How do institutional factors impact on international deal successions and returns in the M&A sector? : An event study of the abnormal returns and their relation to regulatory policies under different M&A waves.

Master Thesis

How do institutional factors impact on international deal successions and

returns in the M&A sector?

-An event study of the abnormal returns and their relation to regulatory

policies under different M&A waves

By

Jing (Jennifer) Yang

10475842

Supervisor: Prof. Jens Martin

August 2014

Amsterdam Business School

UVA

Master in International Finance

Abstract

This paper examines the possible influential factors specifically known as the regulatory policy issued by government to block proposed foreign acquisitions within the mergers and acquisitions (hereinafter referred to as M&As) sector in France, and to what extent such a policy narrows down deal successions that would ultimately result in shortened returns for both acquiring and target companies. To analyze, the ANOVA statistics has been deployed to test the abnormal returns of acquiring and target companies. Hypothesis tests have been conducted to detect difference in returns between acquirers and target firms and whether their returns are significantly outweigh the market returns impacted by an M&A event. The event window is set to different date ranges between [-2; +2], [-7; +7] and [-14; +14] for time analysis of the daily stock returns. The time-series analysis applied to the years 1990 to 2014 to explore further evidence of the positive correlation between macro-economic trends and M&A waves. In spite of the regulatory policy rules on deal successions, the test results show that the policy has no major impact on daily stock returns for both acquiring and target companies in France. Precluding the test of the long-term return performance, the short-run test upholds the Central Limit Theorem and the Market-Return model is applied. In addition, the relatively less than expected successful M&A volumes compared with the U.S. in the given time horizon does give rise to questions on the institutional factors of France.

Table of Content

1. Introduction ………4

2. Literature Review………9

3. Hypothesis and Methodology………14

4. Data ……….19

5. Empirical results and discussion ………..………..22

6. Conclusion ..………..31

7. Appendix ………34

8. References ..……….……….62

Introduction

The field of M&As continues to experience changes along with economic development on a worldwide scale. A mostly of U.S. concentrated field evolved into cross-boarder activities with the emergence of the emerging market acquirers and targets. As corporate strategy merits, companies that engaged in M&A activities are seeking complementary to their existing competencies, in the belief that accretive synergies coming from shared capital and resources will result in more investment and competitive advantages. At the same time, the corporate governance improvements coupled with the increasingly tough regulatory environment have caused mega-sized deals to be scuttled. A recent article was spotted in the Financial Times about France’s socialistic government taking action to block foreign corporate acquisition from the U.K. by arming itself with new powers to block takeovers1_{. As this case indicates, institutional} factors such as government imposing political power to barricade foreign acquisitions would have a direct impact on M&A deal succession. By 2013, France ranked third among top acquirer and fourth place among top target of $99 bln and $57.9 bln respectively in EMEA region2_{. Yet some would believe that the deterred mega deals} would not necessarily hurt the volumes per se, as the higher regulatory costs may encourage smaller companies to merge. In this paper, we investigate the impact of regulatory policies on M&A deal successions by answering the following specific research questions: have both acquiring and target companies experienced significant

1

Financial Times May 15, 2014 article title “France takes ‘nuclear weapon’ power to block foreign takeovers” 2

Bloomberg Global Financial Advisory M&A Rankings 2013

3

abnormal returns in post-M&A activities? And what is the role of government regulations in terms of changes in macro-economy and merger waves in France?

Previous studies have found that the abnormal returns to the target companies increase positively after an M&A announcement, and zero to negative abnormal returns to the acquiring companies (Hackbarth & Morellec, 2008; Loughran & Vijh, 1997). Moreover, the abnormal returns for the acquiring companies worsen in long-term post-merger periods and during the periods where merger activities are high (Gugler et all, 2012). This syndrome have been explained by Shleifer & Vishny ( 2003) and Rhodes-Kropf & Viswanathan (2004) as the overvaluation of the firms in the growth financial market climate. While the determinants of these results rest on the long-term performance in low and high M&A period, as an extended study, the daily abnormal return is applied to observe returns to the target companies for which are being partially acquired and whether the successful M&A events will propagate positive abnormal returns to target companies in France without governmental interventions. To link between returns and political factor is by the rate of abnormal returns. The lower rate of the abnormal returns is, the lower significance of regulatory policy on M&A is as to which the political factor would not influence much on the level of wealth creation. Moreover, this research provides comparison analysis between the returns of domestic and non-French acquirers’ returns as well as the returns to the listed targets. Since a bulk of researches have given the evidence of overpaid stock price destroys in bidders’ value (Madura and Wiant 1994), therefore, by taking into the political risk factor in the offer would help

acquiring companies to avoid overpaid in the premium, overprice the stock offer and more accurate to capture the expected return given by this risky factor.

Characterized by different periods of high volume activities on the market, M&A has experienced six merger waves in the U.S3. The sixth wave took place in period 2003- 2008 during the boom in the private equity sector, and it ended with the MBS housing bubbled. While whether the current growth trend indicates the seventh merger wave is still a puzzle, the cross-border M&As continue to recovery and include a total of 2537 deals worth $ 753bn which took place in the first half of this year4_{. Year to date, M&A is} up 61 percent to almost $ 2 trillion compared to a year ago and it remains on pace to top the last six years since 20075_.

So far, there has not been any study proving merger waves in France. However, by examining the M&A activities for the period 1990-2014 it proves the existence of pattern-like merger waves in France6. There have been four significant merger waves in the given time period. The years 1998-2001 and 2003-2007 saw the highest merger waves, and during which M&A volumes were consistent with the booming economic growth. And in the post-crisis period, even though the stock market has begun to pick up, the business climate in the M&A sector remains sluggish. Noteworthy is that for the years 2012 to 2013 M&A activities followed the downward trend in contrast to the

3

1 Institute of Mergers, Acquisitions and Alliances see in Appendix 1

4

FT August 4, 2014 article titled “ Cross-border M&A deals at post-crisis high” 5

Bloomberg news brief August 11, 2014 6

INSEE-National Institute of Statistics & Economic Studies. Website http://www.insee.fr/en/ see in Appendix 2

5

growing stock market. And from late 2013 M&A have started to pick up at a very slow pace. This evidence of inconsistency in French M&A has been seen between 1994 and early 1996 when the market experienced a high level of M&A activities under a negative stock-return market. As such, the inconsistency of the French business market trends implies the existence of institutional factors that affects M&A climate and which can be in form of catalyzing business volume through relaxed regulatory policies in passive economic view and to restrict business activities when government sees threats to national security and interests etc.

