University of Amsterdam, Faculty Economics and Business BSc Finance and Organization, Finance track
Bachelor Thesis
Relevance of M&A deals in the pharmaceutical industry
Pharmaceutical industry structure analysis on the cumulative abnormal return of the acquiring company around the announcement of a M&A: Is there value creation for the acquiring pharmaceutical company around the announcement of a M&A deal and which factors influence the cumulative abnormal return?
Robert Meij, 10015469 June 2015 Dr. J.E. Ligterink
2 Abstract
This study examines the short-term performance of 60 pharmaceutical acquisitions from 1998 to 2014 and find evidence that on average the acquirers realize neutral returns. These returns are correlated with the method of payment stock and the size of the acquiring company. The returns are not correlated with the information on the research and development activities from the target firm and have no significant effect from the mortgage crisis. A index of 14 pharmaceutical companies is created to determine the movement in internal research and development activities from 1994 to 2014. This study finds that the pharmaceutical
companies experience declining internal productivity and are outsourcing their research en development activities to secure their pipelines.
Keywords: Acquisitions, R&D Capital, Pharmaceutical Industry, Cumulative Abnormal Return
3 Contents
1 Introduction ... 5
2 Literature Review... 8
2.1 Pharmaceutical Industry: Investment in R&D Capital or M&A’s ... 8
2.2 General Motives for M&A Deals ... 9
2.3 Previous Research on Short-term Returns for the Shareholders of Acquiring Company ... 10
2.4 Deal Characteristics Influencing ’s of Acquiring Pharmaceutical Companies ... 11
2.5 Hypothesis Following from Literature ... 12
3 Data ... 15
3.1 Design Sample... 15
3.2 Measuring Relative Deal Size Value ... 16
3.3 Data Sources ... 16
3.4.1 Subsample 1: Method of Payment ... 18
3.4.2 Subsample 2: Mortgage Crisis ... 18
3.5 Table Statistics of Subsamples ... 19
4 Methodology ... 20
4.1 Benchmark ... 20
4.2 Calculating Short-term Abnormal Stock Return ... 21
4.3 OLS regression on ... 22
4.4 Calculating the Significance of the ... 23
4.5 Testing for Difference in Means Between Two Subsamples ... 24
5 Results ... 25
5.1 Results from Testing the H1: > 0 ... 25
5.1.1 Subsample 1: Method of Payment ... 26
5.1.2 Subsample 2: Mortgage Crisis ... 27
5.1.3 Testing for difference ’s in Subsamples ... 28
5.2 Results from the OLS-regression on with determinants ... 29
6 Conclusion ... 31
7 Discussion ... 33
8 References ... 34
4 9.1.2 Movement R&D Capital in Pharmaceutical Industry for Acquiring Companies .... 36 9.1.3 R&D Expenditure Target Firms as Percentage of Market Capitalization ... 38 9.2 Covariance tables ... 39
5 1 Introduction
Existing literature is divided about the value-effects from mergers and acquisitions (M&A) on the acquiring company. Bruner (2002) states that it is popular to say that M&A deals have a negative value effect for the acquiring firm. Goergen & Renneboog (2004) report that 11 of the 20 studies described in their article find small negative returns for the acquiring
companies around the announcement of a M&A deal. Furthermore Andrade, Mitchell and Stafford (2001) have reported that the shareholders of the target firm earn significant positive abnormal returns around the announcement of a M&A. The authors state furthermore that the shareholders of the acquiring company earn significant negative abnormal returns around the announcement of a M&A, but that the target firm and acquiring firm both create a significant combined positive wealth effect. This study will investigate if a acquiring firm in the
pharmaceutical industry also has negative value effects around the announcement of a M&A. These significant combined positive wealth effects are a reason, for by example firms in the pharmaceutical industry, to apply a M&A. With tinning pipelines and the constant search for innovation, the pharmaceutical industry is a competitive industry. In the last 10 years this has lead to several M&A’s within the pharmaceutical industry. Furthermore M&A deals are popular in the pharmaceutical industry. For example: Hornke and Mandewirth (2010) state that the global M&A transaction volume fell from 4200 billion USD in 2007 to 1300 billion dollar in 2009. They state that the pharmaceutical industry on the other hand had a strong M&A year in 2009, for example: Pfizer acquiring Wyeth for 65 billion USD.
According to Shahrur (2005) M&A’s are beneficial due to increased synergies. The author supports the findings of Andrade, Mitchell, & Stafford (2001) and states that there are no short-term positive value effects for the acquiring company and its shareholders. This study looks at the short-term value effect, because according to the market efficiency test, there will be no abnormal return in the long run, because all the information is then incorporated in the stock price.
In 2010 Astellas Pharma acquired OSI with a deal value of 400 million USD. The cumulative abnormal return for the (1,+1) event window for Astellas Pharma amounted -1.9%. This value is supported by existing literature that state that M&A’s have a negative cumulative abnormal return for the acquiring company. The research question is therefore: Is there value creation for the acquiring pharmaceutical company around the announcement of a M&A and which factors influence the cumulative abnormal return? For example the method of payment could be an important factor influencing the cumulative abnormal return.
6 Myers and Majluf (1984) state that firms use a stock payment if the shares are overvalued and use cash payment when the shares are undervalued. This means that a stock payment implies a negative cumulative abnormal return, because the market sees it as a signal that the shares are overvalued.
This study will investigate M&A deals financed with cash or shares within the sample. The M&A deals will also be analyzed according to deal size relative to the value of the
acquirer (based on market capitalization), before or after the mortgage crisis and the
percentage R&D capital according to market capitalization. The last determinant R&D capital is important when analyzing the cumulative abnormal return within this study. In the
literature review and appendix the movements from R&D capital within the pharmaceutical industry will be discussed and analyzed.
Furthermore this study will gather empirical evidence to answer the question if there is a significant cumulative abnormal return for the acquiring firm around the announcement of a M&A in the pharmaceutical industry. This study analyses the cumulative abnormal returns for the acquiring firm around the announcement of a M&A with a sample of 60 M&A’s. The data and information about the M&A deals is obtained from Compustat WRDS and Zephyr. Other investigation have shown that negative cumulative abnormal returns can be the result of benchmark errors, therefore this study takes a sample of 100% acquisitions within the American pharmaceutical industry (deal value is a minimum of 350 million).
