What is the effect of different deal characteristics on Cumulative Abnormal Return of Mergers and Acquisitions in The Netherlands?  

What is the effect of different deal characteristics on

Cumulative Abnormal Return of Mergers and

Acquisitions in The Netherlands?

Student: Esmée Grim Student number: 11269022 Faculty: BSc Economics and Business Track: Economics and Finance Supervisor: Mr. Lopez de Silanes Date: 29th _{of June 2020} Abstract Mergers and acquisitions are a widely used method to expand businesses. In earlier papers, the effects of many deal characteristics have been researched, but less specific for the Netherlands in the period of 2002-2019. In this paper, the effects of 100% acquisitions and partial acquisitions with final stake 100% on the CAR of the bidding company are researched, together with payment method and industry linkage for different periods around the announcement date. The share payment and industry linkage are only of significant positive influence on the CAR of the bidder in the long-term performance and the run-up period to the announcement date. This result contradicts the research of Sherif (2012) where stock payment was found to be of negative influence on the abnormal return of mergers and acquisitions in the UK. And remarkably, Corhay and Tourani Rad (2000) found a negative effect of industry linkage for Dutch companies in 1990-1996. In this research, 100% acquisitions are found to be creating a significantly higher CAR in event windows close to the announcement date, which is in line with the research of Chen (2008) which stated that full acquisitions are performed to create synergies and

Statement of Originality This document is written by student Esmée Grim, who declares to take full responsibility 2for the contents of this document. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Table of Contents

1. Introduction ... 4 2. Literature ... 6 2.1 Mergers and Acquisitions ... 6 2.2 Abnormal Return ... 6 2.3 Different effects of the transaction ... 7 2.4 Payment method ... 7 2.5 Industry linkage ... 8 2.6 Partial acquisitions ... 9 3. Hypotheses ... 10 4. Data Collection and Research Methodology ... 11 4.1 Data Collection ... 12 4.2 Event Study ... 13 4.3 Event windows ... 14 4.5 Regression ... 14 5. Results ... 15 5.1 Cumulative Abnormal Returns ... 15 5.2 Basic Regressions ... 16 5.3 Regressions with control variables ... 19 6. Robustness check ... 22 7. Conclusion ... 24 8. Reference list ... 26

1. Introduction

Many researchers have done research on the Cumulative Abnormal Return, also known as CAR, and which factors have an influence on the Cumulative Abnormal Return. The geographical areas and outcome vary between these papers. The effect of the payment method on the Cumulative Abnormal Return shows different significant effects. Dutta, Saadi and Zhu (2012) found a negative effect of cash payments on the CAR, however other researchers found a positive effect or no effect when a transaction was completed by a cash payment. Also, the industry in which the involved companies are active was found to be of influence on the CAR. In the research of Bhara and Huang (2013), the conclusion is that when the two companies in the acquisition are active in the same industry, the Cumulative Abnormal Return is higher in comparison to acquisitions with companies that are active in different industries. But a research about acquisitions in the Netherlands, 1990 to 1996, from Corhay and Tourani Rad (2000) found that industry linkage is negative for the CAR. Due to many different outcomes, this research will be investigating the effect of payment method, industry linkages and the effect of full acquisitions compared to partial acquisitions with final stake 100%, for the CAR of the bidding company in the acquisition process. In the research of Chen (2008) the reasons for acquisitions were tracked down, with the conclusion that full acquisitions are made for synergy reason, but is this showing through in the CAR of the acquiring firm? The Netherlands, by itself, is not often investigated, even though there are many Dutch companies acquiring other companies. For this reason, the focus of this paper is on the Dutch companies that act as the acquiring firm. The paper starts with an outline of the prior investigated variables that might have and influence on the Cumulative Abnormal Return. Thereafter the hypotheses will be described. The data collection, research method, and regression results are explained, and in addition a robustness check is performed to give assurance about the robustness and predictiveness of the original regressions. Lastly a conclusion will be made about the results of the regression, the limitations of this research are given and potential further research that could be done.

For this research, acquisitions in the period from 2002 to 2019 are used. All transactions have a final stake of 100% of the shares, the acquiring companies are originated in The Netherlands and the deal has been completed. The acquiring companies are active in different industries, from construction to software and the banking sector. Firstly, the Cumulative Abnormal Returns of the acquiring companies have been calculated for different periods of time, from run-up period to long-term periods after the announcement date, via an event study. Secondly, the regressions for the different event windows are performed and tested for a significance level of 10%, 5% and 1%. These regressions are firstly performed without the control variables, year fixed effects and industry fixed effects. Later the control variables are added to the regressions. Lastly a robustness check is performed by using the Buy-Hold Abnormal Return to conclude that the regressions are showing the same outcomes when the dependent variable is determined in another way. For the results of the regressions, different periods can be categorized. For the run-up period, the share payment and the industry linkage dummies showed a significant positive effect on the CAR of the bidding companies. The same outcome was found for the long-term period after the announcement date. For companies that are active in the same industry an increase of around 7% of the CAR of the bidding company was found in the run-up and long-term periods. For the short event windows around the announcement date a significant effect was found for the 100% acquisitions. The Cumulative Abnormal returns for the acquiring company tend to be 4% higher on average when the transaction was a full acquisition in comparison to a partial acquisition that is completed to 100% possession of the shares. Which is confirming the research of Chen (2008) where he found the reason for 100% acquisitions is mostly about the synergies that are arising.