Three hypotheses are tested to examine the relationship of stock price reactions to M&A announcements with a sample of 33 companies whose M&As took place between 1990 and 2014 in France across different period of time. Year 1990 is the in the starting period of the fifth merger wave known for cross-border and mega-sized M&As. In order to test the significance, the event window is set to +/- two days to absorb stock price changes. And additional event windows are used for examining the change in stock return over a longer period of time. The first hypothesis is to test whether the targets earn positive abnormal returns in post-event period that have been partial acquired less than 50% shares. For Hypothesis 2 is to test whether the acquirers earn higher abnormal return in the post-event period. With Hypothesis 3 we expect to know the difference in returns when acquire listed targets and non-listed targets. Time-series analysis helps to examine the macro-economic patterns along with merger waves.

The paper is structured as follows. Section 2 gives an introduction to the relevant theories about both M&A event impact and the political risk factor on company returns and their relationships to the returns. Based on these theories, methodologies and hypotheses to be tested are put forth in Section 3. Section 4 describes the data selection procedure used in the empirical test. The results of the test are presented and discussed in Section 5. Section 6 concludes and summarizes the main findings of the paper.

Literature review

Developed by Chang (1998), Moeller et al (2004) and Fuller et al (2002) stating that the M&A phenomenon observed in U.S. is that acquirers get zero or negative average announcement period cumulative abnormal returns (CARs) when acquiring listed targets and positive average CARs when acquiring unlisted targets. Furthermore, the article of Faccio et al (2004) shows that acquirers of listed targets earn an insignificant average abnormal return of -0.38%, while acquirers of unlisted targets earn a significant average abnormal return of 1.48%. This phenomenon remains persistent throughout time and countries after controlling for the so-called Listing-related (LR) effect, which is explained as acquirer’s size, mode of payment, pre-announcement leakage of information about the transaction, whether there happen to be a block holder, whether it is a cross-border deal and other variables. The cross-sectional regressions prove two factors among the LR effect that are consistently significant are the acquirer’s size and whether the target is listed. Since the two factors appear to be universal and outweighing other LR effect variables in terms of dependency and level of significance and which can also be used in experimenting the M&As in France. Furthermore, Faccio et al (2004) disapproves the existence of institutional factors would have impact on CAR, instead it indicates the best scenario to better CAR is when the acquirer’s size is fairly small combining an unlisted target. However, reiterating the M&A trends in history we may see that institutional factors greatly have certain degrees of effect in CAR.

To further strengthen the relationship between the regulatory policies and M&A

successions and to a lesser extend the returns, the meta-analysis by Datta et al (1992) indicates the impact of regulatory policies influence the targets’ returns while the presence of multiple bidders and the type of acquisition influence the bidders’ return. Citing two regulatory changes in the article of Datta et all (1992) namely the 1968 Williams Amendment and the 1969 Tax Reform Act. For the 1968 Williams Amendment the required 10 trading days to evaluate tender offers had allowed additional bidders to enter the process, thereby the increased level of competition had directly impacted on the bid price to go up, which benefited the target’s shareholders at the expense of the bidding firms. As for the 1969 tax reform it prohibited the interest deduction on convertible bonds issued to finance a merger and also taxed negotiable bonds given to the seller as installment payments. Both policies had favored the targets’ returns but inversely to the bidders’ returns. These two regulatory changes happened in the third merger wave that is characterized by the U.S. companies were obsessed with entering new markets and it was ended with the crash of share prices. Perceived in merger waves that the level of policy restrictiveness can largely affect the returns for both acquiring and target companies.

Further, Martynova and Renneboog (2006) explains M&A returns in relation to merger waves. For example, the fifth merger wave is known for cross-border and mega-sized mergers. And this was the first time continental European firms participated openly towards cross-border M&As. The total value adds up to US $5.6 trillion, more than eight times the combined total of the fourth wave. Moreover, the relatively relaxed

regulatory policy seen by lower premiums in Continental Europe lead to a positive 0.5% significant effect on takeovers than a bid on a U.K. target with higher premiums but zero or negative significance over the short term period. The difference in level of premiums reflects a more strict takeover legislation in the U.K.

In addition, when the market is being over optimized given by overpaid premiums or overvalued stocks the abnormal return to the acquiring companies typically is hard to realize. According to Gulger et al (2012), as the volume of M&A increases the optimism in markets also increases. Duchin and Schmidt (2013) suggests that long-term post-merger performance of post-mergers initiated during post-merger waves are worse off than for deals undertaken outside merger waves. Therefore, to connect with the similar study we can use the time-series analysis to investigate the implication in returns under merger and outside the merger waves in France.

When looking at the deal succession factors, the study developed by Barattieri et al (2014) states that the succession of M&As have much to do with investment policies and it is a state-dependent factor which relates to the composition of gross domestic product (GDP) in the target economy. It explains that countries with large shares of manufacturing and services in GDP are less inclined to restrict M&A activities. Extracting the method by Borchert (2014) the Services Trade Restrictiveness Index (STRI) is used in this article that explains the effect of policy restrictiveness according to the segments of market, geography, culture and economic development sector. And the important result

highlights are that 1) institutional and cultural factors such as common legal system or religion affect M&A in services more than in manufacturing; 2) cultural barriers affect M&A in services more than in manufacturing e.g. a common language is more important for M&A in telecommunications, banking and insurance sectors; 3) developing countries tend to have more restrictive policies than developed countries and it is more likely to attract M&As in countries with more regulated corporate governance and investor protections.

Furthermore, developed by Eckbo (2000) the theory indicates that none of the factors such as direct foreign investment controls, horizontal product market relationships and acquisition propensities have influential factors in M&A returns. But it is the differentials between domestic and international deal types that influence the returns. The study shares the similar opinion as in Faccio et al (2004) that the bidders’ return is at the greatest when it involves stock payment and bidders have the smallest equity size relative to the target and it is a domestic acquisition.

