Determinants of the return and probability of an upmarket M&A
Guido Zwier 6153119 University of Amsterdam
MSc Business Economics, Finance track Master Thesis
July 2015
Thesis supervisor: Mr. Ligterink
Abstract
In the world of mergers and acquisitions there has been a growing number of emerging market firms acquiring developed market firms over the last decade. This phenomenon is called an upmarket M&A. This paper analyses the Cumulative Abnormal Returns and the likelihood of a diversified and upmarket M&A in the period between 2005 and 2014. Surprisingly, there is a statistically significant and positive Cumulative Abnormal Return (CAR) for emerging market firms in an upmarket M&A; 4.6% and 5.2% in a twenty- and thirty-day window. This paper takes into account agency problems, method of payments and potential synergies to determine the effects of the probability of an upmarket M&A.
2 Statement of Originality
This document is written by Student Guido Zwier who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Table of Contents
1. Introduction 4
2. Literature Review 5
2.1 Emerging markets 5
2.2 Cross-border and upmarket M&As 5
2.3 Stock price returns on announcement 6
2.4 Reasons for upmarket M&As 7
2.5 Free cash-flow theory and agency problems 8
3. Hypothesis and methodology 9
3.1 Hypotheses 9
3.2 Methodology and Empirical model 11
3.2.1 Cumulative Abnormal Return model 11
3.2.2 Probit regression model 12
4. Data 14
5. Results 17
5.1 Cumulative Abnormal Return on announcement 17
5.2 Determinants of the Cumulative Abnormal Return 18
5.3 Method of payment and the probability of Upmarket M&A 20 5.4 Controlling Shareholder and the probability of a diversified M&A 22
6. Robustness test 24
7. Summary and conclusion 26
8. Reference list 28
4
1. Introduction
This research will investigate the determinants for the expansion of emerging market firms to developed market firms. To see what factors affects this upmarket M&A, there will be looked at the probability and the Cumulative Abnormal Returns of this event. The goal of this research is to obtain new empirical result about M&As from emerging market firms. In these markets is limited empirical and literature information available. The United Nations published in their World Investment Report an increase of emerging market investments in developed markets (2014, p. 44). 53% of the cross-border deals came from emerging markets in 2007. The United Nation report implies that there is an upward expansion trend of emerging markets firms globally.
Testing if the Free Cash-Flow Theory on Takeovers of Jensen (1986) holds in upmarket M&As is another contribution to existing theory. Jensen (1986) implies that M&As based on diversification are value destroying because managers of firms with excess cash are likely to invest in low-benefit takeovers as result of agency problems. So M&As based on diversification lowers the total gain of a firm. Stucchi (2012) found that diversification M&As are not value destructive because emerging market firms may obtain more experience and knowledge in other foreign markets. These advantages can also be used in the home country.
The general reasons for an industrial diversified M&A are to create value and synergy advantages (Chari et al., 2009; Andrade et al., 2001; Erel et al., 2012). Bertrand and Betschinger (2012) found that emerging market firms want to obtain knowledge from the advanced market firm and implement this in the emerging home country instead or improve the targets firms business. The emerging market firm will improve profitability and their technological and economic
development. So the research question of this paper is:
Which factors affect the return and probability of an upmarket M&A focused on industrial diversification?
In this paper the Cumulative Abnormal Returns (CARs) of 29 upmarket M&As will be analyzed. Also the factors that affect the CAR will be estimated by an OLS regression. Besides that, the likelihood of an upmarket M&A and a diversified M&A will be tested in a sample of 2,219 worldwide M&As with a probit model with multiple regressors. The most important factors that will take into account are the methods of payment, sector diversification, protection of investors, controlling shareholders and deal value.
The results suggest that the upmarket Cumulative Abnormal Returns are positive and significant. 4.6% in a twenty- and 5.2% in a thirty-day event. The effect of a diversified deal has a negative relation with the returns. The effect of a cash or debt payment has a positive relation compared to a stock payment on the CARs and makes a upmarket M&A more likely. The presence of
5 a large controlling shareholder in the acquiring firm will increase the probability of an upmarket M&A.
The structure of the paper is as follows. Section 2 will discuss the related literature. Section 3 explains the hypotheses and the methodology and section 4 the data. Sections 5 and contains the results and robustness check. Section 7 concludes this paper and answers the research question.
2. Literature review
This literature part will consist of five subsections. The first subsection is introducing the emerging markets. The second subsection is about general M&As and the upmarket M&As. The strategies and synergies will be discussed. Section 2.3 is about the theory of the stock market returns after a M&A announcement. After that, the sector diversification in M&As will be discussed and finally the Free cash-flow theory of Jensen (1986) and the agency theory.
The goal of this literature review is to gain insights in the reasons for an upmarket M&A,
determinants that affect the probability of an upmarket M&A and the choice for a sector diversified deal.
2.1 Emerging markets
At the end of April 2015, the S&P Dow Jones published a fact sheet about emerging markets. They selected 23 countries as emerging market economies (S&P Dow Jones, 2015). These emerging and developed market countries are shown in appendix 1. The country classification requirements to be an emerging or developed market are based on quantitative and qualitative criteria. Quantitative criteria for Emerging market, stated in appendix 2, are based on several factors, for example debt rating, currency and a market developing ratio of over 5%. These criteria should be in combination with a GDP per capita lower than $15,000 to be named emerging market (S&P Dow Jones, 2014). The qualitative criteria are based on opinions of global investors. These criteria are about regulations, shareholder protection, legislation in foreign exchange trading, operating and trading costs and transparency of financial data (S&P Dow Jones, 2014).
