Long-term performance of technology acquisitions and the role of
corporate governance: the influence of the financial crisis
by
Timon Schuurmans
University of Groningen Faculty of Economics and Business
MSc Finance
June 12, 2019
Supervisor: Dr. J.V. Tinang Nzesseu
Student number: 2944588 Address: Prinsesseweg 44a Postal code: 9717 BK Groningen
2
Long-term performance of technology acquisitions and the role of
corporate governance: the influence of the financial crisis
Abstract
This research examines whether the increase in technology acquisitions after the financial crisis is driven by value creation, and if this can be explained by changes in corporate governance as a result of the financial crisis. Two methods for long-term value measurement, the buy-and-hold abnormal return and the calendar time approach, suggest that there is a positive transition in value creation after the financial crisis. A regression analysis concludes that only board
independence is positively related to overall value creation, but there is no evidence that changes in corporate governance explain the transition in value creation after the financial crisis.
3 1. Introduction
Technological progress has led to increasing technology- and industry-convergence (Kim et al., 2015), resulting in the blurring of classic industry borders. The impact of technological progression leads to increasing demand for technology acquisitions, especially after the financial crisis. Last year, technology mergers and acquisitions (further referred to as M&As) were worth 240.1 billion dollars. Furthermore, technology was the number one reason to engage in M&As in the last few years, leading also to an increase in cross-industry transactions (PwC, 2018).
Earlier findings indicate that technology acquisitions are mostly driven by positive market sentiment (Kohers and Kohers, 2001), but could destroy shareholder value due to long-term underperformance. Previous research on both short-long-term and long-long-term value creation of M&As (Alexandridis, Mavrovitis, and Travlos, 2012; Bruner, 2002; Burner, 2005; Rau and Vermaelen, 1998) indicates that the abnormal returns to the acquirers’ shareholders are mostly negative, and can be explained by agency and hubris motives of managers. Therefore, control and monitoring of the decisions made by managers should lead to an increase in value creation of M&As (Fama and Jensen, 1983). Control on managerial decision-making is mostly referred to as corporate governance, a system which received increased attention in the last decades.
One of the most impactful events on corporate governance is the financial crisis of 2008. The world economy faced the consequences of massive financial misconduct that resulted in the financial crisis. While the financial crisis was mainly caused by the misconduct of financial companies, companies across all industries faced the consequences. The public lost its
confidence in corporations, and reforms driven by legislation were conducted, such as the Dodd-Frank act, aimed at improving the internal control process. Additionally, the crisis led to
shareholders being more engaged, increasing shareholder activism. The changes in legislation and corporate culture resulted in reforms in corporate governance mechanisms, such as the internal control mechanism, executive compensation and board independence (Bainbridge, 2012).
The question that therefore arises is: Is the recent increasing trend in technology
acquisitions driven by value creation as a result of reforms in corporate governance mechanisms due to the financial crisis?
4
executive compensation, and ownership structure, as reforms of the corporate governance system after the financial crisis were mainly focused on these mechanisms (OECD, 2009).
Most literature on M&A value creation is focused on the short-term abnormal returns of M&As, which is often biased by market sentiment. However, a recent finding on short-term value creation of Alexandridis, Antypas, and Travlos (2017) indicates that short-term abnormal returns could differ after the financial crisis. The short-term abnormal returns after the financial crisis are significantly positive, in contrast to negative abnormal returns in the earlier period. Due to the possibility of market sentiment, a long-term abnormal return approach could give further insight in whether there is indeed positive value creation after the financial crisis. Therefore, the value creation of technology acquisitions is estimated by using a term approach. The long-term approach is based on a three-year horizon, calculating the buy-and-hold returns of
technology acquiring firms and comparing them to an industry matched benchmark, equal- and value-weighted market benchmark, and a matched size and book-to-market benchmark, as well as using a Fama and French 3-factor regression (Fama and French, 1993) with both equal- and value-weighting of portfolios, to examine the abnormal return that a portfolio of technology acquiring firms generates. Both approaches are used to estimate the value creation of completed technology acquisitions in the time period of January 2001 till December 2015. To determine how changes in value creation are influenced by the changes in corporate governance, a distinction will be made for the period after June 2009 (NBER, 2010), examining the change after the financial crisis. The relation with changing corporate governance is examined by using variables for board independence, executive compensation and ownership concentration in an ordinary least squares and fixed effects regression analysis.
This research is further organized as follows. In Section 2, the relevant literature underlying the decision to conduct M&As, the measurement of M&A value creation, and the influence of corporate governance mechanisms is discussed. Section 3 discusses the
methodology, explaining how to measure the long-term value creation of technology acquisitions and the influence of changing corporate governance. Section 4 presents the data set used in this research. Section 5 contains the results estimated by both long-term value measurement
5 2. Literature review
The underlying motives for engaging in technology acquisitions can be based on two views. Firstly, the neoclassical view suggests that managers use M&As to create value because of changes in the economic, regulatory, and technological environment (Gort, 1969; Harford, 2005). This value creation benefits the shareholders, because managers of acquiring companies are acting in their interests (Berkovitch and Narayanan, 1993; Fama and Jensen, 1983). This view is supported by Bruner (2005), who states that shareholders from acquiring firms benefit from M&As, at least to the point that there is a value maintaining proposition, in which the shareholders retrieve their required rate of return. The value creation of M&As results from synergy effects. Three ways of value creation through synergy effects can be defined (Chatterjee, 1986). Firstly, operational synergies could lead to cost reduction because of economies of scale and scope (Porter, 1985). Technology acquisitions are mainly conducted to increase product developments, efficiency in the production process, or to grow within or outside the original operating market. Secondly, financial synergies could lead to risk diversification and reductions in the cost of capital. Lastly, collusive synergies could occur when conducting a horizontal M&A, because this could lead to higher market power.
However, value creation could not be the only motive to engage in M&As. The
behavioural view suggests that there are two motives for engaging in M&As; the agency motive and hubris motive. The agency motive argues that there is an information problem due to the separation of ownership and control (Berle and Means, 1932; Coase, 1937; Fama and Jensen, 1983). Managers can act in self-interest because of information asymmetry, and purposely maximize their own utility instead of the utility of shareholders (Jensen and Meckling, 1976; Schleifer and Vishny, 1989). Managers can do this because they possess the residual control rights. The value maximization of managers’ own utility is the consequence of the bounded rationality of shareholders (Grossman and Hart, 1986). Bounded rationality leads to an
6
The agency theory argues that managers act in their own interest on purpose. However, it could also be that value destroying M&As are conducted unintentionally. The motive supporting this behavioural view is the motive of Roll (1986). Roll (1986) defines the hubris hypothesis, which entails that hubris by an individual decision-maker leads to overvaluation of a possible target company. Acquiring companies will pay too much for a target company, resulting in unintentional value destruction for the acquirers’ shareholders.
The three motives examined above generate insight in the explanation of M&A
outcomes. On the one hand, the synergy motive indicates that M&As are undertaken because of value creation, leading to positive abnormal returns to the acquirers’ shareholders. On the other hand, the motive of agency and hubris suggest that the M&As are undertaken due to self-interest and overestimation of own abilities by managers. This leads to negative abnormal returns to the acquirers’ shareholders because of value destruction. An important part of the underlying motives is the actual measurement of value creation in M&As, which will be discussed in the next section.
