The French Companies and Their Stock Market Performance at the Launch of Corporate Acquisition Programs

This paper presents an econometric analysis of the profitability generated at the initiation of the corporates acquisition programs by French investors. These programs generally launched by the firms directors, need to have a high frequency of acquisition attempts, in order to cover fixed costs of the program and generate profits, which they increase the value of the firm and enrich the shareholders. Our empirical study applied on a sample of French companies of various economics branches gives a controversial result. Indeed, the sample of study made up of 46 firms of any size over a period of 11 years from 1997 to 2007, shows that only the hypothesis of the announcement effect was been verified. In the sub-sample of the studied companies, the acquisition programs of two “SMEs” namely Guerbet and Sartorius had a positive value, what proves their financial performance. These results show that the corporate acquisition programs are projects of value-creating investment for French small and medium enterprises. However, generally, for the sample of study, the French companies considered as frequent bidders on French acquisition market indicate clearly that firm’s acquisition programs during the study period were destructive of value. It means they do not maximize the value of the firm nor the stockholder’s wealth.


Introduction
One of the aim objective of merger and acquisition (M&A) operations for the acquiring firms is to operate an external expansion with the aim of getting bigger and developing. Our empirical study [5] in which we evaluate the profitability of corporates acquisition programs; try to specify stocks price reactions at the announcement of M&A operations. It is a question particularly to verify if the partially anticipated events by the French acquirers generates consistent returns, which can be compared with those of the American case 1 .
The actual financial literature about this subject largely try to calculate the gains associated to the acquisition attempts for both, acquirers and target firms. Another point treated in this literature focus on the identification of the impact of these operations on the shareholders wealth.
Song and Walkling [11] argues that at any point in time, a firm's market value is a weighted average of the values of the firm under current and alternative management where the weights are the probability of the firm being controlled by the various managerial teams. Acquisition attempts will occur when the expected gain from an acquisition attempt exceeds the cost.
Malatesta and Thompson [8] provide evidence on the acquisition programs profitability for American case. Both researchers are the first ones who established an econometric model, allowing estimating the value of an acquisition program. Previously, Schipper and Thompson [10] realized an empirical study in which they measure the impact of acquisitions activity on firm value by differentiating between specific merger events and programs of acquisition activity. Through a sample of conglomerate acquirers, they find significantly positive abnormal returns associated with the announcement of acquisitions programs and significantly negative returns associated with certain institutional laws.
In another type of studies published in the volume 11 of the "Journal of Financial Economics", Asquith, Bruner and Mullins [4] examines the effect of mergers on the wealth of bidding firms' shareholders. They clarifies some interesting points about acquisitions activity. They feel that the bidding firms gain significantly during the twenty-one days leading to the announcement of each of their first operation within four merger bids. These results have not a support to the capitalization hypothesis that acquirers' gains are capitalized at the beginning of merger programs.
In the same volume of the journal reserved to the wealth effects of mergers and the market for corporate control, Malatesta [9] examines the net effects of the long-run sequence of events leading to merger, and of merger per se, on shareholder wealth. The author find that the long-run wealth effect of the event sequence culminating in merger is significantly negative for acquiring firms.
This study examines the profitability of corporates acquisition programs by distinguishing between types of effects. Specifically, an acquisition program has an announcement effect and an economic impact for each acquisition attempt made by the acquiring firm. The study determines the value of the both effects and the abnormal returns for the acquisition attempts on the non-event periods, for the total sample of big and SME French companies. The results shows that for the total sample which account 46 French acquirers, acquisition programs are losing. However, by considering the subsamples of the study, we find that the French SME perform better than, the big companies, and create value expressed in positive returns generated from their programs.
The next section explores theoretical considerations about the hypothesis studied within the framework of the estimation of corporates acquisition programs. We describe the methodology used to evaluate reaction to acquisition programs announced by French acquirers. In this section, we present the mathematical development of the model that we apply on a sample of French acquirers in order to determine if corporates acquisition programs at the announcement date have a positive net present value or no. We reserve the following section to discuss the results concerning the specificity of French small and medium-sized enterprises, which are included in the initial sample. The final section summarizes the results of the study.