2. Literature

2.1 Mergers and Acquisitions

There are multiple ways to expand an organisation, of which acquisitions are most often used. Instead of acquiring a company, another possibility to grow a business is by merging with another company. All the former shareholders of the two companies get shares in the new firm (Schoenberg, 2006). When a firm, the acquirer, purchases another firm, the target firm, we are talking about an acquisition. The acquiring firm is interested to grow and expand the business and finds a target firm to take over. The acquirer bids a price to buy the firm, if this bid is accepted, the target firm will become possession of the bidding firm. At least the majority of the shares of the target firm will be bought by the acquiring firm, so they have the power to make decisions in the organisation and do not need approval from the other shareholders. When two firms decide to combine their knowledge and other assets, this will generate synergies (Mariana, 2015). The same research states that synergies arise when the intangible or tangible assets of the two firms are combined and create more value than the total value of these assets when they are used separately. After the merger or acquisition, the two firms will be more efficient or less costly in running their business. Therefore, the synergies create extra value for the new entity or for both the acquired and acquiring firm, which implies that the shareholder’s value will increase (Campa, 2004).

2.2 Abnormal Return

When a merger or acquisition is announced, the market will react immediately, showing an increase or decrease in the share price of the companies involved. This represents the expectation of the market about the synergies (Luh Putu and Dony, 2017) that will arise, from positive to negative values. These expectations are not always true and can widely vary between companies. In some situations, there are rumours about an organisation taking over another organisation. In this case the market already reacts to the rumours with a change in stock prices. On the actual announcement date of the merger or acquisition, there will be

little to no reaction of the market, as the market had already anticipated on the new situation and possible synergies. Abnormal return is the difference between the stock’s actual return and the expected return, also known as the benchmark (Bodie et al., 2014). The benchmark is based on historical data and recent developments, the market return is mostly used as the benchmark. Abnormal returns are positive when the merger or acquisition creates positive synergies, the value of the firm increased more than the value of the rest of the market. (Bodie et al., 2014).

2.3 Different effects of the transaction

According to Moeller et al. (2004) the weighted average abnormal return is 1,1%, which indicates that there are positive synergies present. But concluding from this research, the average abnormal returns are around 2% higher for smaller companies than for large companies. Beside this, Putu and Dony (2017) have found that a positive Net Profit Margin of the acquirer a year before the acquisition, has a slightly positive effect on abnormal returns. However, a higher firm value of the acquiring firm has a negative effect on the abnormal return (Putu and Dony, 2017 & Li, Li and Wang, 2016). Large abnormal returns are earned by acquiring companies that make successful bids, in contrast with negative abnormal returns in case of unsuccessful bids (Dodd et al., 1977). Another research, from Campa and Hernando (2004), found that only the target firms had a positive cumulative average abnormal return, while the acquiring firms had zero to negative abnormal return. Wong (2009) came with the opposite conclusion, the abnormal returns of target firms are negative, while the returns of the acquiring firm are positive.

2.4 Payment method

There are different characteristics of M&A’s that can cause these big differences in the abnormal return, for example method of payment, is the transaction paid with cash or in a different way? When taxes are not taken into account Modigliani and Miller (1958) state that the market value of a firm is not influenced by the way the firm decides

to finance their investments. However currently, corporate tax must be taken into account when looking at the optimal way of financing investments. Dutta, Saadi and Zhu (2012) have found that when an acquisition or merger is paid by cash it will generate lower abnormal return, compared to other payment methods. But another research found a different result, stock has a significant negative effect on the abnormal return after the M&A announcement. (Sherif, 2012). Sherif (2012) looked at the effect of the size of the acquiring firm and target firm on the Cumulative Abnormal Return in the same study.

2.5 Industry linkage

Another characteristic that could influence the abnormal return is industry linkage (Bhara & Huang, 2013). Industry linkage is the extent to which the two companies involved in the acquisition process are active in the same industry. Bhara and Huang (2013) concluded that when the two firms operate in the same industry, they will have higher returns than firms that are active in two different industries. Into this same regression Bhara and Huang (2013) took the payment method and the status of the acquiring company into account. The return will be lower when there is no industrial connection between the firms, since they need different knowledge and specific skills for each of the firms and will not be able to understand earlier strategic decisions (Ellwanger and Boschma, 2015). This research was performed on Dutch acquisitions, where the industry and the proximity of the companies where used as independent variables to determine the Cumulative Abnormal Return. On the other side, Corhay and Tourani Rad (2000) have investigated Dutch firms acquiring foreign companies from 1990 to 1996 and found a negative effect of industry linkage. The cumulative abnormal was found to be lower when Dutch companies acquired companies in the same line of business in comparison to when the companies are performing totally different business activities (Corhay and Tourani Rad, 2000). Together with the industry linkage the international exposure was tested for a significant effect in this research.