2.1. M&A value measurement
The measurement of value creation in M&As can be divided into two streams of
literature. The first stream measures the abnormal returns of the M&A around the announcement date, measuring short-term performance. The second stream measures the abnormal returns in a period up to five years after the announcement or completion date to determine long-term performance.
Short-term announcement studies often lead to a positive stock market reaction for the target company but have mixed results on the stock returns of the acquirer (Bruner, 2005; Moeller, Schlingemann, and Stulz, 2003; Shah and Arora, 2014). Confirming this relationship, Bruner (2002) concludes from 130 empirical studies that there are zero short-term announcement returns for acquiring companies. However, Kohers and Kohers (2001) find that specifically the acquisition of high-technology firms generates positive short-term returns, which is mainly due to the positive sentiment that these firms possess. Using the announcement abnormal return is a short-term metric for valuing an M&A, and this has received some criticism. It is not likely that these announcement returns will be carried over in the long-term, because they could be
7
The second stream of literature focuses on the long-term value measurement of M&As. Most of the literature indicates that there is also a negative long-term value effect (Franks, Harris, and Titman, 1991; Rau and Vermaelen, 1998; Alexandridis, Mavrovitis, and Travlos, 2012), following the theory of agency and hubris motives. This indicates that the agency and hubris motives are the dominant effect that explain the M&A behaviour of managers best. Therefore, we hypothesize that there should be negative value effects to the acquirers’ shareholders when conducting a technology acquisition, due to agency and hubris motives, leading to the first hypothesis.
Hypothesis 1: Long-term value creation for the acquirers’ shareholders is negative when conducting a technology acquisition.
The financial crisis, ending in June 2009, had a major impact on corporate governance, leading to reforms in legislation and increasing importance of managerial control (Bainbridge, 2012). The world economy shattered due to the interconnectedness of the financial markets. The crisis was mainly caused by misconduct of financial companies, which was possible due to a lack of control and monitoring. However, companies across all industries faced the consequences. The public lost its confidence in corporations, and reforms driven by legislation were conducted, such as the Dodd-Frank act, aimed at improving of the internal control process. Additionally, the crisis led to shareholders being more engaged, increasing shareholder activism. The reforms in legislation and corporate culture resulted in changes in corporate governance mechanisms, mainly focused on the internal control mechanism, executive compensation and board
independence, because these mechanisms were the main cause of the financial crisis (OECD, 2009). The reforms and increased focus on these mechanisms should imply that companies made changes regarding their monitoring and control process. These changes should lead to better control on the decisions made by managers, leading to less value destroying M&As.
The change in corporate governance mechanisms should positively impact the control on managerial decision-making, leading to less value destroying M&As. This relationship is
supported by the short-term announcement study of Alexandridis, Antypas, and Travlos (2017), which indicates that announcement returns after the financial crisis are positive, contradicting outcomes of earlier research. They argue that these positive returns could be due to the emphasis that is placed on corporate governance after the financial crisis. Additionally, Ratcliffe,
8
technology acquisitions conducted after the financial crisis has a positive transition, due to changes in corporate governance. These changes in corporate governance result in less agency and hubris problems, which should imply that decisions are made in the interest of shareholders. This leads to the second hypothesis.
Hypothesis 2: Long-term value creation for the acquirers’ shareholders is not negative when conducting a technology acquisition after the financial crisis, due to better control and monitoring of managers.
2.2. Corporate governance mechanisms
In this section, the corporate governance mechanisms that could impact the value creation of M&As are presented, and the expected effects of these mechanisms are hypothesized. The expectation is that the effects of these variables become more significant after the financial crisis. The definition of corporate governance, according to Schleifer and Vishny (1997), pp 737, is: “the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment”. The reason why corporate governance is needed, is explained by the
behavioural motives of M&As, which indicate that managers could act in their own interests instead of the interests of shareholders. Managers can act in their own interests due to the separation of ownership and control, incomplete contracts and information asymmetry (Fama and Jensen, 1983). Corporate governance is a mechanism to overcome this problem, because it leads to monitoring and control of managers. In the next sections, the corporate governance mechanisms that were focussed on after the financial crisis, and which could influence the value creation for shareholders in technology acquisitions, will be examined.
2.2.1. Board independence
9
many others. This should implicate that the presence of more independent board directors leads to better control on decisions made by the managers, and result in less value destroying M&As.
Additionally, the chairman position of the board of directors can be a proxy for board independence. The chairman often has a great influence on the decision-making process. This indicates that when the CEO, which is a dependent director, has the chairman position (often referred to as CEO-duality) the control on the decision-making process is less independent, leading to decisions that are less in the interests of shareholders. Although empirical research is not conclusive about the relation between board independence and abnormal returns, this research follows the theory that more independent boards lead to increased abnormal returns. Masulis and Cong Wang (2007) confirm that short-term announcement returns of M&As are higher when there is a separation of the CEO and chairman position. Hillier and McColgan (2006) find that there is an increase in the independence of board members, and a separation between the CEO and the chairman position in companies. This separation leads to decisions that are made in the interest of shareholders. After the financial crisis, legislation led to the inclusion of more independent board members in the board of directors (Bainbridge, 2012). Furthermore, the Dodd-Frank act requires companies to disclose why a duality position is chosen, leading to less of these positions. This implicates that the decision to engage in M&As after the financial crisis is better monitored, which should lead to value creation, leading to the third hypothesis.
Hypothesis 3: There is a positive relation between board independence and long-term value creation for the acquirers’ shareholders when conducting a technology acquisition after the financial crisis.
2.2.2. Ownership concentration
The ownership concentration of a company could be a corporate governance mechanism that is of great influence (Schleifer and Vishny, 1997). Dispersed ownership of a company often leads to more agency problems, due to the inability of small shareholders to incentivize
10
share prices. Empirical literature (see, e.g., Edmans, 2014) proxies for a five-percentage stake in a company to be considered a block holder, because this is the percentage at which the
shareholder has requirements for disclosure in the US.
Although large block holders can increase the value maximization behaviour of managers, there are also costs to holding a large fraction of a company. Large block holders incur more risks, as a large part of their portfolio is invested in one company, leading to less diversification. Additionally, minority shareholders could be expropriated by large shareholders, because minority shareholders do not have enough voting power (Schleifer and Vishny, 1997). In recent years, an increase in block holders in the form of institutional investors can be
observed. The research of Kohers and Kohers (2001) finds that more block holders in the form of institutional ownership in the acquiring firm has a positive effect on post-M&A returns to the acquirers’ shareholders. In line with this research, Moeller and Schlingemann (2005) find that large block holders have a positive effect on short-term abnormal stock returns of M&As. After the financial crisis, an increase in active ownership is observed. This active ownership could imply that there are more block holders present in the ownership structure, leading to better monitoring of decisions. Additionally, these block holders are more incentivised to engage in active ownership, because of their large stake. Large block holders are expected to have a positive effect on long-term M&A value creation for the acquirers’ shareholders, because they have more incentive to monitor managers, leading to the fourth hypothesis.
Hypothesis 4: There is a positive relation between concentrated ownership and long-term value creation for the acquirers’ shareholders when conducting a technology acquisition after the financial crisis.
2.2.3. CEO compensation
11
compensation scheme of CEOs is linked to long-term performance. Additionally, Datta, Iskandar-Datta, and Raman (2001) find that equity-based compensation schemes result in
positive announcement returns for the acquirers’ shareholders.On the contrary, CEOs with larger managerial power often engage in larger value destroying M&As, because short-term
compensation generally rises with larger M&A deals (Grinstein and Hribar, 2004; Sapp, 2008). The notion that is important here is that equity-linked compensation generally increases the value creation of managerial decisions. The financial crisis was predominantly caused by
short-termism. Therefore, more attention has been given to equity-linked compensation after the financial crisis (OECD, 2009). This change in compensation after the financial crisis should impact the decisions of managers whether to engage in M&As. Equity-linked compensation could increase value creation of technology acquisitions, leading to the fifth hypothesis.