Theoretical Considerations on Acquisition Programs
The corporates acquisition programs were little handled by researchers because they are a very specific subject in finance. Indeed, directors must publicly announce the acquisition programs in order to be able to capitalize the gains, which are associated to it, at the time of their launch. The public announcement of this type of information, which is the launch of acquisitions programs, is very important for shareholders of acquiring firms. It allows their enrichment by the increase in share prices [10].
In this case, the very important assumption implied that acquisitions operations as mergers and tender offers are comprised partly of individual events undertaken within structured acquisitions programs. However, the empirical studies shows that the results of acquiring firms are expected by acquirers to gain from acquisitions activity, although some of the evidence is without supports [8].
The effect of an acquisition program must be profitable to the acquiring firm and to these shareholders at the same moment of the launch of said program. For example, if a firm were to announce an event concerning the structure of its capital, by then the reaction of its shares should take place at the announcement of the event, not at any later date [8].
There is a difference between individual acquisitions and programs of acquisition activity. Generally, for individual acquisition operations, the methodology used in the estimation of the returns of the shareholders and the variation in the firm value of the both target and bidding companies is the "event study methodology". Fama, Fisher, Jensen and Roll (1969) to call back what they said, we would refer to an article [7]. These researchers developed this technique in order to capture the effect of an event on stock prices by calculating the abnormal returns caused by this event. Studies that have examined the gains of mergers and tender offers applies the method based on the estimation of abnormal stock returns at the time of the specific event and surrounding this event date.
One of the preliminary hypothesis on which is based the evaluation of an acquisition program is relative to the valuemaximizing behavior [10]. Within the same framework, several hypothesis have been put forward in empirical studies, which estimate the value of the acquisition programs. In Malatesta and Thompson [8], the capitalization hypothesis were treated and they find that is consistent with constant announcement effects for successive acquisition attempts. The net present value of all the acquisition attempts made by the acquiring firm within the program is fully capitalized in firm value at the time of the initiation of the program. The frequency of acquisition attempts influences the net present value of the program, also the expected economic impact imputed to future attempts raise the acquisition program value, and the fixed costs are attributed to the program.
The estimation of the capitalized value of acquisitions programs is very important for two reasons [10]. First, as Malatesta [9] have pointed out, the fixed costs of an acquisition program will not affect the price reaction around an acquisition event. Further, if acquisition program requires any initial expenses, price reactions to individual event announcement as take-over or merger could indicate a positive return to those expenses. For that purpose, the price reaction to an acquisitions program will reflect the cost of the initial outlay as well as the expected return on it. The second and equally important reason is that once the expected value of an acquisitions program is capitalized, variations in the value of an acquiring firm surrounding event announcements will reflect only the surprise associated with the terms of the individual event. This surprise is measured relative to the expectations, which were capitalized at the time the acquisitions program was publicly announced.
The two main studies of the profitability of corporates acquisition programs within the framework of M&A operations are the ones of Shipper and Thompson [10] and, Malatesta, and Thompon [8]. The results of the first study indicate that the acquiring firms anticipate gains from acquisitions activity. The second study approve this result. Indeed, Malatesta and Thompson [8] find that the average estimated economic impact of an acquisition attempt exceeds 4 million dollars, but they cannot conclude that an acquisition program is desirable for firms in general.

Analysis of French Acquisition Programs Profitability
Our questions relative to the profitability of the corporates acquisition programs were studied in the context of French M&A operations realized between 1997 and 2007. All researchers who have studied the announcement returns to shareholders of bidding firms reveal that bidder (acquirer) shareholders gain when buying a private firm or subsidiary but get almost nothing or lose when buying a public firm [6]. In the continuation of this idea, we try to verify the profitability of the French bidders concerning the corporate acquisition programs. In this section, we describes our sample of firms involved in a process of acquisition operations in series called acquisition program, then we present the empirical methodology used and hypothesis. In the third part of this section, we present the results for the entire sample of 46 acquiring firms.