To calculate the abnormal returns this study uses the ‘Event Study’ method described by Binder (1998). According to MacKinlay (1997) an event study is useful when using financial market data. The article states the effect of an event is reflected immediately in the stock price. The reason for taking a sample period from 1994 to 2014 is to measure the impact from the mortgage crisis, from 2007 to 2008. Hornke and Mandewirth (2010) state that the mortgage crisis is still influencing the pharmaceutical industry. They state that a combination of demographic increasing health costs and national budget deficits results in minimizing the costs of the national health system. Because of regulatory means, the pharmaceutical industry product pricing is therefore decreasing. Hornke and Mandewirth state furthermore that the drug pipelines are low and many blockbuster drugs will soon disappear as revenue for the innovative pharmaceutical companies. They state that the pressure is high on the innovative pharmaceutical companies to create innovative drugs and they need to increase their efficiency according to synergies (with for example M&A’s) and R&D expenditure.
7 The next section will review the literature according to M&A’s and will develop the hypothesis. Thereafter section three will describe the criteria for the data selection and the sample. In the fourth section the methodology used in this thesis will be discussed. In section five, the results around the announcement of a M&A for the acquiring pharmaceutical
company will be analyzed. At last section six will provide the conclusions for the above described research question.
8 2 Literature Review
This section provides an overview from the existing literature considering the abnormal stock performance around the announcement of a M&A for the acquiring firm. First, paragraph 2.1 will describe the tradeoff between investing in R&D capital or M&A’s for companies in the pharmaceutical industry. Thereafter, paragraph 2.2 discusses some general motives behind M&A’s. After that Paragraph 2.3 discusses other short-term event studies on the effects of M&A’s. In paragraph 2.4 describes the factors influencing the cumulative abnormal returns ( ’s). Finally paragraph 2.5 discusses the hypothesis following from the literature.
2.1 Pharmaceutical Industry: Investment in R&D Capital or M&A’s
According to Higgins and Rodriguez (2006) the development of new drugs in the pharmaceutical industry has declined since the end of the 90ies, because more exclusive drugs fell of patent than got replaced by new drugs. The authors state that R&D expenditure for US pharmaceutical companies grew from 6,8 billion USD in 1990 to 21,3 billion USD in 2000. This is due to the fact that the cost of a developing a new drug grew from 231 million USD in 1987 to 802 million USD in 2000 (Higgins & Rodriguez, 2006; cited in DiMasi, 2001). The increasing R&D investment in the pharmaceutical doesn’t increase productivity of new developed drugs. According to Higgins and Rodriguez (2006) pharmaceutical companies can enhance their R&D investment when acquiring other pharmaceutical companies. The authors state that the outsourcing of R&D investment could have an impact on R&D
productivity. Gilbert, Henske and Singh (2003) state that the pharmaceutical industry needs a new approach. The authors describe that the pharmaceutical companies need to focus on investing in R&D capital and not in M&A deals. The article states M&A deals will not improve returns.
On the other hand Joglekar and Paterson (1985) state that the R&D capital in the pharmaceutical industry needs to be highly profitable to bear the risk when creating an innovative drug. They state that creating an innovative drug requires on average 12 years of R&D expense and the drug has a patent of 10 years. The authors state that a investing in R&D capital will not always provide a high return for the innovative pharmaceutical companies.
Furthermore DiMasi, Hansen, Grabowski and Lasagna (1991) state that product innovation in the pharmaceutical industry is risky and R&D capital represents a high
9 percentage of the book value. The authors state that R&D capital is important when analyzing the return on R&D investment. DiMasi, Hansen, Grabowski and Lasagna state furthermore that R&D capital influences the international competiveness and the economic performance of the companies in pharmaceutical industry. They state that the higher R&D costs and risks are one of the main factors underlying the trend towards more M&A’s in the pharmaceutical industry.
According to Hornke and Mandewirth (2010) pharmaceutical companies are focusing on creating synergies and ‘purchase of drug pipelines’ when acquiring another
pharmaceutical firm. The M&A deals like Merck/Serono (2006 for 10,1 billion USD),
Merck/Millipore (2010 for 5.4 billion USD) and Astellas Pharma/OSI pharmaceuticals (2010 4 billion USD) are acquisitions of small biotech companies. The authors describe the main reason for the acquiring of this biotech companies as the financial need for product
development. The next section describes the general motives behind M&A deals.
2.2 General Motives for M&A Deals
There are many reasons why M&A’s occur and economic theory has several explanations. For example: Devos et al. (2009) describes three general motives: diversification, increasing market power and synergies. A successful synergy creates efficiency for the firm, which is supported by Berk & Demarzo (2007). The synergies can be separated in two parts:
economies of scope and economies of scale. Companies could for example combine overhead costs, which is an example of economies of scope. An example to create economies of scale is to make more efficient use of the production capacity and produce at a higher volume. When trying to create market power after a M&A deal, firms form a monopoly or oligopoly and create extra value.
A M&A could also occur if a manager engages in empire-building, this could lead to a misalignment between the shareholders and the managers of the acquiring firm. These misalignments are described as managerial hubris (Chen, Roll, & Ross, 1986). The author describes this as a possible reason for a negative for the acquiring company around a M&A. Another misalignment between the manager and the shareholders is defined as the ‘free cash flow theory’ described in Jensen (1986). The author describes the free cash flow-theory as the excess of what is required to fund the projects, which have a positive net present value. He states that the manager has more interest in investing in more projects instead of
10 paying out dividend and repurchase stock. The free cash flow gives the manager an incentive to grow the firm beyond the optimal size and this lead to a downward market price for the acquiring company after the announcement of a M&A.
Managers will apply a M&A to improve the performance of the acquiring firm, but that will not always lead to positive effects for the shareholders. Therefore the next section will provide more information about the short-term effects around the announcement of a M&A deal.
2.3 Previous Research on Short-term Returns for the Shareholders of Acquiring Company
To make a statistical analysis whether M&A’s create ore destroy value for the acquiring pharmaceutical company, this study uses the event study. According to Bruner (2002) this give insight about market based returns for target firm shareholders, buyers and a
combination of the two. In 11 of the 20 studies described by Bruner, the acquiring firm gains a significantly negative around the announcement of a M&A deal. This is based on empirical evidence, where the average abnormal stock market reaction around the
announcement of a M&A deal is provided.