Hypothesis 5: There is a positive relation between equity-linked CEO compensation and the long-term value creation for the acquirers’ shareholders when conducting a
12 3. Methodology
To test the relation between changes in corporate governance and long-term value creation of technology acquisitions, an event study will be used based on previous work by Laabs and Schiereck (2010), Kohers and Kohers (2001), and Rau and Vermaelen (1998). First, the long-term value creation of technology acquisitions needs to be examined. This research will follow the two approaches that are most common to long-term value measurement of M&As. These two approaches are the buy-and-hold abnormal return (BHAR) approach and the calendar time (CTIME) approach. Thereafter, the relation with corporate governance will be examined.
3.1. BHAR approach
The BHAR approach is a cross-sectional approach that calculates the abnormal returns of an acquiring company by computing the three-year holding period return (HPR). This HPR starts at the first month after an M&A is announced and ends 36 months later. After these three years, the total three-year HPR for every acquiring company will be calculated as
𝐻𝑖 = ∏36𝑡=1(1 + 𝑅𝑖,𝑡)− 1 , (1)
where Hi is the holding period return for company i, and Ri,t is the monthly return for firm i at period t. What follows is the calculation of the BHAR, calculated as
𝐴𝑖 = 𝐻𝑖− 𝐻𝑏 , (2)
where Ai is the buy-and-hold abnormal return for company i, HPRi is the three-year holding period return for company i, and HPRb is the three-year holding period return of the matched benchmark for company i.
13
benchmark, composed of all companies listed on the CRSP database that have data reported on Compustat. These benchmarks will be used as a control to see how the acquiring firms fared relatively to the market. Lastly, the HPR of the acquirer will be compared against a size and book-to-market benchmark (Rau and Vermaelen, 1998; Kohers and Kohers, 2001). These benchmarks are made based on book and market values of firms listed on both the CRSP and Compustat databases. All companies that have conducted a technology acquisition will be deleted from the benchmark. The average BHAR computed by this method compares how well the acquirer of a technology company fared relatively to companies with the same size and book-to-market ratio that did not acquire a technology company. To compute the benchmark, the method of Kohers and Kohers (2001), Rau and Vermaelen (1998), and Ikenberry (1995) will be used. The total sample of benchmark firms will be divided into 50 size and book-to-market portfolios. Firstly, company data on size and book-to-market values are gathered from Compustat. Secondly, ten size deciles are calculated by comparing the market values of non-merging companies in the sample, which results in ten portfolios based on size. After that, all ten size portfolios will be divided into book-to-market quintiles based on the book-to-market values of the non-merging companies, which results in 50 size and book-to-market portfolios. The technology acquiring firms will be matched to the size and book-to-market portfolio based on their own size and book-to-market value. The significance of the HPRs and BHARs are tested using standard t-tests for the mean values, and a generalized sign test for the median HPRs.
The BHAR approach uses a cross-sectional regression, which examines every acquisition independently, not accounting for multiple acquisitions made by a single firm and time dependence of acquisitions. This brings methodological problems with assumptions of
stationarity, normality, and time dependence of observations, as argued by Rau and Vermaelen (1998). Lyon, Barber, and Tsai (1999) indicate that the BHAR approach suffers from two biases. These biases are the bad model bias and the cross-sectional dependence bias. The bad model bias indicates that it is not possible to identify the consequence for a technology acquiring company if it had not acquired a company exactly. Cross-sectional dependence occurs between firm
observations because the HPRs could be calculated in the same time frame, creating a dependence between these HPRs. Additionally, the calculation of HPRs of multi-acquiring companies results in the use of overlapping returns, leading to autocorrelation. However, no solution to account for these two biases are present in the literature. Previous research confirms the drawbacks of the BHAR approach. Fama (1998) and Mitchell and Stafford (2000)
14
Using a large data sample accounts for some of these drawbacks, but to further substantiate the measurement of value creation, the CTIME approach is used.
3.2. CTIME approach
The CTIME approach in this research is based on the Fama and French 3-factor model (Fama and French, 1993). Fama and French (1993) empirically extended the ‘alpha’ model of Jensen (1968) by adding a “small minus big” (SMB) term and a “high minus low” (HML) term, explaining differences in stock returns of companies. The extension leads to the three-factor model as calculated as
(𝑅𝑎𝑟,𝑡 − 𝑅𝑓,𝑡) = 𝛼𝑝− 𝛽1(𝑅𝑚𝑘𝑡,𝑡− 𝑅𝑓,𝑡) + 𝛽2∗ 𝐻𝑀𝐿𝑡+ 𝛽3∗ 𝑆𝑀𝐵𝑡+ 𝜀𝑝𝑡, (3) where Rar,t is the monthly abnormal return of a rolling portfolio containing technology acquiring companies, minus a portfolio of industry matched returns. Every acquiring company enters the acquiring portfolio in the first month after a technology acquisition and leaves the portfolio 36 months after the acquisition. The portfolio of the industry matched benchmark follows the same methodology. The monthly risk-free rate (Rf,t )is based on the 1-year rate on US government bonds. The return on the market (Rmkt,t) is based on the returns of an equal- or value-weighted market portfolio of all companies listed on both CRSP and Compustat. The SMB-term (SMBt) represents a zero-investment portfolio that is short in big capitalization stocks, because these are deemed to be over-priced, and long in small capitalization stocks, as these are under-priced in the long-run. The HML term (HMLt) represents a zero-investment portfolio that goes short in high book-to-market stocks and long in low book-to-market stocks, following the same argumentation as the SMB term. Regressing Eq. (3) results in an estimate for α, which indicates whether there are monthly abnormal returns for the portfolio of acquiring companies compared to the industry benchmark that cannot be explained by the market portfolio. The significance of this α will be estimated by using a standard t-test. The αp is estimated with both an equal- and value-weighted acquirers and market portfolio. Both methods of weighting could have
implications for the risk of a portfolio. An equal-weighted portfolio could be overinvested in small and middle capitalization stocks, increasing the risk of the portfolio. A value-weighted portfolio suffers from a loss in diversification, as most of the funds are invested in a few large companies. This also increases the risk of the portfolio. Therefore, both weighting methods are used for the estimation.
15
will be corrected for by using robust standard errors. The presence of autocorrelation will be tested for by using a Durbin-Watson and augmented Dickey-Fuller test, and if needed corrected for by using Newey-West standard errors for autocorrelation and heteroskedasticity. Using both the BHAR and CTIME approach generates a more precise measure of the long-term value creation of technology acquisitions.
