University of Groningen Master Thesis How are cross-border M&A activities with British companies affected by the threat of the Brexit?

(1)

Student Number: S3399133 RENOULT Sonia

June 2019

University of Groningen

Master Thesis

How are cross-border M&A activities with British

companies affected by the threat of the Brexit?

Author: S.B.M Renoult Supervisor: Dr. P. Rao Sahib University of Groningen

(2)

2

Abstract

What are the effects of the Brexit referendum on cross-border mergers and acquisitions? This question is addressed in this thesis using a panel data analysis of cross-border M&As between 2011 and 2018. The results show that the Brexit referendum has a negative impact on the likelihood of M&A completion. Secondly, I study the effect of the Brexit referendum on M&A activity with target companies located inside the European Union. The goal of the study is to examine how political events can influence cross-border M&A activities. Lastly, the final model focuses on the currency fluctuation effect as it became an important factor for cross-border M&As, especially between companies located in the Euro-area and British companies due to exchange rate volatility. The results of this thesis does find a currency effect when investigating cross-border M&As between Euro and British firms.

Sonia Renoult s.renoult@student.rug.nl Student number: 3399133

(3)

3

1. Introduction

Along with globalization, merger and acquisition (M&A) deal increased continuously reaching a 17-year high record with 3,774 announced deals totaling $890.7 billion in the first three months of 2018 (Amaro, 2018). This rising M&A trend is driven, for an important part, by the search for growth opportunities in countries around the world that display low economic growth rates (compared to periods prior the financial crisis in 2008) and have exceptionally low interest rates level that encourages investment (PwC, 2016). More specifically, cross-border mergers and acquisitions (CBM&As) have risen over the last decades, reaching a record of £1.02 trillion in 2007. Europe is following the same trend, as European CBM&As showed a significant increase since 2001. In 2003 for example, companies invested over $297 billion in CBM&As and 40 % of this activity was attributed to the European Union (Armitstead, 2006). The value of European deals in the first half of 2018 was up by 16% compared to the same period in 2017 (CMS, 2018). This increase in the amount of CBM&As, as well as the increased relative weight of European markets, attracts attention to the underlying determinants of this trend. The European business environment entered into a process of economic integration with the European Commission attempting to encourage standardization and transparency to support the development of a single market for M&As (Moschieri and Campa, 2008). The reason for the European Commission to promote the single market and standardization is to help reduce costs, enhance competition and facilitate the spread of innovations (European Commissions, n.d). Additionally, the adoption of a single currency eliminated a significant portion of the transaction costs (Campa and Hernando, 2008) associated with CBMAs coupled with industry dynamics such as technological innovation, fostered European companies to take part in M&As during the past decade.

(5)

5 pound hit its lowest level compared to the US dollar for 31 years. The currency fell about 13% on a day-to-day basis (Reuters, 2016). Some analysts and academics expect the M&A activity to slow down in the UK relative to the rest of Europe (PwC, 2016; Deloitte, 2016). British firms are allocating most of their resources and attention on ‘’investment in existing operations’’ and ‘’improving working capital management’’. This internal focus led to a sharp decline of UK respondents to pursue M&A deals from 65% to 45% in the following 12 months (EY, 2018a). Although, as this thesis will show, later on, the decline in M&A activity in the UK appears more like a deal pause than an active retreat.

The objective of this thesis is to build on the paper of Erel, Liao and Weisbach, (2012) by analyzing similar and additional determinants of the likelihood of M&A completion but mainly to extend the literature by studying the impact of a political event over CBM&As. Here, the political event of interest is the Brexit referendum which given the time frame, not many empirical studies examined the consequences of it over M&A transactions. For simplicity and clarity, I will refer to the result of the Brexit referendum by BR16. The purpose is to test whether this political event had indeed, an impact on CBM&A activity in the UK market but also how it might have led to major changes in the other CBM&A determinants. For this purpose, I first analyze whether the number of M&A deals decreased after the announcement of the Brexit. To do so, I perform several probit regressions while estimating additional independent variables influencing the completion of such deals. The results show that the Brexit referendum has a negative impact on the likelihood of M&A completion between the UK and firms from other geographical regions. This result holds when using alternative econometric models and when clustering the sample per industry. Furthermore, the Post-BR16 period (between 2016 and 2018) shows several changes regarding the determinants of both domestic and CBM&A deals. However, when performing a similar test on M&A deal value, the Brexit variable was found insignificant in explaining the deal size of CBM&As between the European and UK firms. Last empirical model is going to investigate another aspect of CBM&As by focusing on the currency fluctuation effect. In fact, the European minister of Hesse in Germany, Lucia Puttricht, one of her main concerns regarding the Brexit for Mittlelstand firms is the weakening of the pound (The Economist, 2019). The depreciation of the pound became a real concern for companies based in the UK as well as for European companies selling their products over there since they would become less competitive. The pound became weaker due to the traders selling British assets, but then also made takeover targets cheaper (The Economist, 2018). Then, a weaker pound against the euro might also bring growth opportunities for companies looking to acquire British firms. Currency fluctuation became an important factor due to the exchange rate volatility. As we can see in figure 1 below, the pound weakened against the Euro since the Brexit announcement in 2016. Therefore, it is interesting to observe whether companies inside the Euro-area started to increase their M&As activities with UK companies once the exchange rate of GBP/EUR decreased. This thesis contributes to two broad strands of literature. The first explains the variations of the different CBM&As determinants regarding the different periods (before and after the Brexit was announced). The second links political event to the global economy through CBM&A channel.

(6)

6 Research Question 1: What is the impact of the Brexit referendum on the likelihood of

cross-border M&As completion and its determinants?

Research Question 2: Did the Brexit referendum affect the value of M&A deals between firms

in the UK (as purchasers) and the European Union?

Research question 3: Did the bidding companies located in the Eurozone increase their M&A

activities with British companies after the Brexit announcement?

Figure 1: Daily historical foreign exchange rate of £1 in currency units for the Euro, January 2014 to December 2017. Source: Bank of England: Daily spot exchange rates against Sterling. Available at

https://www.bankofengland.co.uk/boeapps/database/Rates.asp?Travel=NIxIRx&into=GBP

The remainder of the thesis proceeds is set out in the following way. The next section reviews the previous literature relating to cross-border mergers and acquisitions, including relevant papers on factors determining the activity and the likelihood of completion of M&A deals. Third section develops the data and methodology of the study. The findings are presented and discussed in the fourth section. Last section provides a summary and the conclusions found in this thesis. 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 2-1-2014 2-1-2015 2-1-2016 2-1-2017

(7)

7

2. Literature review

This section will present an overview of the existing literature and its limits with respect to the determinants of cross-border M&As.

Despite the fact that European companies surpassed their US counterparts in terms of volume of M&As in 2007 (Moschieri and Campa, 2008), the literature focuses mainly on domestic deals between publicly traded firms in the United States and little on European acquisitions (Mateev and Andonov, 2018). This focus can be explained because of the dominance of the US deals in the worldwide M&A market by constituting 33% of the number of deals. However, in the meantime, cross-border M&As also increased and accounts now for about 30% of the total global M&A market (JPMorgan, 2019), Therefore, I will contribute to the literature by extending the research on European cross-border M&As.