2.2 Cross-border and upmarket M&As
The general cross-border theory suggest that M&A deals result in higher synergy gains due the complementary resources and capabilities across countries and there is a potential redeployment for firms active in the same sector. Overall, this can be a cost synergy and a productivity synergy (Bertrand & Betschinger, 2011). But Bertrand and Betschinger (2011) found also that emerging market firms cannot fully utilize the value of a cross-border M&A deal because of lower M&A
6 experience and capabilities compared to the non-emerging market firms. But the high technology sector benefits more from a cross-border deal than from a domestic deal. Reason for this is that in abroad markets new opportunities are available.
The mergers and acquisitions (M&As) from an emerging market (EM) firm to a developed market (DM) firm is called an ‘upmarket M&A’ (Stucchi, 2012). Firms have in general four different strategies when considering an upmarket M&A. Stucchi (2012) divides these strategies into exploitation or augmentation strategy on marketing activities and technological activities. The difference between exploitation and augmentation is that exploited firms want to improve their current expertise and with augmentation they want to develop more knowledge and skills in their expertise and beyond this. Marketing activities are distribution and service skills, customer attention and brand loyalty. Technological activities are R&D, patents and licenses and the developing of technology.
The first strategy is the upstream acquisition strategy where the EM firm focused on the EM market, exploits their marketing activities and augments their technological activities. This strategy is focused on industry/sector diversification. The second strategy is the exploitative acquisition strategy. This strategy exploits both marketing and technological skills. The market focus is here also on the EM and no focus on either industry or location diversification. The augmenting strategy focuses on the DM and location and industry diversified. In here, the focus is on marketing and technological augmentation. The last strategy is the downstream acquisition strategy where marketing
augmentation and technological exploitation is the case. The market focus is on DM and location diversification.
All these strategies have in common that upmarket M&As for diversified firms can have experience and knowledge advantages (Stucchi, 2012). Bertrand and Betschinger (2011) describe that emerging market firms use upmarket M&As in a more aggressive way than advanced market firms. They try to compare for their competitive disadvantages with these M&As because of limited resources. They found that these upmarket deals could end up counterproductive for the EM firms. Information asymmetry, overestimated synergies, overpayment of the AM target or reputation problems could be the reason for this.
2.3 Stock price returns on announcement
The stock price return on an announcement is in general an indicator how successful the shareholders think a M&A is. Chari et al. (2009) investigate these stock price returns based on Cumulative Abnormal Returns (CARs) in different time periods. They tested their model on an overpayment assumption of Dyck and Zingales (2004). They assume that acquiring firms did not paid too much because the CARs are positive. In the case of Chari et al. (2009), they investigates the
7 three-day event abnormal stock return of acquiring firms in M&A deals from developed markets to emerging markets. DM-EM gives a 1.2% positive return.
An empirical upmarket M&A result is from Erel et al. (2012). They tested several determinants on the
probability of a M&A and they study the effect of currency movements. In these currency
movements, they studied the differences in stock and market returns between acquirer and targets before and after an M&A announcement. There is a big difference between DM-EM M&As and EM-DM M&As. The stock returns are positive, 12.8% for EM-DM-EM and 9.5% for EM-EM-DM. The currency returns is 34.2% when DM firms acquire EM firms. When EM firms acquire DM firms, the currency depreciate: -23.3%. Overall, they found that acquiring firms are more likely from the country which is better economic developed and has higher accounting standards (Erel et al, 2012). Also it finds that in M&A deals, the market-to-book ratio of the acquirer is 9.93% higher at time of the deal than the target (Erel et al., 2012).
This result is in line with findings from Rossi and Volpin (2004): they found in the general cross-border M&A theory that targets are more likely from countries with less corporate governance regulations. These regulations, also called the investors protection, consist of accounting standards, shareholders protection and which type of law is used. In their Tobit model on the dependent variable number of M&A deals, they found a 12 point increase with higher accounting standards. 1% increase in shareholder protection leads to a 4% increase in total M&A deals. Also the civil law is positive: 7.5% if the common-law is adjusted. So, better investor protection will lead to more M&As.
Andrade et al. (2001) found for the period 1973-1998 that the three-day event abnormal stock returns after a merger announcement is combined (target and acquirer) positive (1.8%). In the period of 20 days before the merger announcement and the merger itself it is 1.9%. The acquiring firm has a negative return of 0.7% in the one-day window. Even in longer period of twenty-day window it is negative: 3.8%. The target company has always a positive return: 16% in the one-day window and 23.8% in the longer period.
2.4 Reasons for upmarket M&As
Stucchi (2012) divide reasons for emerging market firms to take-over a developed market firm into three different categories. First is the resource based antecedents where the focus of an EM firm lies on the market(s). The market factors competitiveness, consumers, infrastructure, regulation and diversification are important. Location diversification develops knowledge about the new markets and consumers. But also industry diversified firms can gain product development advantages. Last of the resource based antecedent is that firm size and age are important. Small firms are less likely to attract bigger firms and firm size is important to attract managers with AM
8 experience (Stucchi, 2012). The age is also important. The more mature firms are more likely to do acquisitions and can better develop the resource based antecedents.
The second category is the institution based antecedent. In here, the firms government, ownership control and financing an acquisition. Rossi and Volpin (2004) found that the probability of an all cash bid increases in cross-border M&As, but decrease if the target becomes greater in size. Jensen (1986) found that cash and debt acquisitions are more successful than stock acquisitions. Linn and Switzer (2001) suggest that the post M&A stock return is larger for cash based M&As than for stock based M&As. Reason for this is that the bidden probably has private information about the synergies and cash bids are likely more to improve these synergies. Also the performance of a firm will increase more after a cash M&A than a stock M&A.