3.3. Regression analysis
To test the effect of corporate governance mechanisms on the value creation for the acquirers’ shareholders in technology acquisitions, an ordinary least squares (OLS) and fixed effects (FE) model are used. The regression model is estimated as follows
𝐴𝑖,𝑡 = 𝛼0+ 𝛽1𝐵𝑖,𝑡+ 𝛽2𝐷𝑖,𝑡+ 𝛽3𝐸𝑖,𝑡+ 𝛽4𝑂𝑖,𝑡+ 𝛽5𝐵𝑎𝑖,𝑡+ 𝛽6𝐷𝑎𝑖,𝑡+ 𝛽7𝑂𝑎𝑖,𝑡+ 𝛽8𝐸𝑎𝑖,𝑡 + 𝛽9𝐴𝐶𝑡+ 𝛽10𝑅𝐴𝑖,𝑡−1+ 𝛽11𝑅𝐸𝑖,𝑡−1+ 𝛽12𝑆𝐼𝑖,𝑡−1+ 𝛽13𝑀𝑉𝑖,𝑡−1+ 𝛽14𝐵𝑀𝑖,𝑡−1+ 𝛽15𝐷𝑉𝑖,𝑡+
𝛽16𝑆𝑇𝑖,𝑡+ 𝛽17𝑆𝑖+ 𝛽18𝑃𝑖,𝑡+ 𝛽19𝐶𝐵𝑖,𝑡+ 𝛽20𝐶𝐼𝑖,𝑡+ 𝛽21𝐵𝑈𝑡+ 𝜀𝑖,𝑡 , (4) where the dependent variable (Ai,t) is based on the industry abnormal returns of the BHAR approach as estimated by Eq. (2).In the fixed effects model, the constant terms (α0,i) are entity-specific and capture the heterogeneities between the different firms.
The independent variables in these regressions are the four measures of corporate
governance, namely board independence (Bi,t), measured by the ratio of independent directors to the total board size, and a dummy variable for CEO-duality, displaying whether the CEO is also the chairman of the board (Di,t), CEO equity compensation (Ei,t), as measured by the ratio of equity compensation to total compensation gathered from the database of BoardEx, and the presence of block holders (five percent or more ownership) in the acquirers’ ownership structure (Oi,t). To estimate differences between time periods, a dummy variable for the time period after the financial crisis, starting July 2009, will be included (ACt). Four interaction variables on the change in corporate governance mechanisms after the crisis are used. These variables are constructed by interacting the after crisis dummy with the four respective corporate governance measures, resulting in four variables (Bai,t, Dai,t, Eai,t, and Oai,t).
book-to-16
17 4. Data selection
The sample used for this event study is retrieved from the M&A database of Zephyr. The time period of the post-M&A performance will be set at three years, following earlier research (Laabs and Schiereck, 2010; Rau and Vermaelen, 1998). The time period in which acquirers must have completed a technology acquisition is January 2001 till December 2015. The year 2001 is chosen because this date is after the bursting of the dot-com bubble, which could influence the results. A dummy variable in the regression analysis for the year 2001 is added to see whether there is still influence of the bubble. The year 2015 is chosen because of the three-year available stock data that is needed for the estimation of the long-term value creation.
The focus of this study will be based on companies in the US. The US is chosen because most of the technology M&As are conducted by US firms and information on US firms is best available. Additionally, corporate governance mechanisms differ across legal systems (Doidge, Karolyi, and Stulz, 2004). For example, the UK and the US have a common law system. This common law system provides more legal protection to shareholders, while countries like
Germany and Japan have a civil law system that has weaker shareholder protection. By focusing on one country, the differences between governance systems could be neglected. The data on corporate governance mechanisms is retrieved from the database of BoardEx.
Additionally, the M&As used for this study must meet the following criteria. Firstly, the target of the M&A must be operating in the technology industry. A technology company will be qualified by their primary three-digit US SIC code. According to Kile and Phillips (2009), the code combinations of high-technology companies are 283, 357, 366, 367, 382, 384, 481, 482, 489, 737, and 873. Secondly, deals must be completed and have three years of post-M&A stock returns for the acquirer reported on the database of CRSP. Thirdly, the acquiring company must have financial statements reported on the research database of Compustat. Fourthly, the
acquiring company must be a publicly traded company on one of the US stock exchanges of the NYSE, Amex or Nasdaq. Fifthly, the technology acquisitions must at least have a deal value of one million dollars. Deals that have a value below one million dollars do not have a significant impact on the share price of the acquirer. Lastly, one hundred percent of the target company must be acquired with the M&A. The total amount of technology acquisitions in the specified time period is 2861. Combining the data set of Zephyr with the data sets of BoardEx, Compustat, and CRSP results in a final sample of 1024 technology acquisitions made by 330 firms. The
descriptive statistics of the final sample are reported in Table 1.
18
financial crisis. Differences in the values of these variables could have an impact on the effect. Therefore, the difference will be shortly discussed. The variable for duality in the period up to and including June 2009 is 71.6%, while the duality percentage after this date is 54.8%. This indicates that there is a difference of 16.8%. Additionally, the variable for the ratio of
independent board directors has a value of 71.1% in the period up to and including June 2009, while this value is 78.8% after the crisis. This is a change of 7.7%. Overall, there is a clear indication that the board independence increased, with less CEO-duality positions and more independent boards. This difference should have a positive effect on value creation, following the theory presented. When looking at the equity-linked compensation variable, there is only a marginal change in the average value. The value in the period up to and including June 2009 is 77.05%, while the value after this date is 78.26%. Lastly, the variable for block holders had a value of 4.64 in the period up to and including June 2009, while it decreased after the financial crisis to 4.40. This change is not in line with the expectation that there should be more block holders after the financial crisis to increase shareholder activism. However, it could be that the activism of block holders is not due to the total amount of block holders, but rather in the changing behaviour of existing block holders.
The variable for equity-linked compensation is worth further discussing. This corporate governance measure is reported for only 388 observations by 94 firms. The sample size for this measure is low and could bias any results reported for this measure. Therefore, this measure is discussed separately. The variables for ROE, ROA, and BTM have been winsorized at the one percent level, such that extreme outliers of these values do not impact the results of the
19 Table 1: Descriptive statistics of the data set
VARIABLES Mean Median Min Max Number of
observations
Number of firms
DEAL VALUE (in millions, $) 873557.70 78400 1.5 200000.00 1024 330
MARKET VALUEt-1 (in millions, $) 27860.48 2575.57 7.57 626550.35 1024 330
SIZEt-1 (in millions, $) 15041.80 1672.20 5.66 272315.00 1024 330
BTMt-1 44.54% 35.45% 1.58% 193.13% 1024 330 ROAt-1 5.64% 6.88% -44.31% 24.86% 1024 330 ROEt-1 11.19% 11.34% -60.27% 89.80% 1024 330 BLOCK HOLDERSt (>5%) 4.52 4 1 16 1024 330 BOARD INDEPENDENCEt 75.10% 77.78% 27.27% 100.00% 1024 330 EQUITYt 77.66% 82.00% 24.00% 100.00% 388 94 DUALITYt 62.89% 1 0 1 1024 330 STOCK 19.14% 0 0 1 1024 330 SERIAL 54.59% 1 0 1 1024 330 PUBLIC 15.13% 0 0 1 1024 330 CROSSINDUSTRY 74.12% 1 0 1 1024 330 CROSSBORDER AFTERCRISIS 25.88% 51.95% 0 1 0 0 1 1 1024 1024 330 330 This table describes the descriptive statistics of the final sample, which is composed of 1024 technology
acquisitions by 330 firms. For each variable, the mean, median, minimum value, maximum value, and total observations are reported. Variables that are reported are: the deal value of the acquisition (DEAL
VALUE) measured in millions of dollars, the size of the acquirer (SIZE) measured by the total assets in
millions of dollars, the market value of the acquirer (MARKET VALUE) measured in millions of dollars, the book-to-market ratio (BTM) where the book equity = total assets – total liabilities – preferred stock + deferred taxes + convertible debt (Kayhan and Titman, 2004), the return on assets (ROA) = net income / total assets, the return on equity (ROE) = net income / total equity, the amount of block holders (+5% ownership) (BLOCKHOLDERS) present in the acquiring firms, the ratio of independent directors to total
board size (BOARD INDEPENDENCE), the equity-linked compensation (EQUITY), a dummy for the presence of CEO duality (DUALITY), whether the deal is financed with shares (STOCK), whether the deal
is done by a serial acquirer (five or more acquisitions) (SERIAL), whether the target company is a public company (PUBLIC), whether the deal is across different industries (CROSSINDUSTRY), whether the deal
20 5. Results
First, the overall value creation of technology acquisitions in the estimated time period will be examined. Thereafter, a distinction between the time period after the financial crisis will be made. Lastly, a regression analysis will be used to estimate the effect of changing corporate governance mechanisms.