2.1 Motives for cross-border M&A’s

According to Shimizu et al. (2004), CBM&As are deals containing “an acquirer firm and a target firm whose headquarters are located in different home countries”. CBM&A activity is a primary mode by which multinational companies engage in foreign direct investment (Blonigen and Pierce, 2015). Multiple studies try to explain the reasons for M&A swings. In the paper of Martynova and Renneboog (2008), they examine different M&A waves and underlying motives. They argue that takeovers occur due to external economic, technological, financial, political, and regulatory factors. Most often, the waves occur in periods of economic recovery and go along with rapid credit allocation. In the paper of Piesse, Lee, Lin, Kuo. (2013), there are various motives stated for M&As with the most common ones being the efficiency theory and agency theory. According to the efficient theory, mergers are planned in order to achieve synergies while the agency theory refers to both parties pursuing their optimal goals that affects at the end the interests of both parties. The former motif referring to synergies has been quoted several times by previous studies as the principal reason to merge or acquire another company (e.g. Trautwein, 1990; Glaister and Ahammad, 2010; Whitaker, 2012). Synergies refer to the combination of operations and activities such as marketing and R&D, which allow the firm to reduce its costs through economies of large-scale production. CBM&As can create value by allowing firms to integrate lines of production across borders and/or generate positive externalities such as transfer of technology and knowledge between the acquirer and target firms (Andrade, Mitchell, and Stafford, 2001).

(8)

8 Furthermore, the acquisition might appear as a more appropriate choice compared to Greenfield investment when the company is looking to penetrate a foreign market faster (Shimizu et al., 2004). Especially, if the target market displays a high growth rate, the choice of acquisition allows the acquirer to penetrate it more quickly. Finally, CBM&As benefit bidding companies by increasing their market power. Companies can increase their market power by either acquiring a company in the same industry or a business in a highly related industry, or either by acquiring a supplier or distributor (Hitt et al., 2001). A company obtaining more market power in its industry has the ability to then affect and/or control prices.

Prior studies devote considerable effort to understand the different M&A activity waves. I will contribute to this ongoing discussion by analyzing other important sources of disparity in M&A deals: political event, currency appreciation, and volatility. To pursue the analysis regarding the different determinants of CBM&As, a review of the main independent variables used in the models, are then going to be discussed.

2.2 Determinants of CBM&A completion and activity

A large amount of studies has depicted the different factors that influence M&A decisions but it is difficult to cover all takeovers. Therefore, I only can review some of the most common factors that have been found relevant in previous studies in influencing M&A activity.

2.2.1 Method of payment

First, it is important to know that there are several methods of payments. It refers to the resources and financing tools used when a company is acquiring the ownership and control rights of a target company (Zhang J., Zhang Y., 2011). M&A deals can be financed using three alternatives: cash payment, shares payment, and leveraged buyout. Cash payment simply relates to the purchase of a certain amount of assets or stocks from the target company using a certain amount of cash. Another acquisition currency is the stock payment where the purchasing company issues new stocks to buy the stocks or assets of the target company. Finally, a company can also purchase another company through debt. A leveraged buyout is a payment method where the purchasing company increases its debts to finance the acquisition (Zhang J., Zhang Y., 2011). The choice of the payment methods of M&A depends upon different factors such as ownership structure, financial leverage, deal size, tax considerations or growth opportunities (Uysal, 2010; Swieringa and Schauten, 2007; Ayers, Lefanowicz and Robinson, 2004).

(9)

9 1999). In terms of frequency, another research pointed out a striking trend about acquisitions method of payment. In the 1990s, about 60% of the large deals were paid entirely in cash while 10 years later, the trend reversed with 50% of the value of all large deals was paid entirely in stock, and only 17% entirely in cash (Rappaport and Sirower, 1999). Therefore, I would expect by including a variable accounting for the method of payment (CASH) to obtain significant but negative coefficient, as it appears that equities financing is predominant in M&A deals.

2.2.2 Industry regulation

Government intervention for some specific industries has also been under study regarding their consequences on acquisition deals. Government regulation is seen as bringing higher costs and risks for investors due to taxes and the transfer of assets for the regulated sectors (e.g. Doh, Teegen & Mudambi, 2004). Deregulation seems to be an important determinant of M&A activity as it removes artificial constraints on the entrance of the new firms in the market (Huyghebaert and Luypaert, 2010). As a result, whether an industry is regulated or not, the firm decisions are influenced concerning the ownership structure and their investment strategy (Doh, 2000; Doh et al., 2004). In regulated industries such as water supply, telecommunication, railways, defense, and others are perceived to be strategic in European countries (García-Canal & Guillen, 2008) as the governments want to keep control over those domains to set pricing and quality standards. Therefore, we could expect that CBM&As in non-regulated industries are more likely to be completed than CBM&As in regulated industries because of the flexibility in decision making. The European Union also aimed at liberalizing the markets and created the Single European Act to have an open market for some of the strategic industries. Public control and ownership of key industries like in railways, telecommunications, electricity, gas, and water were supposed to serve the interest of the public interest by preventing excessive competition and reduce the conflict of interest between the private and public sector in those key sectors (Moschieri and Campa, 2014). Although, despite the setting of common rules for corporate takeovers among European countries, the implementation of this new legislation is subject to nationalistic biases (European-Commission, 2007) which leads to dissimilarities in terms of regulatory approaches, ownership structures, and business practices. There are differences even inside the European Union especially between the UK and some European countries such as Germany or France. For instance, concerning the acquisition deals, public tender offers are more frequent in the UK where deals through private negotiations occur more often in the rest of Europe (Moschieri and Campa, 2008). Thus, to account for the regulatory difference between countries, I integrated an independent variable (REG) which is a dummy variable that equals to one if the firm operates in a regulated industry and zero otherwise.

2.2.3 Legal status

(10)

10 geographical dispersion of potential exchange partners, acquiring public targets becomes more attractive due to data availability. However, the volume of acquisitions involving privately held targets surpasses that of publicly traded firms (Capron and Shen, 2007). The authors found that between 60 and 75 percent of the acquired firms in the US between 2000 and 2004 were private companies. Their results also demonstrate that acquirers favor private targets that operate in similar industries and rather turns to public targets for entering new businesses or industries with a high level of intangible assets. Given the percentage of M&A deals involving private companies as targets, I would expect a similar finding in my models. Other studies also demonstrate how the legal status of the firm is an important factor influencing the completion of an M&A deal. In the paper of Moeller, Schlingemann, and Stulz, (2004) their results show that acquirers of unlisted targets earn a significant average abnormal return while acquirers of listed companies earn zero or slightly negative average abnormal returns. Therefore, it appears that the wealth of shareholders of acquiring firms is greater when it concerns an unlisted target compare to acquisitions of listed targets. There is also a listing effect in corporate acquisitions, which is not simply due to an institutional or regulatory feature (Faccio, McConnell, and Stolin, 2006).

2.2.4 Shareholder protection

(11)

11 evidence that low investor protection in the bidding firm decreases the cross-border effect in target abnormal returns1.