The empirical results of aforementioned articles have shed light on the determinants of M&A returns for both acquirers and targets and what are the valid and relevant factors to examine M&A returns in a country-specific event setting, to note the French regulatory policies to M&A returns as studied in this research. Nevertheless, it is worth mentioning that when we evaluate the M&A returns we measure the returns to shareholders since all motives are for ultimate goal of enhancing profit to shareholders

for most companies. According to the Efficient Market Hypothesis (EMH) developed by Fama (1969) states that prices on capital markets fully reflect available information. The price change in stocks should be in line with the investors’ perceptions on the information at hand. And according to the Strong Form of the EMH that test can be done on the shareholders’ return of with/without deals undertaken. However, it is not possible to conduct the test since it is impossible to examine the exact stock price regardless the M&A (Bruner, 2004). And the Weak Form is where the information known is historical prices and the share price does not include the unrelated factors so it cannot be tested. Under the Semi-strong Form the market incorporates all related information into the stock price. One could conduct a test to compare the returns of M&A with a benchmark such as CAC40 index in this case. The reason the test is considered semi-strong is because benchmarks have certain level of error e.g. the weighing method. However, by using a large sample size and select the relevant benchmark the error can be minimized. Whereas the long-run performance have witnessed negative abnormal returns to acquiring companies and positive returns to target companies (Gugler et all, 2012), the short-term event window [-2,2] of daily abnormal returns under the Strong Form theory would be sufficient to reflect the economic outcome of M&A announcement.

Hypothesis and methodology

Hypothesis

Based on the theories mentioned in the previous chapter, three hypotheses have been employed to prove statistical significance on our presumptions. The tests are conducted under different event windows in order to capture more accurate statistical evidence. Since the long-run performance of the abnormal returns to both acquirers and targets evolves significantly under several control variables such as payment method, countries, size of the companies, high or low merger waves, macro-economic factors of the targets etc., the study of abnormal returns in the short-term event period is considered more appropriate as the stock price is able to absorb the economic outcome of M&A announcement immediately given in the five day event window period [-2,+2]. The long-term event window can be misleading since external events can interfere (Khotari and Warner 1997).

As mentioned earlier, the significance of level of political risk has much linkage with the abnormal returns of the targets. If targets in France perform a negative to zero rate of returns on average, it is meaningful to say that the existence of political risk would not result in a significant loss in M&As for the abnormal returns cannot be realized regardless of an M&A. Therefore, in applying the following three hypotheses we expect to have a comprehensive view on the abnormal returns through both the targets and the acquirers in France.

Hypothesis 1: Do the targets’ shareholders earn abnormal returns in the post-event period in France?

H0: The targets’ shareholders do not earn abnormal returns in the post-event period given by ;

H1: The targets’ shareholders earn abnormal returns in the post-event period given by

Hypothesis 2: Do the acquiresr’ shareholders earn abnormal returns in France in the post-event period?

H0: The acquirers’ shareholders do not earn abnormal returns in the post-event period given by

H1: The acquirers’ shareholders do earn abnormal returns in the post-event period given by CAAR>0

Hypothesis 3: Do the international acquirers’ shareholders receive higher abnormal returns than the domestic ones?

H0: The international acquirers’ shareholders do not receive higher abnormal returns than the domestic acquirers given by CAARintl< CAARdom

H1: The international acquirers’ shareholders do receive higher abnormal returns than the domestic acquirers given by CAARintl>CAARdom

Methodology

In this research we use the event study as the method to assess the impact of M&A announcement on companies’ returns. Developed by Fama et al (1969) , the method has been widely approved by academic researchers and scholars (MacKinlay 1997) in examining the changes in stock prices to a particular event. An event study measures the market-based returns to shareholders and it uses the abnormal return of a specific security as an indicator of wealth creation or destruction in the short-term event window. The abnormal return is calculated as the actual return of a security minus the normal return of a security over the event window (MacKinlay 1997). The formula is given as follows:

Rit = E[Rit|Xit] + εit

Rit is the actual return 
; E[Rit|Xit] is the e

εit is the abnormal return.

The actual return is what investors or shareholders actually received from investments. The normal return is the return expected by shareholders regardless of the M&A event. Under the Market Return model, the normal return measured by two parameters α and β that can be estimated by using the Ordinary Least Square (OLS) data analysis tool. Instead of using the Market-adjusted model which is a preferred model when there is only a short time-series of pre-event data available, in this research we use the Market-return model for more accuracy in the normal Market-return rate. The formula given in below indicates α and β estimators in the Market return model:

Rit = α + β *Rmt + εit,

Where ERit is the normal return of the security in the time period t; Rmt is the return of the market portfolio in the time period t, in this case it is the daily return of CAC40 and can be calculated using the formula:

And the actual return of the security can be calculated using the below formula:

To test the statistical significance we use the single factor ANOVA program to compute for each security. Further, we also apply the rate of the cumulative abnormal return (CAR) to capture the abnormal returns throughout the event window for all days.

∑

=

−

ˆ

)

2

,

2

(

ˆ

AR

C

ε

The cumulative average abnormal return (CAAR) is the average abnormal returns cumulated throughout the even window. CAAR will be used to compare the results given by the ANOVA test in order to provide strong argument to the economic outcome.

∑

=

→

−

ˆ

)

2

2

(

ˆ

AAR

C

ε

In this research, there are several variables not considered in our hypothesis, but can have major impact on the abnormal returns such as the payment method and company’s size etc. Here we have limited the control variables to the following: bidder ‘s

public status, target’s company public status, partial interest acquired (below 50% controlling state) and country region of the target. By controlling these variables we expect to test the difference in statistical significance between the abnormal returns of international acquirers and domestic acquirers in France as well as the returns of the target companies.

To further strengthen the causal relationship of the research questions and the findings, the Times-series analysis is applied. By use of this method we expect to have the comparison result between deals completed and being withdrawn in France and the likelihood of the political interference by examining any misalignment in business volumes given under high and low economic development patterns.

