross-sectional deal and bidding firm characteristics determine the gains associated with M&A activity. All works highlight that the mixture of transaction motives, which impacts both the transaction success of the bidding firms and the response of the rivals, differs across subsamples of bids constructed based on these characteristics. However, although a huge amount of empirical evidence exists regarding the impact of cross-sectional determinants, so far no work has investigated how the point in time when the deal is announced influences the gains associated with a transaction announcement.

Considering the dynamics within industry-specific merger waves, the returns associated with a transaction should not be independent of the point in time of the transaction for the following reasons. Within industry merger waves, industries experience a significant consolidation and restructuring process in a very short time period. As the industry structure highly influences the motive for a transaction, the mixture of motives change as an industry-specific rapid consolidation process is ongoing. Therefore, both the transaction success and the associated intra-industry effects should depend on the point in time of the deal within an industry merger wave.

Based on this intuition, the overall objective of this work is to investigate how and why the point in time of a deal influences both the gains to the bidding firms and the response of the competitors to a transaction. Thus, I will document a completely new determinant of the gains associated with M&A transactions and I will show that not only cross-sectional deal and bidding firm characteristics determine the success of a transaction, but that time-series determinants are also important.

To accomplish the basic research objective, the empirical investigation answers three basic questions. At the beginning of my examinations, I investigate how the gains associated with a transaction change during an industry-specific merger wave. Therefore, the first question is the following:

Question1: How does the point in time of a deal within an industry merger wave determine both the transaction success and the intra-industry effects of a transaction?

The results show that both the gains to the bidding firm as well as the response of the industry rivals decrease significantly towards the end of an industry merger wave.

Based on these basic findings, I investigate why the gains associated with a transaction change during an industry-specific merger wave. The financial literature has examined extensively how cross-sectional determinants, like acquiring firm and deal characteristics, influence the transaction success and the stock market gains of the competitors. Thus, the basic results may be due to changing cross-sectional deal or bidding firm characteristics. Therefore, the second question is the following:

Question2: Do changing firm and deal characteristics explain why both the transaction success and the intra-industry effects depend on the point in time of the deal within an industry merger wave?

I find that deal and bidding firm characteristics change towards the end of a wave. However, these changing cross-sectional determinants cannot explain why both the transaction success and the intra-industry effects change dramatically towards the end of a merger wave. This implies that the gains associated with M&A transactions decrease because the underlying mixture of transaction motives changes. A number of hypotheses have been offered to explain why acquisitions should be associated with negative wealth effects. If some of these theories apply only or at least more to bids at the end of a wave, the basic result of decreasing wealth effects within industry merger waves can be explained. Therefore, the third question is the following:

Question3: How does the mixture of transaction motives change towards the end of an industry merger wave?

1.2 Organization of the dissertation

The dissertation is divided into six sections. Based on the theoretical content and motivation of Section 2, Sections 3 to 5 empirically answer the three research questions that are outlined in the prior section. Section 6 concludes.

Section 2 provides an extensive review of the financial literature. The objective of this section is threefold. First, it gives a broad understanding for the major issues in the field of mergers and acquisitions. Second, the review provides the theoretical motivation for the following empirical investigations. Finally, it provides the methodological basis for the research design of the empirical examinations of Sections 3 to 5.

Sections 2.1 and 2.2 illustrate the major sources of M&A activity. First, in Section 2.1, I describe the major theories for why single transactions occur. A number of transaction motives have been offered by the financial literature. These motives directly determine the transaction success. Broadly, they can be divided into value-decreasing and value-increasing motives, depending on the associated wealth effects to the merging firms. Section 2.2 reviews the characteristics and causes of market-wide M&A activity. The level of M&A activity fluctuates heavily, whereby the mass of deals occur within a few industry-specific merger waves. The financial literature offers two major hypotheses for this wave behaviour of mergers and acquisitions. Recent empirical evidence shows that the causes of the industry merger waves are exogenous shocks that force the industry to restructure within a rapid consolidation process.

Sections 2.3 to 2.5 document the empirical evidence regarding the wealth effects of mergers and acquisitions. Section 2.3 reviews the empirical investigations that examine the transaction success of the merging firms. It starts with a description of the standard event study methodology as the dominating approach for measuring the transaction success. Then, it summarizes the major empirical results of a number of studies. The financial literature has shown that the average gain to bidding firm shareholders is essentially zero. Thus, transactions that are undertaken for good reasons are cancelled out by mergers driven by less benign reasons. However, the gains to the bidding firms are not constant across all types of bids. A number of studies provide strong evidence that acquiring firm and deal characteristics influence the transaction success. These cross-sectional determinants of the wealth effects to bidding firms are outlined in Section 2.4. The financial literature has also shown that major M&A transactions are not only associated with significant wealth effects to the merging firms, but also influence the wealth of rival firm shareholders. Section 2.5 starts with a review of the theoretical foundations regarding the intra-industry effects of M&A transactions. Then, it reviews the major empirical studies that investigate the wealth effects of rival firm shareholders. Finally, it outlines the major cross-sectional determinants of the returns to rival firm shareholders.

Section 2.6 summarizes the major findings of the review and theoretically motivates the empirical investigations of the subsequent sections.