2.2.5 Tax divergence

Taxes affect the incentives of agents and firms in all areas of economic activity, and M&As is no different. Devos et al (2009) found that many transactions are driven by taxation purposes. I would expect to observe a tax effect on CBM&As as those deals have higher transaction costs compared to domestic deals due to the international setting, different taxation and legal fees across countries (Xie, Reddy, Liang, 2017; Bris et al, 2008). It is actually one factor that can explain the relative weight of domestic deals compared to cross border deals in the total number of M&As completed. A taxation effect has been established from previous studies (Devreux and Griffith 1998; Di Giovanni, 2005). In the former paper, the authors analyze the determinants of a US company to choose to move in France, Germany and the UK and found that the first investor’s choice is influenced by the effective average tax rate. Therefore, all else being equal, one would expect that a country displaying a lower tax rate would attract more foreign investment. Other papers (Erel, Liao and Weisbach, 2012) also found that tax rates affect cross-border decisions as acquirers are more likely to be from countries with higher corporate income tax rate than the target firm. Thus, I am including in the deal size model, the domicile corporate tax rate of the target country (TAX)i, as it is a common proxy for tax effects used by the prior studies. Although the variable TAX is only going to be included in the second model (the sample including solely European countries) as in the first sample for the ‘’likelihood of M&A completion’’ model, many data concerning domicile tax rate were missing for several countries. Finally, a country displaying a low statutory corporate tax rate or labor tax is an attractive target for global investors to make mergers or acquisitions with (Ciobanu and Dobre, 2015). Therefore, because of the possibility that international tax difference motivates cross-border M&A, it appears to be a relevant factor to include in this thesis.

2.2.6 Political uncertainty

The uncertainty about the host government policies is a major concern for companies considering CBM&As. At the macro level, political shocks can lead to a diminution of investments (Pindyck and Solimano, 1993; and Alesina and Perotti, 1996). At the micro level, firms could reduce their capital expenditures in the presence of political uncertainty (Julio and Yook, 2012). Previous studies found that cross-border M&A activity is negatively correlated with political uncertainty (Bonaime, Gulen and Ion, 2016; Lee, 2018). Therefore, we could expect that an event such as the BR16 had an important impact over M&As in the UK and for other geographical regions. The intensity of these implications will obviously depend on the final outcome, whether it is a hard or soft Brexit. For now, only assumptions can be made regarding the real effects of the Brexit. There is still a lot of uncertainty about the legal implication and the future regarding the relationship with the EU. Various financial advisers (PwC 2016, Deloitte, 2016) expect the M&A activity to diminish in the UK as well as in other EU regions. Potential buyers postpone deals, as they are uncertain regarding the finalization of UK-EU cross-border deals that are currently not completed and the impact that might have on

(12)

12 the exchange rates. For this reason, I found this topic interesting to study since it is a current economic-political issue that affects several countries (in particular for European countries). Especially as we know little about how political uncertainty influences cross-border deals due to few studies investigating this issue. The study of Cao, Li, and Liu (2017) are one of the few papers that I could find regarding the evaluation of a specific political event effect on CMB&A deals. To test the impact of a political event over cross-border acquisitions the authors use as a dependent variable ’Pre-election’, which is a dummy variable equals to one if the observation year is the year just before the target nation election year, and zero otherwise. Their results show that firms are more likely to purchase a foreign target in the year before the domestic national election for international diversification due to political uncertainty. I use a similar approach in this thesis where the variable PostBR16 is a dummy variable equals to one if the year announced M&A deal is between 2016 and 2018, and 0 otherwise. However, I did not find any studies focusing on a specific political shock that directly affects the investment decision-making of firms affecting several countries at a time. Therefore, this thesis will contribute to the existing literature by studying an important political event such as the Brexit and investigate its impacts on domestic and CBM&A activities.

2.2.7 Currency fluctuation

(13)

13 the currency also appreciated relative to the pound2_{. However, the logic by which valuation}

differences can lead to an increase of cross-border mergers and acquisitions depends on whether participants believe these movements are going to be temporary or permanent (Erel, Liao and Weisbach, 2012).

To recap the main independent variables that are going to be tested in section 4, Table 1 provides an overview of the literature review for each showing its expected coefficient signs.

Table 1: Hypotheses Cross-border M&As

Independent variable Coefficient

PostBR16 Negative

Euro value Positive

Currency volatility Negative

Method of payment Negative

REG Negative

Status Negative

Protection Positive

TAX Negative

(14)

14

3. Methodology and Data

This section outlines the data and methodology used for the dependent and independent variables. The largest data sample on cross-border M&As of this thesis consists of 52 countries (including the United Kingdom). The second model concerns cross-border M&As with the European Union and the last model refers to only firms based in the UK and inside the Euro-area.

3.1 Data

The data sources in this thesis come mainly from the database Zephyr that is commonly used for corporate finance research. This database contains information on M&A, IPO, private equity and venture capital deals. Because I wish to extend the paper of Erel, Liao and Weisbach, (2012) and study transactions that are clearly motivated by changes in control, I solely focus on mergers and acquisitions. Thus, samples include only M&A announced deals between 2011 and 2018. Rumored type of deals has been removed to avoid any confusion regarding the interpretation of the results. After excluding IPOs, joint-venture, MBOs, and other deals I end up with a sample of 20,031 observations when analyzing M&A transactions with a target located in different geographical regions (including the UK) and the UK as the bidding side. Furthermore, M&A deals that were missing relevant values have been removed from the sample, which ended up at 2,504 observations.

3.1.1 Likelihood of M&A completion model

The sample in this section regroups data of domestic (inside the UK) and CBM&As when British firms are the purchasing companies. The targeting firms are located in several regions around the world3. The dependent variable for this analysis is the likelihood of M&A completion. Therefore, the first model is going to analyze the various explanatory variables on the probability of completion of the announced deals. The dependent variable (Completed) is a dummy variable equals to one whether the deal announced during the specific time period was completed and zero otherwise4.

The first explanatory variable refers to the Brexit referendum– PostBR16 that is a dummy variable that equals to one if the M&A deal occurred between 2016 and 2018 and zero otherwise. Another indicator variable refers to the method of payment (CASH) where the variable equals one when the M&A deal is paid solely in cash and zero otherwise. Then, the legal status of the target firm (Status) refers to whether the company is listed (1) or unlisted (0) company. Finally, I estimate the shareholder protection (Protection)i-j variable that is measured by using an index from The World Bank referred to as the Strength of Investor Protection Index. It measures the degree to which investors are protected through disclosure of ownership and financial information. The index ranges from 0 (little to no investor protection) to 10 (greater investor protection).

Macroeconomic indicators are also important factors into determining the size and the volume of M&A activity (Rossi and Volpin, 2004; Erel, Liao and Weisbach, 2012). Existing literature

3_{The target companies come from several geographical regions: Africa, Middle East, North America, Oceania,} South and Central America, Far East and Central Asia, Scandinavia, and the European Union.

(15)

15 agree on a positive relationship between GDP per capita and CBM&As. As the authors Neto & Bandao (2009) explain it, high economic growth stimulates firms to invest abroad to avoid the domestic market saturation. I selected the constant 2010 GDP per capita in Euro to reflect the true growth by removing the inflation effect. Therefore, I included several economic indicators similar to the ones used in previous studies. To account for the change in economic conditions, I included the real GDP growth (annual percentage in EUR) and the annual constant 2010 GDP per capita (in EUR) to proxy for the economic size of the countries5. Those data have been collected from the OECD website. Furthermore, other control variables concerning firm characteristics are also influencing M&A deals (Erel, Liao and Weisbach, 2012). Therefore, I included similar control variables for every estimation like the total assets (FirmSize) and the return on assets (ROA) which refers respectively to the firm size and its profitability. The dataset of both variables come from the data source Orbis that contains various information on around 300 million businesses.