The last category is the industry based antecedent. Emerging market firms try to increase their knowledge of the advanced markets. With this knowledge they can compete themselves in these markets (Stucchi, 2012). Others found that the knowledge of the AM market will be used for the EM market, but in the AM market there will be no improvement (Bertrand & Betschinger, 2012).
2.5 Free cash-flow theory and agency problems
Jensen (1986) investigates the effect of agency cost on the payout to shareholders during a M&A. He links the Free Cash-Flow theory to takeovers. This theory predicts that a M&A destroys and not create value. Managers in firms with low debt and high cash flows do not create shareholder value because they are more likely to invest in low-benefit or value destructive takeovers. Diversification in industry lowers the total gain of a firm. A solution for this agency problem is the increasing of the firm’s debt. Increasing firm’s debt is efficient for solution making. It avoids firms to invest in low-return projects. But this debt can create a new crisis in the firm, and can lead to cuts in the expansion and fire sales of firms divisions. Jensen (1986) indicates that successful takeovers are done with targets which are poorly-managed and badly performs. Also targets with outstanding performances and which pay no dividend to their shareholders. Jensen (1986) finds that acquirers should perform exceptionally well before an M&A. So the Free cash-flow theory of Jensen (1986) is contradictory with the idea of the upmarket M&A theory of Stucchi (2012).
Like Jensen (1986), Rau and Vermaelen (1995) find that there arise agency problems when a company with a high market-to-book ratio acquires a company with a low market-to-book ratio. Possible reason for this problem is managerial hubris. They found in the short-run a positive abnormal return for the bidders, but in the long-run a negative abnormal stock return. These negative long-term performances can be blamed to two factors (Duchin & Schmidt, 2013). First, the high cost that will arise of external monitoring of the managers. Because of this high cost, external monitoring will not be done and managers can get away with negative M&A deals. An alternative of
9 external monitoring to avoid agency problems is internal monitoring of the managing board (Lane et al., 1998). In the research ‘Agency problems as antecedent for unrelated M&A’, the researchers investigate the effect of large block shareholders in the firm. They found that managers in a firm without large block shareholders are involved in more unrelated acquisitions. So controlling
shareholders will restrain managers from doing diversified acquisitions. It is interesting to test if the presence of large block holders also affecting the probability of a diversified (upmarket) M&A.
Second factor for negative long-term performance is managerial herding. Managers are selfish and take risks for their own career (Duchin & Schmidt, 2013). Lane et al. (1998) suggest that managers are only self-interested if there is no large block holder in the firm. Block holders try to increase their own shareholder value at the cost of other shareholders and the managers.
But Bertrand and Betschinger (2011) do not support the Jensen’s Free Cash Flow theory. They found no supporting evidence that Russian managers abuse excess cash for unprofitable investments. Besides that, they found no evidence for blaming agency problems to the negative long-term performance of acquisitions. However, they did find that diversified M&As could destroy firm value.
3. Hypotheses and Methodology
The literature review supports the following described four hypotheses. The empirical models of these hypotheses will be describes in section 3.2 and tested in section 4.
3.1 Hypotheses
H1. Upmarket M&As gives a negative Cumulative Abnormal Return to the acquiring firms shareholder.
Based on general merger results, Andrade et al. (2012) found that the acquiring firm a negative CAR had in a one-day and in a twenty-day window. Stucchi (2012) claims that upmarket M&As are not value destroying. So this could result in a positive return. Bertrand and Betschinger (2011) did not found empirical evidence in Russia to support this hypothesis. However, Jensen (1986), Rau and Vermaelen (1995) and Duchin and Schmidt (2013) support this hypothesis. In this case a low book-to-market firm acquires a high book-to-market firm. Also, agency problems, method of payment and an incomplete monitoring system are possible explanations. But based on other CAR studies, Chari et al. (2009) found that for the acquirer and the target the CARs are positive and significant in case of a M&A. They declare that the acquirers positive CAR will reject their
overpayment hypothesis. This hypothesis implies that a negative CAR for the acquiring firm means that the acquiring shareholders think the acquiring firm pays too much for the target firm.
10
H2. An industry diversification based upmarket M&As gives a negative Cumulative Abnormal Return for the acquiring firms shareholder.
The Free Cash Flow Theory on takeover suggests that a M&A based on industry
diversification is value destructive (Jensen, 1986). Bertrand and Betschinger (2011) described the advantages of a M&A without diversification. So, diversified deals will miss the advantages of cost synergies, production synergies and potential redeployments.
However Lane et al. (1998) suggest that problems of a value destructive or unrelated M&A is based on if the company has large block holders or not. Such a large controlling shareholder can exercise pressure in the acquiring company and could prevent the managers to do value destructive or unrelated M&As. This controlling shareholder can monitor the managers and their decisions. So well good monitored managers will only do profitable M&As. Berstrand and Betschinger (2011)
investigate the abuse of excess cash in Russia. They found no evidence for negative long-term performance of upmarket M&As because of agency problems.
H3. Diversified based upmarket M&As paid with cash or debt are more likely than diversified based upmarket M&As paid with stocks.