The around the announcement of a M&A is the real return less the benchmark return. This study defines the value effects according to the measurement methods provided by Bruner (2002), which defines the value of the acquiring firm after a M&A in three categories:
Value conserved: The investment return is equal to the required return, so the M&A
investment has a net present value (NPV) of zero and the investors gets his required return, which he would also get without a M&A.
Value created: When the returns on investment exceed the required return the net present value becomes bigger than zero. Because of the competition in the market this event is rarely to occur.
Value destroyed: When the returns on investment are less than the required return the net present value is less than required, for investors this is a negative value event.
11 When the value is determined, according to Bruner it is important to examine the combined net economic combined gain from a M&A for the target and bidder firm. The author states that the biggest challenge lies in measuring the net economic gain from a M&A. This is caused the difference in size between the buyer and the seller. The buyer is substantially larger than the target firm and a large percentage gain for the target firm can be offset by a small percentage loss by the acquiring firm. The article uses the findings of 20 studies where a combined portfolio of buyer and target firms are examined by measuring the absolute dollar value of returns. In most of the studies described in Bruner a positive significant combined return was found. This means that overall there is value creation after a M&A for the buyer and target firm.
According to Andrade, Mitchell, & Stafford (2001) the target firms are the winners of a M&A, with an average abnormal return of 16% within the event window. Furthermore the authors found a combined positive abnormal return from 1.8% for 3688 mergers during the period from 1973 to 1998. This suggests that M&A’s create value for the shareholders and the acquiring firm. The change in percentage could be caused by the difference in firm value according ore a negative abnormal return for the acquiring company during the event window. The article found that on average the acquiring firm has a negative cumulative abnormal return of 0.7% during the tested event window of a M&A.
2.4 Deal Characteristics Influencing ’s of Acquiring Pharmaceutical Companies To examine which factors influencing the from the acquiring pharmaceutical
companies this study investigates relative deal size value, influence of mortgage crisis, R&D capital and the method of payment (cash or stock). Existing literature suggests that payment with stock is costly and that paying with cash is neutral (Bruner, 2002). The results from existing literature find that stock based deals are associated with significantly negative returns at the deal announcement (Bruner, 2002; cited in Asquith, Bruner, & Mullins, 1987). They find that the announcement of the payment with shares could be taken as a signal that managers believe that the firm’s shares are overpriced.
Furthermore Fuller, Netter and Stegemoller (2002) found a significant negative for the acquiring firm around the announcement of a M&A, when the relative deal value (deal value as percentage of acquiring firm) is small. The authors state that when the
12 relative deal value is large, the for the acquiring firm around the announcement of a M&A becomes positive.
According to Hornke and Mandewirth (2010) the mortgage crisis is still influencing the pharmaceutical industry. The combination of increasing health costs and national budget deficits results in minimizing the costs of the national health system. Therefore the
pharmaceutical has because of regulatory means a decreasing price level within the industry. As described in paragraph 2.1 there is are doubts within the pharmaceutical industry whether to invest in R&D capital or to invest in M&A deals. In the appendix section the movement in R&D capital for a sample of pharmaceutical companies is described. Furthermore this section tries to provide some potential explanations what the relation between R&D intensity from the target firm and the short-term effect of for the acquiring company is.
2.5 Hypothesis Following from Literature
In this section the academic literature on M&A’s is used to develop the hypothesis of this thesis on the stock performance of the acquiring firm around the announcement of a M&A. The second hypothesis is based on determining the effects of the deal characteristics influencing the abnormal returns. The determinants to be tested are deal size, method of payment, R&D capital and the influence of the mortgage crisis.
The above-described literature suggests that the market based return of the acquiring firms show significant negative ’s around the announcement of a M&A (Bruner, 2002) and value creation after a M&A is unsecure for the acquiring company. According to Hornke and Mandewirth (2010) M&A deals in the pharmaceutical industry are also applied to create new products and therefore value. For example: Astellas Pharma acquired OSI
Pharmaceuticals for the molecule Enzulatamide when creating their new drug Xtandi (see appendix). After the launch of Xtandi the stock price grew 1 Jan. 2012 from 30,9 USD to 42,47 USD at 1 Jan. 2014. This in an increase of 37,44% in two years. If investors know what is in the pipeline of the target firm, they will respond on the announcement of a M&A. To answer the question if M&A’s create value for the acquiring pharmaceutical company the following hypothesis will be tested.
13 H1: The cumulative average abnormal return ( for the acquiring pharmaceutical firm will have a significant positive value around the announcement of a M&A deal
( .
The above-described literature suggests that the cumulative average abnormal return is influenced by the mortgage crisis and method of payment. Therefore, this study will divide the sample into two subsamples and test if there is a difference in means for the subsample before the mortgage crisis and after the mortgage crisis. Hornke and Mandewirth (2010) state that unless the mortgage crisis, the pharmaceutical industry still had a strong year in 2009. This study will test if the ’s differ before and after the mortgage crisis. This study will expect that the M&A deals financed with cash will be more value creating than the deals financed with stock.
Based on hypothesis 1, this gives the following sub hypothesis:
H1: - 0 and
H1: - > 0
This study uses the CAPM model, described in section 3, to examine whether the acquiring firm gains from a M&A. The firms tested in this study are matched on the basis of size and industry. The above-described hypothesis will try to reject the hypothesis described by Mueller (Bruner, 2002; cited in Mueller, 1984). He states that the ’s of the acquiring firms are significantly negative around the announcement of a M&A (H1: ).
The second part of this study will examine whether which determinants influence the ’s around the announcement of a M&A deal for the acquiring pharmaceutical firm. The influence of relative deal size, method of payment and market capitalization will be used as control variables in the model described in section 4.3. The influence of the mortgage crisis, R&D expenditure from the target firm and will be tested on the of the acquiring firm around the announcement of a M&A. Given the previous described literature, this study will expect that the mortgage crisis has influence on the of the acquiring firm around the announcement of a M&A. The second hypothesis in this thesis will be determined as follows:
14 H2: The from the acquiring pharmaceutical company around the announcement of a M&A is influenced by the mortgage crisis.
The above described literature expects that the method of payment is influencing the The third and fourth hypothesis of this study to be tested in this study is as follows:
H3: The from the acquiring pharmaceutical company around the announcement of a M&A is negatively influenced by method of payment stock.