Data
Forty six (46) French companies constitute the empirical framework of our study. They are active firms on the French acquisition market; they are known to be frequent companies regarding corporates acquisition. The acquiring firms sample analyzed in this study comes primarily from the AMF [12] files, which, are on its Web site.
The constitution of the final sample of the study begins with the inventory of all the acquisition operations made on the French market during the observation period (1997-2011), the French acquirers made 1050 acquisition operations. These operations concern only four types of acquisition, which are takeover bid, public offer of exchange, mixed public offer and merger.
The basic selection criterion is concerning the number of acquisition attempts made by the acquirer. Every acquiring firm having made at least an acquisition attempt during the eleven years of the study period is held in the sample. We strictly apply this selection criterion because it is impossible in this case (French case) to verify the real public announcement of the acquisition programs in the financial press. Two firms are concerned in every acquisition operation, the acquiring firm and the acquired (target) firm. Our empirical analysis touches only the French acquiring firms. In final our sample study covers over 11 years, from 1997 to 2007 and they contains 46 French frequently acquirers. These firms represents practically all the economic sectors of France. The entire 46 firm sample, which announced and subsequently carried out acquisition programs launched 97 acquisition operations as shown in table 1. They represents on average 2.11 attempts by firm during the study period. This represents 95% of all the acquisition attempts counted during observation period (1997-2011) as it shown in table 2 and figure 1. We consider the acquisition operations made by the acquiring firms as acquisition attempts since they can be successful or unsuccessful.
The stocks, which make up this sample, represent almost all the branches of industry in France. Nevertheless, the sector of the real estate stays the most dominant sector accounting 12 securities. Among which of the stocks which have a great fame in the French economy, as for example: Vinci, Saint-Gobain, Bouygues and Lafarge.
All the financial and stock exchange data concerning the 46 French acquirers about which we evaluate the profitability of their acquisition programs are obtained from Datastream database. This data ranges from stock price, dividend, CAC 40 stock index, market value and FRANCE TREASURY BILL 1 MONTH for interest rate. These data is gathered every month and we calculate the zero beta portfolio return after having calculate the monthly stock return for each firm. We use the zero beta return to calculate the stock and market risk premium and it instead of interest rate.
The sample of departure contained 185 companies; we studied 91 companies considered as being frequent on the corporates acquisition market. From this group we kept only 46. We eliminated from the sample of study the firms, which does not satisfy selection criteria, they are among 45. The main cause of the elimination is the lack of data, what represents 23 companies (25%) of the number of studied companies.
In table 1, we present the distribution of the acquisition attempts made by 46 companies making up our sample of study, during the observation period, which begins on 01/01/1997 and ends 31/12/2011.
The last column, presents us the sum of all the acquisition attempts according to four types of acquisition, realized every year by the set of the companies composing the sample. The number of the attempts during the last 4 years of observation reveals a low rate of acquisition. Indeed, between 2008 and 2011, the set of the firms of the sample made only 5 attempts (4.90%).

Empirical Methodology and Assumptions
The fundamental idea according to which Malatesta and Thompson (1985) have developed their econometric model for assessing profitability of an acquisition program is diverted from estimation of a classical investment. In the launch of an acquisition program, the acquiring firm has fixed costs that she hope to cover with future cash flows generated by their acquisition attempts. Successive acquisitions also increase firm size and diversity and thus may generate additional integration costs [1]. Considering the various types of cash flows, the net present value of the acquisition program for an acquiring firm , is given by .
(1) Where denotes the program's fixed cost that contains all kinds of expenses in order to prolong the program. In addition, for each individual future attempt, the economic impact is indicated by . Acquisition attempts within the program have a constant frequency per period, the risk free interest rate equal to .
The approach that I use in our empirical study in order to determines if French acquisitions programs are profitable or not begin by the estimation of the stock rate of return for each firm within the sample and the market return (2) where is the realized rate of return to stock at time , ! " are the stock price as they are taken from Datastream database at time ! 1 successively, and $ represents the dividend on the stock at time .
in which % is the return on the market at time , and + ,-,./ , + ,-,./ are the stock index CAC 40 (Paris Stock Index) at time ! 1 successively. These two calculations allow us to begin the estimation of the multiple regression coefficients of the principles four models. The basic model serving to verify the hypotheses of the study is in the form of a multiple regression with two independent variables, the first one is the market risk premium weighted by firm value " at the time 1. The second variable is a binary,! equal to one for a firm , which makes an acquisition attempt during period and zero otherwise 1 ̃ 3 " 4 5 6 " 7 ̃% 8 5 9 ! 5 :̃ (4) where 1 ̃ 3 " is the stock risk premium weighted by the firm value expressed by the market capitalization at the time 1. The market risk premium " 7 ̃% 8 is one of the two independent variables of the regression equation; this one also is weighted by the firm value. The second independent (explanatory) variable of the multiple regression is the binary variable ! , which is equal to one if the firm take an acquisition attempt and zero when the firm does not take attempt during given period.
The main objective of our study is to estimate the gains or the losses of the French corporates acquisition programs, which spread out over several years. This evaluation is completed by the calculation of three coefficients 4 , 9 ! 4 + 9 of the eq. (1). The analysis on which is based our empirical study is applied through four regressions, among which the three coefficients of two of them are expressed in excess euro returns, and the two other regressions are in excess rate of return form, as it is shown below 1 ̃ − 3 " = 4 + 6 " 7 ̃% − 8 + 9 ! + :̃ (5) 1 ̃ − 3 = 4 + 6 7 ̃% − 8 + 9 ! + :̃ (6) 1 ̃ − ̃; 3 " = 4 + 6 " 7 ̃% − ̃; 8 + 9 ! + :̃ (7) 1 ̃ − ̃; 3 = 4 + 6 7 ̃% − ̃; 8 + 9 ! + :̃ Both equations in which appear the variable defining the firm value " are the ones which give the results in excess euro returns, and the two others which are without this variable, their results are in excess rate of returns. Another specification concern both risk premium. Two equations eq (5) and eq. (6) are expressed in interest rate and the two others eq. (7) and eq. (8) use the zero beta return to calculate both risk premium, that of the stock and that of the market.
We treat three hypotheses in this study. The first one concern the variations of the abnormal returns that should be negative in non-event period. The second one predicts that the expected economic impact is positive and the last one it is about the expected announcement effect that should be positive. An overview on these hypotheses is presented in table 3 below Table 3. Hypotheses of the empirical study.