2.6 Partial acquisitions

The acquiring company can take over another firm completely, but a firm can also opt to firstly buy a majority of the shares. In the latter, the acquiring firm is having the power to make decisions but does not have to pay the full amount the company is worth. Later it is possible to expand the holding to a larger part of the shares, or even 100% of the shares. Akihigbe et al. (2007) have investigated partial acquisitions and found a positive linkage for the possibility that the target firm will be involved in a full acquisition later in time by the same firm or another third party. Chen (2008) found that a partial acquisition is done for other reasons than where a full acquisition is made for. A partial acquisition is a strategic decision, while a full acquisition is performed to gain control over the other company and create synergies. In practice it is seen that a partial acquisition is completed to a 100% possession of shares of the acquired firm, even though a partial acquisition was intended for strategical benefits. Because a full acquisition is mostly used to generate synergies, this would suggest that the Cumulative Abnormal Return around the announcement of a full acquisition is higher than for a partial acquisition. But what is the effect of the second deal including the same companies on the abnormal return? Is there a difference in CAR between a full acquisition and an acquisition that is completed after the majority of the shares was already in possession of the acquiring firm? Therefore, together with the little researches for this specific geographical area, the research question will be the following. What is the effect of different deal characteristics on the Cumulative Abnormal Returns of M&A’s made by companies originated in The Netherlands?

3. Hypotheses

Regarding the prior studies that focus on different determinants for the Cumulative Abnormal Return, this paper will take three different aspects into account, acquisition form, payment method and industry linkage. Since the found effects are not unambiguous throughout prior studies, these will be tested for the Dutch acquirers with the following regression: Cumulative Abnormal Return = a + b1D1+ b2D2 + b3D3 + b4D4 + b5V5 + Year fixed effects + industry fixed effects + εi Where, a = the constant coefficient bi = the regression coefficient, i = 1,2,3,4,5,6 D1 = 100% Acquisition Value of 1 when the deal is an 100% acquisition at once Value of 0 when the deal was a partial acquisition and now completed to full acquisition D2 = Cash payment Value of 1 when the deal is paid with cash Value of 0 when the deal is paid with different method than cash or multiple methods D3 = Shares payment Value of 1 when the deal is paid with shares Value of 0 when the deal is paid with different method than shares or multiple methods D4 = Same industry Value of 1 when both firms do business in the same industry Value of 0 when firms do business in different industries V5 = the natural logarithm of deal value in euros εi = the error term The dummy variable for acquisition distinguishes the acquisitions where a company is acquiring another company completely at once, so takes 100% of the shares, and completing an earlier partial acquisition to 100% of the shares. Is the reaction of the market more positive when the firm is already partly in possession of the acquirer? Or is the bid for 100% of the shares creating more synergies and thus a higher CAR?

Hypothesis 1: it is expected that there is no difference in CAR caused by a 100% acquisition, and thus b1 = 0. The cash payment has been researched multiple times and in different ways but has not resulted in an unambiguous effect. The same can be concluded for a deal that is paid with shares. Hypothesis 2: The effect of a cash payment on the CAR is expected to be zero, and thus b2 = 0 Hypothesis 3: The effect of a shares payment on the CAR is expected to be zero, and thus b3 = 0 The last dummy variable is the industry dummy, is it positive for companies to take over a company within the same industry? Or causes it more synergies when taking over a company in another working area? Hypothesis 4: the expectation is for the industry linkage to have no effect on the CAR when taking over another company, and thus b4 = 0 In this regression the natural logarithm of the deal value is used as a control variable. Deal value is one of the deal characteristics and can be of influence on the outcome of the merger or acquisition in general, and therefore influence the height of the Cumulative Abnormal Return. This variable will not be used to find a significant effect of deal value on the Cumulative Abnormal Return, it only serves the purpose of the control variable to keep the conditions the same among throughout the research. Therefore, no hypotheses will be stated for b5. The Year fixed effects and the industry fixed effects are added to the regression to create a more robust and a better predicting regression model.

4. Data Collection and Research Methodology

In this section the method for the research conducted will be described. Firstly, the data collection and the reasons for the criteria of the sample will be explained. Secondly, a description on how the dependent variable and the Cumulative Abnormal

will be used for the regression, together with the criteria for the significance, will be described.