H4: The from the acquiring pharmaceutical company around the announcement of a M&A is positively influenced by method of payment cash.
Furthermore this study will investigate if the R&D expenditure from the target firm is influencing the of the acquiring firm around the announcement of a M&A. Given the literature described regarding M&A deals in the pharmaceutical industry to create innovative drugs, the fifth hypothesis to be tested is:
H5: The from the acquiring pharmaceutical company around the announcement of a M&A is positively influenced by R&D intensity of the target firm.
15 3 Data
This section will give insights in the data set used to test the hypothesis in the previous section. First, the sample design will be provided and after that the different data sets used in this study will be discussed.
3.1 Design Sample
To answer the research question, this study will investigate M&A’s during a period from 1 Jan. 1998 until 31 Dec. 2014. The sample consists of M&A deals meeting the following restrictions:
1. The acquiring firm is a pharmaceutical firm 2. The target is a pharmaceutical firm
3. The deal value has to be at least three-hundred million dollar 4. The target is a public firm
5. The acquiring firms are listed on the Nasdaq and NYSE stock exchanges with daily and monthly stock prices provided by CRSP.
6. The buyer and target firm manufacture basic pharmaceutical products and pharmaceutical preparations.
7. The deals are included in the sample when the acquiring firm acquires at least 100% of the target’s stock after the M&A.
8. The firms are active on the US stock market. Only American based pharmaceutical companies are included in the sample.
9. Method of payment is cash ore with shares
10. R&D intensity for the target firm is included in the model
The next sections will provide more information about how relative deal size is defined and describes the data sources used in this study.
16 3.2 Measuring Relative Deal Size Value
To examine the effects of the deal value on the profitability of M&A deals, this study
measures the deal value as percentage of the market capitalization of the acquiring firm. The deal values are provided by the Zephyr database and the market capitalization is calculated with monthly stock data provided by Compustat WRDS. The formula for relative deal size is provided below:
Relative deal size = deal value / market capitalization of acquiring firm.
3.3 Data Sources
This study uses the Zephyr database to identify the pharmaceutical M&A deals. To obtain monthly and daily stock prices the CSRP database is used. The acquirers monthly and daily stock prices are downloaded and converted into Microsoft Excel. There they are used to calculate the market capitalization and the daily and monthly returns. The results of the restrictions used in the Zephyr database are a sample of 60 pharmaceutical M&A deals. In the table below a quick overview from the M&A deals used in the sample is provided. On the vertical-axis the amount of M&A’s in the corresponding year (horizontal-axis) is displayed.
The result of the table shows that large M&A deals (deal value > 300 million USD) in the pharmaceutical industry have increased since 1998. The most M&A deals used in this sample
0 1 2 3 4 5 6 7 8
Mergers & Acquisitions
17 were in 2006, one year before the mortgage crisis. At the beginning of the mortgage crisis in 2007 there were only two large M&A deals in the pharmaceutical industry (Eli Lilly & Company acquiring ICOS corporation for 2.3 billion USD and GlaxoSmithKline acquiring Reliant Pharmaceuticals for 1,65 billion USD). The deal values have a range from 350 million USD (Astrazeneca acquiring Novexel FR SA for 345 million USD) to 90.000 million USD (Pfizer INC acquiring Warner-Lambert Company). The median of the deal values is 1.825 million USD and have an average value of 6375,65 million USD (standard deviation of 15956,6 million USD). The next table will provide the summary statistics of the M&A deals.
Table 1
Summary Statistics of Specific
M&A Deals
Number Percentage%
Deal
size Less than 5000 Mil USD 45 75,00%
5000-7500 Mil USD 7 11,67%
More than 7500 Mil USD 8 13,33%
Total 60 100,00% Method of payment Cash 56 93,33% Stock 8 13,33% Combined 4 6,67% Market Capitalization
(acquiring firm) Less than 20000 Mil USD 15 25,00%
20000-80000 Mil USD 21 35,00%
More than 80000 Mil USD 24 40,00%
Total 60 100,00%
Relative Size (deal value/ market cap.
acquiring firm) Less than 5% 31 51,67%
5%-20% 17 28,33%
More than 20% 12 20,00%
Total 60 100,00%
Crisis Before crisis 26 43,33%
After crisis 34 56,67%
Total 60 100,00%
18 The M&A deals in this sample are more cash-financed than stock financed acquisitions. Furthermore the market capitalization of the acquiring firms is equally divided in the sample. Because the first hypothesis described in this sample is on testing the impact of the mortgage crisis on the of the acquiring firm around a M&A the sample will be divided in two subsamples to test the first hypothesis. According to existing literature method of payment is also influencing the of the acquiring firm around a M&A. therefore the first
hypothesis will also be tested on this two different subsamples. The next section describes how the subsamples are divided and the descriptive statistics will be discussed.
3.4.1 Subsample 1: Method of Payment
To obtain the method of payment per M&A deal the information is gathered from the Zephyr database. This study divides the method of payment into two different groups.
1) All cash financing is when the method of payment was cash only. 2) All stock financing is when the method of payment was stock only.
In table 2 the differences for method of payment according to mean and median for deal size, market capitalization and relative size are displayed.
3.4.2 Subsample 2: Mortgage Crisis
To test the influence of the mortgage crisis on the profitability of M&A deals in the
pharmaceutical industry, this study divided the sample into two different groups. This study uses 1 Sept. 2007 as the beginning of the Mortgage Crisis.
1) Before the crisis is when the announcement of a M&A was before 1 Sept. 2007 2) After the crisis is when the announcement of a M&A was after 1 Sept. 2007.
In table 2 the differences in the period before and after the mortgage crisis according to mean and median for deal size, market capitalization and relative size are displayed.
19 3.5 Table Statistics of Subsamples
Table 2
Descriptive Statistics of Subsamples
Panel A: Subsample Method of Payment
Deal Size Market. Cap. Relative Size
Method of payment mean median mean median mean median
stock 11111,09 4000 93725,24 76476,27 18,94% 9,25%
cash 6737,649 1825 84302,4 69483 12,19% 3,86%
combined 21940 8180 40085,15 21179,29 44,44% 41,30%
Panel B: Subsample Influence Mortgage Crisis
Deal size Market. Cap. Relative Size
mean median mean median mean median
before the crisis 7.073,62 2.100,00 85300,87 71616,75 12,69% 3,90% after the crisis 5326,423 1750 83623,39 67349,25 11,17% 3,82%
* Deal size and market capitalization are in mil USD Panel A:
For the cash sample the mean and median for deal size is 6,373 billion USD and 8,2 billion USD. For the stock sample the mean and median for deal size is 11,1 billion USD and 4 billion USD. The relative size has an average of 18,94 % as relative size for stock payments and an average of 12,19% for cash payments. The results derived in this sample suggest that only large pharmaceutical companies apply large acquisition. Furthermore large acquisitions are mostly financed with stock or a combination of stock and cash.