Section 3 is the starting point of the empirical investigation. The objective is to examine whether the point in time of the deal within an industry merger wave determines both the transaction success and the intra-industry effects of a transaction.

First, in Section 3.1, I explain in detail how I derive the sample of bids that occur within industry merger waves. Based on a wave identification process developed in the recent financial literature, I identify 17 industry merger waves that occur in the 1990s. For the empirical investigations, I concentrate on a sample of 1,012 bid announcements that take place within these waves.

In Section 3.2, I examine how the transaction success of the merging firms and the response of the industry competitors differ across different stages of industry merger waves. The results show that mergers are associated with positive wealth effects to bidding and rival firm shareholders at the beginning of industry merger waves.

However, the gains to both bidders and rivals become significantly negative towards the end of a wave.

Based on these findings, Section 4 investigates whether the decreasing gains to both bidders and rival firms are due to changing bidding firm and deal characteristics towards the end of industry merger waves.

Section 4.1 compares bidding firm and deal characteristics at different stages of merger waves. The comparisons show that cross-sectional determinants of the transaction success often change substantially as a wave is ongoing.

Then, in Section 4.2, I investigate whether these changing determinants can explain the results for bidders obtained in Section 3. Section 4.3 repeats the investigations regarding the abnormal returns to rivals. For both bidders and rivals, I analyze whether the results obtained in Section 3 hold across subsamples of bids. Additionally, I estimate regressions that control for a variety of bidding firm and deal characteristics and use different variables to capture the point in time of a deal within an industry merger wave. For all subsamples of bids, the abnormal returns to bidders and rivals are significantly lower at the end of an industry merger wave. In all regressions, the coefficient for the point in time of the deal within an industry merger wave is always highly significant. Thus, the results of Section 4 imply that the pattern of abnormal returns within industry merger waves is not due to changing bidder and deal characteristics, but exists because the underlying transaction motives differ between early and late wave stages.

Based on these results, Section 5 examines how the mixture of transaction motives changes towards the end of a wave. The financial literature proposes a number of hypotheses for why mergers should be associated with negative wealth effects. Section 5.1 starts with a discussion of the rationales for why some of these transaction motives could play more of a role at late wave stages. The theoretical discussion highlights that among the proposed theories, only the competitive advantage theory can explain the results to both bidding firms and rivals.

Therefore, in Section 5.2, I empirically investigate the importance of the competitive advantage theory at the beginning and the end of industry merger waves. The results show that the theory only applies to deals at late wave stages. I provide strong evidence that this is the reason for why the point in time of a deal within an industry merger wave determines both the gains to the bidding firm and rival firm shareholders.

For completeness, Section 5.3 investigates the importance of agency-related problems and the hubris theory at different stages of industry merger waves. The results show that agency-related problems are not more severe at the end of a merger wave. Furthermore, I only find very slight indications that bidding firm managers become more prone to hubris as a wave is ongoing.

Finally, Section 6 concludes. I start with a summary of the major results of the empirical investigation of this dissertation Then, I discuss in detail the implications of my dissertation for further research.

2 Literature review

2.1 Motives of M&A transactions

The financial literature has shown that an acquiring firm can seek a merger for a number of reasons. These underlying transaction motives highly influence the transaction success leading to either value-increasing or value-decreasing deals. First, Section 2.1.1 describes the major theories that state why mergers are associated with positive wealth effects for bidding firm shareholders. Second, Section 2.1.2 illustrates the hypotheses that have been offered for why the stock price reaction of firms announcing an acquisition can be negative.

2.1.1 Value-increasing motives

A large body of research suggests that value-increasing takeovers are driven by efficiency considerations. In addition to this efficiency view, there is a longstanding proposition that horizontal takeovers often occur because the merging firms want to expropriate wealth from customers and suppliers. For example, according to antitrust authorities merging firms may gain because horizontal mergers enable them to engage in anticompetitive collusion.[1] Further, a horizontal takeover can increase the buying power of the merging firms. This enables them to put pressure on their suppliers.[2] This section describes the major motives for value-increasing mergers.[3]

2.1.1.1 The productive efficiency hypothesis

Managers of firms undertaking horizontal mergers and acquisitions often cite the existence of synergies as the main argument to justify a transaction. In the context of mergers and acquisitions, the term synergies refers to the ability to create a combination that is more profitable than the individual parts of the original firms.[4]

Gaughan (2002) states that the main types of synergies are cost-reducing operating synergies, revenue-enhancing operating synergies, and financial synergies. Cost- reducing operating synergies can be either a result of economies of scale or can be due to economies of scope. The term economies of scale refers to the decreases in per-unit costs that are due to an increase in the size of a company’s operations. It is reasonable that mergers and acquisitions of manufacturing firms are often motivated by the pursuit of scale economies. These firms often have high per-unit costs for low levels of output due to the high fixed costs of operating the manufacturing facilities. The term economies of scope refers to the reduction in total costs due to the ability of a firm to utilize one set of inputs to provide a broader range of products or services. An example is the use of the same equipment for various products.[5]

The merging firms may achieve revenue-enhancing operating synergies by the fusion of distinct attributes of the merger partners to generate a significant revenue growth. An example where revenue-enhancing operating synergies can be achieved is the merger between a “company with a strong distribution network merging with a firm that has products of great potential, but questionable ability to get them to the market. [6]