Data

To form the hypotheses test, a complete financial dataset needs to be fulfilled. By using Thomason One database the daily stock price of both acquirers and targets can be obtained. The information collected will include name of the M&A companies, Datastream codes, form such as merger or acquisition, percentage of interest acquired and announcement date. Further on, the stock prices for each acquiring and target company is collected from Datastream from two days before to two days after the M&A announcement. There are a total of 33711 deals took place with 27736 completed M&As took place between January 1990 and August 2014 in France. With this dataset, we can filter out the data designated for each hypothesis test. First, we draw the data of the listed targets whose shares are being partial acquired and it gives a total number of deals and from which we randomly selected 33 companies across time. As the acquirers control state to the targets should be less than 50% it allows us to test the abnormal return as well as CAR of the targets. Second, we abstract the data of the acquirers and separate them between the international and domestic ones. Within the dataset there are a total of 33 international acquirers and a total of 14 domestic acquirers in the sample. With this dataset we are able to calculate the abnormal returns of the acquirers as a whole and further to test whether the international acquirers earn high AR than the domestic ones. All these data have been filtered under the completed deals status meaning we neglect the ones under announced or rumored only. In addition, with ANOVA we are able to calculate the standard deviation, min and max of the returns for both the acquirers and the targets. As the standard deviation measures how close the

data to the mean, by calculating the min and max of the AR it helps to understand on which day in the event window is most or least affected by the M&A announcement.

Moreover, the normal return needs to be collected for the same days as the stock prices in order to compare the values for each day. Due to the non-synchronous trading of the daily stock returns, the return on a security and the return on the market index can be measured over a different trading interval. When drawing the data, we found out that the difference in trading interval is limited to only 1-2 days and thereof negligible. Under the OLS we can calculate both alpha and best estimator via the data regression tool pack.

Last but not least, we also use the data from INSEE-National Institute of Statistics & Economic of France. The site consists of the complete information on the French business climates over the period of 1990 and 2014. With this information, we are able to draw the merger waves signaled as high and low in stock returns. The business climate is defined as the volume of business activities such as M&As and with this information we are able to form the patterns throughout the whole period indicated the M&A waves in France. Since stock return of INSEE is calculated on a monthly basis, therefore in CAC40 we also need to download the monthly stock return to match the results. Since both returns are measured under different scale, after normalizing we are able to compare the results under the same scale. As shows in the chart (Appendix 2) four merger waves took place in France, with the highest one during year 1998 and year

2001. In recent history, France has experienced two times economic recession periods from 2008 to 2009 and from 20012 to 2013 . When there is contradiction in the trends namely as a low merger wave under an stock price booming period or a high merger wave under a decreasing trend that to a lesser extent it can denote to the role of regulatory policy established or other bureaucratic facts in the given period.

Finally, it is noteworthy that when use the daily stock returns we should bear in mind of a couple of potential problems. For instance, the daily stock return suffers from non-normality and the distributions of daily stock returns are fat-tailed compared to a normal distribution (Fama 1976). Since the event study focuses on the cross-sectional sample of mean of security excess returns, according to the Central Limited Theorem (Billingsley 1979) it guarantees that if the excess returns in the cross-section of securities are independent and identically distributed drawings from finite variance distributions, the distribution of the sample mean excess return converges to normality as the number of securities increases (Brown 1984). Developed by Brown (1984) It indicates the non-normality of daily returns has no obvious impact on event study methodologies. Although daily excess returns are also highly non-normal and volatile to changes when new information comes in, there is evidence that the mean excess return in a cross-section of securities converges to normality as the number of sample securities increases. Therefore, by selecting a large sample from Thomas One the statistical distribution will be not encounter non-normality problem.

Empirical results and discussion

Empirical Results

The following tables show the test results of the ANOVA calculation on the abnormal returns of both acquirers and targets at 5% significant level. Hypothesis 1 predicts the significant level in returns to the target companies in France in the given period. Under the null hypothesis we expect to conclude that the targets’ shareholders do not earn abnormal returns in the post-event period. The abnormal returns of the targets group calculated are insignificant increase over the event window surrounding at zero percent (table 2). Around the announcement day the abnormal normal of the targets have witnessed a slight increase from 0.22% to 0.88% given in the average column. F test shows that the critical value 2.38 is bigger than the F value obtained 1.48 and the F ratio do not fall within the critical region. Therefore we do not reject the null hypothesis and conclude that the targets do not earn abnormal returns in the post-event period in France. This conclusion can also be interpreted by the P value. Since P value at 0.20 is greater than alpha of 0.05. therefore, it is not significant enough to reject the null hypothesis.

Table 2- ANOVA test of 5-day event abnormal returns of the target firms

Table 2 displays the test result of the targets over a five-day event window [-2;+2]. The Alpha level is set at 0.05. A initial total of 132 companies being tested. P-value is not significant at 0.21 greater than Alpha level 0.05 . The abnormal return under the [-2;+2] event window is between -9.91% and 10.86% , and any rate outside this range should be not be considered. F crit is greater than F obtained, therefore we do not reject the null hypothesis.

Now when comparing with a [-7; +7] event window the table in below shows a different result.

Table 3- ANOVA test of 15-day event abnormal returns of the target firms

Table 3 shows the test result of targets over a [-7; +7] event window. The Alpha level is set at 0.05. A initial total of 33 companies being tested. P-value is not significant at 2.313. Between groups the degree of freedom is 14, and within groups it is 480. The F crit of 1.72 is much lower than F obtained 24.87 which leads to REJECT the null hypothesis, indicating that the target companies do earn abnormal returns out of M&A.

Hypothesis 2 predicts the abnormal returns of the acquirers in France and test whether by the impact of the M&A announcement the companies are able to generate positive and significant from zero rate returns. Given as in Table 4 the F value obtained is 0.0033 that is smaller than the F critical at 2.428. This means that the F value does not fall into the critical region. Therefore we do not reject the null hypothesis, concluding that the acquirers’ shareholders do not earn abnormal returns to in the post-event period in France. CAAR T+2 is at -0.0403 compared with 0.0396 at T-2. The P value is insignificant at 0.9999.

Table 5 shows the comparison test results between the international acquirers and the domestic acquirers. Given by Hypothesis 3 that we expect to test the abnormal returns of whether the international companies earn higher rate of return than the domestic ones surrounding the event window. Table 5 displays the test results of the international acquirers. The F value 1.577 is smaller than F critical 2.479. And 0.187 of the P value also close to the test significance of 0.10. Under [-7;+7] event window (see Appendix 2 & Appendix 3) the test significant level of the international acquirers’ shareholders earn higher abnormal returns at relatively significant level than the domestic do. Moreover, the lower table in Table 6 indicates the test results of the domestic acquirers. Since the F critical is much higher than the F value obtained, therefore, we can say that the domestic acquirers’ shareholders do not earn abnormal returns in the post-event period in France. To answer the hypothesis 3 question, since the significant level of the international acquirers are higher than the domestic ones, we do reject the null hypothesis and conclude that the international acquirers’ shareholders actually earn abnormal returns than the domestic ones in the post-event period in France.