Panel B:
The M&A’s after the crisis have a mean and median of 5,3 billion USD and 1,8 billion USD. For the M&A’s before the crisis the mean and median amount 7,1 billion USD and 2,1 billion USD. The average deal size has an average of 12,7% as relative size before the crisis and 11,17%. This suggests that the M&A deals in the pharmaceutical industry have remained at a constant level, if we look at relative deal size. This supports the findings of Hornke and Mandewirth (2010) who state that the pharmaceutical industry will keep the M&A deals on a constant level.
20 4 Methodology
This section describes the methodology used in this research. For this study two event windows are used to define the CAR. The first one is set from -1 to +1 around the
announcement of the M&A. The second one is set from -20 to +20 around the announcement of a M&A. The second window is to see if there are effects from insider trading. Section 4.1. describes the method to derive the normal stock return. Thereafter, section 4.2 describes the definition of the short-term abnormal stock return and how to calculate the abnormal return. In section 4.3 the OLS regression is described.
4.1 Benchmark
This study creates a benchmark using the market model described by MacKinlay (1997). The article states that the market model has potential improvements over the constant mean model. The market model is described as follows:
- is the return of stock i at time t
- is the alpha of the market model of stock i. - is the beta of the market model of stock i - is the period-t return on the market portfolio
- is the zero mean disturbance term for firm i at time t
This study uses for the return on the market portfolio the return on the S&P 500 index with data provided by Compustat. When the is calculated the abnormal returns can be calculated. The next section describes how the abnormal returns are calculated.
21 4.2 Calculating Short-term Abnormal Stock Return
This study will evaluate the short-term abnormal returns for pharmaceutical companies. The following formula is used when calculating the abnormal returns:
- is the realized stock return for firm i on day t.
- is the expected stock return for firm i in year t, the expected return is calculated with an event window of -541 to -21. This is a window of two years to calculate the expected return.
After calculating the abnormal returns ( ) the cumulative abnormal returns are calculated by the following formula:
- is the cumulative abnormal return for stock i - is the event window with (-1,+1) and (-20,+20)
- is the summation of the abnormal returns calculated for stock i within time τ
To calculate the cumulative average abnormal returns ’s) for the sample first the average abnormal return needs to be calculated:
Where N is the number of acquiring firms in the sample.
After this the is defined as:
22 The is calculated to evaluate the performance of the short-term stock price returns. The will be examined according to the different subsamples described in section 3. This study will provide information over difference in ‘s during different time periods. The next section will describe the OLS regression on the and which determinants
influence the value of the .
4.3 OLS regression on ’s
This study constructs the following model to test hypothesis (1) and (2) described in section 2. The motivation for the independent variables is described in section 3.1.2. The model is described as follows:
- denotes the cumulative abnormal return for firm i around the announcement of a M&A
- is the alpha of firm i
- is the dummy variable for a cash acquisition by firm i (denotes 1 if M&A is financed with cash)
- is the dummy variable for a share acquisition by firm i (denotes 1 if M&A is financed
with shares)
- is a control variable for the natural logarithm on the deal size of firm i - is a dummy variable for the effect of the mortgage crisis (denotes 1 if M&A announcement is after begin of crisis)
- is a control variable for the natural logarithm on the size of market capitalization of firm
i in month t
- is the independent variable for the effect of the R&D intensity (%) from the target firm I
in year t
- is the control variable for the effect of relative deal size (%) for firm i in month t - is the error term for firm i
23 According to Stock & Watson (2012) the OLS-estimator used in this study chooses the regression coefficients so that the estimator regression line is close to the observed data set. This is measured by taking the sum of the squared mistakes made in predicting the (Y) given the independent variables described above (X). After the regression is made, the - and the standard error of regression-value provide how much the OLS regression line fits the data.
4.4 Calculating the Significance of the
To test the hypothesis H1: τ > 0, this study uses the following formula:
τ
τ
With τ as the standard deviation, the variance is given by:
τ
The assumption of the investigation of Ikenberry, Lakonishok and Vermaelen (1995) is that the t-test is prone to event-induced volatility. This would mean that the test has low
24 4.5 Testing for Difference in Means Between Two Subsamples
To test for difference in means between the two samples first the standard error (SE) needs to be computed. The SE is defined as followed:
Where is the standard deviation of sample 1, is the standard deviation of sample 2, is the size of sample 1 and is the size of sample 2.
After that the test statistic can be computed. The test statistic is a t-score and is defined as followed:
25 5 Results
This section describes the results from the empirical research after the effects of horizontal M&A deals in the pharmaceutical industry. The previous chapter described the research methods used for the empirical research in this study. In this section the first paragraph describes the empirical data for testing hypothesis 1 described in section 2.4. After that Hypothesis 2 will be tested with empirical data provided by an OLS regression.
5.1 Results from Testing the H1: > 0
According to the literature review, the empirical evidence used in other studies report small negative returns for the acquiring companies around the announcement of a M&A deal. The table below provides the results for the τ’s of the acquiring pharmaceutical companies
around the announcement of a M&A deal.
The value for the measured is 0,19% and has a t-value of 0,47, which is not significant at the 5th and 10th significance level. The is 0,53% which is more
than two times higher than the and has a t-value of 0,5137. It has also no
significance at the 5th and 10th percentage-level. This means that the H0 is not rejected for both ’s. The insignificant t-value can be explained the fact that this study uses a relative small sample (the factor becomes bigger when taking a bigger sample). Another reason is provided by Stock & Watson (2012). They state that if the sample is not randomly assigned, the result can get biased. The sample used in this study contains only horizontal M&A deals in the pharmaceutical industry. Therefore this study needs to investigate whether the sample in this study is biased or not. The conclusion on the results of the empirical research provided above is that the has no effect on the acquiring pharmaceutical company around the announcement date of a M&A deal at the 5th and 10th percentage significance level.