It has often been argued that not only operating synergies motivate major M&A transactions, but also that corporate mergers often have a positive influence on the cost of capital of the merging partners. Especially, if there is no perfectly positive correlation between the income streams of two firms, the bankruptcy risk of the new combined entity may be reduced. This may lead to decreasing costs of capital of the merging firms.[7] An obvious weakness of the risk reduction argument is that, in perfect capital markets, stockholders can achieve the preferred degree of risk on their own by rearranging their portfolios.[8] However, even in efficient financial markets, there may exist some sources of financial synergies as it can be argued that larger companies enjoy better access to financial markets and tend to experience lower costs of raising capital.[9]

Theoretically, synergies are one of the main arguments of managers to justify mergers and acquisitions. A number of recent empirical studies examine in detail the importance of synergies as a reason for why an acquisition occurs. The results do, indeed, provide strong evidence that synergies are one of the primary motives for value- increasing transactions.[10]

2.1.1.2 The collusion hypothesis

If collusion is the underlying motive of a horizontal takeover, the merging firms benefit at the expense of their customers and suppliers. The collusion argument presumes that companies have a high incentive to coordinate the production rates of the individual firms within an industry, if the costs of monitoring the cartel agreement are relatively low.[11] Eckbo (1988) states that a horizontal merger may reduce these monitoring costs because it reduces the number of independent producers in the industry. “The fewer the firms in the industry, the more visible are each producer’s actions, and the higher the chance of detecting members who try to ‘free-ride’ (or cheat) on the cartel by secretly increasing output.”[12]

The financial literature has studied extensively the importance of collusion as a key transaction motive. For market-wide samples of deals, Stilman (1983), Eckbo (1983), and Eckbo and Wier (1985) show that collusion is not a key motive for horizontal takeovers.[13] The results of recent works by Fee and Thomas (2004) and Sharur (2005) also provide evidence that is inconsistent with the view that major horizontal mergers have collusive, anticompetitive effects.[14]

However, industry-related work confirms the importance of collusion as a primary driver of value creation for the merging firms. For example, Kim and Singal (1993) and Singal (1996) provide very strong evidence that the collusion motive is a key source of M&A activity in the airline industry.[15]

2.1.1.3 The buying power hypothesis

A horizontal takeover can benefit the merging firms because it increases their buying power. Large buyers have an advantage as it is easier for them to put pressure on their suppliers. Thus, an increased concentration within the buying industry can enable the buyers to pay lower input prices.[16]

According to Ellison and Snyder (2001), the theoretical research on the buyer-size effect can be divided into two categories. Although they concentrate on different economic environments, both categories of buyer power models imply that the merging firms benefit at the expense of firms in their supplier industry.[17] The first category models the buyer size effect in industries that involve competing suppliers. These models suggest that suppliers have to compete more aggressively for the business of the larger post-merger buyer. This intensified post-merger competition, which leads to lower input prices in the buying industry, is the source of gains of the merger.[18] The second category involves theoretical models that investigate the buyer size effect when there is a monopoly supplier. These studies show that under certain conditions, a monopoly supplier charges a lower input price to a larger buyer as well.[19]

In a recent empirical study, Fee and Thomas (2004) conclude that a large percentage of value-increasing transactions is motivated by these buying power considerations. Additional empirical evidence regarding the importance of increased buying power is given by Schumacher (1991) and Ellison and Snyder (2001).[20]

2.1.2 Value-decreasing motives

Empirical studies show that about 50% of all transactions are associated with negative wealth effects to acquiring firm shareholders.[21] Based on these empirical results, the financial research offers a number of hypotheses for why value-decreasing transactions occur. The major theories are illustrated in this section.

2.1.2.1 The hubris hypothesis

One theory for why the acquirers might be overpaying for their targets is the hubris hypothesis of Roll (1986). According to Roll (1986), managers often believe that their valuation of a potential target is superior to that of the market. However, if the market is efficient, managers are simply overconfident in their abilities to estimate the fair value of the target. Roll (1986) states that this overconfidence of the acquiring firm’s management leads to an overpayment for the assets of the target. Consequently, overconfident managers undertake value-decreasing transactions. This reduction of acquiring firm shareholder value might even occur, if a lot of synergies can be realized after the transaction.[22]

Recent studies support the intuition that managerial hubris is a major motive for value-decreasing transactions. Hayward and Hambrick (1997) measure hubris by variables such as the company’s recent performance or the importance of the CEO. They find that overconfident managers pay significantly higher premiums.[23] Malmendier and Tate (2005) measure overconfidence by the options the management has left unexercised. Their results show that overconfident managers make significantly more acquisitions and that these acquisitions are associated with a significantly lower transaction success.[24] The importance of the hubris hypothesis is also highlighted by Moeller, Schlingemann, and Stultz (2004). They show that large bidders pay significantly higher premiums, which leads to a lower average transaction success. According to Moeller, Schlingemann, and Stultz (2004), large bidders pay higher premiums because the managers of large firms are more often prone to overconfidence than the managers of small companies.[25]

2.1.2.2 Agency theories

Agency-related problems can lead to takeovers that are primarily motivated by the self-interest of the bidder management.