4.1 Data Collection

For this research only mergers and acquisitions regarding Dutch listed companies will be considered. A widely used database for merger and acquisition data is Zephyr, which was also used for the data of this research. The following requirements for the data list are in place. i. The deal includes a Dutch listed company as acquirer ii. The deal is an acquisition iii. The deal has taken place between 2002 and 2019 iv. The deal is completed v. The deal includes a 100% take-over of the targets shares in the final stage. When taking data from 2002 to 2019 it is possible to catch two merger and acquisitions waves in the data set (IMAA-institute, 2020). Zephyr provides the deal value, announcement date, industry linkages, by BvD sector description, and the payment method for the list of companies that fit all criteria listed above, 175 transactions in total. In the period of 2002 to 2019 more than 175 acquisitions have taken place, due to the missing information in Zephyr the complete observations are down to 175 transactions. The acquisitions on the list vary from a deal value of €47 euros for a 100% take over to an acquisition with a deal value of €61 900 000. The deals are categorized in eight industries in which they are active and one extra sector marked as other. In table 1 the division of the different industries in the data set is shown. And in table 2 the division of the acquisitions over the years is listed. Table 1 Division over the industries Industry Number of acquisitions Banking 22 Business Services 30 Communication 10 Construction 7

Food 28 Industrial 38 Retail 5 Software 26 Other industries 9 Table 2 Division over the years Year Number of acquisitions 2002 10 2003 7 2004 10 2005 17 2005 19 2007 21 2008 21 2009 5 2010 7 2011 4 2012 12 2013 2 2014 11 2015 7 2016 5 2017 7 2018 8 2019 1

4.2 Event Study

At first the event studies will be performed, to see the reaction of the market and to be able to calculate the Cumulative Abnormal Return of the bidding companies around the announcement date of the acquisition. Wharton Research Data Service is

used to perform the event studies and calculate the Abnormal Return by subtracting the market return, Rmarket, from the company specific return, Ri.

ARi = Ri – Rmarket The Cumulative Abnormal Return is formed by adding up the abnormal returns of all days in the event window. This is performed for different event windows around the announcement day. CARi = ∑ ARi

4.3 Event windows

The average Cumulative Abnormal Returns of the acquiring firms will be investigated for different periods in time. Firstly, the results will be found for different run-up periods, thereafter short event windows and lastly for the long-term period after the announcement. The event windows are indicated as follows. i. Forty days before the announcement date ii. Ten days before the announcement date iii. Five days before the announcement date iv. Five days before and five days after the announcement date v. Three days before and three days after the announcement date vi. Two days before and five days after the announcement date vii. One day before and one day after the announcement date viii. Forty days after the announcement date ix. Sixty days after the announcement date x. Ninety days after the announcement date For all event window results the mean of the CARs of the biddings companies will be calculated and put through the general t-test to see if the value is significantly different from zero.

4.5 Regression

Lastly, the regressions will be performed with the different variables included in this research, using the Ordinary Least Squares Model (OLS). The regression will be performed for all event windows mentioned earlier. The event windows are different in length and in timing, in this way the investigation of the possible effects of the variables

on the Cumulative Abnormal Return of acquiring companies is optimal. The effects will be tested to a significance level of 10%, 5% and 1%.

5. Results

In this section the results of the t-tests and the regressions will be described. Firstly, for the Cumulative Abnormal Returns over different time periods is taken the average and the t-value is calculated. Secondly, the coefficients of the regressions will be given, together with the standard deviations and the significance level. Lastly, the results of the regressions with all control variables included will be described and the coefficients, standard deviations and the corresponding significance level will be given.

5.1 Cumulative Abnormal Returns

Table 3 Cumulative Abnormal Return of bidding companies for the run-up period with corresponding t-test value, tested against zero

Interval (days) CAR t-value

-40 to -1 2.0472% 1.835 ** -10 to -1 0.6283% 1.094 -5 to -1 0.4785% 0.734 *** significant at 1% level ** significant at 5% level * significant at 10% level Table 4 Cumulative Abnormal Return of bidding companies for the event windows around the announcement date with corresponding t-test value, tested against zero

Interval (days) CAR t-value

-5 to 5 0.4786% 0.735 -3 to 3 1.0348% 1.911 ** -2 to 5 0.7689% 1.374 * -1 to 1 0.6882% 1.674 ** *** significant at 1% level ** significant at 5% level * significant at 10% level As can be seen in table 3, not all Cumulative Abnormal Returns of the bidding companies for the acquisitions in the Netherlands are significantly different from zero in the run-up periods. Only for the long run-up period, forty days before the announcement, the CAR is found to be significantly different from zero. When analysing the CARs of the short event windows, table 4, more significance

significant, except for the [-5, 5] window. For the [-3, 3] and [-1, 1] event windows the CAR is found to be almost 2%, implicating the return is 2% higher than normal around the acquisitions. Table 5 Cumulative Abnormal Return of bidding companies for the long-term period with corresponding t-test value, tested against zero

Interval (days) CAR t-value

-1 to 40 -0.6774% -0.596 -1 to 60 -2.5131% -1.5025 * -1 to 90 -2.6658% -1.462 * *** significant at 1% level ** significant at 5% level * significant at 10% level The Cumulative Abnormal Return for the long-term period after the announcement is found to be significantly negative. The return for the acquiring companies is lower than normal in the two and three months after the acquisitions. This cannot be seen in the short window after the announcement date, where a much lower negative abnormal return is found.

5.2 Basic Regressions

In the following regressions all independent variables are included, except for the industry fixed effects and the year fixed effects.