The choice for an offer with cash or debt, comparing with a stock offer depends on private information about synergies (Linn & Switzer, 2001). Jensen (1986) and Rossi and Volpin 2004) found that cash bids are more successful and likely that stock bids. Interesting to see if this is the case in diversification based upmarket M&As. Because the reason of these types of M&As are not always based on synergies according to the upmarket strategies of Stucchi (2012). Andrade et al. (2001) also looks as the effect in mergers of a stock or no stock deal and the Cumulative Abnormal Returns of this choice. The return for stocks deals is in a oneday and in a twentyday event window negative, -1.5% and -6.3%. No stock deals have a better return. 0.4% in the one-day event and -0.2% in a twenty-day event. With these results a firm will rather choose no stock deal then choose the stock deal.
H4. The presence of a large block shareholder is negatively affecting the probability of a diversified (upmarket) M&A.
With the presence of a large block holder, the firm is internal monitored on agency problems (Lane et al., 1998). Managers will be more controlled on diversified M&A deals. The results of Lane et al. (1998) suggest that firms without a large block shareholder are more involved in diversification based M&A deals. Reason for this is the monitor function that a large block shareholder has on managers. Managers are selfish and will take excessive risk for success and their rewarding (Duchin & Schmidt, 2013).
11
3.2 Methodology and Empirical model
This section will discuss two empirical models that will test the hypotheses discussed in section 3.1. The first model tests an event study about the Cumulative Abnormal Return (CAR) of an acquiring firm after the news of an upmarket M&A announcement. The intention of this model is to test whether an announcement gives a positive or negative shareholder return. The CAR variable will also tested as dependent variable in an OLS regression, to test what effect a diversified deal has on the CAR. This model test the first two hypotheses. The second model is a probit regression on a binary dependent variable upmarket M&A and discussed in section 3.2.2. This model test the hypotheses 3 and 4. The purpose of this model is to see if the choice of payment is in line with the Free Cash-Flow theory on takeovers of Jensen (1986). Also if there is a causal effect between diversification and the existence of a controlling shareholder in the acquiring firm (Erel et al., 1998)
3.2.1 Cumulative Abnormal Return model
The first empirical model is based on a stock price reaction on a M&A announcement. This sample contains the upmarket M&As from the total sample of 2,119 observations. The weekend announcements or missing data announcements are canceled out of the sample. For in total 29 announcements is the reaction measured by a CAR. 27 different firms are used. These CARs are based on the Chiari et al. (2009) papers and the book of Stock and Watson (2012). Specifically, the data that is used are 29 upmarket M&A with the [-30/+30], [-20/+20] [-3,+3] and [-1/+1] Cumulative Abnormal Returns.
Mitchell and Netter (1994) define first the stock returns and market returns:
[ ( ) ] (1)
Where
P₁ = price at the end of a trading day P₀= price at beginning of a trading day DIV₁= dividend paid during period.
Thus the stock and market return is simply the difference between the price between the end and the beginning of a trading day. In the upmarket M&A sample is no dividend payment of stock split registered.
Next the normal return model is:
12 With this normal return model Stata estimated the square root of the Mean Square Error (RMSE) and the price prediction Pest. The RMSE is the forecast error of the uncertainty of the predictions of the regression coefficients and the uncertainty of the future unknown value of the error term (Stock and Watson, 2012,p. 581). After this, according to the event window, the Abnormal Return ARit is:
P (3)
The Abnormal Return in the event window predicted as the difference between the actual return of the stock and the benchmark. Bodie, Kane and Markus (2011, p. 382) suggest that any Abnormal Return a poor measurement is for the total impact of the announcement release. The Cumulative Abnormal Return is the sum of all abnormal returns over the event window and a better
measurement.
With this Abnormal Return is the Cumulative Abnormal Return (CAR) calculated:
( ) (4)
Finally, to calculate if the CAR is significant or not, the t-value is calculated:
( ) (5)
To see what variables have effect on the CAR, a standard Ordinary Least Squares (OLS) regression will be done:
CAR = α + β1 LnDeal Value + β2 Method of payment + β3 Diversified deal + (6) β4 Investor Protection Index + β5 GDP growth + β5 Controlling Shareholder Acquirer +
β6 Control Variables.
The control variables are the year dummies to test the year fixed effects and the firm fixed effects test the initial stake of the acquirer in the target. In this OLS regression are robust standard errors added to get more reliable estimations. The tables and results are discussed in the next section.
3.2.2 Probit regression model
The second empirical model is based on the regression of Erel, Liao and Weisbach (2012). The reason for this is to see what factors (especially emerging market factor) affect the firms’ take-over chance of one country to acquire a firm from a cross-border country. The regression is adjusted for this research. To see what affect the binary dependent variable upmarket M&A affects, a probit regression is done. The data that is used are the 2,129 M&A observations.
13
Share Blockholder, Shareholders Protection index, Control Variables) =
Φ (α + β1 GDP growth + β2 Diversification + β3 LnDeal Value + β4 Method of payment + β5 Controlling Shareholder + β6 Shareholder Protection Index + β7 Initial Stake + β8 Year + β9 Listed Acquirer
The variables Initial Stake, Year and Listed Acquirer are the control variables.
Formula 8 tests the effect of a controlling shareholder on the choice of a diversification based M&A. Based on the results of Lane et al. (1998), firms with controlling shareholder are less involved in diversified deals.
PR(Diversification = 1 | GDP growth, Domestic deal , Ln Deal Value, Method of payment, (8)
Share Blockholder, Shareholders Protection index, Control Variables) =
Φ (α + β1 GDP growth + β2 Domestic deal + β3 LnDeal Value + β4 Method of payment + β4 Controlling Shareholder + β5 Shareholder Protection Index + β6 Initial Stake + β7 Year + β8 Listed Acquirer
The variables Initial Stake, Year and Listed Acquirer are the control variables. The model is also adjusted with an upmarket M&A dummy instead of the Domestic deal dummy.