Jensen (1986) asserts that managers of firms with large free cash flows often do not pay out these cash flows to their shareholders, as this would decrease the resources under their control. Using the cash to finance acquisitions can increase the managers’ private benefits, whereas the payment of cash to stockholders reduces managers’ power.[26]

Amihud and Lev (1981) state that managerial motives are also the source of many conglomerate mergers. They posit that managers want to engage in these diversifying transactions to decrease their undiversifiable employment risk. Thus, agency-related problems can lead to conglomerate mergers even though the risk reduction will not be beneficial to stockholders.[27]

Shleifer and Vishny (1989) describe how managers can entrench themselves by making manager-specific acquisitions that increase the firm’s dependence on the management. This increased dependence reduces the probability of being replaced or increases the possibility of extracting higher wages.[28]

The basic idea of all these explanations is that managers extract value from the acquiring firm shareholders. Thus, agency-related problems lead to value-decreasing bids and the more severe the agency problems, the higher is the shareholder value decrease.[29]

2.1.2.3 Overvaluation theory

The preceding sections show how the quality of the management influences the transaction success. However, the overvaluation theory posits that also exogenous factors like the company’s valuation level can cause value-decreasing transactions.

Shleifer and Vishny (2003) develop a model that demonstrates in detail how overvaluation can be a motive for takeovers. They posit that managers rationally value firms, but investors do not. To protect their shareholders from the loss in wealth that will come when the market lowers the valuation, managers exchange their overvalued shares for the real assets of another company, which is presumably less overvalued.[30] In a different theoretical model of Rhodes-Kropf and Viswanathan (2004), the motivation for the acquiring firm is essentially the same as in the work of Shleifer and Vishny (2003).[31]

Bids that are driven by market overvaluation are associated with negative information signaling effects. Therefore, the market corrects the overvaluation of the bidding firm’s equity. This correction leads to negative returns to the acquiring firm shareholders at the announcement of an overvaluation driven transaction. “Takeover bids are salient events that call attention to the firms involved. To the extent that more careful analysis helps investors correct misvaluation, the stock price reaction will tend to oppose the prior misvaluation.”[32]

In a recent empirical study, Dong, Hirshleifer, Richardson, and Teoh (2006) empirically investigate how irrational market misvaluation affects firms’ takeover behaviour. Their results are consistent with the view that overvaluation is a primary motive for M&A activity.[33] Ang and Cheng (2006) provide further evidence that confirms the hypothesis that overvaluation is an important motive for M&A transactions.[34]

2.1.2.4 Competitive advantage theory

Bradley, Desai, and Kim (1983) show that bidding firm shareholders do not lose when the target rejects the initial bidder and remains independent. However, when the target accepts a rival bidder’s offer after rejecting the first bid, there is a significant wealth reduction for the shareholders of the initial bidder.[35] According to Bradley, Desai, and Kim (1983), these results suggest that “when a firm loses the competition for a target firm to a rival bidding firm, the market perceives it to have lost an opportunity to acquire a valuable resource. Perhaps the transfer of control of the target resources to another firm places the firm at a competitive disadvantage vis-a-vis the successful bidding firm.”[36]

Based on this intuition, Akdogu (2003a) develops a theoretical model where the target is modelled as possessing special resources or technology that gives the successful acquirer a competitive edge over its rivals. Thus, it is costly for the bidders to lose the target to one of the competitors. The implication of her model is that it can be rational to overpay for a target, to avoid that a competitor acquires the target and gets a competitive advantage through the acquisition. This overpayment leads to a wealth reduction of the acquiring firm shareholders.[37]

Another pre-emptive theory of mergers that is very similar to the theoretical model of Akdogu (2003a) is offered by Molnar (2003).[38] He also concludes that bidders experience negative returns because they are willing to overpay to avoid losing the target to a rival. This alternative would bring even more negative returns.[39]

2.1.3 Summary

A number of theories have been offered for why mergers and acquisitions occur. This section reviews the most often cited motives for mergers and acquisitions. The underlying motive for why a transaction occurs highly influences the transaction success. An overview of the respective hypotheses, their basic intuitions, as well as the implied effect on the acquiring firm value is given in Table 2.1.

It becomes obvious that essentially two kinds of factors determine the acquisition motive. First, firm-specific factors highly influence the decision to acquire. For example, a bad management might seek acquisitions that primarily benefit the bidding firm management and not the bidding firm shareholders. Second, exogenous factors might play an important role for explaining why mergers occur. Examples include a changing industry structure that allows more collusion among the companies or a market-wide overvaluation that may enable managers with superior information to use their stock as an acquisition currency.

The underlying set of motives determines the average success of an M&A transaction. This mixture of motives differs substantially across subsamples of bids constructed based on cross-sectional deal and bidding firm characteristics. Therefore, acquiring firm and deal characteristics are key determinants of the transaction success of the merging firms. A detailed discussion on how the importance of the different motives varies across subsamples of bids is given in Section 2.4.