5.1. Overall value creation
The expectation is that the overall effect of value creation is negative, following earlier findings on negative long-term value creation of M&As, due to agency and hubris motives. First, the BHAR approach will be used to estimate the overall value creation, using Eq. (1) and Eq. (2). The results of this approach are reported in Table 2. This table reports the comparison of the average and median three-year HPR of the acquiring firms (1) after conducting a technology acquisition, relatively to the HPR of the matched industry benchmark (2), a value-weighted (3) and equal-weighted (4) benchmark of firms listed on CRSP, and the matched size and book-to-market ratio benchmark (5).
The mean BHAR against all benchmarks is positive, ranging from 4.90% to 10.24%, and significant at least at the 5%-level. The significance of the mean BHARs indicates that the average value creation of technology acquisitions across the whole time period is positive. However, the mean values might be impacted by outliers. Therefore, the median BHARs are also included. When looking at the median BHARs, only the median difference against the size and book-to-market benchmark (5) is negative, but insignificant, while the median is positive and significant for the BHAR against the industry (2) and value-weighted (4) benchmark. Overall, this implicates that companies conducting technology acquisitions seem to create value. The outperformance is contradictorily to the expectation of hypothesis 1 that there is negative value creation.
To further check for the value creation across the whole time period, the CTIME
approach is used. This approach is reported in Table 3. The abnormal returns of the portfolio of acquirers against the industry matched portfolio, both equal- and value-weighted, is compared against a value-weighted portfolio of all firms available on both CRSP and Compustat in
21
22 Table 2: Comparison of the HPR of technology acquirers against the industry, equal-weighted, value-weighted and size and book-to-market benchmark.
(1) (2) (3) (4) (5)
Acquirers Industry Equal-weighted Value-weighted SizeBTM
Mean HPR 38.62% 28.99% 28.27% 30.28% 33.93% (t = 15.20)*** (t = 17.37)*** (t = 28.33)*** (t = 32.97)*** (t =30.53)*** Median HPR 28.45% 20.92% 31.78% 27.93% 35.42% Mean BHAR 9.63% 10.35% 8.33% 4.69% (t = 3.70)*** (t = 3.45)*** (t = 4.38) *** (t = 1.98)** Median BHAR 4.67%** 2.19%* 2.98% -4.16% Observations 1024 1024 1024 1024 1024 Significance levels: *** p<0.01, ** p<0.05, * p<0.1
This table visualizes the mean and median three-year holding period return (HPR, as calculated by Eq. (1)) based on variable (1) the technology acquiring firms, (2) the industry matched benchmark, (3) the equal-weighted benchmark of firms listed on CRSP, (4) the value-weighted benchmark of firms listed on
CRSP, and (5) the matching size and book-to-market benchmark. Furthermore, the mean and median values of the buy-and-hold abnormal returns (BHAR, as calculated by Eq. (2)) are visualized by comparing the HPR mean and median of the acquiring firms (1) to the HPR means and medians of the
benchmarks (2), (3), (4), and (5). T-statistics of the means are reported in between brackets with the corresponding significance level denoted by *. To estimate the significance of the median values a sign
23 Table 3: Returns calendar-time approach.
DV: Rap Equal-weighted portfolio Value-weighted portfolio
VARIABLES (1) (2) (3) (4) Constant 0.001 0.002 -0.003 -0.003** (0.002) (0.001) (0.002) (0.001) MktRF-vw -0.107*** -1.067*** (0.035) (0.046) MktRF-ew -0.148*** -1.013*** (0.032) (0.026) SMB -0.318*** -0.272*** -0.403*** -0.081*** (0.037) (0.037) (0.033) (0.023) HML 0.050 0.047 0.042 0.091** (0.045) (0.040) (0.071) (0.036) Chow-break test 6-2009 0.001*** 0.001*** 0.002*** 0.402 Observations 216 216 216 216 Durbin-Watson statistic 2.018 2.044 1.868 1.600 Dickey-Fuller test 0.000*** 0.000*** 0.000*** 0.000*** Adjusted R-squared 0.454 0.491 0.815 0.915
Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
This table reports the results of the regression model of Eq. (3). The dependent variable (Rap) in regression (1) and (2) is an equal-weighted rolling portfolio of technology acquiring firms that join the portfolio in the first month after an acquisition and leave after 36 months, minus the risk-free rate. The dependent variable (Rap) in regression (3) and (4) is a value-weighted rolling portfolio following the same
methodological. The variable MktRF represents the market risk factor, which is the return on the market minus the risk-free rate, the variable SMB represents a zero-investment portfolio that goes short in big
capitalization stocks and long in small capitalization stocks. The variable HML represents a zero-investment portfolio that goes short in high book-to-market stocks and long in low book-to-market stocks.
The constant is referred to as ‘alpha’, which measures the abnormal returns of the portfolio. In the first two regressions (1) & (2), the market factor is based on an equal-weighted market portfolio of firms listed
on CRSP. In the regression (3) & (4), the market factor is based on a value-weighted portfolio of firms listed on CRSP. A Chow-break test is included to examine whether a break is present at 6-2009.
24
5.2. Difference between time periods
The total effect could be impacted by the difference between time periods, as indicated by hypothesis 2. To test if there is a difference, a separation between the different time periods is made. This implies that we expect a difference between acquisitions done in period one, ranging from January 2001 to June 2009 and period two, ranging from July 2009 to December 2015. Therefore, both approaches of estimating the abnormal returns for the acquiring companies are performed on the two time periods. Additionally, a Chow-break test, included in Table 3, is used on the CTIME regressions of the whole period to identify whether there is indeed a break point at June of 2009. The Chow-break tests of regression (1), (2) and (3) indicates that there is a difference between time periods, significant at the 1%-level, while only the regression of (4) does not indicate a significant difference between time periods. Overall, the results strongly suggest that there is a difference in value creation.
Next, an estimation of the abnormal returns differentiated by periods is done to determine whether there are differences in value creation of technology M&As after the financial crisis. The results for the two time periods using the BHAR approach are reported in Table 4. Panel A displays the results for the time period up to and including June 2009. Panel A indicates that there are mixed results on the significance of the mean BHAR. While the BHAR of the industry (2), equal-weighted (3) and size and book-to-market (5) benchmark are all negative, only the abnormal return against the equal-weighted (3) benchmark is significant, while the abnormal return against the value-weighted (4) benchmark is significantly positive. Furthermore, the median BHARs against all benchmarks are negative and significant, with only the abnormal returns based on the value-weighted (4) benchmark being not significant. From the results above we can conclude that there are mixed results on the significance of the BHAR approach, but that most of the results indicate a negative or an insignificant negative BHAR. The median values all indicate a negative effect, with only one being not significant. These results weakly support hypothesis 1, as there seems to be negative abnormal returns in the period up to and including June 2009, but these are mostly insignificant.