Table 6

Coefficients from the run-up regressions with corresponding standard deviations

-40 to -1 -10 to -1 -5 to -1

100% acquisition 6.5714 ** 3.031 0.6627 1.865 1.3213 1.064 Same industry -2.0293 2.548 2.0809 * 1.434 0.1465 0.821 Cash payment -2.6745 2.854 -1.5110 1.636 -0.1495 0.929 Shares payment -0.3926 4.274 2.3063 2.415 3.7747 *** 1.367 Ln (Deal Value) -0.5990 0.451 -0.2849 0.259 -0.0461 0.148 Constant 6.0182 6.080 1.9207 3.458 -1.004 1.984 Observations 175 172 170 R-squared 0.0467 0.0467 0.0502

Industry fixed effects No No No

Year fixed effects No No No

*** significant at 1% level ** significant at 5% level * significant at 10% level As described in table 6, many deal characteristics are not found to be of significant influence. In the period before the announcement there is no information available for the market, but in some cases, there are rumours about a possible take-over and thus about possible synergies. For the short run-up period, five-day period before the announcement date, a highly significant positive effect of a share payment on the CAR shows up. The ten-days period before the announcement gives a slightly significant positive effect of takeovers within the same industry, while this is found to be negative in the longest run-up period. In the forty days event window, the full acquisitions are of significant positive effect on the Cumulative Abnormal return of bidding company at the 5% significance level. The CAR is 6.5% higher for a full acquisition compared to a partial acquisition that is completed with 100% of the shares in possession of the acquirer. Table 7 Coefficients from the event windows around the announcement date with corresponding standard deviations

-5 to 5 -3 to 3 -2 to 5 -1 to 1

100% acquisition 5.1084 *** 1.882 2.6019 * 1.590 4.2017 *** 1.637 1.3183 1.219 Same industry 2.6064 * 1.452 1.7125 1.226 1.8810 1.263 1.5940 * 0.940 Cash payment -1.1057 1.626 -1.2677 1.373 -0.4361 1.414 -1.1841 1.053 Share payment 3.4623 2.435 3.3875 * 2.057 2.3237 2.118 0.6461 1.577 Ln (Deal Value) -0.2793 0.257 -0.1598 0.217 -0.2429 0.223 -0.2428 0.166 Constant -2.4767 3.4644 -0.2382 2.926 -1.6067 3.013 1.7123 2.243 Observations 175 175 175 175 R-squared 0.0918 0.0696 0.0691 0.0444

Industry fixed effects No No No No

Year fixed effects No No No No

*** significant at 1% level ** significant at 5% level * significant at 10% level

In Table 7 the results of the short event windows are shown, where the 100% acquisition shows a significant positive effect. For the [-5, 5] and [-2, 5] event windows, the significance levels are the highest. The three days before and after show a lower significance level. The [-1, 1] window is showing no significant effect at all. For the same industry dummy the results are also different throughout the periods. Where there is found a significant effect for the [-5, 5] and [-1, 1] window, the industry linkage does not show any effect for the three days before and after window and the two days before and five days after window. Lastly the [-3, 3] event window shows a significant positive effect of a share payment compared to other payments. The coefficients of the cash payments are negative but not found to be of significant influence on the CAR in this research. Table 8 Coefficients from the long-term performance regressions with corresponding standard deviations

-1 to 40 -1 to 60 -1 to 90

100% acquisition 4.9360 3.327 9.9416 ** 4.804 8.1327 5.398 Same industry 5.0336 ** 2.570 7.5146 ** 3.711 5.9397 4.170 Cash payment -3.2191 2.904 7.1897 * 4.193 4.2114 4.711 Shares payment 2.8674 4.321 16.8300 *** 6.239 9.5905 7.010 Ln (Deal Value) 0.2542 0.460 0.1311 0.664 0.2081 0.747 Constant -9.3114 6.143 -24.6602 *** 8.869 -20.1024** 9.965 Observations 174 174 174 R-squared 0.0581 0.0980 0.0417

Industry fixed effects No No No

Year fixed effects No No No

*** significant at 1% level ** significant at 5% level * significant at 10% level For the long-term performance regression significance is found in the constant coefficients. In Table 8 the coefficients are listed and are negative, which implies that the CAR of bidding firms is negative when all independent variables are equal to zero for the long-term period. Variables that are of significant effect on Cumulative Abnormal return are the industry dummy, which causes a positive effect on CAR for the forty and sixty days after

the announcement. For the window of sixty days after the announcement date more variables show a significant effect. The 100% acquisitions perform significantly better than the acquisitions that are completed in multiple deals for this time period. The CAR is around 10% higher in the case of a 100% acquisition in one deal. Lastly the payment method, there is a positive effect found for the cash payments and the share payments for the CAR of two months long-term period.