Table 4- ANOVA test of 5-day event abnormal returns of the acquiring firms

Table 4 shows the test result of a [-2; +2] event window of the acquiring firms. The Alpha level is set at 0.05. A initial total of 33 companies being tested. P-value is 0.999 not significant from 0.10 . Between groups the degree of freedom is 4, and within groups it is 164 . The F crit of 2.428 is higher than F obtained 0.003 which leads to NOT REJECT the null hypothesis, indicating that the target companies do NOT earn abnormal returns out of M&A.

Table 5- ANOVA test of 5-day event abnormal returns of the INTL acquiring firms

Table 5 shows the test result of international or non-French acquirers under the [-2; +2] event window. The Alpha level is set at 0.05. A initial total of 18 companies being tested. P-value is not significant at 0.18. Between groups the degree of freedom is 4, and within groups it is 85. The F crit of 2.479 is much lower than F obtained 1.577 which leads to NOT REJECT the null hypothesis, the target companies do NOT earn abnormal returns out of M&A.

Table 6- ANOVA test of 5-day event abnormal returns of the domestic firms

Table 6 shows the test result of domestic acquirers under the [-2; +2] event window. The Alpha level is set at 0.05. A initial total of 15 companies being tested. P-value is not significant at 0.999. The F crit of 2.5026 is much lower than F obtained 0.007. Thus we do NOT REJECT the null hypothesis, indicating that the target companies do NOT earn abnormal returns out of M&A.

Discussion

The empirical results given above prove that both acquirers and targets in France do not generate significant abnormal normal returns to the shareholders in the post-activity period for CAARs are less or very close to zero. And within the acquire group we concluded that the international or non-French acquirers do not earn higher abnormal returns than the French acquirers by having lower CAAR. The results have matched with our expectation and which are also in line with the findings of Andrade et al (2001) which states it is mostly unprofitable for the acquirers’ shareholders presented by negative abnormal returns around the event day and domestic bidders outperform

international bidders as in Canada’s case by Eckbo et al (2000). Given in the short event period, the stock prices are more volatile surrounding the event window other than during the longer period of time. This can be viewed as insider trading incidence when new information is leaked and stock price begins to fluctuate according to the investors’ expectation. However, as the Central Limited Theorem holds the daily abnormal returns does not suffer from the fat-tail event. Further, the pre-acquisition performance of the acquirers and the targets are not significantly related to the M&A returns in the post-event period seen as irregular changes throughout the even window.

Chart 1 in below plots the trends of business climate and CAC40 stock returns during 1990 to 2014. In most periods, the stock markets fluctuate at the similar level compared to the business climate. However, the period of early 2012 to 2013 the graph displays a reversed pattern given under decreasing business volumes within a growth market scenario. This arises the question of whether the country risk factor such as regulatory policies may play a role causing the business volume to decrease. Or it is just a momentum when the global financial market starts picking up from a depressed period. Or it is under a bubble where the stocks are being overvalued. Nevertheless it is worth mention that due to the facts that in the post financial crisis period more and more companies and financial institutions have increased their levels of corporate governance and become more careful towards credit risk exposures.

The decreasing levels of business climate are strongly linked with market pessimism. According to Puri & Robison, 2007) that market participations are likely to undertake more risk in the booming period other than during recessions which contributes to increased investment incentives and thus higher willingness to take part in M&As. And the reversed pattern implies the impact of regulatory policies effect on business investments when business climate have should experience an upward trend.

Chart 1-Comparative analysis of stock market returns and the business climate over 1990-2014

Finally, the below chart displays the different returns among CAC40, CAAR targets and CAAR acquirers. Given in the availability of data, the CAAR targets experienced some extreme values which are being excluded in this analysis. Therefore, for the target we

only focus on the less than [-8; +8] event window. Since both CAAR targets and CAAR acquirers have been performed below the market with -5.5 and -4.0 respectively, therefore, it supports our hypothesis that given in the event window we do not see any significant returns impacted from M&As in France.

Chart 2- Comparative analysis of CAC40, CAAR targets and CAAR acquirer over [-14;+14] event window -7 -6 -5 -4 -3 -2 -1 0 1 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 CAC40 CAAR Target CAAR Acquirer 29

Conclusion

The influence of regulatory policies has captured more and more attention in recent cross-border business environment. Given by its strong presence in M&As, as part of the national interest and security purview governments have been striving to protect their nation’s high value companies from being acquired by competitors in the market or by political conflict countries through increased state approval right, especially on mega-sized deals. This political factor has negatively affected M&A successions when shareholders foresee wealth creation whereas the government sees a threat. In this paper, we examine stock price reactions to M&A announcement surrounding a five-day event window. A total of 33 companies have been collected from Thomason one database with 18 non-French acquirers and 15 French acquirers between 1990 and 2014. Three hypotheses have shown the test results of whether the targets’ and the acquirers’ shareholder earn abnormal returns in the given period and whether the international or non-French acquirers gains more abnormal returns than the French ones. The results present non-significant abnormal returns gained to the acquirers’ and the targets’ shareholders. And among which the domestic acquirers earn slightly better rate of returns than the non-French acquirers do. The test results also infers that the impact of regulatory policies on M&As is insignificant as explained by the fact that the wealth creation from M&As fluctuates at zero percent and regardless of the state’s disposals that both the bidders and the potential targets would not experience significant loss when deals are scuttled. However, testing under a longer-term event window we expect to witness different test results.

Previous theories state that bidders’ abnormal return become negative and positive to target companies in the long run. Further it concludes that bidders’ return is at the greatest when it is a domestic acquisition under several other control variables by Faccio et al (2004). This result has matched with our findings on the French market in the short-term scope. To summarize, we can claim that there has been no significant impact found on government interventions to M&As successions in France. This is proclaimed under the fact of insignificant CAARs in the given period. Second, during early 2012 and 2013 there has been contrary development in trends. Although the temporary misalignments restored at later stage to match with stock market returns, we see a potential country risk factors e.g. political, regulatory or even bureaucratic problems during the reversed development periods. As indicated by Barattieri et al (2014) that M&A succession has much to do with investment policies differentiating among high and low GDP economies and such influence is stronger in low growth GDP economy. France ranks at the seventh while U.S. is at the second place in year 2013 and when comparing M&A deals U.S. has scored $ 1.223 billion that is nearly 10 times higher than what France achieved measured by deal value. Finally, this research is subject to certain limits. Given by the short-term event window we are not able to test the asserted theories of previous studies done on the long-run performance of the abnormal returns. It is interesting for investors at large to gain insights into the long-term test results of M&As with knowing the fact that it takes certain period of time to realize the synergy from M&A event. Nevertheless, since this paper can be seen as a pioneer study done on the political influence on France’s M&As in specific and it can be subjective when

adapting researches done on other countries such as U.S. given as differences in M&A market maturity and company structure etc. This paper has as its main purpose to provide guidelines for those plans to invest in France through M&A form. And by taking into account of the political risk factor, it helps bidders to make more accurate offer and ultimately resulting in better investment returns.