Table 2.1: Transaction motives and the gains to bidders

This table summarizes the predictions of the transaction motives regarding the signs of the gains to bidding firms. The table also briefly describes the reason for the impact of the transaction motive on the announcement gains to the bidding firm shareholders.

illustration not visible in this excerpt

The transaction motives that are outlined in this section do not only influence the transaction success. They also impact the intra-industry effects of a transaction. This topic will be discussed in detail in Section 2.5.1.

A variety of potential sources of gains and losses in M&A transactions have already been formulated by the financial literature. However, in a recent work, Mentz (2006) highlights that industry-specific motives also have significant influence on both the average transaction success and the intra-industry effects.[40] His intuition is supported by an empirical study for a sample of mergers in the automotive supply industry. The empirical research on the importance of industry-specific sources of gains and losses in M&A transactions is still at the beginning. Future findings in this research topic will probably extend the list of major transaction motives.[41]

2.2 Aggregate and industry-specific merger waves

The previous section describes the sources of individual M&A transactions. This section outlines the econometric characteristics and the key drivers of market-wide M&A activity.

The level of M&A activity is not constant over time, but fluctuates heavily. At both the industry and overall market level mergers tend to come in waves, where most of the deals occur. Over the last decades, aggregate merger waves were not independent of the overall market valuation, but always coincide with stock market booms. Section 2.2.1 reviews econometric studies that investigate these two key facts about the time-series behaviour of M&A activity.

Until recently we did not know the drivers of market-wide merger waves. The financial literature has developed two major hypotheses that seek to explain why mergers come in waves. Section 2.2.2.1 reviews these studies. Harford (2005) investigates in detail the relevance of the respective theories. His results are outlined in Section 2.2.2.2.

2.2.1 The time-series behaviour of M&A activity

One of the most striking characteristics of mergers is that they tend to come in waves. Hence, there are oscillations between low and high levels of merger activity.

An extensive literature exists that tests whether the claim that mergers come in waves can be established econometrically. Town (1992) formally tests the wave hypothesis by modelling M&A activity as a two-stated Markov regime-switching model. In a different study, Golbe and White (1993) provide a test of merger wave behaviour based on sine- curve estimations for the M&A time series. Both studies provide strong evidence that the time series pattern of aggregate merger activity does conform to waves.[42] Further formal tests of the wave hypothesis are presented by Linn and Zhu (1997) for the United States, and by Resende (1999) for the United Kingdom.[43]

Financial researchers also show a high correlation between the evolution of the stock market and merger activity. That is, merger waves are correlated with increases in share prices and valuation ratios. Nelson (1959) is the first to document the correlation between merger activity and share prices. However, numerous subsequent studies offer confirmation of his findings. For example, further evidence is presented by Melicher, Ledolter, and D’Antonio (1983) and Geroski (1984).[44]

2.2.2 Reasons of the wave pattern

The previous section shows that we know a lot about the time-series pattern of M&A activity. However, until recently there has been only limited evidence regarding the causes for this wave behaviour of M&A activity. Neoclassical explanations of merger waves argue that merger waves result from shocks to an industry’s economic, technological, or regulatory environment.[45] In contrast, Shleifer and Vishny (2003) and Rhodes-Kropf and Viswanathan (2004) highlight the fact that merger waves are correlated with high stock market valuations. They develop behavioural models in which merger waves result from managerial market timing.[46] Section 2.2.2.1 outlines the two potential explanations for merger waves.

These main hypotheses of merger waves offer an extensive set of empirical predictions. Based on the respective theoretical hypotheses, Harford (2005) examines in detail whether merger waves can be explained by neoclassical theories or whether such clustering is due to managerial timing of stock market overvaluations. The results of his study are outlined in detail in Section 2.2.2.2.

2.2.2.1 Potential explanations of merger waves
2.2.2.1.1 Neoclassical explanations of merger waves

Mitchell and Mulherin (1996) provide strong evidence that not only aggregate merger activity clusters in time, but that there is also industry-specific clustering in M&A activity. Especially, they show that aggregate waves are characterized by very strong inter-industry differences in the rate of takeover activity.[47] Based on these results, they suggest that future research should highlight the importance of industry-level factors to further understand the causes of these clusters in M&A activity.[48] Mitchell and Mulherin (1996) base their study on a sample of bids in the 1980s. In later studies, Mulherin and Boone (2000) and Andrade and Stafford (2004) confirm the industry-level clustering of M&A activity for the 1990s.[49]

The neoclassical hypothesis of mergers posits that the causes of these industry- specific merger waves are economic, regulatory, or technological shocks. It states that takeovers are often the only possibility for industry structures to respond to these exogenous shocks that require the reallocation of assets within the industry.[50] Harford (2005) summarizes the neoclassical hypothesis as follows: “Summarizing, under the neoclassical hypothesis of merger waves, once a technological, regulatory or economic shock to an industry’s environment occurs, the collective reaction of firms inside and outside the industry reallocates assets through mergers and partial firm acquisitions. This activity clusters in time, as managers simultaneously react and compete for the best combination of assets.”[51]

Critics of the neoclassical hypothesis often posit that the clustering of mergers is only one characteristic of the time-series pattern of M&A activity. Additionally, these merger waves are correlated with stock market booms. This association is entirely ignored by the standard neoclassical industry-shocks hypothesis.[52]