Hypothesis
Coefficient Coefficient sign Definition of the hypotheses Expected abnormal return of an acquisition attempt for a non-event period < @ : 9 = > 0 9 (+) Economic impact of an acquisition attempt < B : 4 + 9 = 11 − 3 > 0 4 + 9 (+) Announcement effect of an acquisition attempt According to the event study methodology, the abnormal return is defined as the return measured at the time of the occurrence of an event in the company. The abnormal return is defines as the difference between the normal (observed) return and the theoretical return which represents the return that should have taken place and occur in the absence of the event. We can measure the normal return by means of a model of asset pricing such as the market model (Single Index Market Model) or CAPM (Capital Asset Pricing Model).
We do the measure of corporate acquisition programs profitability consisted mainly of French takeovers and mergers by using the partially anticipated events model and our principle hypothesis concerning the French case is that the acquisition programs are profitable for French investors. This hypothesis is specific to the French acquisition market.
However, there are three other hypotheses attached to the model that we applied to the sample of French companies to determine if the acquisition programs are profitable or not. The first hypothesis < " implies that the non-event period is the non-announcement period this means that is the time when the acquirer do not launch acquisition attempt. The meaning of this first hypothesis is that the expected abnormal value change in non-announcement periods is negative. In every period, the frequency of acquisition attempt is positive and because the expected economic impact of an acquisition attempt is positive, then the variation of the abnormal returns in non-announcement periods 4 must be negative. In the second hypothesis < @ , since 9 is equal to it follows that 9 that measure the economic impact must be positive. The economic impact is defined as the capitalized value of future net cash flows resulting from the event's occurrence. It is the net present value of the event. The announcement effect of acquisition attempts is tested by the hypothesis < B , which states that the announcement effect must be positive if the expected economic impact is positive. The announcement effect represents the change in firm value attributable to the resolution about the uncertainty of the event occurrence.

Empirical Results
For each of 46 companies constituting our sample, the parameters of every model 1, 2, 3 and 4 are considered by the ordinary least squares approach (OLS) applied to 132 months of study period (01/01/1997 to 31/12/2007). Table 3 reports the main results of our study. In this table, I present for the four models, the average estimation of the coefficients, the average of t-student test and the significant percent different from zero of the coefficients. In the panel A, the model 1 test the first hypothesis that 4 < 0, I find that only one case on 46, this result was obtained and which is significant at the level of 0.05. This unique negative alpha is the one of the firm Unibel, an SME company specializes in the food industry. On the other hand, in 32 cases we reject the null hypothesis in favor of the alternative 4 > 0 for a significance level of 0.1. The number of non-significant cases is 13 cases on 46 studied what represents approximately 28% of the total sample.
The average of 4 C for the entire sample is 169.033 million € what is against the hypothesis which predicts a negative sign. The explanation of this result according to which the French financial market is not efficient is not plausible, but there are other factors, which can explain this, as for example the competitiveness in the French corporates acquisition or what the launch of acquisition programs is expensive to the French acquiring firms. The results given in panels B, C and D confirm the result of the basic model shown in panel A for both types, results expressed in excess euro returns or results in excess rate of return form.