25 Table 4: Comparison of the HPR against the industry, equal-weighted, value-weighted, and size and book-to-market benchmark differentiated by time periods.
Panel A (1) (2) (3) (4) (5)
January 2001 – June 2009
Acquirers Industry Equal-weighted Value-weighted SizeBTM
Mean HPR 21.05% 26.20% 28.75% 13.44% 23.70% (t = 5.47)*** (t = 8.98)*** (t = 13.59)*** (t = 9.66)*** (t = 12.41)*** Median HPR 6.34% 10.67% 21.15% 10.11% 19.55% Mean BHAR -5.15% -7.70% 7.60% -2.65% (t = -1.34) (t = -2.18)** (t = 2.15)** (t = -0.76) Median BHAR -5.19%* -12.40%*** -2.35% -12.01%*** Observations 493 493 493 493 493 Panel B (1) (2) (3) (4) (5) July 2009 – December 2015
Acquirers Industry Equal-weighted Value-weighted SizeBTM
Mean HPR 54.93% 31.57% 31.71% 42.03% 43.43% (t = 17.19)*** (t = 18.20)*** (t = 50.91)*** (t = 73.73)*** (t = 41.34)*** Median HPR 50.66% 29.31% 34.10% 41.76% 43.68% Mean BHAR 23.35% 23.22% 12.90% 11.50% (t = 6.83)*** (t = 7.32)*** (t = 4.08)*** (t = 3.59)*** Median BHAR 14.01%*** 17.17%*** 7.18%** 6.11%** Observations 531 531 531 531 531 Significance levels: *** p<0.01, ** p<0.05, * p<0.1
This table visualizes the mean and median three-year holding period return (HPR, as calculated by Eq. (1)) based on variable (1) the technology acquiring firms, (2) the industry matched benchmark, (3) the equal-weighted benchmark of firms listed on CRSP, (4) the value-weighted benchmark of firms listed on
CRSP, and (5) the matching size and book-to-market benchmark. Furthermore, the mean and median values of the buy-and-hold abnormal returns (BHAR, as calculated by Eq. (2)) are visualized by comparing the HPR mean and median of the acquiring firms (1) to the HPR means and medians of the
benchmarks (2), (3), (4), and (5). T-statistics of the means are reported in between brackets with the corresponding significance level denoted by *. To estimate the significance of the median values a sign test is used. Panel A shows the estimation of the values for period one (January 2001 – June 2009), Panel
26
To further examine the positive value creation after the financial crisis, a differentiation between periods is made for the CTIME approach. When looking at the CTIME approach regressions on the first period in Table 5, regressions (1) and (3) against the equal-weighted market portfolio show an insignificant return, indicating that no significant effect can be
examined on the abnormal returns for the acquirers of technology acquisitions in the period up to and including June 2009. Regressions (5) and (7) against the value-weighted portfolio show a clear negative and significant alpha, indicating underperformance of technology acquirers in the period up to and including June 2009. The negative alpha of the value-weighted portfolio partially supports hypothesis 1 that states that there is negative value creation for the shareholders of the technology acquiring firms.
When looking at the period after July 2009, the abnormal returns on regression (2) and (4) against the equal-weighted benchmark show that there are positive and significant monthly abnormal returns to the acquirers of technology companies, indicating that there are monthly abnormal returns not explained by the model of 0.002% to 0.003%. When we look at regression (6) and (8) against the value-weighted market portfolio, no significant abnormal returns are measured, neither negative nor positive. When comparing the regressions within each market portfolio between periods, there seems to be a positive transition. For the equal-weighted acquirers’ portfolios (regressions (1)-(4)), the abnormal returns are insignificant in period one, and become positive and significant in period two against both the equal- and value-weighted market portfolio. The results for the value-weighted acquirers’ portfolio (regressions (5)-(8)) are somewhat different in the effect and significance, when compared against the equal-weighted acquirers’ portfolio. The abnormal return in period one is negative and significant, while they become insignificant in period two. While there are some differences between the equal- and value-weighted regressions, the overall positive transition in value creation is supported,
27 Table 5: Calendar-time approach comparison between periods.
DV: Rap Equal-weighted acquirers’ portfolio Value-weighted acquirers’ portfolio
(1) (2) (3) (4) (5) (6) (7) (8)
VARIABLES Period 1 Period 2 Period 1 Period 2 Period 1 Period 2 Period 1 Period 2
Constant 0.000 0.002* 0.001 0.003* -0.008** 0.000 -0.005** -0.001 (0.003) (0.001) (0.003) (0.001) (0.004) (0.002) (0.002) (0.002) MktRF-vw -0.160*** -0.039 -1.147*** -0.981*** (0.051) (0.033) (0.066) (0.052) MktRF-ew -0.187*** -0.063** -1.026*** -0.990*** (0.044) (0.031) (0.034) (0.037) SMB -0.347*** -0.204*** -0.286*** -0.187*** -0.414*** -0.302*** -0.075** -0.056 (0.041) (0.035) (0.042) (0.037) (0.045) (0.048) (0.029) (0.044) HML 0.072 0.000 0.069 0.000 0.057 0.025 0.104** 0.080** (0.068) (0.037) (0.060) (0.036) (0.103) (0.048) (0.050) (0.040) Observations 101 114 101 114 101 114 101 114 Dubrin-Watson statistic 1.983 1.891 2.258 1.869 1.981 1.989 1.517 1.728 Dickey-Fuller test 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** Adjusted R-squared 0.519 0.241 0.559 0.255 0.845 0.779 0.938 0.862
Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
This table reports the results of the regression model of Eq. (3). The dependent variable (Rap) in (1), (2), (3), and (4) is the abnormal return of an equally weighted rolling portfolio of technology acquiring firms that join the portfolio in the first month after an acquisition and leave after 36 months minus a portfolio of
industry matched firms that follows the same methodology, minus the risk free rate. In (5), (6), (7), and (8), the dependent variable (Rap) is the abnormal return of a value-weighted rolling portfolio following the
same methodology. The variable MktRF represents the market risk factor, which is the return on the market minus the risk-free rate, the variable SMB represents a zero-investment portfolio that goes short in
big capitalization stocks and long in small capitalization stocks. The variable HML represents a zero-investment portfolio that goes short in high book-to-market stocks and long in low book-to-market stocks. The constant is referred to as ‘alpha’, which measures the abnormal returns of the portfolio. In regressions (1), (2), (5), and (6) the market factor is on a value-weighted market portfolio of firms listed on CRSP. In
the regression (3), (4), (7), and (8), the market factor is based on an equal-weighted portfolio of firms listed on CRSP. In (1), (3), (5), and (7) the regression is based on period 1: January 2001 – June 2009, in
28
Lastly, a comparison of the three-year BHARs between estimation years is made, to see whether there are substantial changes between the two periods. These results are added in Table 6. The results of Panel A indicate that there is clear underperformance in the first two years after a technology M&A in the period up to and including June 2009, measured by the negative and significant mean BHARs and negative medians. This negative value creation contrasts with the BHARs in Panel B. In period two, after June 2009, all mean BHARs are positive and significant. The outperformance of technology acquirers is the highest in the second year after the