2.2.2.1.2 Behavioural explanations of merger waves

The possible impact of misvaluation on the decision to acquire is outlined in Section 2.1.2.3. However, the theoretical models of Shleifer and Vishny (2003) and Rhodes- Kropf and Viswanathan (2004) not only show that overvaluation can be a major motive for single transactions, but also demonstrate how periods of overvaluation can cause merger waves.[53]

As outlined in Section 2.1.2.3, Sheifer and Vishny (2003) assume that managers behave rationally and takeovers are purely driven by irrational stock market overvaluation. Based on these assumptions, Shleifer and Vishny (2003) conclude that periods of high market valuations can cause merger waves, as these high valuations can lead bidders with overvalued equity to use the stock as an acquisition currency to buy less overvalued targets.[54]

The model of Rhodes-Kropf and Viswanathan (2004) confirms the importance of valuation waves that drive merger activity. They also state that “mergers occur more often the more the market is overvalued.”[55] In addition to Shleifer and Vishny (2003), they highlight the fact that the market has to stay overvalued after the first mergers occur. “On average, after a merger, firms are correctly priced. Hence, we should not expect a wave or any ex post drift in prices. In fact, a second merger is less likely than the first because of the market correction, but if the market is still overvalued, then a merger is more likely than it would be in expectation unconditionally.”[56]

In a recent study, Rhodes-Kropf, Robinson, and Viswanathan (2005) empirically investigate the predictions of the theoretical models. Their results show that aggregate merger waves indeed occur when market valuations are high relative to true valuations. Thus, their findings confirm the theoretical assumption that merger waves are caused by an overvalued market.[57]

2.2.2.2 Neoclassical versus behavioural explanations of merger waves

The previous section outlines the two major explanations of merger waves. Harford (2005) investigates whether merger waves are due to exogenous shocks or to managerial timing of market overvaluations. He criticizes that studies that examine the importance of the explanations tend to only provide evidence consistent with one explanation, rather than considering both neoclassical and behavioural hypotheses and then formally rejecting one theory.[58] Harford (2005) argues that both theories can explain the reasons for both aggregate and industry-specific merger waves. However, he states that the neoclassical hypothesis can only explain aggregate merger waves, if it is modified by a capital liquidity component. This capital liquidity argument asserts that exogenous shocks only generate industry-specific merger waves when sufficient capital liquidity is present to accommodate the reallocation of assets. Thus, due to the importance of the capital liquidity argument, industry merger waves will cluster in time to create aggregate merger waves, even if industry shocks do not cluster in time.[59]

To test whether aggregate merger waves are due to a clustering of exogeneous industry-specific shocks or to overall market timing, Harford (2005) establishes a detailed framework for testing the hypotheses. He asserts that the predictions of the two hypotheses for the causes of merger waves can be outlined as follows.

The neoclassical hypothesis posits that industry merger waves occur when economic, regulatory, or technological shocks force industries to restructure. Thus, clearly identifiable exogenous shocks should be observed preceding waves.[60] Aggregate merger waves are caused by a clustering of many industry-specific merger waves, which is a consequence of macro-liquidity factors. Due to the capital liquidity component of the neoclassical hypothesis, the average valuation level should be high and the credit constraints are predicted to be low.[61] Finally, partial-firm transactions should also spike during industry merger waves, as the industry-specific response to an exogenous shock likely involves not only firm-level transactions.[62]

The behavioural hypothesis asserts that merger waves are a result of managerial timing of stock market overvaluations. Thus, exogenous shocks should not systematically precede a wave, as there is no economic driver of a wave. Merger waves will only occur, if the overall market valuation is high. That is, they will only take place in times of high average market to book ratios or following abnormally high stock returns. The returns after the merger wave should be abnormally low, as the market will tend back to a fair valuation level.[63] High valuation levels will especially lead to merger activity, if there is a sufficiently large pool of potential targets with less overvalued market prices. Thus, not only the absolute average valuation level, but also the cross- sectional dispersion in the valuation ratios should be abnormally large.[64] In addition, partial-firm transactions for cash should not be common, especially not if the firms are bidding for other firms with stock. The reason is that the behavioural hypothesis claims that merger waves are driven by the acquisition of real assets with overvalued stock.[65] Finally, the post-merger operating performance of mergers that occur within merger waves should be worse than the post-merger operating performance of mergers outside of waves.[66]

Table 2.2 summarizes the predictions of both the neoclassical and behavioural hypotheses and shows the findings of Harford’s (2005) empirical investigations.

Harford’s (2005) results can be summarized as follows. All theoretical predictions of the neoclassical hypothesis of merger waves can be empirically validated. Thus, industry merger waves are caused by economic, regulatory, or technological shocks. However, these shocks only propagate waves, if efficient capital liquidity is present. This macro-level component causes industry merger waves to cluster in time, even if industry-specific shocks do not. Thus, aggregate merger waves are due to a clustering of multiple industry-specific merger waves.

Table 2.2: The neoclassical and behavioural hypothesis of merger waves

This table summarizes the predictions of both the neoclassical and the behavioural hypothesis of merger waves. The table also shows the empirical findings of the investigations of Harford (2005).

illustration not visible in this excerpt

Source: Adapted from Harford (2005), p. 536.