The daily abnormal returns has its aim to mitigate the impact of eternal events affecting the stock price seen in a long-term horizon e.g. macro-economic changes, merger waves etc. However, the daily abnormal returns may still encounter insider trading surrounding the M&A announcement date.

Further research

For future research, it would be interesting to test if the abnormal returns are indifference given by payment method, size of companies, abnormal returns under high and low merger waves under long-term performance.

Appendix 1

Appendix 2

Table 7- ANOVA test of 15-day event abnormal returns of the INTL acquiring firms

The below table shows the test result of international or non-French acquirers under the [-7; +7] event window. The Alpha level is set at 0.05. A initial total of 18 companies being tested. P-value is close to significant at 0.11. Between groups the degree of freedom is 14, and within groups it is 255. The F value of 1.496 is very close to F critical value of 1.7307 which leads to NOT REJECT the null hypothesis, the target companies do NOT earn abnormal returns out of M&A.

Appendix 3

Table 8- ANOVA test of 15-day event abnormal returns of the domestic acquiring firms

This Table shows the test result of domestic acquirers under the [-7; +7] event window. The Alpha level is set at 0.05. A initial total of 15 companies being tested. P-value is not significant at 1 greater than 0.05. The F crit of 1.7391 is much higher than F obtained 0.011. Thus we do NOT REJECT the null hypothesis, indicating that the target companies do NOT earn abnormal returns out of M&A.

Appendix 4

Target returns summary ARs,CAR, CAAR over [-10;+10] event window

Appendix 5

Acquirer returns summary ARs,CAR, CAAR over [-14;+14] event window

Appendix 6

CAC40 stock returns over 1990-2014

0 1000 2000 3000 4000 5000 6000 7000 8000

Adj Close

Returns

Appendix 7

Comparative analysis of CAC40 and business climate

01/03/1990 1.201469378 -1.152297239 01/04/1990 1.104651301 -1.105299441 01/05/1990 1.104651301 -1.160772252 01/06/1990 0.911015148 -1.214704152 01/07/1990 0.911015148 -1.426579471 01/08/1990 0.52374284 -1.573736512 01/09/1990 0.233288609 -1.461249978 01/10/1990 -0.057165622 -1.492068207 01/11/1990 -0.057165622 -1.566031955 01/12/1990 -0.347619853 -1.511329599 01/01/1991 -0.44443793 -1.372647572 01/02/1991 -1.025346392 -1.329502052 01/03/1991 -0.541256007 -1.358008913 01/04/1991 -0.638074084 -1.294215181 01/05/1991 -0.734892161 -1.382201223 01/06/1991 -0.638074084 -1.376730987 01/07/1991 -0.541256007 -1.292289041 01/08/1991 -0.541256007 -1.279730613 01/09/1991 -0.44443793 -1.298144505 01/10/1991 -0.638074084 -1.388287823 01/11/1991 -0.638074084 -1.368255974 01/12/1991 -0.734892161 -1.283814029 01/01/1992 -0.638074084 -1.200527766 01/02/1992 -0.347619853 -1.232116451 01/03/1992 -0.347619853 -1.163777029 01/04/1992 -0.638074084 -1.162082027 01/05/1992 -0.44443793 -1.264321499 01/06/1992 -0.541256007 -1.376730987 01/07/1992 -0.928528315 -1.430508795 01/08/1992 -1.122164469 -1.39059919 01/09/1992 -1.4126187 -1.386207592 01/10/1992 -1.606254854 -1.363864377 01/11/1992 -1.606254854 -1.297297003 01/12/1992 -1.896709085 -1.363248012 54

01/01/1993 -2.090345239 -1.20029663 01/02/1993 -2.187163316 -1.163545892 01/03/1993 -2.187163316 -1.234736 01/04/1993 -1.993527162 -1.273489922 01/05/1993 -1.896709085 -1.209388007 01/06/1993 -2.187163316 -1.121556056 01/07/1993 -1.799891008 -1.020934541 01/08/1993 -1.703072931 -1.099443978 01/09/1993 -1.606254854 -1.047515263 01/10/1993 -1.509436777 -1.102911028 01/11/1993 -1.218982546 -0.981101981 01/12/1993 -0.928528315 -0.930097813 01/01/1994 -0.638074084 -1.004292697 01/02/1994 -0.541256007 -1.124637879 01/03/1994 -0.153983699 -1.059842554 01/04/1994 0.136470532 -1.164701576 01/05/1994 0.330106686 -1.270947418 01/06/1994 0.52374284 -1.129954024 01/07/1994 0.52374284 -1.134499712 01/08/1994 0.717378994 -1.280732206 01/09/1994 0.911015148 -1.260392175 01/10/1994 1.007833225 -1.206306184 01/11/1994 1.007833225 -1.27926834 01/12/1994 0.911015148 -1.3434473 01/01/1995 1.007833225 -1.35962687 01/02/1995 0.814197071 -1.295987229 01/03/1995 0.814197071 -1.250530342 01/04/1995 0.620560917 -1.227801899 01/05/1995 0.717378994 -1.296526548 01/06/1995 0.620560917 -1.249374658 01/07/1995 0.52374284 -1.277573337 01/08/1995 0.233288609 -1.35076663 01/09/1995 -0.057165622 -1.331042964 01/10/1995 -0.347619853 -1.320025447 01/11/1995 -0.638074084 -1.286356532 01/12/1995 -1.025346392 -1.171558632 01/01/1996 -1.025346392 -1.194826394 01/02/1996 -1.025346392 -1.153221786 01/03/1996 -0.638074084 -1.074635304 01/04/1996 -0.541256007 -1.102911028 01/05/1996 -0.734892161 -1.092432831 55