Overall, the effect of each coefficient is difficult to interpret in a probit model. The coefficient affects the probability of an upmarket M&A or a diversified deal via the z-value Φ or cumulative standard normal distribution function (Stock and Watson, 2012, p. 432). The results can better be interpreted as positive of negative related with the dependent variable and the statistical significance determined the reliability of the coefficients.
It is also interesting to see what the effect of the variables is on the M&A price. What effect has an upmarket deal comparing to a downmarket deal. This robustness test will look at the assumption of Bertrand and Betschinger (2011) that emerging market firms cannot fully utilize the value of a cross border deal. Reason for this is that emerging market firms has a low experience and capabilities compared to developed countries. With a OLS regression this will be tested:
LnDeal Value= α + β1 Diversified deal + β2 Domestic deal + β3 Emerging Market firm + (9) β4 Method of payment + β4 Controlling Shareholder + β5 Shareholder Protection Index +
14 The control variables are the year and the firm fixed effect. The year dummies and the firm fixed effects are the initial stake of the acquirer in the target and if the acquirer is a listed company. Also in this OLS regression are robust standard errors used to get reliable estimations. Section 6 discussed the results of this robustness test.
4. Data
The used data sample is generated from the Bureau van Dijk dataset Zephyr. The dataset contains completed acquisitions from the period 2005-2014 all over the world. The period is based on the announcement date per deal. Since mergers give a direct positive change to the financial statements of a company they are excluded (Bertrand & Betschinger, 2011). Also the number of mergers is relative low, so this will not make biased estimations. Based on the research of
Renneboog and Martynova (2008b), deals where the acquirer has more than 50% of control of the target and acquirers with less than 10 Mln Euros in operating revenue are deleted. Besides that, they should be both independent corporations of each other, no mother-daughter constructions. The sample excludes the deals with Foreign Direct Investments (FDI), nationalizations and privatizations. To avoid the appearance of variable bias, incomplete and missing data will be excluded. There are 2,129 observations in this dataset. 626 (29.4%) observations are cross-border deals and 40 (1.88%) of them are upmarket deals. The overview of the total value of the M&A deals per year and the number of deals per year is show in figure 1. There can be observed that the lowest value of the deals was the year 2009 and the next years, during the financial crisis. But in 2014 the value boomed over 60 billion dollar total deal value. In 2014 increased also the number of deals to around 300. Appendix 3 provides a table with all cross-border countries in terms of acquirer and target.
To provide information about the upmarket M&As, the dummy variables Emerging Market Firm and
Advanced Market Firms are made based on the report of the S&P Dow Jones (2014). Appendix 1
contains the emerging market countries and developed market countries based on this report. With this data the dummy variables Upmarket M&A and Domestic Deal are created.
From all these upmarket M&As, the stock prices and index prices around the announcement date are generated from the databases Compustat Global and Datastream and linked by ISIN numbers of the acquiring firms.
The variable GDP growth is GDP growth per country per year, based on the announcement date. This data is merged to each deal and per year linked to the acquiring companies’ country. GDP growth is derived from the website of the International Monetary Fund from the World Economic Outlook Database April 2015.
15
Figure 1 Numbers of M&A and Deal Value per Year
This figure shows the total deal value per year and the total number of M&A deals per year in the sample that is used.
Also year dummies are created to cancel out possible time effects which can lead to bias estimations. Dummy variable Diversified deal is based on the Zephyr major sectors. The M&A deal is called
diversified if the acquirer major industry sector is not the same as the target major industry sector. In this sample are 36.45% of the deals based on diversification.
LnDeal Value is the logarithm of the Deal Value in thousands of Euros. The Deal Value is the value of
the payment the acquirer paid for the target. This payment can differ in Cash payments, Debt
payments, Share payments or Other methods. These variables are divided in dummies. Possible other
payments are earn-outs, loan notes, deferred payments. 52% of the total deals are cash payments, 6.9% are debt payments, 35.18% are share payments and just 5.17% are other ways of payments. To evaluate the effect of an acquirers major shareholder on a M&A, the variable Acquirer
Shareholder Direct % and the dummy variable Controlling Shareholder >50% are created. The Acquirer Shareholder Direct % gives the percentage of share of the acquiring firm hold by the largest
shareholder. The dummy variable gives 1 if there is a majority shareholder with more than 50% of the shares. 39.6% of the acquiring firms do have a Controlling Shareholder.
16 The variable % Initial Stake Acquirer in target is the percentage of stake the acquirer hold before the M&A which are all smaller than 50%.
As described earlier, Rossi and Volpin (2004) found evidence about corporate governance regulations. In line with these regulations, Djankov, La Porta and others (2008) created a
shareholders protection index per country and some large cities. This index reflects problems for shareholders like: self-dealing of managers, excessive compensation, tunneling problems and controlling shareholders power abuse. This index is built on several factors. The first one is the conflicts of interest regulation index. This index is the sum of the disclosure index, the extent of director liability index and the ease of shareholder suits index and divided by three. These three indexes can have a value from 0 to 10 and the higher the number the better the regulation. The second index is the extent of shareholder governance index. This index is the average of the extent of shareholder right index, shareholder governance index and strength of corporate governance
structure index. The average of the two main indexes is the strength of Minority Investor Protection Index or Investor Protection Index and has a value between 0 and 10.