26
5.1.1 Subsample 1: Method of Payment To determine if the method of payment has an effect on the τ the sample is divided into
two components: stock payment and cash payment. Within every two subsamples the t-test described in section 5.1 is conducted to test hypothesis 1: H1: >0. The table below provides the empirical results.
The overall observations count to 64, 4 M&A deals in this sample where a combination of stock and cash.
The results indicate that the τ stays insignificant whether we divide the sample into
subsamples for payment with cash or payment with stocks. The for method of payment stock are closer to zero (0,05% and 0,13%). This might be the reason that a stock payment implies overvaluation of stock which results into negative according to Bruner (2000). The results are not significant : 0,36 and 0,43), so for the pharmaceutical companies tested in this study H1: >0 is not accepted.
27
5.1.2 Subsample 2: Mortgage Crisis To test whether the Mortgage crisis has a significant effect on τ the sample is again
divided into two subsamples. Also in this section the t-test described in section 5.1 is conducted to test H1: >0. The results are provided in the table below.
The results for this subsample have also insignificant τ ‘s before and after the crisis.
The for the subsample before the Mortgage crisis have t-values who are not different from the sample after the crisis. This indicates that dividing the sample into two subsamples, before and after the Mortgage Crisis, don’t result in significant t-values to accept H1:
>0. To test whether the dummy variables used in creating the above described subsamples have a significant effect on the , this study uses the OLS regression described in section 4.3. The empirical results are provided in the next section.
28 5.1.3 Testing for difference ’s in Subsamples
In the table below the difference in ’s, the SE and the t-values in the two subsamples are displayed.
As we can see from table 6, the first sub hypothesis H1: -
0 is not accepted at the 5th and 10th percentage significance level
for the two event windows (t-value: -0,08 for (-1,+1) and t-value: -0,05 for the (-20,+20)). This means that the mortgage crisis has no influence on the around the
announcement of the acquiring company.
The second hypothesis H1: - > 0 is not accepted, with t-values 0,08
for (-1,+1) and 0,1 for (-20,+20). The empirical evidence in this study doesn’t support existing literature. The findings of Bruner (2002) described in the literature section described that an acquisition paid with stock is costly and an acquisition paid with cash could result in positive returns. For the sample tested in this study this statements doesn’t hold when dividing the sample in two subsamples. Therefore the next section will provide an OLS-regression on the with the determinants described in section 4.3.
29 5.2 Results from the OLS-regression on with determinants
In this study the ’s of the target firms are regressed on the determinants described in section 4.3. The below provided tables shows that the market capitalization of the acquiring company is an important determinant of the (t-value 2,17) and has a coefficient of
-009. The table below shows that the method of payment stock (t-value 1,86) is also a important determinant of the .
* The deal characteristics market capitalization and deal value are a natural logarithm.
The provides no significant empirical t-values for the determining variables.
The is lower than the from the (21% against 26%). According to Stock and Watson (2012) the model provided by with the same determinants has smaller explanatory power than the model provided by with the same determinants. Therefore this study will test the hypothesis described in section 4.3 only on the .
The second hypothesis: H2: The from the acquiring pharmaceutical company around the announcement of a M&A is influenced by the mortgage crisis, is rejected at the 5th and 10th percentage significance level for . The dummy variable mortgage crisis
has a coefficient of -0,009 and a t-value of -1,11. In this sample the mortgage crisis has no influence on the of the acquiring pharmaceutical companies around the announcement
30 of a M&A. This supports Hornke and Mandewirth (2010) who state that unless the mortgage crisis didn’t have an impact on M&A deals in the pharmaceutical industry.
Furthermore the dummy determinant payment with stock has a negative coefficient: -0,03. The empirical results in this study therefore suggest that a payment with stock results in a -0,03 effect on . This implies that if the acquiring company announces that the M&A deal will be financed with stock (or stock and cash) the will decrease with 3,0%.
The third hypothesis: H3: The from the acquiring pharmaceutical company around the announcement of a M&A is negatively influenced by method of payment stock, is accepted at the 10th percentage significance level on . This supports the literature described by Bruner (2002) that stock payment has a negative effect on the of the acquiring firm.
In the descriptive statistics table 1 93,33% of the M&A deals are financed with cash in this sample. Furthermore the dummy determinant method of payment cash has an
insignificant coefficient of 0,01 (t-value 0,51). The fourth hypothesis: H4: The from the acquiring pharmaceutical company around the announcement of a M&A is positively influenced by method of payment cash, is therefore rejected at the 5th and 10th percentage significance level for the . This in line with the findings of Bruner (2002), who states that a M&A deal financed with cash is neutral.
The fifth hypothesis: H5: The from the acquiring pharmaceutical company around the announcement of a M&A is positively influenced by R&D intensity of the target firm, is rejected at the 5th and 10th percentage significance level for . The
independent variable R&D intensity has a coefficient of -0,07 and a t-value of -1,21. Existing literature like Hornke and Mandewirth (2010) suggests that pharmaceutical companies acquire for example biotech firms to secure their pipeline. The R&D intensity of the target firm should therefore have a significant influence on the of the acquiring
pharmaceutical firm around the announcement of a M&A.
The insignificant t-values for these determinants might be due to the non-random sampling in this study. According to Stock and Watson (2012) random sampling is important,
otherwise the will be biased. This might lead to insignificant t-value in the determining variables.
31 6 Conclusion
Companies in the pharmaceutical industry invest a large amount in research and development. The companies depend on the R&D productivity to secure their pipeline that creates long-term value. This study tries to provide insights on the effects of acquisitions from
pharmaceutical companies to expand their pipeline and outsource R&D investment. This study examined the short-term cumulative abnormal returns for the acquiring firm around a M&A. A sample of 60 100% acquisitions within the pharmaceutical industry during the period from 1 Jan. 1998 to 31 Dec. 2014 with a minimum deal value of 350 million USD is analyzed. The methodologies used are: 1) The ‘Event’ Study; 2) the Capital Asset Pricing Model 3) an OLS-regression on the short-term cumulative abnormal returns for the acquiring firms around a M&A deal.