3.1.2 Deal size model

Regarding the second model, I test different determinants influencing the M&A activity between the British and European firms. In the paper of Rossi, S., and Volpin, P. (2004), as a proxy for the deal size as dependent variable, the authors estimated different alternatives. They used the value of all completed deals divided by GDP and the value of completed deals among traded companies divided by stock market capitalization. However, as the sample in this thesis does not only regroup listed companies, I could not take the last option using the stock market capitalization and clustering my sample by country will reduce drastically the sample size and display less information. Thus, as a proxy for the deal size, I simply used the deal value of the individual M&A deals across countries. The reason for taking the deal size instead of the number of CBM&As is that deal value contains more information because of the magnitude of the deals. The sample includes M&A deals that occurred between the companies based in the UK and companies that are inside the European Union. In this regression, we will be adding another explanatory variable such as the target industry regulation (REG) where the regulated industries are Banks; Insurance companies; Post and telecommunications; gas, water, electricity; health care; airline; and oil. The content of the list is mainly constituted from the one made by the paper of Moschieri and Campa (2014) investigating the European market. Thus, the dummy variable sets up to one whether the deal involves a target company among those regulated industries and zero otherwise. Another variable is added to the model related to the tax system of the country: income tax rate differences (TAX)i-j. This variable refers to the difference between acquirer (i) and target (j) countries of domicile in corporate tax rates obtained from the OECD. In addition, control variables are included in the model to control for macroeconomic factors and firm characteristics. As the deal size depends on the size of the company, I included the variable FirmSize that refers to the total assets of the company. Additionally, the profitability of the firm might be correlated with the previous variables, thus the return on asset (ROA) of the company is also included as a proxy. To control for the changes

(16)

16 in macroeconomic conditions, I include the GDP growth and GDP per capita variables that have been used in the previous model.

3.1.3 Likelihood of M&A completion model with currency effect

In this last regression, the focus is going to be on the currency fluctuation effect. The aim is to provide new information and evidence on the effect of the exchange rate volatility on CMB&As in the Eurozone and observe whether the Pound-to-Euro exchange rate movement constitutes an incentive for Euro firms to increase their M&A activity with the UK. Thus, in this model, the bidding companies are based inside the Euro-area and the target are located in the UK. The main explanatory variables are the value of the euro against the pound sterling (Eurovalue) and the annual average of currency volatility (Currency Volatility). More precisely, the former variable is calculated using the annual percentage change of EUR/GBP exchange rate. When the annual percentage change is positive then the dummy variable Eurovalue equals to one and zero if negative. When the dummy variable equals to one it means that the Euro currency appreciated relative to the pound. Regarding the latter variable, the Currency Volatility is calculated using a similar approach than Erel, Liao and Weisbach, (2012) by taking the standard deviation of the monthly bilateral real exchange rate between 2011 and 2018. To calculate the real exchange rate volatility, I obtain the monthly EUR/GBP exchange rate historical data from Datastream and calculate the standard deviation for each year to obtain the annual data. In addition, the control variables used previously in the models to capture country and firm characteristics are also included in this regression.

3.1.4 Descriptive statistics

This subsection gives more insight into the data to better understand the trend in the M&A activity focusing on the UK. First, it is important to notice that there is no multicollinearity issue in the models. Table 11 in the appendix, provides the correlation coefficients for all the variables involved in the first regression model. There is only one correlation coefficient higher than 0.5 between the logarithm of the size between the acquirer and the target company which is the control variables. In table 12, displaying the correlation coefficients for the deal size model, there are two variables showing a correlation higher than 0.5. If there is a presence of multicollinearity in the regression model, the coefficients become unstable and the standard errors of the coefficients could be inflated. However, when using the VIF (Variance inflation factor) command in Stata to detect multicollinearity, none of the VIF values are greater than 10. Finally, in table 13, the correlation coefficients of the variables included in the last model do not show any large correlations between the variables.

(17)

17 than the targets which make sense since they have more financial resources available to invest. Another interesting finding from the descriptive statistics is about the method of payment. Cash only deals represent solely about 35 percent while in the model including the currency effect it is about 80 percent. Thus, it appears that deals between the UK and Euro firms are more frequently completed using cash. First, this can be explained by the content of the sample. The data in table 8, refers to the whole sample including the UK as both purchasing and targeted countries completed by other country targets located around the world. While table 10, the targeted firms are solely based in the Eurozone. Since CBM&As were found to be often achieved in cash (Moschieri & Campa, 2008; André and Ben-Amar, 2009) it could explain why the number of deals in the last model display a higher number of M&A completed in cash as the whole sample also included UK firms as targets. Another possible explanation relates to the legal form of the company. The variable Status in the last model was omitted because only unlisted targets were present in the sample. This finding is consistent with the paper ofAndré and Ben-Amar (2009) since the authors found that bidders acquiring unlisted targets were more likely to be deals completed in cash. The, it could also contribute to the difference in the method of payment’s use between the models.

3.2 Methodology

This subsection describes the methodologies used in this thesis. Firstly, I am going to determine whether BR16 affected the likelihood of M&A completion deals. Secondly, a similar regression is run but testing the effect of BR16 on M&A deal value. Finally, the last empirical model is going to focus on the currency value and volatility effect on the likelihood of completion of M&As.

In order to test whether the Brexit referendum had an influence on the probability of success of the deal, I created a dummy variable PostBR16 that equals one if the deal occurred between 2016 and 2018 and zero otherwise. As the dependent variable for this model is binary, I estimate the equations using a probit model. Since it is not obvious how to decide which model to use between logit and probit regression in practice (Maddala, Lahiri and Maddala, 2010), I present also the results using a logit model in the appendix (Table 16). In addition, a target country fixed-effect 𝑢𝑖 is included in the model to capture the effect of potential omitted variables that affect the other variables in the model. For instance, there are systematic differences across countries like in the financial environment (i.e. Bankruptcy laws) that might correlate with the variables.

𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑑i, t = 𝛼 + 𝛽1(𝑃𝑜𝑠𝑡𝐵𝑅16)i, t + 𝛽2(𝐶𝐴𝑆𝐻)i, t + 𝛽3(𝑆𝑡𝑎𝑡𝑢𝑠)𝑖, t + 𝛽4(𝑃𝑟𝑜𝑡𝑒𝑐𝑡𝑖𝑜𝑛)𝑗 − 𝑖 + 𝑢𝑖 + 𝜀it

(18)

18 domestic M&As gathering a total of 773 deals and cross-border M&As with 165 observations. The goal is to compare those two periods and detect any major changes regarding the M&A activity determinants.

Second part of the thesis is to test the relevant determinants on the volume of M&A activity. The deal size model is an ordinary least square (OLS) regression, which is a method used to estimate the unknown parameters in a linear regression model. In this section, the sample only includes cross-border M&As with companies located inside the European Union as targets and British companies remain the acquirers. The objective is to determine whether BR16 has an impact on the cross-border M&A deal values and if so, analyze how the factors influencing the deal size change after the vote.