01/06/1996 -0.734892161 -1.19089707 01/07/1996 -1.025346392 -1.2103896 01/08/1996 -0.928528315 -1.085421684 01/09/1996 -0.831710238 -1.079489175 01/10/1996 -0.831710238 -0.944505335 01/11/1996 -0.734892161 -0.944505335 01/12/1996 -0.541256007 -0.789720783 01/01/1997 -0.250801776 -0.719455222 01/02/1997 0.039652455 -0.681779938 01/03/1997 0.039652455 -0.695031776 01/04/1997 -0.057165622 -0.737869113 01/05/1997 -0.057165622 -0.526456067 01/06/1997 0.136470532 -0.358958996 01/07/1997 0.039652455 -0.594102078 01/08/1997 0.330106686 -0.410887711 01/09/1997 0.620560917 -0.618140296 01/10/1997 0.52374284 -0.529460844 01/11/1997 0.717378994 -0.418129994 01/12/1997 0.911015148 -0.284687066 01/01/1998 0.911015148 -0.09222723 01/02/1998 1.007833225 0.257097389 01/03/1998 1.298287455 0.26142735 01/04/1998 1.007833225 0.384885172 01/05/1998 1.201469378 0.509922429 01/06/1998 1.298287455 0.489782717 01/07/1998 1.201469378 0.172293329 01/08/1998 1.201469378 -0.264770786 01/09/1998 0.911015148 -0.014388089 01/10/1998 0.620560917 0.232504442 01/11/1998 0.52374284 0.308995285 01/12/1998 0.233288609 0.547173963 01/01/1999 0.330106686 0.424779369 01/02/1999 0.330106686 0.505630991 01/03/1999 0.136470532 0.665477437 01/04/1999 0.233288609 0.623826601 01/05/1999 0.426924763 0.766607453 01/06/1999 0.330106686 0.647533523 01/07/1999 0.814197071 0.807141128 01/08/1999 1.104651301 0.808836131 01/09/1999 1.395105532 1.037815567 01/10/1999 1.491923609 1.386832003 56

01/11/1999 1.685559763 1.861972039 01/12/1999 1.78237784 1.631983305 01/01/2000 1.491923609 2.041210855 01/02/2000 1.588741686 2.114473488 01/03/2000 1.685559763 2.217460303 01/04/2000 1.588741686 2.222499083 01/05/2000 1.588741686 2.238123925 01/06/2000 1.685559763 2.31204915 01/07/2000 1.685559763 2.375943042 01/08/2000 1.685559763 2.099511238 01/09/2000 1.78237784 2.20046405 01/10/2000 1.588741686 1.838673458 01/11/2000 1.395105532 1.837394502 01/12/2000 1.298287455 1.892921244 01/01/2001 1.201469378 1.406755988 01/02/2001 1.201469378 1.262657657 01/03/2001 1.298287455 1.616743691 01/04/2001 0.717378994 1.473562202 01/05/2001 0.52374284 1.297235709 01/06/2001 0.426924763 1.189510592 01/07/2001 0.233288609 0.884279154 01/08/2001 0.136470532 0.414054626 01/09/2001 -0.057165622 0.616122044 01/10/2001 -0.44443793 0.71995636 01/11/2001 -0.541256007 0.834384442 01/12/2001 -0.44443793 0.709023594 01/01/2002 -0.347619853 0.709886504 01/02/2002 -0.250801776 0.883262152 01/03/2002 -0.057165622 0.70969389 01/04/2002 0.136470532 0.564771171 01/05/2002 0.330106686 0.274579029 01/06/2002 0.233288609 -0.097250601 01/07/2002 -0.057165622 -0.135133908 01/08/2002 -0.153983699 -0.588747411 01/09/2002 -0.250801776 -0.301683319 01/10/2002 -0.347619853 -0.165613136 01/11/2002 -0.153983699 -0.368042669 01/12/2002 0.039652455 -0.465143202 01/01/2003 -0.347619853 -0.606760666 01/02/2003 -0.347619853 -0.711242164 01/03/2003 -0.831710238 -0.452977706 57

01/04/2003 -1.122164469 -0.423638753 01/05/2003 -0.734892161 -0.352487168 01/06/2003 -0.734892161 -0.255278771 01/07/2003 -0.44443793 -0.177347176 01/08/2003 -0.347619853 -0.313278677 01/09/2003 -0.347619853 -0.129748423 01/10/2003 -0.250801776 -0.090000613 01/11/2003 0.039652455 0.012554747 01/12/2003 0.233288609 0.074607249 01/01/2004 0.620560917 0.141636896 01/02/2004 0.52374284 0.06442953 01/03/2004 0.52374284 0.102220382 01/04/2004 0.620560917 0.098637763 01/05/2004 0.620560917 0.147453837 01/06/2004 0.52374284 0.081279396 01/07/2004 0.717378994 0.040583925 01/08/2004 0.717378994 0.076279138 01/09/2004 0.717378994 0.127291011 01/10/2004 0.814197071 0.163448497 01/11/2004 0.52374284 0.215384917 01/12/2004 0.717378994 0.286675183 01/01/2005 0.620560917 0.374098792 01/02/2005 0.620560917 0.405394703 01/03/2005 0.620560917 0.285149681 01/04/2005 0.52374284 0.446190333 01/05/2005 0.330106686 0.529877232 01/06/2005 0.426924763 0.701218877 01/07/2005 0.52374284 0.660862407 01/08/2005 0.52374284 0.81546205 01/09/2005 0.426924763 0.689438609 01/10/2005 0.717378994 0.790337489 01/11/2005 0.620560917 0.904226252 01/12/2005 0.814197071 1.083557523 01/01/2006 0.911015148 1.123975629 01/02/2006 0.717378994 1.293784067 01/03/2006 0.717378994 1.26878278 01/04/2006 0.814197071 1.069835706 01/05/2006 1.007833225 1.097402612 01/06/2006 0.911015148 1.130886617 01/07/2006 1.104651301 1.250784934 01/08/2006 1.104651301 1.316250556 58