29 Table 6: Comparison of the 3-year holding period between periods.
Panel A BHARind BHARind Panel B BHARind BHARind
January 2001 – June 2009
Mean Median July 2009 –
December 2015 Mean Median Year 1 -6.21% -3.92% Year 1 4.23% 1.36% (t=-3.17)*** (t=2.38)** Year 2 -4.57% -0.81% Year 2 8.65% 5.41% (t=-1.93)* (t=5.11)*** Year 3 1.24% -0.53% Year 3 5.77% 4.35% (t=0.65) (t=3.25)*** Observations 493 Observations 531 Significance levels: *** p<0.01, ** p<0.05, * p<0.1
This table reports the buy-and-hold abnormal return (BHAR, as calculated by Eq. (2)) based on the industry benchmark. The left column shows the three-year BHARs differentiated by years of technology acquisitions completed in the period of January 2001 – June 2009. The right column shows the three-year
30
5.3. Regression analysis
Based on the results presented above, it seems that there is indeed a significant difference between time periods. To examine whether the difference between time periods is due to the change in corporate governance mechanisms, the overall effect of the different corporate governance mechanisms and the change between time periods will be estimated by conducting an ordinary least squares (OLS) and a fixed effects (FE) regression, as presented in Eq. (4). The expectation is that the variables for the ratio of independent directors (BOARD
INDEPENDENCE), equity-linked compensation (EQUITY) and block holders (BLOCK HOLDERS) show a positive and significant sign, while the variable for duality (DUALITY)
should show a negative and significant sign. We expect that the variables for the effect of these mechanisms after the financial crisis (BOARD INDEPENDENCEafter, EQUITYafter, BLOCK
HOLDERSafter, and DUALITYafter) show a significant hypothesized effect. This would indicate
that the value creation of technology acquisitions after the financial crisis is explained by changes in corporate governance mechanisms.
The results of these two regressions are added in Table 7. The OLS regression (1) indicates that the dummy variable (AFTERCRISIS) for the period after the crisis is insignificant and positive, indicating no significant outperformance after the financial crisis. When looking at the corporate governance variables mentioned above, no significant impact of the mechanisms on the abnormal returns are observed. Additionally, the corporate governance mechanisms in the period after the financial crisis are all negative but insignificant, indicating that there is no change of the impact of corporate governance mechanisms after the financial crisis. However, the model has a low adjusted R2.
A better regression to fit the data would be a FE regression, to account for the
heterogeneities between firms. This FE regression (2), significant at the 1%-level, is displayed in Table 7. Using this model increases the adjusted R2 to 21.7%. The dummy variable for the period after the financial crisis (AFTERCRISIS) is significant at the 10%-level and indicates a positive value creation, supporting the earlier results of the BHAR and CTIME approach. This positive and significant variable supports hypothesis 2 that there is a positive transition in value creation after the financial crisis. When looking at the corporate governance mechanisms, board
independence (BOARD INDEPENDENCE) is positive and significant at the 1%-level. The significant positive sign indicates a positive relation between board independence and abnormal return. However, there seems to be no influence of the variables for block holders (BLOCK
HOLDERS) and duality (DUALITY). Furthermore, there is no difference observed in changes of
31
insignificant variables BLOCK HOLDERSafter, DUALITYafter, and BOARD
INDEPENDENCEafter. The insignificance of these variables is in contradiction with hypothesis
3 and 4, that stated that the effect of these mechanisms after the crisis explain the positive transition in value creation.
The significance of some control variables in the FE-regression of Table 7 generates some additional insight. We observe that the deal value (DEALVALUE) of technology M&As is positive and significant at the 5%-level, indicating that bigger deals tend to increase the value creation of technology acquisitions. However, the market value (MARKET VALUE) of the acquirer is negative and significant at the 1%-level, indicating that companies with a high market value tend to destroy value when conducting technology acquisitions. The book-to-market (BOOK-TO-MARKET) ratio is not significant, indicating that the over- or under-pricing of companies does not impact the value created. Public (PUBLIC) deals have a negative value effect significant at the 1%-level, indicating that public deals destroy value. Lastly, the dummy variable on cross industry (CROSSINDUSTRY) acquisitions is positive and significant at the 5%-level.
As stated earlier, the equity compensation variable is only reported by a small sample of firms. When using the sample of firms that have values for the equity compensation, no
differences in outcomes occur when conducting a FE regression. The variable for equity-linked compensation is insignificant, indicating that there is no positive relation between equity-linked pay structures and M&A value creation.
32 Table 7: OLS and FE regression
Dependent variable: BHARind (1) (2)
VARIABLES OLS FE BOARD INDEPENDENCE 0.392 2.006*** (0.324) (0.455) BLOCK HOLDERS 0.002 0.057 (0.015) (0.047) DUALITY 0.039 0.112 (0.087) (0.139) AFTERCRISIS 0.512 1.128* (0.369) (0.587)
BOARD INDEPENDENCE after -0.067 -0.922
(0.447) (0.697)
DUALILTY after -0.128 -0.049
(0.111) (0.154)
BLOCK HOLDERS after -0.037 -0.003
(0.024) (0.030) ROA -0.758 -1.686 (0.316) (0.356) ROE 0.484 1.270 (0.763) (0.898) TOTAL ASSETS 0.000 0.000*** (0.000) (0.000) DEAL VALUE 0.051*** 0.044** (0.019) (0.020) MARKET VALUE -0.073** -0.654*** (0.030) (0.132) BOOK-TO-MARKET 0.006 -0.353 (0.0134) (0.293) CROSSBORDER -0.077 -0.064 (0.055) (0.056) CROSSINDUSTRY 0.094 0.143** (0.064) (0.069) STOCK -0.169** -0.157 (0.083) (0.092) BUBBLE -0.245 0.283 (0.178) (0.182) PUBLIC -0.159** -0.237*** (0.067) (0.083) SERIAL 0.043 o. (0.058) Constant -0.378 2.858** (0.325) (0.995) Significance of FE-regression 0.000*** Observations 1,024 1,024 Adjusted R-squared 0.047 0.217 Number of firms 330
Robust standard errors in parentheses, *** p<0.01, ** p<0.05, * p<0.1. This table reports an OLS (1) and FE regression (2), as calculated by Eq. (4). The dependent variable is the buy-and-hold abnormal return on the matched
33 6. Discussion
The results presented above generate insight in the value creation of technology
acquisitions, and whether changing corporate governance mechanisms affect this phenomenon. When examining the value creation of technology acquisitions, there seems to be a positive difference in abnormal returns after the financial crisis, as measured by both the BHAR and CTIME approach. The BHAR approach finds mostly insignificant or a significant negative value creation in the period up to and including June 2009, while this becomes significantly positive after this date. The CTIME approach finds a transition in value creation. The equal-weighted portfolios have insignificant abnormal returns up to and including June 2009, and significantly positive abnormal returns after this date, while the value-weighted portfolios have significant negative returns in the period up to an including June 2009, and insignificant value creation afterwards. The positive transition in value creation is in contradiction with earlier research on long-term value creation of (technology) M&As by Kohers and Kohers (2001), Rau and Vermaelen (1998), and Alexandridis, Mavrovitis, and Travlos (2012) that predominantly find negative long-term abnormal returns. However, these results are in line with the short-term announcement returns study of Alexandridis, Antypas, and Travlos (2017), which suggests that there are positive abnormal returns for the period after the financial crisis. The results suggest that there is at least a difference in value creation for technology acquisitions that are conducted after the financial crisis.