To do so, I will take additional variables similar to the ones used in Erel, Liao and Weisbach (2012) study such as the (TAX)i variable which refers to the corporate income tax rates of the targeted country. To include only important deals, the deal values are filtered by only including transactions greater than 10 000 euros and removing the deals without any value, which reduced the sample to 106 observations6.

𝑙𝑜𝑔(𝑆𝐼𝑍𝐸)i, t = 𝛼 + 1(𝑃𝑜𝑠𝑡𝐵𝑅16)i, t + 𝛽2(𝑅𝐸𝐺)𝑖, t + 𝛽3(𝐶𝐴𝑆𝐻)i, t + 𝛽4(𝑆𝑡𝑎𝑡𝑢𝑠)𝑖, t + 𝛽5(𝑃𝑟𝑜𝑡𝑒𝑐𝑡𝑖𝑜𝑛)𝑗 − 𝑖 + 𝛽6(𝑇𝐴𝑋)𝑖 + 𝑢𝑖 + 𝜀it

Finally, in the last section, the effect of currency fluctuation is going to be the focus of attention. The independent and control variables are the same than the ones used and described in the previous sections. Although, two additional variables are added, Eurovalue and Currency Volatility that are the main independent variables for this model. The purpose is to evaluate whether the currency appreciation of the euro against the pound (GBP) increases the likelihood of M&A completion when the acquiring firms are located in the Eurozone and the targets in the UK. In all models, I controlled for outliers and removed it from the concerned variables to avoid any biases in the results.

𝐶𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑑i, t = 𝛼 + 𝛽1(𝐸𝑢𝑟𝑜𝑣𝑎𝑙𝑢𝑒)i, t + 𝛽2(𝐶𝑢𝑟𝑟𝑒𝑛𝑐𝑦 𝑉𝑜𝑙𝑎𝑡𝑖𝑙𝑖𝑡𝑦)i, t + 𝛽3(𝐶𝐴𝑆𝐻)𝑖, t + 𝛽4(𝑆𝑡𝑎𝑡𝑢𝑠)𝑖, t + 𝛽5(𝑃𝑟𝑜𝑡𝑒𝑐𝑡𝑖𝑜𝑛)𝑗 − 𝑖 + 𝛽6(𝑅𝐸𝐺) + 𝑢𝑗 + 𝜀it

(19)

19

4. Results

In this section, the main results of the regressions will be given and analyzed. At first, the trend in the UK M&A market is going to be examined in terms of value and total number of deals. Then, first regression model will follow including the indicator variable ‘’Completed’’ as dependent variable (Table 3). In table 4 the results from the deal size model are presented. Finally, the last estimation results in table 5 refers to the likelihood of M&A completion with the currency effect.

4.1 Brexit announcement effect on Cross-border M&A activities

When looking at table 2, we can observe some interesting patterns regarding M&A activity in the UK. The total number of M&A deals decreased from 2016 to 2017. However, the decline did not come from less cross-border M&A activities but rather within the UK market. Throughout the years, domestic M&As represented the highest percentage share of the total number of M&A deals in the UK. Most M&A deals in Europe remain domestic, with 81% of European deals being completed between domestic companies in 2008 with the UK being the country with the largest proportion of domestic deals (89%) (Moschieri and Campa, 2008). As we can see in the table below, most of the cross-border deals involve companies in neighboring countries. UK domestic deals started to decrease slightly since 2012 to shift toward cross-border M&As especially with European countries. The M&A deals in the UK involve more and more European countries, in particular with The Netherlands where the number of M&As increased by 54% and Germany with 28% from 2012 to 2018. We can also observe this trend in figure 2 below, where the fraction of CBM&A deals in the UK with European companies increased throughout the years to reach about 20% of the total number of M&A. While in other geographical regions such as with the US, the number of deals fell in 2018. This is in line with the findings of Moschieri & Campa (2014) where the M&A activity between 1995 and 2007 show a decrease in domestic transactions and an increase in intra-European transactions.

Figure 2: This figure plots the deal value of both domestic and CBM&As between 2011 and 2018 between the UK (as the purchase country) and countries located inside the European Union. Bars represent values in a given year

0 5 10 15 20 25 30 35 2011 2012 2013 2014 2015 2016 2017 2018 0% 5% 10% 15% 20% 25% To tal v al u e o f M &A d e al s (i n M ill io n s o f E u ro ) % o f C BM &A d e al s

Figure 2: Total value of M&A deals

(20)

20 while the solid line represents the fraction of cross-border mergers and acquisitions relative to the total deal value of all M&As in a given year in the UK as the purchasing country. Source: Zephyr

Although, the interesting pattern is between 2016 and 2017. The total number of M&As shrank during this period (for both domestic and cross-border M&As), which could be attributed to the BR16 effect. International firms might have decided to suspend their acquisitions in the UK until there is a clearer picture of how the UK is going to leave the EU (with or without a deal). For instance, before the referendum to leave the EU occurred, Japanese investment into the UK and the rest of the EU were growing at a similar pace between 2015 and 2016. However, from mid-2016 Japanese investment was still coming into the UK but at a much slower rate (half that pace) than in the rest of the EU (Maidment, 2019). Figure 2 also displays an increase in the volume of M&A activities from 2011 until 2015 but a deep fall in 2016. This decline is mainly coming from a drop in domestic M&As as we can see in table 15 in the appendix, where the number of domestic M&As in the UK fell by 19% in 2017 compared to the previous year. In this same table, we can see that it does not appear to be a global trend as the UK is the country showing the biggest drop during this period, thus this decline could be explained by the BR16 effect.

(21)

21 Table 2: Number of M&As by Country Pair between 2012 and 2018

The table represents the number of announced M&A deals for British companies. The rows represent the country of the target companies7_.

Acquirer country: UK Target 2012 2013 2014 2015 2016 2017 2018 AE 2 6 5 6 7 6 10 AM 0 0 0 2 1 0 0 AR 3 5 2 1 0 2 2 AT 6 2 1 4 5 3 1 AU 28 28 34 41 20 40 29 BB 1 1 1 1 1 1 0 BE 12 8 6 14 6 9 11 BG 2 22 50 5 3 2 0 BH 0 0 0 0 0 1 1 BM 2 2 1 1 2 1 1 BR 11 13 13 13 7 7 6 BS 0 0 1 1 0 0 0 BW 0 0 1 0 2 0 1 BZ 1 0 0 1 1 0 0 CA 14 20 21 18 20 12 17 CD 1 1 0 0 0 1 0 CL 4 4 3 2 2 2 2 CN 13 13 6 4 8 2 4 CO 0 3 3 4 0 1 2 CY 1 1 2 1 1 3 2 CZ 0 3 2 6 28 20 7 DE 39 36 44 37 35 39 50 DK 7 6 9 6 10 8 6 EC 0 0 1 0 1 0 0 EE 1 2 0 2 1 0 0 EG 0 1 1 2 1 0 1 ES 16 17 21 24 19 27 21 ET 1 0 0 1 0 0 0 FI 10 7 6 7 4 5 4 FR 26 20 22 26 16 27 20 GB 1949 2068 2004 2047 1863 1482 1665 GE 0 0 0 2 2 2 1 GH 0 2 0 2 0 0 0 GI 0 0 0 1 0 0 0 GN 0 0 0 1 0 0 0 GR 0 1 3 4 2 0 0 GT 0 0 0 1 0 0 0 HK 4 5 8 2 5 1 7 HR 0 0 1 2 0 2 0 HU 4 1 2 6 1 3 1 ID 1 1 3 2 1 4 1 IE 18 21 22 25 24 25 29

(22)

(23)

23 ZA 9 24 18 14 12 9 5 ZM 0 1 0 0 1 1 1 ZW 1 1 1 0 0 0 0 Total 2450 2602 2657 2642 2386 2039 2185 Domestic M&As 79.55% 79.48% 75.42% 77.48% 78.08% 72.68% 76.20%

4.2 Likelihood of M&A completion model

The starting point of the empirical part of this thesis is to determine whether the BR16 affected the likelihood of completion of M&A activities. The main variables of interest are PostBR16, REG, CASH, Status, Protection. The other variables are the control variables.