For some country were the data of the protection index missing, so instead the values of the largest available city are used: Bangladesh- Dhaka, Brazil-Sao Paulo, China- Shanghai, India- Mumbai,
N Mean Median
Standard
Deviation Minimum Maximum
40 1.88% 0 0.1358 0 1 626 70.60% 1 0.706 0 1 776 36.45% 0 0.3645 0 1 1,107 52.00% 1 0.4997 0 1 147 6.90% 0 0.2536 0 1 749 35.18% 0 0.4776 0 1 110 5.17% 0 0.2214 0 1 2,129 - 2010 2.9154 2005 2014 2,129 4.67% 0% 11.8643 0 49.94 843 39.60% 0 0.4892 0 1 2,129 2.52% 2.36% 2.9534 -8.86% 15.24% 2,129 675.21 96.4 2745.25 0.00128 60736.53 2,129 6.17 6.7 2.4108 0 9.7
Table 1. Descriptive statistics
This table presents descriptive statistics per variable out of the total 2,129 M&A deals from 2005-2014. Upmarket M&A Domestic Deal Diversified Deal Cash Payment Debt Payment Shareholder Protection Index Shares Payment Other Payment Announced years Initial Stake Controlling shareholder GDP Growth
17 Indonesia- Jakarta, Japan- Tokyo, Mexico- Mexico City, Nigeria- Lagos, Russian Federation- Moscow and United States of America- New York City. The data is derived from the Doing Business website of the World Bank Group.
The dummy acquirer listed is created to look if listed firms act different than non-listed firms. An overview of the descriptive statistics is shown in table 1.
5. Results
This section will present and discuss the results of the tests. These tests are described in section 3.2. The first subsection discuss the Cumulative Abnormal Returns for upmarket M&As in a different time period, the second section will discuss the the OLS regression on this CAR. The final two subsections will test the probability of an upmarket M&A and a diversified deal in a probit model.
5.1 Cumulative Abnormal Return on announcement
Table 2 characterizes the Cumulative Abnormal Return of the upmarket M&A announcement for the acquiring party. 27 firms are used for this upmarket event. The CARs are calculated in 4 different time periods: [-1,+1], [-3,+3], [-20,+20] and [-30,+30]. The time period [-1, +1] means that the CAR is calculated for 1 day before the announcement date till 1 day after the announcement date. In the first days there is a small return for the shareholders of the acquirer. In the three-day window there is positive return around 0.4%. The first two CARs are not statistically significant. In the twenty- and thirty-day events the returns are positive and significant. In a twenty-day event the return is 4.57% and for the thirty- day event the return is 5.22%. This is not what was expected in the hypothesis. Andrade et al. (2012) had in the one-day event a -0.7% CAR and in the twenty-day event a -3.8% CAR. Jensen (1986), Rau and Vermaelen (1995) and Duchin and Schmidt (2013) assumed that agency problems, method of payment and an incomplete monitoring system in these emerging market countries would let to a negative return. But in case whether the shareholder is satisfied with the M&A, Chari et al. (2009) and Dyck and Zingales (2004) look at the overpayment theory. According to this theory, there is no overpayment in table.
So, with these CAR results, there can be assumed that the shareholders of the acquiring firm think that an upmarket M&A will give the firm synergies (Stucchi, 2012). According to Bertrand and Betschinger (2012) these synergies are knowledge for the home country and an improving of the technological and economic development.
18
5.2 Determinants of the Cumulative Abnormal Return
But it is also interesting to see which factors have effect on these positive Cumulative
Abnormal Returns. Especially the effect of the choice for a sector diversified deal. Table 3 reflects the OLS regressions on the dependent variable CAR. The value between the brackets is the t-value. The first two columns reflect the CAR on the one-day event. Like table 2, the variables are not very significant.
Columns 3 and 4 directly reflect the assumption of the overpayment theory of Chari et al. (2009). If the value of the deal increases with 1%, the CAR will drop with 0.00527%. So the higher the price, the lower the CAR will be. The diversified deal variable is in this event window negative with the year and firm fixed effects but not significant. The only significant variables are the method of payment and the controlling shareholder with control variables.
More interesting is the twenty-day event in column 5 and 6. As for the three-day event the deal value and the method of payment is statistically significant. If the payment is done with shares, the CARs are negative. This is notable because shareholders prefer to earn (cash) dividend and not to see that potential dividend flow out to another firm (Jensen, 1986). Column six has a significant negative effect of a diversified deal on the returns. So this is in line with the formulated hypothesis.
Overall, the effect of a diversified deal, a controlling shareholder, a share payment, increase in investor protection index and an increase in deal value does lower the Cumulative Abnormal Return in a twenty-day event. In case of a cash payment and the growth in GDP are positive related to the dependent variable.
Period CARs -0.0002 0.0044 0.0457*** 0.0522*** 27
The reported *** give the significance of the CARs at 1%, respectively.
Table 2. Cumulative Abnormal Returns
[-30, + 30] [-1, + 1]
No. Firms
Announcement Period Cumulative Abnormal Returns for Upmarket M&A acquirers
[-3, + 3] [-20, + 20]
19 Like column 6, column 8 is also completely significant. In this year and firm controlled model are all relations between the CAR and the variables the same, except the deal value. In this case the deal value is positive related to the CAR. The method of payment is as expected the same for the merger theory of Andrade et al. (2001) where the stock payments are negative and the no stock deals in the short run positive long run small negative.