The behavioural hypotheses is motivated by the relation between asset values and aggregate merger activity. However, there is no abnormally high dispersion in pre-wave returns. Post-wave returns are not abnormally low and the post-merger operating performance is not worse in waves. Thus, the positive relation between the valuation level and merger activity reflects a capital liquidity effect rather than any overvaluation effect.[67]

[...]


[1] See Eckbo and Wier (1985), p. 119.

[2] See Snyder (1996), p. 749.

[3] In recent studies Fee and Thomas (2004) and Sharur (2005) distinguish between three major motives for value-increasing mergers. I rely on their differentiation. See Fee and Thomas (2004), p. 427; Sharur (2005), p. 65.

[4] See Mentz (2006), p. 27.

[5] See Gaughan (2002), p. 120-123; Mentz (2006), p. 28.

[6] Gaughan (2002), p. 121.

[7] See Lewellen (1971), p. 522; Higgins and Schall (1975), p. 93.

[8] See Lev and Mandelker (1972), p. 95.

[9] See Trautwein (1990), p. 284.

[10] Berkovich (1993) states that synergies are the main reasons for why a transaction takes place. However, his conclusion is based on an empirical design that does not try to distinguish between

different motives for value-increasing mergers, but for all mergers. In a recent paper, Sharur (2005) concentrates on the motives for value-increasing transactions and concludes that the productive efficiency is the primary motive for these bids. See Berkovich (1993), p. 361; Sharur (2005), p. 95.

[11] See Eckbo and Wier (1985), p. 119; Eckbo (1988), p. 11.

[12] Eckbo (1988), p. 11.

[13] See Stillman (1983), p. 235-239; Eckbo (1985), p. 348; Eckbo and Wier (1985), p. 139. Song and Walkling (2000) provide further evidence that is inconsistent with the collusion hypothesis. See

Song and Walkling (2000), p. 170. A detailed description of the respective studies and their results is given in section 2.5.

[14] See Fee and Thomas (2004), p. 457; Sharur (2005), p. 95.

[15] Kim and Singal (1993) show that mergers in the airline industry lead to higher prices on routes affected by the merging firms in comparison with a control group of routes that were unaffected. The

stock market tests of Singal (1996) confirm these results. See Kim and Singal (1993), p. 567; Singal (1996), p. 265.

[16] See Sharur (2005), p. 67. The industrial organization literature has examined extensively how downstream industries are influenced by a change in the structure of upstream industries. A brief overview of this work is given by Ellison and Snyder (2001). See Ellison and Snyder (2001), p. 1-2.

[17] See Ellison and Snyder (2001), p. 1-2.

[18] See Snyder (1996), p. 753-760.

[19] In this context, the existence of a buyer size effect depends on the surplus function of the monopoly supplier. See Chipty and Snyder (1999), p. 331-332; Ellison and Snyder (2001), p. 1-2.

[20] See Fee and Thomas (2004), p. 457. Schumacher (1991) finds that industry profits are negatively correlated with the buying power of downstream industries. Ellison and Snyder (2001) empirically confirm the importance of buying power effects in the wholesale pharmaceuticals industry. See Schumacher (1991), p. 283; Ellison and Snyder (2001), p. 23.

[21] Section 2.3 gives a detailed overview of the empirical studies that investigate the transaction success of the bidding firms.

[22] See Roll (1986), p. 199-200.

[23] See Hayward and Hambrick (1997), p. 122-123.

[24] See Malmendier and Tate (2005), p. 2679-2695.

[25] See Moeller, Schlingemann, and Stultz (2004), p. 220-222.

[26] See Jensen (1986), p. 323-329. Harford (1999) empirically confirms the importance of Jensen’s free cash flow theory. Harford (1999) shows that cash-rich acquirers have significantly lower abnormal returns and that the abnormal stock price reaction is decreasing in the amount of excess cash held by the bidder. See Harford (1999), p. 1976-1987.

[27] See Amihud and Lev (1981), p. 615. Based on the theoretical work of Amihud and Lev (1981), Morck, Shleifer, and Vishny (1990) empirically investigate the importance of the target relatedness for bidder returns in acquisitions. They find that bidders earn significantly lower returns in diversifying bids than in bids for related targets. See Morck, Shleifer, and Vishny (1990), p. 41-42.

[28] See Shleifer and Vishny (1989), p. 137. For example, firms invest into businesses that are highly related to the manager’s personal background.

[29] See Berkovich (1993), p. 350.

[30] See Shleifer and Vishny (2003), p. 297-301.

[31] See Rhodes-Kropf and Viswanathan (2004), p. 2690-2693.

[32] Dong, Hirshleifer, Richardson, and Teoh (2006), p. 754.

[33] Dong, Hirshleifer, Richardson, and Teoh (2006) base their conclusions on a large set of empirical results. They show that misvaluation influences the method of payment chosen, the mode of

acquisition, the premia paid, target hostility to the offer, the likelihood of offer success, and bidder and target announcement period stock returns. See Dong, Hirshleifer, Richardson, and Teoh (2006), p. 737-749.

[34] See Ang and Cheng (2006), p. 205-215.

[35] See Bradley, Desai, and Kim (1983), p. 198-204.