Table 3 depicts the results of the probit models. First models (1) and (2) have been estimated in order to determine whether the BR16 changed the likelihood of M&A completion for both domestic and cross-border deals. Since the dummy variable PostBR16 is highly significant, it confirms my first assumption that the Brexit announcement did have an impact on the completion of M&A transactions. In addition, the variable displays a negative coefficient which means that the Brexit decreases the likelihood of M&A completion. Thus, this finding is in line with the previous literature (Bonaime, Gulen and Ion, 2016; PwC, 2016) arguing that political uncertainty is strongly negatively associated with M&A activity.

Then, when looking at the variable Status it is consistent with the findings of Capron and Shen (2007) as it is negatively significant for most of the estimated regressions in table 3. In other words, the completed deals concerns more often unlisted target companies than public ones. Furthermore, the control variables Log(FirmSize)i and Log(FirmSize)j are also significant for most of the models with the former one showing a negative coefficient while the latter variable displays a positive coefficient. Those results were expected as bigger firms, in terms of total assets, are often the acquiring companies and the smallest firms the targets (Erel, Liao and Weisbach, 2012). Finally, larger differences in GDP growth increases the likelihood of M&A completion between the two countries. In other words, it appears that firms based in countries displaying higher growth rate are more likely to be the purchasing company than the firms located in a country with lower growth.

After testing whether BR16 has an impact of the M&A completion, I examine deeper the different determinants regarding the time period and run additional models by splitting the sample into two different periods: Pre-BR16 period (2011-2015) and Post-BR16 period (2016-2018) to examine which determinants of M&A changed from one period to another. First, when looking at the domestic M&As, we can observe that there are no major changes in the explanatory variables before and after the Brexit announcement. However, there are some major changes regarding the cross-border M&As for several variables.

(24)

24 a country with low investor protection, target shareholders are likely to prefer cash over equity financing due to being minority shareholders that might raise the risk of expropriation. But in the table below, the variable Protection is not significant and even display a negative coefficient which does not support their finding. However, what is surprising is the fact that the variable is only significant regarding the Post-BR16 period. As mentioned in the literature review, several factors are influencing the decision of payment method in M&As such as the deal size, tax considerations or growth opportunities that are influencing the decision for cash or non-cash deal. Here, as the variable only shows significance in the Post-BR16 period we could assume that it is due to the risk uncertainty of the Brexit that drives target international firms to negotiate for cash deals only. Since in stock transactions, the acquired company will have to share the synergy risk in proportion to its percentage of the combined company owned. Additionally, a cash deal is more straightforward and is paid directly in cash with the current exchange rate. Thus, as businesses are uncertain about the Brexit and its future consequences, it might be preferable to trade in cash rather than equities due to volatility risk.

(25)

25 Table 3: Determinants of the CBM&As

The table presents the results of the probit regression models using different samples. The dependent variable is a dummy variable set equal to one when the anounced M&A deal has been completed and 0 otherwise. The independent variables are: PostBR16, a dummy variable that equals 1 if the M&A deal occurs in the period between 2016 and 2018; CASH, a dummy variable that equals one if the method of payment is 100% cash and 0 otherwise; Status(i), a dummy variable that equals one if the legal status of the target company is a listed company and 0 otherwise; Log(FirmSize)j is the logarithm of the total assets of the acquiring firm; Log(FirmSize)i is the logarithm of total assets of the target firm; ROA(j), the return on assets of the acquiring firm; Protection(j-i), the score difference between the acquirer (j) and target (i) firm for an index measuring the strenght of shareholder protection of the country. The

GDP growth, ROA, and logFRIMSIZE are included in all regressions as control variables.8_Robust standard errors are shown in parentheses.

Full sample Domestic M&As Cross-border M&As

2011-2018 Year<2016 Year>=2016 Year<2016 Year>=2016

Completed (1) (2) (3) (4) (5) (6) PostBR16 -0.77*** (0.17) -0.84*** (0.18) CASH 0.06 (0.18) -0.01 (0.19) -0.13 (0.35) 0.01 (0.26) 0.41 (0.64) 5.30*** (0.70) Status(i) -0.58** (0.29) -0.65** (0.30) -0.74* (0.42) -0.48 (0.46) -0.84 (0.68) -6.20*** (1.07) Protection(j-i) -0.05 (0.09) -0.02 (0.49) 0.12 (0.16) -0.27 (0.35) Log(FirmSize)j 0.09** (0.04) 0.16*** (0.05) 0.17* (0.09) 0.20*** (0.07) 0.07 (0.10) -0.25** (0.12) Log(FirmSize)i -0.15*** (0.05) -0.18*** (0.05) -0.23** (0.10) -0.18** (0.07) -0.21 (0.15) -0.07 (0.14) ROA(j) -0.004 (0.005) -0.004 (0.01) -0.01 (0.01) -0.01 (0.01) -0.00 (0.01) -0.09 (0.10) GDP growth(j-i) 0.05 (0.07) 0.31** (0.14) -0.10 (0.08) -0.14 (0.11) Constant 2.55*** (0.38) 1.26* (0.69) 2.60*** (0.33) 0.78 (0.64) 3.11*** (1.00) 4.60*** (1.74) Pseudo R2 0.17 0.24 0.17 0.10 0.17 0.32 N observations 1002 912 575 247 129 45 Target-country FE NO YES NO NO NO NO

***, **, * indicate significance at 1%, 5% and 10% levels, respectively

(26)

26

4.3 Deal size model

In this subsection, I present the results from the second model regarding the different factors of M&A activity. The sample is constituted of companies based in the UK and firms located inside the European Union. As previously, the British firms are the purchasing firms but also the targets as well as the European firms. Thus, this model includes both domestic and cross-border M&A deals. The dependent variable refers to the deal value of the individual M&A deal (Dealsize). However, the main explanatory variable PostBR16 is insignificant in both regressions so it does not appear that the Brexit referendum led to a decline in terms of deals size. Thus, no further regressions have been pursued regarding the time period. This result follows the continuous rise in the value of M&As in the European regions by 96 percent reaching $767 billion year-on-year during the first six months of the year 2018 according to Reuters (2018). This article highlights the fact that even if the number of M&A transactions fell by 18 percent in Europe, the total deal value rose which indicates that companies engage on fewer M&A deals but bigger takeover attempts. The global M&A market was largely driven in 2018 by ‘’megadeals’’ meaning deal values greater than $10 billion (JPMorgan, 2019).