In economic theory, this table does satisfy the hypothesis for the longer event days. A diversified deal lowers the CAR. Shareholders probably think that firms could use synergies due
[-30, + 30] [-30, + 30] (1) (2) (3) (4) (5) (6) (7) (8) -0.00162 -0.00245 -0.007*** -0.00527*** -0.00575*** -0.000675*** -0.00428 0.00269* (-0.71) (-0.98) (-3.58) (-3.13) (-2.91) (-0.40) (-2.55) (1.82) 0.0157 0.03 0.219*** 0.236*** 0.0247*** 0.0861*** 0.00133 0.0632*** (0.80) (1.23) (17.58) (13.04) (3.23) (9.55) (0.17) (8.11) 0.00998 0.0266 0.230*** 0.245*** (0.46) (0.85) (15.97) (10.76) -0.261*** -0.145*** -0.486*** -0.346*** (-10.14) (-5.64) (-16.99) (-12.36) 0.00645 0.00683 0.00139 -0.00393 -0.00140 -0.0429*** -0.0540*** -0.0745*** (0.67) 0.55 (0.16) (-0.36) (-0.16) (-4.81) (-6.81) (-10.26) 0.00395** 0.00474 0.00652*** 0.00106 0.00191 -0.00489** -0.000754 -0.0119*** (2.52) (1..32) (4.42) (0.32) (1.47) (-2.04) (-0.66) (-5.64) 0.00274* 0.00276* 0.000794 0.00204 -0.00644*** 0.00261** -0.00519*** 0.00366*** (1.85) (1.94) (0.52) 1.49 (-4.78) (2.38) (-4.49) (3.82) -0.00188 -0.0179** 0.000461 -0.0356*** 0.0111 -0.0176*** 0.0164*** -0.0158*** (-0.25) (-2.08) (0.07) (-4.56) (1.61) (-2.78) (2.79) (-2.94) -0.0340 -0.0495 -0.168*** -0.200*** 0.122*** -0.000628*** 0.160*** 0.0554*** (-0.91) (-1.19) (-5.66) (-5.28) (4.23) (0.03) (6.46) (2.77)
No Yes No Yes No Yes No Yes
No Yes No Yes No Yes No Yes
87 87 203 203 1189 1189 1769 1769
0.1093 0.3430 0.4413 0.5917 0.2191 0.4257 0.449 0.6027
Table 3. OLS regression on Cumulative Abnormal Returns
This table test the effect of the variables deal value, method of payment, sector diversification, investor protection, GDP growth and shareholder control on the Cumulative Abnormal Return (CAR). This OLS regression test on four different event windows: [-1, +1], [-3, +3], [-20, +20] and [-30, +30]. Each event window has two columns. In one column is the year and firm fixed effects not included and in one it is. Year fixed effects are the dummy variable for each year and firm fixed effects are the percentage initial stake of the acquirer in the target. To prevent omitted variable bias for the method of payment, the dummy variable share payment is in the first four columns omitted and the dummy variable debt payment in the last four columns omitted. The R-square is the measure of fit for each OLS regression. Robust standard errors are used to take into account for better estimations. The reported *, ** and *** give the significance of the variable at 10%, 5% and 1%, respectively.
Dependent variable: Cumulative Abnormal Return [-1, + 1] [-1, + 1] [-3, + 3] [-3, + 3] [-20, + 20] [-20, + 20]
Acquirer Controlling Shareholder Log Deal Value
Cash Payment
Constant
Year Fixed effects Firm Fixed effects
N R-squared Debt Payment Shars payment Diversified deal Investor Protection Index GDP Growth
20 resources and complementary resources in a non-diversified deal. The Free Cash-Flow of Jensen (1986) holds in terms of shareholders return because diversification in industry lowers the total gain of the firm.
5.3 Method of payment and the probability of Upmarket M&A
Table 4 reflects the results of the hypothesis that upmarket M&As based on diversification paid with cash of debt are more likely than a upmarket M&A paid with shares. The table exist of six differ probit regressions. The standard errors are calculated as robust standard errors. The value between the brackets is the t-value.
Column 1 shows that without the variable diversified deal the methods of payments are all significant. In this case, cash and debt payments have higher coefficients to test the likelihood of the upmarket M&A than the share payment. Column 2 adds year and firm fixed effects to the model. The variable diversified deal is also added and negative related for significance level of 10% to the
upmarket likelihood. Also the effect positive effect of a controlling shareholder is positive significant. In economic theory this means that the presence of a controlling shareholder the probability of an upmarket M&A increases.
Column 3 includes the variables Investor Protection Index and GDP growth as control variables. These variables are both significant. Column 3 concludes the same as column 1 and 2 for the method of payment.
Columns 4-6 are comparable to columns 1-3, but here are only the cross-border deals included. The same conclusions can be drawn, but only the not significant variable deal value is negative instead of positive.
This table 4 can be described in economic theory that diversification a negative effect had on the likelihood of an upmarket M&A. In upmarket M&As, firms prefer to acquire another company in a sector that is familiar. A cash or debt payment has a higher likelihood than a stock payment. This conclusion is in line with Rossi and Volpin (2004), but note that the likelihood of shares also positive is. It is in this regression not possible to see if cash of debt payments are more successful are than debt payments, but Table 4 confirms this supposition of Jensen (1986). Besides that, perhaps there can be an effect with the fact that foreign shareholder get shares in their firm. But this is not
supported by literature. Linn and Switzer (2001) also conclude that returns for cash based M&As are higher than for stock based. Table 3 also concludes this. The investor protection index, measured for the acquiring firm, is significant negative related to upmarket M&As. In this case, the higher the protection of investors, the less likely an upmarket M&A will happen. In the whole sample conclude there is a (little) positive relation between deal value and upmarket M&A. In the cross border sample
21 this is negative related. Overall, this table tend to conclude that upmarket M&As are more institution based than resource or industry based (Stucchi, 2012).