Furthermore, the empirical results from this study are compared with table 4 provided in Goergen and Renneboog (2004). The article used 142 observations in the empirical research and found also a insignificant τ’s for the -60,+60 window and -40,0 window (-0,48%
and 0,40%). The in this study includes 41 trading days and has a insignificant
value of 0,46%. The τ’s obtained in the empirical investigation in this study are close to
the ones founded by Goergen and Renneboog. Gregory and McCorriston (2002) investigated the wealth effects for cross-border acquisitions by UK firms. The article found no gain or loss for the acquiring company. This supports the rejection of the first hypothesis in this study, which stated that there is no short-term value effect for the acquiring company.
This study analyzes which factors influence the short-term value effects ( and
) for the acquiring pharmaceutical firms. This study found a significant effect of
the determinants market capitalization of the acquiring firm and method of payment stock on the (t-value: -2,17 and 1,86). The rejection of the fourth hypothesis (payment with cash has a positive influence on the ’s) in this study furthermore suggests that payment with cash is neutral, which is supported by Bruner (2002). In this study the determinants relative deal size, R&D expenditure from the target firm and deal value have no significant influence. The empirical results in Goergen and Renneboog (2004) shows that there is no significant effect from relative deal size. The coefficient on relative size is 0,01 (t-value: 0,673). The empirical results in this study show a coefficient of 0,01 (t-value: 0,26). Goergen and Renneboog (2004) used M&A deals with a minimum value of 100 Mil USD. The
32 To sum up, this study shows that M&A deals for the acquiring pharmaceutical companies generate neutral short-term abnormal returns. The OLS-regression suggests that if the deal is financed with cash the short-term effect will be less. The effect for financing the deal with cash is neutral. Whether the deal is before or after the Mortgage Crisis has no significant effect and the R&D expenditure of the target firm shows no significant effect on the short-term return for the acquiring pharmaceutical firm.
33 7 Discussion
This study tries to address with methodologies like the ‘Event Study’ and afterwards an OLS-regression on of the acquiring pharmaceutical firm around a M&A, if it results in significant effect. As described in the result section this investigation didn’t result in significant The main point of the insignificance is the relatively small amount of observations (N=60). A second point for future studies is that this study should implement for example a control group of 60 M&A deals with a minimum deal value 350 million USD. When for example including the dummy variable pharmaceutical, the empirical data could show a difference in for pharmaceutical firms compared to the random control group.
The second deficiency that future studies need to supplement in their empirical research is to create a smaller timeframe around the mortgage crisis. This study investigates pharmaceutical M&A deals that were announced from Jan. 1998 to Dec. 2014. If the sample is taken from Jan. 2006 to Dec. 2010, the effects of M&A deals 2 years before the Mortgage Crisis could be compared with M&A deals 2 years after the Mortgage Crisis. This might give more insights in the influence of the Mortgage Crisis on the pharmaceutical industry.
According to Goergen and Renneboog (2004) the agency problems described in the literature section (managerial hubris and free cash flow theory) occur when the manager does not own equity in the acquiring firm. The article states when the managers own a large equity stake in the bidder firm, the for the acquiring firm around the announcement of a M&A will be significantly positive. The role of the managers from the acquiring company within the pharmaceutical M&A deals is not tested in the empirical research of this study. Therefore future studies should implement the role of the managers in the acquiring firm in their empirical research.
At last, because Compustat WRDS did not provide all the values from R&D expenditure from the target firms, this might be the reason for the determinant R&D intensity in the OLS regression to be insignificant. This study used the R&D expenditure from the financial annual statement from the target firm 1 year before the announcement of the M&A deal, this
insufficient to determine the R&D capital from the target firm. This study would therefore be more comprehensive if the analysis on the R&D capital from the target firm would be more sufficient.
34 8 References
Andrade, G., Mitchell, M., & Stafford, E. (2001). New Evidence and Perspectives on Mergers. SSRN Journal SSRN Electronic Journal, 1-32.
Binder, J. (1998). The Event Study Methodology Since 1969. Review of Quantitative Finance and Accounting, 11, 111-137.
Bruner, R. (2004). Does M&A Pay? A Survey of Evidence for the Decision-Maker. Journal of Applied Finance, 12(1), 48-68.
Berk, J., & DeMarzo, P. (2007). Corporate finance. Boston: Pearson Addison Wesley.
Chen, N., Roll, R., & Ross, S. (1986). Economic Forces and the Stock Market. The Journal of Business, 59(3), 383-403.
Chan, L., Lakonishok, J., & Sougiannis, T. (2001). The Stock Market Valuation of Research and Development Expenditures. The Journal of Finance, 56(6), 2431-2456.
Devos, E., Kadapakkam, P., & Krishnamurthy, S. (2007). How Do Mergers Create Value? A Comparison of Taxes, Market Power, and Efficiency Improvements as Explanations for Synergies. Review of Financial Studies, 22(3), 1179-1211.
DiMasi, J., Hansen, R., Grabowski, H., & Lasagna, L. (1991). Cost of innovation in the pharmaceutical industry. Journal of Health Economics, 10(2), 107-142.
Fuller, K., Netter, J., & Stegemoller, M. (2002). What Do Returns to Acquiring Firms Tell Us? Evidence from Firms That Make Many Acquisitions. The Journal of Finance, 57(4), 1763-1793.
Forbes Magazine. (2015). Retrieved from http://www.forbes.com/
Gilbert, J., Henske, P., & Singh, A. (2003). Rebuilding Big Pharma's Business Model. The Business & Medicin Report, 21(10), p.1-10.
Goergen, M., & Renneboog, L. (2004). Shareholder Wealth Effects of European Domestic and Cross-Border Takeover Bids. European Financial Management, 10, 9-45.
Gregory, A., & Mccorriston, S. (2002). Foreign acquisitions by UK limited companies: Short- and long-run performance. Journal of Empirical Finance, 99-125.
Higgins, M., & Rodriguez, D. (2006). The outsourcing of R&D through acquisitions in the pharmaceutical industry. Journal of Financial Economics, 80, 351-383.
Hornke, M., & Mandewirth, S. (2010). Mergers & Acquisitions (M&A) in the pharmaceutical industry: The wheel keeps on turning. Journal of Business Chemistry, 7(2), 67-68.
Ikenberry, D., Lakonishok, J., & Vermaelen, T. (1995). Market Underreaction to Open Market Share Repurchases. Journal of Financial Economics. 39(2-3), 181-208.