Although, it is interesting to observe which determinants are explaining the value of M&A deals. First, when looking at the CASH variable, the coefficient is negative in contrast to the previous model where it was positive. Thus, here cash deals relate to a lower M&A deal value. As we can see in table 9 presenting the summary statistics of the model, only about 35% of the M&A deals are using cash only as a method of payment. This finding is in line with the paper of Moschieri & Campa (2014) where the authors also found cash deals to be negatively significant which they partly explained by the fact that ‘’30% of all announced foreign deals are cash-only’’ when considering European cross-border. Another explanation may be that equity prices kept increasing (see figure 3) in the UK, which increases the viability of equity financing (di Giovanni, 2003). Second, concerning the importance of firms size Log(FirmSize), the results are in line with the reasoning that bigger are the firms (in terms of total assets), either the bidder or target firm, the greater is the M&A deal size. Third, when a country fixed effect is included in the model, the variable TAX(i) becomes negatively significant. This finding is consistent with previous studies (di Giovanni, 2005; Erel, Liao and Weisbach, 2012) stating that a country with a low corporate taxes level will be an attractive target country to purchase assets in. Finally, the variable REG has been found negatively significant which confirms my previous hypothesis regarding the regulated industries and is in line with the paper from Moschieri and Campa (2008). Firms operating in non-regulated industries are more involved in M&A deals than firms in regulated industries are. This can be explained due to the government regulation that represents higher costs and risks for investors due to taxations and to the transfer of assets for the regulated sectors (e.g. Doh, Teegen & Mudambi, 2004).

(27)

27 Blyth, a partner at PwC underlined the fact that if the UK leaves the EU without a withdrawal agreement and transition period in place, cross-border mergers will no longer be possible for UK companies after Brexit (Blyth, 2019). Therefore, I would expect the BR16 to have a stronger effect on CBM&As and lead UK companies to increase their M&A deals with European companies in case of a ‘no-deal Brexit’. Nevertheless, as we can see in table 18, the PostBrexit16 variable remains insignificant. Thus, this result does not support the hypothesis.

Table 4: Determinants of M&A activity

Determinants M&A value

(1) (2) PostBR16 0.05 (0.09) -0.07 (0.10) REG(i) -0.32** (0.16) -0.39** (0.16) CASH -0.29*** (0.08) -0.28*** (0.08) Status(i) -0.60** (0.27) -0.5 (0.23) Protection(j-i) 0.06 (0.10) 0.09 (0.11) TAX(i) -0.02 (0.01) -0.06*** (0.02) Log(FirmSize)j 0.46*** (0.02) 0.51*** (0.03) Log(FirmSize)i 0.35*** (0.02) 0.30*** (0.02)

The table presents the results of the regression models using different samples. The dependent variable is the deal value of each M&A deal that occurred between 2011 and 2018. Independent variables are: PostBR16, a dummy variable that equals 1 if the M&A deal occurs in the period between 2016 and 2018; REG, a dummy variable that equals one if the target company operates in a regulated industry; CASH, a dummy variable that equals one if the method of payment is 100% cash and 0 otherwise;

STATUS, a dummy variable that equals one if the legal status of the target company is

a listed company and 0 otherwise; Protection, the score difference between the acquirer (j) and target (i) firm for an index measuring the strenght of shareholder;

TAX(i), the target (i) countries of domicile in corporate income tax rates. Several

control variables are included in the regressions such as Log(FirmSize)j and

Log(FirmSize)i that respecctively refers to the total assets of the acquirer and target

(28)

28 ROA(j) -0.002 (0.003) -0.001 (0.003) Log(GDPcapita)j-i -0.27 (0.42) 3.8** (1.78) GDP growth(j-i) 0.001*** (0.03) -0.06 (0.04) Constant 1.67*** (0.43) 3.61*** (0.80) R-squared 0.63 0.65 N observations 992 992 Target-country FE NO YES

Time fixed-effect NO YES

***, **, * indicate significance at 1%, 5% and 10% levels, respectively

4.4 Likelihood of M&As completion model with currency effect

This section analyses the regression results of the likelihood of M&A completion as the dependent variable. The independent variables in this model are similar than the ones used in the previous sections but this time I am focusing on two additional variables, which are the Eurovalue and Currency volatility. In this model, the sample only includes firms located in the Eurozone and in the UK as I want to analyze the currency effect on the decision-making of firms based in the Euro area embarking on CBM&As with UK companies as targets.

First, the results show a currency effect. By including a country fixed effect for the acquirers, the Eurovalue variable becomes significant. When the Euro appreciate vis-à-vis pound sterling, the likelihood of M&A completion increases. In other words, firms from the Eurozone when the Euro currency appreciates over the pound are more likely to be the acquiring firms while UK firms the targets. Thus, the exchange rate of EUR/GBP movements have an impact on M&A deals completion. This result is consistent with Erel, Liao and Weisbach (2012) as they also found a currency effect where firms from countries whose currencies appreciated were the purchasers of firms whose currencies depreciated. Therefore, it is in line with the previous hypothesis that international investors would have an incentive to acquire companies in the UK as the pound is weakening which makes British firms attractive targets. A continued decline in the pound sterling provides opportunities for foreign acquirers (JPMorgan, 2019). Furthermore, the currency volatility shows a large positive value, which could also support this finding if it was significant.

(29)

29 the authors run different models and when testing with cross-border deals as a dependent variable, the shareholder protection becomes negatively significant. ‘’The probability that a completed deal is cross-border rather than domestic is higher in countries with lower investor protection’’ (Rossi and Volpin, 2004). Thus, as the sample is constituted solely of CBM&As that might be the reason why the coefficient of Protection variable is negative.

I performed as the robustness of check for this model, a similar regression model but taking Japanese firms as the bidding companies instead of Eurozone firms9_{. Since it has also been}

found that Japanese investments in the UK decreased from mid-2016 onwards due to Brexit uncertainty (Maidment, 2019). In this case, I measure the annual change percentage of the exchange rate between the pound and the Japanese yen (GBP/JPY). The Japanvalue, that is a dummy variable equals to 1 if Japan yen appreciated and 0 otherwise, is slightly positively significant which confirms my previous finding of the fact that international firms are more likely to acquire a British company if its currency depreciate. Furthermore, when including the exchange rate variable, we can observe that the variable becomes highly significant and since the coefficient is negative it is in line with the finding as the pound weakens compare to the Japanese yen, it increases the likelihood of completion.