01/09/2006 1.201469378 1.392309943 01/10/2006 1.104651301 1.376061032 01/11/2006 1.201469378 1.541031009 01/12/2006 1.104651301 1.592304836 01/01/2007 1.104651301 1.521430615 01/02/2007 1.201469378 1.612221116 01/03/2007 1.298287455 1.863297222 01/04/2007 1.395105532 1.974212026 01/05/2007 1.395105532 1.936405765 01/06/2007 1.491923609 1.702302798 01/07/2007 1.395105532 1.634209922 01/08/2007 1.298287455 1.67503637 01/09/2007 1.298287455 1.776936842 01/10/2007 1.201469378 1.640273409 01/11/2007 1.201469378 1.596750366 01/12/2007 1.201469378 1.023307886 01/01/2008 1.007833225 0.962341726 01/02/2008 0.911015148 0.897939333 01/03/2008 0.911015148 1.120963147 01/04/2008 0.620560917 1.134631032 01/05/2008 0.330106686 0.68820588 01/06/2008 0.136470532 0.655469217 01/07/2008 -0.347619853 0.72499514 01/08/2008 -0.44443793 0.377904844 01/09/2008 -0.638074084 -0.042016631 01/10/2008 -1.509436777 -0.214899188 01/11/2008 -1.799891008 -0.249346262 01/12/2008 -2.187163316 -0.437375978 01/01/2009 -2.38079947 -0.646508476 01/02/2009 -2.574435624 -0.56571849 01/03/2009 -2.864889855 -0.294125148 01/04/2009 -2.768071778 -0.203365466 01/05/2009 -2.477617547 -0.309079693 01/06/2009 -2.187163316 -0.088860338 01/07/2009 -1.993527162 0.086241131 01/08/2009 -1.703072931 0.195545682 01/09/2009 -1.315800623 0.050915736 01/10/2009 -1.122164469 0.106742957 01/11/2009 -0.831710238 0.3041183 01/12/2009 -0.831710238 0.152438685 01/01/2010 -0.638074084 0.128816513 59

01/02/2010 -0.541256007 0.333149072 01/03/2010 -0.541256007 0.212172116 01/04/2010 -0.250801776 -0.026229994 01/05/2010 -0.057165622 -0.076055364 01/06/2010 -0.153983699 0.078228391 01/07/2010 0.136470532 -0.039150536 01/08/2010 0.330106686 0.133732021 01/09/2010 0.426924763 0.22489234 01/10/2010 0.52374284 0.05303449 01/11/2010 0.52374284 0.202764852 01/12/2010 0.717378994 0.357410722 01/01/2011 0.814197071 0.438193003 01/02/2011 0.814197071 0.344836885 01/03/2011 1.007833225 0.43555034 01/04/2011 0.911015148 0.358520178 01/05/2011 0.814197071 0.339466808 01/06/2011 0.911015148 0.101056994 01/07/2011 0.620560917 -0.219460285 01/08/2011 0.136470532 -0.431181514 01/09/2011 -0.153983699 -0.230185029 01/10/2011 -0.250801776 -0.298154631 01/11/2011 -0.44443793 -0.294155966 01/12/2011 -0.541256007 -0.187262941 01/01/2012 -0.638074084 -0.068689808 01/02/2012 -0.638074084 -0.090755659 01/03/2012 -0.44443793 -0.253329518 01/04/2012 -0.347619853 -0.404177042 01/05/2012 -0.734892161 -0.265772378 01/06/2012 -0.831710238 -0.192571381 01/07/2012 -1.025346392 -0.099030354 01/08/2012 -1.025346392 -0.143909399 01/09/2012 -1.218982546 -0.086548971 01/10/2012 -1.4126187 0.012077064 01/11/2012 -1.218982546 0.076633548 01/12/2012 -1.122164469 0.147153359 01/01/2013 -1.122164469 0.139756984 01/02/2013 -1.218982546 0.146244221 01/03/2013 -1.315800623 0.242805435 01/04/2013 -1.4126187 0.313564087 01/05/2013 -1.315800623 0.152014935 01/06/2013 -1.315800623 0.347541184 60

01/07/2013 -1.122164469 0.302153638 01/08/2013 -0.831710238 0.463687382 01/09/2013 -0.44443793 0.584225178 01/10/2013 -0.541256007 0.580619445 01/11/2013 -0.541256007 0.581189582 01/12/2013 -0.541256007 0.480853135 01/01/2014 -0.541256007 0.667580781 01/02/2014 -0.541256007 0.654806625 01/03/2014 -0.44443793 0.728685623 01/04/2014 -0.44443793 0.753478888 01/05/2014 -0.44443793 0.678952707 01/06/2014 -0.638074084 0.542813184 01/07/2014 -0.638074084 0.467054274 61

References

Andrade, G., Mitchell, M. L. & Stafford, E. 2001. New evidence and perspective on mergers. The Journal of Economic Perspectives, Vol.15 (2), pp.103-120

Bruner, R. F. 2004. Applied mergers & acquisitions. John Wiley & Sons. New Jersey.

Chang, S., 1998, Takeovers of privately held targets, methods of payment and bidder returns, Journal of Finance, 53, 773-784.

Eckbo, B. E., and K. S. Thorburn, 2000, Gains to bidder firms revisited: Domestic and foreign acquisitions in Canada, Journal of Financial and Quantitative Analysis, 35, 1-25.

Fuller, K., J. Netter, and M. Stegemoller, 2002, What do returns to acquiring firms tell us? Evidence from firms that make many acquisitions, Journal of Finance, 57, 1763-1793.

Fama, E.F., 1976, Foundations of finance (Basic Books, New York).

Fama, E. 1969. Efficient capital markets: A review of theory and empirical work. The Journal of Finance, Vol 25 (2) pp. 383-417

Gugler, K., Mueller, D. C., Weichselbaumer, M. & Yurtoglu B. B. 2012. Market optimism and merger waves. Managerial and Decision Economics. Vol.33, pp.159-175

Hackbarth, D. & Morellec, E. 2008. Stock returns in mergers and acquisitions. The Journal of Finance, Vol.63(3), pp.1213-1252

Mackinlay, A. C. 1997. Event studies in economics and finance. Journal of Economic Literature, Vol.35(1), pp.13-39

Shleifer, A. & Vishny, R. W. 2003. Stock market driven acquisitions. Journal of Financial

Economics, 2003, Vol.70 (3), pp.295-311

Kothari S. P. & Warner J. B. 2004. Econometrics of Event Studies. Working Paper. Center for Corporate Governance

Website

Trading Economics http://www.tradingeconomics.com/country-list/gdp