35 Jensen, M. (1986). Agency Cost Of Free Cash Flow, Corporate Finance, and Takeovers. SSRN Journal SSRN Electronic Journal, 76(2), 1-15.
Joglekar, P., & Paterson, M. (1986). A closer look at the returns and risks of pharmaceutical R&D. Journal of Health Economics, 5(2), 153-177.
MacKinlay, A. (1997). Event Studies in Economics and Finance. Journal of Economic Literature, 35, 13-39.
Myers, S., & Majluf, N. (1984). Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have. Journal of Financial Economics, 13(2), 187-221.
Shahrur, H. (2005). Industry structure and horizontal takeovers: Analysis of wealth effects on rivals, suppliers, and corporate customers. Journal of Financial Economics, 76(1), 61-98. Stock, J., & Watson, M. (2007). Introduction to econometrics (2nd ed.). Boston:
Pearson/Addison Wesley.
Wharton Research Data Services. (2015). Retrieved from https://wrds-web.wharton.upenn.edu/wrds/ds/comp/index.cfm
36 9 Appendix
This section points out the motives for including R&D intensity in the OLS regression and gives a motivation for taking two different levels of . The next paragraph describes the influence of insider trading on the Thereafter the movement in R&D intensity is pointed out. The last paragraph describes my Internship at Astellas Pharma Inc and provides some additional motives behind this thesis.
9.1 Relevance R&D Intensity in Pharmaceutical Industry
According to Gilbert, Henske and Singh (2003) the pharmaceutical industry needs a new approach. They state the pharmaceutical companies need to focus on their own R&D activity and not on M&A deals. The article states M&A deals will not improve returns. The result of testing the first hypothesis in this study supports the statement of the article. Therefore this study will investigate the movement in R&D capital from 14 acquiring pharmaceutical firms from 1 Jan. 1994 to 31 Dec. 2014 in the next section.
9.1.2 Movement R&D Capital in Pharmaceutical Industry for Acquiring Companies
This study uses the formula provided by Chan et al. (2001) to measure the movement in R&D capital within the pharmaceutical industry. The movement in R&D capital is measured by evaluating the movement in R&D expenditure from 14 pharmaceutical companies. To estimate the R&D capital the following formula is used:
is R&D capital for firm i in year t
is the R&D expenditure for firm i in year t
This formula assumes that the productivity of each dollar spend on R&D expenditure declines with 20 percent a year. This study measured the movement from R&D capital from 1994 to 2014. After measuring the R&D capital it is seen as a percentage of book value and market capitalization. The results are provided in the table below:
37 Appendix Table 1: Movement R&D Capital Sample Pharmaceutical Firms
R&D Capital as %
of: Bookvalue % Change Market Cap. % Change
Year 1999 28,47% 104,00% 2004 25,76% -9,52% 56,71% -45,47% 2009 24,20% -6,06% 59,15% 4,31% 2013 23,16% -4,28% 48,00% -18,86%
* Companies used in Sample: Abbott, Glaxo Smith Kline, Pfizer, Novartis, Johnson & Johnson,
Lilly Eli & Co, Questcor Pharmaceuticals, Novo Nordisk, Teva Pharmaceuticals, Astrazeneca, Gilead Sciences, Baxter International, Merck & Co, Amgen Inc
The above provided table shows that the R&D capital as percentage of book value has an average value of 25,4%. It has decreased by an average percentage of -6,62% since 1999. This might imply that the pharmaceutical firms would invest more capital in M&A deals than in R&D Capital to generate NME.
The movement in R&D capital as percentage of market capitalization shows fluctuations in the sample period. To test whether the mortgage crisis has an effect on R&D expenditure for the sample firms, this study uses the R&D capital in the years 1999, 2004, 2009 and 2013. The reason for the large percentage in 1994 is because the firms Novartis, Novo Nordisk, Glaxo Smith & Kline and Astrazeneca have R&D capital at a high percentage of market capitalization.
The mean of R&D capital as percentage of market capitalization is 66,79% and the
movement has an average change of -20,01%. In 2009 the movement in R&D capital shows a change of 4,31%. This value shows that the mortgage crisis has an effect on the stock values of the pharmaceutical firms used in this sample.
38 9.1.3 R&D Expenditure Target Firms as Percentage of Market Capitalization
A independent variable included in the model affecting the of the acquiring firm after a M&A deal is the R&D expenditure from the target firm as percentage of the market
capitalization of the acquiring company. The data providing the R&D expenditure is derived from the Financial Annual Statement from the target firm one year before the M&A deal. This study uses the Financial Annual Statement one year before the deal because that represents the R&D intensity from the target firm. The results are provided in Appendix Table 2 below.
Appendix Table 2
R&D Expenditure Target Firms
Mean Median
R&D expenditure in Mil USD 333,1202667 30,771
R&D as % of Market
Cap. 1,894% 0,080%
The mean of R&D percentage of the target firm as percentage is 1,894%. This is a relative small amount due the high R&D intensity for innovative pharmaceutical firms. The
explanation is that the acquiring pharmaceutical firms are substantially larger than the target firms.
39
9.2 Covariance tables The tables below provide the covariance tables of the determinants of and
.
Table Covariance Table
CAR(-1,+1) R&D expenditure Payment Cash Payment Stock Mortgage Crisis Lndealvalue Lnmarket cap. Relative deal value
CAR(-1,+1) 1 R&D expenditure -0,04 1 Payment Cash 0,33 0,05 1 Payment Stock -0,33 -0,05 -0,68 1 Mortgage Crisis -0,03 0,13 0,31 -0,25 1 Lndealvalue -0,21 0,13 -0,4 0,53 -0,2 1 Lnmarket cap. -0,33 -0,34 -0,24 0,05 -0,09 0,19 1
Relative deal value -0,07 0,52 -0,23 0,51 -0,08 0,65 -0,36 1
Table Covariance Table
CAR(-20,+20) R&D expenditure Payment Cash Payment Stock Mortgage Crisis Lndealvalue Lnmarket cap. Relative deal value
CAR(-20,+20) 1 R&D expenditure 0,26 1 Payment Cash 0,3 0,05 1 Payment Stock -0,27 -0,05 -0,68 1 Mortgage Crisis 0,27 0,13 0,31 -0,25 1 Lndealvalue -0,01 0,13 -0,4 0,53 -0,2 1 Lnmarket cap. -0,19 -0,34 -0,24 0,05 -0,09 0,19 1