The Specificity of French Small and Medium-Sized Enterprises
All the results on the profitability of the French corporates acquisition programs presented above are relative to the entire sample of study. Now, to seize well the firms, which are more successful than others are, we distributed the 46 companies making up our sample study on subsamples. This work allowed us to constitute eight subsamples of industry: real estate, industrial services, chemistry and oil, technology, banks, media, distribution and health. For example, in the sector of the real estate, we have Vinci, Bouygues and especially Saint-Gobain which is a leader in his domain and very successful. Note that, in our sample there is also a small and medium enterprises "SME" as by example in the subsample of the Healthcare sector. This sub-sample consists of two companies only, Sartorius and Guerbet. For each firm the rates of return is indicated by ̃ and the market return by ̃% . Riskless interest rate is noted by , and in some cases, this interest rate is replaced by the zero beta portfolio return denoted by ̃; . The binary variable ! take two value: (1) if firm made an acquisition attempt during period study and (0) otherwise. The value of the firm at period − 1 is expressed by " . For each of 46 acquiring firms, estimates of 4, 9 ! 4 + 9 are based on 132 monthly periods observation ranged from 1/1/1997 through 31/12/2007, while 4 measures the acquisition attempts abnormal return on non-event periods, 9 measures the economic impact of acquisition attempts and 4 + 9 measures the announcement effect.
Indeed, for this sector and for both firms together the sign of both coefficients 9 ! 4 + 9 is positive. This is in line with the signs waited in both hypotheses, that of the economic impact and the announcement effect. The value of these two parameters, are estimated at 3.57 and 5.28 million €. Thus, there is at least one sector among eight branches of industry, which verified both hypotheses < @ ! < B . This is also an obvious result of the performance of the French EMS within the framework of French corporates acquisition programs.
Guerbet and Sartorius, together has significant results and their acquisition programs are slightly profitable but their financial performance are better than other big firms. Sartorius is listed at the SBF 120 stock index and Guerbet at the CAC SMALL 90 index.

Summary and Conclusions
For more than twenty years, an intensive debate has ensued about acquirers' motivations in mergers and acquisitions (M&A). This is most likely due to early empirical results showing that acquirers' cumulative abnormal returns (CAR) around the announcement date are at best equal to zero or, worse, even negative [7]. Why would firms undertake acquisitions if not to create value? [2].
In this theoretical framework, we try to analyze acquisitions programs, launched by frequently French acquiring firms from 1997 to 2007. We find that only the announcement effect hypothesis appears to apply to French acquisitions programs. Indeed, our results are a little limited by the weakness of the frequency of the acquisition attempts launched by the French acquirers.
In a similar study, Asquith, Bruner and Mullins [4], measure the daily abnormal returns by use of stock returns delivered by CRSP (Center for Research on Security Prices). The authors find a Cumulative Average Abnormal Return (CAAR) equal to 2.80%, on a duration of 21 days, 20 days before until the day of the announcement of the event, which is a merger operation. The comparable result of Malatesta and Thompson [8] is 2.91%, this is the average estimated announcement effect. For our study, we also obtain this estimation, which is equal to 1.40%.
The three rates of return for the announcement effect are close even if the current study result is a little lower than those of the two other researches. Nevertheless, we note that in our case, our result, which we had, has the same sign as the other researches, what comforts us and lets us think that the empirical work was suitably achieved Finally, proved evidence through the four regressions results supports partially the hypothesis that acquisition programs can be perceived as profitable investment projects for the French acquirers essentially those of "SME". This conclusion is justified by the individual results of firms. Otherwise, the sample which consists of 46 French companies considered as frequent acquirers on French acquisition market indicate clearly that firm's acquisition programs during the last decade were destructive of value. It means they do not maximize the value of the firm nor the stockholder's wealth.