5.3 Regressions with control variables

For the following regression all variables are included. The industry fixed effects and the year fixed effects are used as control variables to make the regressions more robust and a better prediction. Table 9 Coefficients from the run-up regressions with corresponding standard deviations

-40 to -1 -10 to -1 -5 to -1

100% acquisition 5.8656 3.680 0.1878 2.058 1.8977 1.296 Same industry -2.4737 2.882 3.5227 *** 1.606 0.2063 0.948 Cash payment -4.2227 3.221 -3.0221 * 1.816 0.0389 1.070 Shares payment -1.9594 4.556 2.8896 2.549 3.9786 *** 1.484 Ln (Deal Value) -1.0464 * 0.545 -0.4730 0.304 0.0074 0.1783 Constant 16.9545 17.882 2.8509 9.920 -5.5132 5.777 Observations 175 172 170 R-squared 0.2124 0.2270 0.2241

Industry fixed effects Yes Yes Yes

Year fixed effects Yes Yes Yes

*** significant at 1% level ** significant at 5% level * significant at 10% level As listed in Table 9, the regressions for the run-up period don’t show many significant effects. For the [-5, -1] window, the share payment is of significant positive effect on the CAR, with a significance level of 1%. A share payment increases the Cumulative

Abnormal Return of bidding companies with almost 4% compared to other payment methods. On the contrary, for the ten-day period before the announcement date the cash payment is of significant effect on the CAR. The coefficient is negative, which causes the CAR to be 3% lower when the deal was paid with cash instead of shares, mixed methods or other payment methods. The industry linkage is showing a significant effect, which causes the CAR to go up when the two companies involved in the deal are active in the same industry. In Table 9 the only significant effect for the forty-day period is the natural logarithm of the deal value, which is a negative effect on the cumulative abnormal return. The earlier mentioned effects were also listed in Table 6, where the fixed effect variables were not included in the regressions, except for the significance of the natural logarithm of the deal value and the cash payment. Table 10 Coefficients from the event windows around the announcement date with corresponding standard deviations

-5 to 5 -3 to 3 -2 to 5 -1 to 1

100% acquisition 5.6626 *** 2.013 3.0860 ** 1.660 4.9694 *** 1.743 1.2468 1.354 Same industry 2.4114 1.577 1.3410 1.300 1.7444 1.365 1.3448 1.060 Cash payment -0.4544 1.763 -0.6234 1.457 0.1147 1.526 -0.7016 1.185 Shares payment 3.3163 2.493 3.2072 * 2.056 2.1003 2.157 0.8705 1.677 Ln (Deal Value) -0.3040 0.299 -0.1317 0.246 -0.3479 0.258 -0.2864 0.201 Constant -7.9899 9.784 -12.0737 8.070 -8.7672 8.469 2.0592 6.581 Observations 175 175 175 175 R-squared 0.3081 0.3240 0.2976 0.2145

Industry fixed effects Yes Yes Yes Yes

Year fixed effects Yes Yes Yes Yes

*** significant at 1% level ** significant at 5% level * significant at 10% level The regressions of the short event windows, listed in Table 10, show slightly different significant variables than the regressions in Table 7. For the industry linkage the significance is not found in the regressions with the fixed effect variables included,

while the earlier regression did show significant positive effects for CAR of the [-5, 5] and [-1, 1] windows. The coefficients are lower, and the standard deviation is slightly higher for the regressions in Table 10 in comparison to the regressions in Table 7. For the share payment, a significant influence is found, in event window [-3, 3] a significance level of 10% for a positive effect of shares payment on the CAR. In Table 7 the same outcome is listed, where also no significant effect for cash payments could be seen. Lastly the 100% acquisitions compared to partial acquisitions that are completed, where a high significance level is found. In Table 7 the 100% acquisition was already found to be of positive influence, but in table 10 the coefficients are higher. For the [-5. 5] and [-2, 5] event windows the 100% acquisition coefficient was found to be significant at the 1% level. Table 11 Coefficients from the long-term performance regressions with corresponding standard deviations

-1 to 40 -1 to 60 -1 to 90

100% acquisition 3.2434 3.574 7.8960 5.040 5.2819 5.923 Same industry 5.6993 ** 2.794 10.5102 *** 2.940 7.5840 * 4.631 Cash payment -2.5410 3.139 6.2429 4.426 2.0129 5.203 Shares payment 1.7336 4.419 14.5703 ** 6.231 7.9857 7.323 Ln (Deal Value) 0.3896 0.527 -0.1991 0.743 0.0161 0.874 Constant 0.5019 17.296 1.6135 24.387 -3.1119 28.664 Observations 174 174 174 R-squared 0.2840 0.3460 0.2397

Industry fixed effects Yes Yes Yes

Year fixed effects Yes Yes Yes

*** significant at 1% level ** significant at 5% level * significant at 10% level In Table 11 the results of the long performance regressions with fixed effect variables included are listed and the outcomes differ from the regressions without the fixed effect

variables. The constant coefficients are not significantly different from zero and also for the cash payments there is no significant effect founded as compared to table 8. The other payment method, shares, is found of significant influence on the CAR for the period of sixty days after the announcement, but only for the 5% significance level. For the 100% acquisitions the coefficient is also lower and less significant than in the regressions without the fixed effect variables, due to higher standard deviations. The industry linkage is showing a significant effect for all long-term periods, with high significance for the forty-and sixty-days period. When industries are performing business in the same industry the Cumulative Abnormal return tends to be 7% higher on average for the long-term periods.