Although the positive transition in value creation of technology acquisitions seems to be present, there is no indication that this positive value transition is due to changes in corporate governance mechanisms after the financial crisis. The effects of corporate governance
mechanisms on the whole time period results only in a significant positive relation between board independence and abnormal returns, as measured by the ratio of independent board members on the long-term value creation of M&As, in line with earlier research by Baysinger and Butler (1985). However, the insignificant relation of duality (Masulis and Cong Wang, 2007), block holders (Moeller and Schlingemann, 2005) and equity-linked compensation (Bliss and Rosen, 2001) are not in line with earlier findings. The expectation that the positive abnormal returns after the financial crisis can be explained by changing corporate governance mechanisms is not supported, contradicting the claims of Alexandridis, Antypas, and Travlos (2017).
34
Lastly, the significance of control variables in the fixed effects regression generate some additional insight in the decision to conduct technology M&As. The effects between value creation and market value (Moeller, Schlingemann, and Stulz, 2005) and public deals (Fuller, Netter, and Stegemoller, 2002) are negative, in line with earlier research. However, there are also some differences in effects of control variables. This research finds that for technology
acquisitions, the deal value positively influences value creation, contradicting earlier research of Alexandridis, Fuller, and Terhaar (2013). Also, cross-industry technology acquisitions are positively influencing long-term abnormal returns of M&As, contradicting research of Morck, Schleifer, and Vishny (1990), and indicating that technology could be better used across industries possibly due to industry and technology convergence (Kim et al., 2015).
Limitations of the methodology in this research are that the results of the BHAR
approach could be biased due to differences in benchmark anomalies when estimating the value creation of acquisitions (Dutta and Jog, 2009). This research tries to account for this bias by using benchmarks based on the industry, similar sized and valued companies, and the whole market. The BHAR approach also has some drawbacks with the data sample used in this research, because it does not account for overlapping estimation periods for the holding period returns between and within companies. Additionally, a limitation of this research with regard to corporate governance mechanisms is the potential endogeneity and measurement error in
35 7. Conclusion
This research finds evidence that the long-term performance of acquirers in technology M&As after the financial crisis is at least nonnegative using both the BHAR and CTIME approach, indicating a positive transition and explaining part of the recent trend in technology M&As after the financial crisis. This positive transition contradicts earlier research on long-term performance of (technology) acquiring firms, which finds significant negative returns. While the trend of technology M&As is simultaneous with reforms and an increase in importance of corporate governance mechanisms, none of the corporate governance mechanisms after the financial crisis examined explain the positive difference in value creation between periods. There seems to be a positive change in returns for technology M&As across industries, as indicated by the significance of the cross-industry variable in Table 7. Earlier research suggests that
diversifying M&As destroy shareholder value, in contradiction to the finding of this research. This finding indicates that technology M&As across industries create value for shareholders, which could be an indication of the technology and industry convergence as measured by Kim et al. (2015) that makes it easier for companies to use technology in their business and go beyond the classic industry borders.
The theoretical implication of this research is that technology M&As after the financial crisis are different from M&As performed up to and including June 2009, and that these M&As do not seem to be driven by the agency (Fama and Jensen, 1983) or hubris motives (Roll, 1986), but that value creation is the dominant motive. However, the notion that the difference in value creation is explained as a result of an increase in the control and monitoring of managers is not substantiated. The practical implication of this research is that acquisitions of external
technology could be a valid option to develop a business in the current technology driven and changing business environment. The classic view that M&As are conducted mainly due to agency and hubris motives is not found, meaning that it could be valuable for shareholders when management proposes a technology acquisition. Additionally, this research finds that imposing legislation and increasing the attention to corporate governance mechanisms, leading to changes in corporate governance mechanisms, does not seem to influence the decision-making behaviour of companies.
Lastly, directions for future research are given. Future research should focus on
36
37 References
Alexandridis, G., Mavrovitis, C., Travlos, N., 2012. How have M&As changed? Evidence from the sixth merger wave. European Journal of Finance 18, 663-688.
Alexandridis, G., Fuller, K., Terhaar, L., 2013. Deal size, acquisition premia and shareholder gains. Journal of Corporate Finance 20, 1-13.
Alexandridis, G., Antypas, N., Travlos, N., 2017. Value creation from M&As: new evidence. Journal of Corporate Finance 45, 632-650.
Amihud, Y., Lev, B., 1981. Risk reduction as a managerial motive for conglomerate mergers. Bell Journal of Economics 12, 605-616.
Bainbridge, S., 2012. Corporate Governance after the Financial Crisis. Oxford University Press, New York.
Baysinger, B., Butler, H., 1985. Corporate governance and the board of directors: performance effects of changes in board composition. Journal of Law, Economics, & Organization 1, 101-124.
Berkovitch, E., Narayanan, M., 1993. Motives for takeovers: an empirical investigation. Journal of Financial and Quantitative Analysis 28, 347-362.
Berle, A., Means, G., 1932. The Modern Corporation and Private Property. Commerce Clearing House, New York.
Bliss, R., Rosen, R., 2001. CEO compensation and bank mergers. Journal of Financial Economics 61, 107-138.
Bruner, R., 2002. Does M&A pay? A survey of evidence for the decision-maker. Journal of Applied Finance 12, 48-68.
Bruner, R., 2005. Deals from hell: M&A lessons that rise above the ashes. John Wiley & Sons, New York.
Chatterjee, S., 1986. Types of synergy and economic value: the impact of acquisitions on merging and rival firms. Strategic Management Journal 7, 119-139.
Coase, R., 1937. The nature of the firm. Economica 4, 386-405.
38
Datta, S., Iskandar-Datta, M., Raman, K., 2001. Executive compensation and corporate acquisition decisions. Journal of Finance 56, 2299-2336.
Doidge, C., Karolyi, G., Stulz, R., 2004. Why do countries matter so much for corporate governance? Journal of Financial Economics 86, 1-39.
Dutta, S., Jog, V., 2009. The long-term performance of acquiring firms: a re-examination of an anomaly. Journal of Banking & Finance 33, 1400-1412.
Edmans, A., 2014. Block holders and corporate governance. Annual Review of Financial Economics 6, 23-50.
Fama, E., Jensen, M., 1983. Separation of ownership and control. Journal of Law and Economics 26, 301-325.
Fama, E., French, F., 1993. Common risk factors in the returns on stocks and bonds. Journal of Financial Economics 33, 3-56.
Fama, E., 1998. Market efficiency, long-term returns, and behavioural finance. Journal of Financial Economics 49, 283-306.
Franks, J., Harris, R., Titman, S., 1991. The post-merger share-price performance of acquiring firms. Journal of Financial Economics 29, 81-96.
Fuller, K., Netter, J., Stegemoller, M., 2002. What do returns to acquiring firms tell us? Evidence from firms that make many acquisitions. Journal of Finance 57, 1763-1793.
Gort, M., 1969. An economic disturbance theory of mergers. The Quarterly Journal of Economics 83, 624-642.
Grinstein, Y., Hribar, P., 2004. CEO compensation and incentives: evidence from M&A bonuses. Journal of Financial Economics 73, 119-143.
Grossman, S., Hart, O., 1986. The costs and benefits of ownership: a theory of vertical and lateral integration. Journal of Political Economy 94, 691-719.
Harford, J., 2005. What drives merger waves? Journal of Financial Economics 77, 529-560. Hillier, D., McColgan, P., 2006. An analysis of changes in board structure during corporate governance reforms. European Financial Management 12, 575-607.