Table 5: Currency fluctuation effect

(1) (2) Eurovalue 0.26 (0.19) 0.70** (0.30) Log(Currency Volatility) 0.07 (0.54) 0.59 (0.66) REG(i) -0.39 (0.37) -0.35 (0.38) 9_{See Table 19 in the appendix}

Table 5 presents the results of the probit regression models focusing on cross-border M&As between companies based in the euro area and Bristish companies. The dependent variable is

Completed which a dummy variable sets equal to one when the anounced M&A deal has been

(30)

30 CASH 0.23 (0.21) 0.10 (0.22) Protection(j-i) 0.10 (0.09) -0.77** (0.35) Log(FirmSize)j -0.04 (0.03) -0.08** (0.04) Log(FirmSize)i 0.04 (0.03) 0.06 (0.04) ROA(j) -0.003 (0.01) 0.001 (0.01) Log(GDPcapita)j-i -8.74*** (3.07) -8.74*** (3.07) GDP growth(j-i) -0.03 (0.06) 0.25*** (0.08) Constant 1.73 (2.28) 0.18 (1.57) Pseudo R2 0.05 0.12 N observations 410 382 Acquiror-country FE NO YES

(31)

31

5. Conclusion

Cross-border mergers and acquisitions play an important role in the allocation of capital and in the integration of global economies. However, some political event might disturb CBM&A deals and contribute to suboptimal capital allocation. Global M&A activity did slow down in 2016 because of geopolitical uncertainty. In this thesis, I study the effect of a specific political event, which is here the Brexit referendum and examine how it has affected cross-border M&As with the UK but also its domestic market. To do so, several probits and OLS regressions have been run using different countries sample with robust evidence supporting the findings. The results indicate that the Brexit referendum did have an impact on both domestic and cross-border M&As activities. The domestic M&A activities in the UK market dropped dramatically in 2016 and 2017. This can be attributed to the Brexit announcement since such a trend has not been observed in the comparison countries used in this thesis. In addition, as observed in figure 2, the total M&A deal value in the UK has also been decreasing in 2016. Although, it appears to be a temporary break as both domestic and cross-border M&As increased again in the year 2018. Furthermore, the results show that BR16 has a negative effect on the likelihood of M&A completion and also changed some determinants’ influences. For instance, the results show that cash-only M&A deals were more likely to be completed after the BR16. This could be explained due to the Brexit uncertainty that leads companies to rather use cash, as it is a more straightforward way of payment by using the current exchange rate so removing the volatility risk. However, despite the uncertainty regarding the Brexit negotiations and its future relationship with the European Union, bidders were still attracted by UK targets with an all-time high of $251 billion in 2018, up 131% from the previous year (JPMorgan, 2019). International companies do not appear to retreat completely their investments from the UK as they are mainly waiting for the final result, whether the UK leaves the EU or not and if so under which conditions (soft or hard Brexit). Also, many companies are still engaging in CBM&A transactions with countries embroiled in trade uncertainties, as those firms are planning to get ahead of potential geopolitical disruption (EYb, 2018). Companies having supply chains across the single market might be consolidating their activities in the UK ahead of Brexit (The Economist, 2018).

Another important finding from this thesis is about the exchange rate movements. Currency fluctuation is a major factor determining the pattern of cross-border M&As. Previous studies found that a currency movement between two countries increases the likelihood that firms in a country where the currency is appreciating are more likely to be the purchasing company of the firms located in the country with the depreciating currency (Erel, Liao and Weisbach, 2012). This thesis supports this finding as it shows that when the Euro currency appreciates over the pound, firms located in the Eurozone are more likely to purchase British firms. Same results have been found when examining the sample with Japanese firms as bidding firms. Therefore, the weakening of the pound appears to give incentives for international firms to increase their investments into the UK as British firms become cheaper to acquire.

(32)

(33)

33

6. Appendix

Appendix A. Data appendix

Table 6: Countries with M&A data

DZ Algeria LS Lesotho

AR Argentina LY Libyan Arab

Jamahiriya AM Armenia LT Lithuania AU Australia LU Luxembourg AT Austria MY Malaysia BS Bahamas MT Malta BB Barbados MU Mauritius BE Belgium MX Mexico BZ Belize MA Morocco BM Bermuda MZ Mozambique BR Brazil NA Namibia BG Bulgaria NL Netherlands

CA Canada NZ New Zealand

KY Cayman Islands NG Nigeria

CL Chile NO Norway CN China PK Pakistan CO Colombia PA Panama CD Congo PE Peru HR Croatia PH Philippines CY Cyprus PL Poland CZ Czechia PT Portugal EG Egypt QA Qatar EE Estonia RO Romania

ET Ethiopia RU Russia Federation

FI Finland SA Saudi Arabia

FR France SN Senegal

DE Germany SG Singapore

GH Ghana SK Slovakia

GI Gibraltar SI Slovenia

GR Greece ZA South Africa

GT Guatemala ES Spain

GN Guinea SE Sweden

HK Hong Kong TW Taiwan, Province of

China

HU Hungary TZ Tanzania, United

Republic of

IS Iceland TH Thailand

IN India TT Trinidad and Tobago

ID Indonesia UG Uganda

IE Ireland AE United Arab Emirates

IL Israel GB United Kingdom

(34)

34

JP Japan UZ Uzbekistan

KZ Kazakstan VN Vietnam

KE Kenya VG Virgin Islands, British

KR South Korea ZM Zambia

LV Latvia ZW Zimbabwe

LB Lebanon

Table 7: Description of the variables

Panel A : Country characteristics variables

Variable Description

Log(GDPcapita)j-i The average difference between acquirer (j) and target (i) countries of domicile in the logarithm of annual GDP (real GDP per capita in EUR (2010)) divided by the population. (Source: Eurostat)

(GDP growth)j-i The average difference between acquirer (j) and target (i) countries of domicile in the annual real growth rate of the GDP (Source: Eurostat & Datastream)

(TAX)i The domestic corporate income tax rates of the

target country (i) (Source: OECD)

(Protection)i-j The difference between the acquirer (j) and target (i) countries’ strength of investor protection index. (Source: The World Bank) Euro Value Dummy variable equals to one if the annual

percentage change of the EUR/GBP exchange rate is positive and zero otherwise (Source: Datastream).

Japan Value Dummy variable equals to one if the annual percentage change of the GBP/JPY exchange rate is positive and zero otherwise (Source: Macrotrends Data).

Log(CurrencyVolatility) The standard deviation of the logarithm of the average annual real exchange rate EUR/GBP (Source: Datastream)

Panel B: Deal characteristics variables

PostBR16 Dummy variable equals to one if the deal

occurred during the period between 2016 and 2018 and zero otherwise.

Log(Deal value) The logarithm of the deal value of the M&A transaction, expressed in Euro. (Source: Zephyr)

Completed Dummy variable equals one if the announced

M&A deal is achieved and zero otherwise. (Source: Zephyr)

CASH Dummy variable equals to one whether the

method of payment is in cash only and zero otherwise. (Source: Zephyr)

Panel C: Company characteristics variables

(35)

35

REG(i) Dummy variable equals to 1 whether the target

company is among the industries regulated. (Source: Zephyr)

Log(FirmSize) Logarithm of the total assets of the firm. (Source: Orbis)

ROA(j) Return on assets as of the percentage of net

income of the acquiring firm. (Source: Orbis)

Table 8: Summary statisitics Model 1

Variable Mean Std.Dev Min Max

Completed .9640 .1862 0 1 PostBR16 .2924 .4550 0 1 CASH .3592 .4800 0 1 Status(i) .0508 0.2198 0 1 Protection(j-i) .3126 .8422 -.3 8 Log(FirmSize)j 12.2293 2.2293 5.6359 21.4038 Log(FirmSize)i 9.1580 2.2586 1.5124 20.4489 ROA(j) 2.1028 14.1190 -87.612 81.039 GDP growth(j-i) -0.0856 1.5440 -22.8 5.85 N observations: 1002

Table 9: Summary statistics Model 2

(36)

36 Table 10: Summary statistics Model

University of Groningen Master Thesis How are cross-border M&A activities with British companies affected by the threat of the Brexit?