[36] Bradley, Desai, and Kim (1983), p. 203.

[37] See Akdogu (2003a), p. 6-9.

[38] See Molnar (2003), p. 7-12.

[39] In an independent work Fridolfsson and Stennek (2001) develop a model that is based on a similar intuition as the models of Molnar (2003) and Akdogu (2003a). See Fridolfsson and Stennek (2001), p. 4-16.

[40] See Mentz (2006), p. 37-41.

[41] See Mentz (2006), p. 188.

[42] In the statistical framework of Town (1992), a merger wave is defined as a large, discrete increase in

the mean of the time series. Thus, Town (1992) models merger activity as a combination of two AR processes with differing means with the transition from a high-wave (high-mean) state to a low-wave (low-mean) state based upon certain probability law. The findings of Town (1992) provide strong evidence that such a model fits M&A time series behaviour well. This highly supports the wave hypothesis. See Town (1992), p. 589-598. In the statistical framework of Golbe and White (1993), the critical parameter for the wave hypothesis is the coefficient estimate for the amplitude of the sine wave. Golbe and White (1993) find that the coefficient is statistically significant, and thus conclude that their model adequately describes the temporal patterns of aggregate merger activity. See Golbe and White (1993), p. 494-498.

[43] See Linn and Zhu (1997), p. 145; Resende (1999), p. 91.

[44] See Melicher, Ledolter, and D’Antonio (1983), p. 429; Geroski (1984), p. 231-232.

[45] These studies date at least back to Gort (1969) and Mitchell and Mulherin (1996). See Gort (1969), p. 625-642; Mitchell and Mulherin (1996), p. 219.

[46] See Shleifer and Vishny (2003), p. 306-308; Rhodes-Kropf and Viswanathan (2004), p. 2705-2708.

[47] Mitchell and Mulherin were not the first who show the strong inter-industry variation in the rate of takeover activity. In an early study, Gort (1969) shows that the M&A activity varies substantially across industries during the 1950s. See Mitchell and Mulherin (1996), p. 201-209; Gort (1969), p. 625-642.

[48] See Mitchell and Mulherin (1996), p. 195.

[49] See Mulherin and Boone (2000), p. 122-123; Andrade and Stafford (2004), p. 5-16.

[50] See Mitchell and Mulherin (1996), p. 219; Harford (2005), p. 532-533.

[51] Harford (2005), p. 533.

[52] See Shleifer and Vishny (2003), p. 296-297.

[53] See Shleifer and Vishny (2003), p. 306-308; Rhodes-Kropf and Viswanathan (2004), p. 2705-2708.

[54] There must be a supply of highly overvalued firms (bidders) as well as of relatively less overvalued ones (targets). Thus, according to Shleifer and Vishny (2003) high valuations should especially lead merger activity when the dispersion of valuations in the market is also high. See Shleifer and Vishny (2003), p. 304.

[55] Rhodes-Kropf and Viswanathan (2004), p. 2706.

[56] Rhodes-Kropf and Viswanathan (2004), p. 2707.

[57] Rhodes-Kropf, Robinson and Viswanathan (2005) use residual income models and valuation multiples to estimate their overvaluation parameters. See Rhodes-Kropf, Robinson, and Viswanathan (2005), p. 570-572.

[58] See Harford (2005), p. 534.

[59] See Harford (2005), p. 532-533.

[60] See Mitchell and Mulherin (1996), p. 219.

[61] See Harford (2005), p. 532-533. Harford (2005) uses the spread between the average interest rate on commercial and industrial loans and the federal funds rate as a macro component which proxies for capital liquidity. See Harford (2005), p. 542-543. He states that there is a positive relationship between this spread and the average valuation level: “A decrease in the rate spread leads to increased economic growth, potentially lower risk-premiums, and as shown later, greater merger and acquisition activity. All of these would have the effect of causing an increase in market-to-book ratios.” Harford (2005), p. 543.

[62] See Harford (2005), p. 549.

[63] See Shleifer and Vishny (2003), p. 306.

[64] See Shleifer and Vishny (2003), p. 304.

[65] See Harford (2005), p. 534.

[66] This hypothesis is only stated by Harford (2005), but none of the theoretical papers that link merger activity to market misvaluation explicitly make any prediction regarding operating performance. However, Harford (2005) states that transactions without a primary operational motive would produce poorer post merger operating performances. See Harford (2005), p. 555-556.

[67] See Harford (2005), p. 537-558.

Excerpt out of 163 pages

Details

Title
The Dynamics within Merger Waves
Subtitle
Evidence from the Industry Merger Waves of the 1990s
College
European Business School - International University Schloß Reichartshausen Oestrich-Winkel
Grade
Summa cum Laude
Author
Year
2008
Pages
163
Catalog Number
V122448
ISBN (eBook)
9783640264759
ISBN (Book)
9783640264469
File size
1470 KB
Language
English
Tags
Dynamics, Merger, Waves
Quote paper
Dr Timo Gebken (Author), 2008, The Dynamics within Merger Waves, Munich, GRIN Verlag, https://www.grin.com/document/122448

Comments

  • No comments yet.
Read the ebook
Title: The Dynamics within Merger Waves



Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free