6. Robustness check

In this section the robustness check will be performed for the executed regressions. The robustness test will analyse whether the effects that were found will remain the same when the dependent variable is determined in a different way. In the earlier executed regressions, the dependent variable is determined by the Cumulative Abnormal Return over different periods before and after the announcement date. To check for robustness the same periods of time will be used, but the Abnormal Return of the acquiring company is calculated in a different way. For the normal CAR the expected daily return is subtracted from the actual daily return that is realised, thereafter the daily abnormal returns are added up for all days in the event window. For the robustness check the buy-hold abnormal return will be used, where the product of the expected daily return plus one over a period of time is subtracted from the product of the actual return plus one over a period of time. 𝐵𝐻𝐴𝑅 = ∏ [1 + 𝑅i, t] −∏ [1 + 𝐸 (𝑅market, t)] For the robustness test selected event windows are chosen to test all regressions for robustness. The control variable, Ln (Deal Value), and the year and industry fixed effects also create more robustness within the regressions, since these variables filter out outside effects that also have influence on the Abnormal Returns. In the regression the following variables are included, the acquisition dummy, the industry dummy, the cash payment dummy and the shares payment dummy, as well as

the year fixed effects, industry fixed effects and the natural logarithm of the deal value. In table 12 the results of these regressions are listed.

Table 12

Regressions for the BHAR with the corresponding standard deviations

-10 to -1 -5 to 5 -1 to 1 -1 to 60

100% acquisition 0.9987 2.159 5.7669 *** 2.152 0.8526 1.457 3.3797 5.425 Same industry 3.4168 ** 1.691 2.7599 1.682 1.2069 1.141 11.1934 *** 4.249 Cash payment -3.2449 * 1.890 -0.4738 1.890 0.0624 1.276 9.1312 * 4.749 Shares payment 3.3738 2.674 3.6097 2.662 1.1640 1.804 20.1573 *** 6.718 Ln (Deal Value) -0.5336 * 0.319 -0.3436 0.319 0.2669 0.216 0.2358 0.804 Constant 1.6793 10.494 -6.3283 10.427 0.7339 7.081 7.0514 26.366 Observations 175 174 175 175 R-squared 0.2276 0.2900 0.1883 0.3973

Industry fixed effects Yes Yes Yes Yes

Year fixed effects Yes Yes Yes Yes

*** significant at 1% level ** significant at 5% level * significant at 10% level The results in table 12 show the same significance coefficients as the original regressions controlled for fixed effects. In the original regression the industry linkages and the cash payment showed a significant effect for the period ten days before the announcement. However, the significance level is lower for the same industry dummy in the BHAR regressions than in the CAR regression. For the short event windows, the 100% acquisitions have a significant positive effect on the CAR and the same results is found for the BHAR. Lastly the long performance regression for BHAR shows the same results as the CAR regression performed earlier. The shares payment and the industry linkages show a highly significant positive effect on the Buy-Hold Abnormal Return and the Cumulative Abnormal Return. For the cash payment no significant influence was found, but for the robustness check regression there is found a positive effect of a cash payment on the

It can be concluded from these results of the regressions that the relation between the different deal characteristics of an acquisition and the Cumulative Abnormal Return is found to be robust.

7. Conclusion

In this research the effect of different variables on the Cumulative Abnormal Return of the Dutch Acquisitions has been tested. The variables consisted of dummy’s which focused on different deal characteristics, cash payment, shares payment, industry linkage and the part of shares that was bought by the acquiring company. All acquisitions used in this research have been required to have a closing share possession of 100% and occurred in the period from 2002 to 2019. In the M&A field plenty of research has been done in the past, especially for different areas over the world. The focus of most researches was on the European Union as a whole. However less research was done on the Netherlands specific. The variables have been researched before but have not shown an unambiguous effect or found to be of no significant effect. The variable regarding the 100% acquisition in one transaction has not been investigated prior to this research, which caused this paper to have a broader area of research. For this research 175 acquisition transactions were included in the dataset, which is a limited amount of observations to use for a regression. The criteria in this research paper for the acquisitions were strict and caused a lower amount of useful observations. In the Netherlands many more acquisitions have taken place in the period 2002-2019, but many of these were partial acquisitions, which did not suit with the 100% acquisition variable criteria or missed crucial information as deal value or payment method. Firstly, an event study was performed to find the CARs for the different event windows. These CARs appeared to be not significantly from zero, only for the short events windows around the announcement date. Secondly, multiple regressions were performed to estimate the effects of the variables on the Cumulative Abnormal Return, excluding and including the industry fixed effects and the year fixed effects. Lastly, a robustness check was performed, to show if the results of the original regressions would remain when the dependent variable is specified in another way.

For the run-up period a significant effect was found for the industry dummy and the shares payment, the other variables did not show a substantial influence on the CAR. If the two companies involved are active in the same industry this caused the CAR to be higher compared to companies that are performing in different industries. The same conclusion can be made for the share payment, when using shares as payment method the Cumulative Abnormal Return will be higher. Furthermore, for the short event windows around the announcement date, the 100% acquisition was found to be of highly significant positive influence on the CAR of the bidding firm. Theoretically, the Cumulative Abnormal Return should occur in these event windows, for which it can be concluded that the 100% acquisitions cause a higher Cumulative Abnormal Return compared to partial acquisitions that are completed in this follow-up deal. Lastly, the long performance regressions have been performed and significant effects were found for the industry linkage and the payment with shares. Both effects tend to have a positive impact on the long-term CAR after the acquisition has been announced. For the acquisitions performed in the same industry an increase of the CAR of almost 10% is measured compared to acquisitions within different industries. For further research the completion of a partial acquisition compared to full acquisition at once could be investigated further. Is there a positive effect when an acquirer has already bought the majority of the shares in an earlier transaction in other geographical areas? This is not often investigated but can be of substantial influence as seen in this research, where a full acquisition shows a positive effect on the CAR, compared to a completion of a partial acquisition. In conclusion, for the run-up event windows and long-term event windows the Cumulative Abnormal Return for acquiring companies is significantly higher for the payment method shares and for acquisitions performed within the same industries. In addition to these effects, the full acquisition was found to be of highly significant impact on the CAR for the short event windows around the announcement date.

8. Reference list

Academic articles: Akhighe, A., Anna, A. D., & Whyte, A.M. (2007). Partial acquisitions, the acquisition probability hypothesis, and the abnormal returns to partial targets. Journal of Banking & Finance, Volume 31, Issue 10, (pp. 3080-3101) Bhabra, H.S., & Huang, J. (2013). An Empirical investigation of mergers and acquisitions by Chines listed companies, 1997-2007. Journal of Multinational Financial Management, Volume 23, Issue 3, (pp. 186-207). Bodie, Z., Kane, A., & Marcus, A. J. (2014). Investments. 10th _{Global Edition, (p. 359).} Campa, J. M., & Hernando, I. (2004). Shareholder Value Creation in European M&As. European Financial Management, (pp.47-81). Chen, S. (2008). The motives for international acquisitions: capability procurements, strategic considerations, and the role of ownership structures. Journal of International Business Studies, Volume 39, (pp. 454–471). Corhay, A., & Tourani Rad, A. (2000). International acquisitions and shareholder wealth Evidence from the Netherlands. International review of financial analysis, Volume 9, Issue 2, (pp. 163-174). Dodd, P. (2009). Merger proposals, management discretion and stockholder wealth. Journal of Financial Economics Volume 8, (pp.105-138). Dutta, S., Saadi, S., & Zhu, P.C., (2012). Does Payment Method Matter in Cross-border Acquisitions? International Review of Economics & Finance, Volume 25, (pp. 91-107). Ellwanger, M., & Boschma, R. (2015). Who Acquires Whom? The Role of Georgraphical Proximity and Industrial Relatedness in Dutch Domestic

M&As between 2002 and 2008. Journal of Economic & Social Geography, Volume 106, Issue 5 (p. 106) Huang, Y. S., & Walkling, R. A. (1987). Target abnormal returns associated with acquisition announcements: Payment, acquisition form, and mangerial resistance. Journal of Financial Economics, Volume 19, Issue 2 (pp. 329-349). Li, J., Li, P., & Wang, B. (2016) Do cross-border acquisitions create value? Evidence from overseas acquisitions by Chinese firms. International Business Review, Volume 25, Issue 2 (pp. 471-483) Mariana, S. (2015). Creating or destroying value through mergers and acquisitions? Economic Science, (pp. 593-600). Moeller, S. B., Schlingemann, F. P., & Stulz, M. (2004). Firm size and the gains from acquisitions. Journal of Financial Economics, Volume 73, Issue 2 (pp. 201-228). Putu, G.G., Dony, A. C. (2017). Abnormal Return and Characteristics of Mergers and Acquistions in Indonesia. Journal of Economics, Business & Accountancy Ventura, Volume 20, Issue 1 (pp. 31-39). Schoenberg, R. (2006). Mergers and Acquisitions: Motives, Value Creation, and Implementation. In R. Schoenberg. The Oxford Handbook of Strategy: A Strategy Overview and Competitive Strategy, (pp. 1-23). Seghal, S., Banerjee, S., & Deisting, F. (2012). The Impact of M&A Announcement and Financing Strategy on Stock Returns: Evidence from BRICKS Markets. International Journal of Economics and Finance, Volume 4, Issue 11, (p. 76). Shah, P., & Arora, P. (2014), M & A Announcements and Their Effect on Return to Shareholders: An Event Study. Accounting and Finance Research, Volume 3, Issue 2, (pp. 170-190).

Sherif, M. (2012). Gains and Payments of Mergers and Acquisitions: Further Evidence from the UK. Corporate Ownership and Control, Volume 9, Issue 3, (pp. 1-29). Wong, A. (2009). The Effects of Merger and Acquisition Announcements on the Security Prices of Bidding Firms and Targets Firms in Asia. International Journal of Economics and Finance, Volume 1, Issue 2 (pp. 274-283). Websites: Zephyr Merger & Acquisition information zephyr-bvdinfo.com Wharton Research Data Services wrds-www.wharton.upenn.edu/ M&A Statistics imaa-institute.org/mergers-and-acquisitions-statistics/