(1) (2) (3) (4) (5) (6) -0.295* -0.386** -0.337* -0.527** (-1.89) (-2.29) (-1.82) (-2.30) 0.0347 0.0195 0.0129 -0.0359 -0.0559 -0.0708 (1.06) (0.61) (0.37) (-0.80) (-1.23) (-0.0708) 3.708*** 3.881*** 5.031*** 4.200*** 4.213*** 4.992*** (48.17) (28.41) (15.62) (38.42) (25.89) (17.78) 3.645*** 3.840*** 4.866*** 4.149*** 4.280*** 4.991*** (16.76) (16.14) (12.69) (15.74) (14.65) (14.13) 2.931*** 3.028*** 3.685*** 3.863*** 3.882*** 4.512*** (14.77) (13.55) (9.38) (14.67) (12.92) (10.71) -0.149** -0.219*** (-2.38) (-2.89) 0.190*** 0.301*** (6.98) (6.49) -0.121 0.264* 0.236 -0.143 0.241 0.373 (-0.85) (1.74) (1.38) (-0.86) (1.26) (1.56) -5.946*** -7.112*** -7.544*** -5.148*** -6.074*** -6.175*** (-15.53) (-13.31) (-9.73) (-10.38) (-9.94) (-5.51)
No Yes Yes No Yes Yes
No Yes Yes No Yes Yes
2,129 2,129 2,129 626 626 626
0.0663 0.1535 0.3053 0.0356 0.1559 0.3982
Dependent variable: Upmarket M&A
N
Firm Fixed effects Log Deal Value
Cash Payment Debt Payment Shares payment Investor Protection Index Pseudo R-squared
This table looks at the likelihood of whether there is an upmarket M&A. The higher this binary variable is the more likely there will be an event involving an upmarket M&A. Columns 1-3 gives the results of the whole sample (2,129 obs) and columns 4-6 gives the results of only cross-border deals (626 obs). To prevent omitted variable bias in the method of payment, the dummy variable 'other' is removed. Year fixed effects are the dummy variable for each year and Firm Fixed effects are the percentage initial stake of the acquirer in the target and the dummy variable listed acquirer. There is no R-squared presented but a pseudo R-squared because a probit model does not measure the fit with a likelihood function (Stock and Watson,2012, p.440). Robust standard errors are used to take into account for better estimations. Because this is a probit model, the coefficients are shown in a cumulative standard normal distribution function. The reported *, ** and *** give the significance of the variable at 10%, 5% and 1%, respectively.
Table 4. Probit regressions on Upmarket M&A 2005-2014
All deals included Only cross-border deals included
Constant
GDP Growth acquirer Diversified deal
Acquirer Controlling Shareholder
22
5.4 Controlling Shareholder and the probability of a diversified M&A
Table 5 reports the results of the hypothesis that the presence of a large shareholder or controlling shareholder negatively affects the probability of an (upmarket) M&A deal based on diversification. The table exist of six differ probit regressions. The standard errors are calculated as robust standard errors. The value between the brackets is the t-value.
Columns 1 and 2 test the whole sample on a sector diversified deal. The fixed effects are included in column 2. In these two regressions is the domestic deal coefficient small negative. This implies that there is little difference between the likelihood of a diversified deal and if the deal is domestic or cross-border. The deal value is significant and negative related to diversification. The methods of payment are positive but not significant. All other variables in these columns are not significant.
In columns 3 and 4 is the upmarket M&A variable significant. The effect of this negative coefficient on the probability that a diversification deal =1 given that Upmarket deal =1 is Φ(-0.495)= 31.10% plus the effects of other variables in column 4. The controlling shareholder dummy is not significant but it is positive. This means in economic theory that the presence of a controlling shareholder in the acquiring firm is increasing the probability of a diversified M&A. This is not what was expected according to Lane et al. (1998).
The probit regressions in column 5 and 6 are done in a cross-border sample. Column 6 adds the year and firm fixed effects results that all the variables are statistically significant except method of payment and GDP growth. In column 6 are the variables upmarket M&A and controlling
shareholder both significant. This effect of both can be calculated as: probability that diversification=1 given that upmarket M&A=1 and controlling shareholder=1 is
Φ(-0.5+0.284)=41.49% plus the effects of other variables. So, in contrast with columns 3 and 4 which was done within the whole sample, the cross-border sample gives statistically significant results that the presence of a controlling shareholder a positive effect has on the probability of a diversified deal.
23 (1) (2) (3) (4) (5) (6) -0.0068 -0.0148 (-0.10) (-0.22) -0.457** -0.495** -0.433 -0.500** (-2.08) (-2.21) (-1.80) (-1.98) -0.0623*** -0.0565*** -0.0620*** -0.0558*** -0.0648** -0.0658** (-4.58) (-4.02) (-4.62) (-4.02) (-2.33) (-2.30) 0.0772 0.0135 0.0877 0.0271 0.257 0.229 0.63 (0.11) (0.71) (0.22) (1.31) (1.13) 0.0816 0.0464 0.0906 0.0583 0.308 0.264 (0.51) (0.29) (0.57) (0.36) ( 1.24) (1.03) 0.109 0.0414 0.105 0.0348 0.459* 0.391 (0.85) (0.32) (0.83) (0.27) (1.95) (1.62) -0.0173 0.018 -0.0179 0.015 -0.0602*** -0.0741** (-1.47) (0.77) (-1.52) (0.64) (-2.95) (-1.97) 0.0017 0.0182 0.01 0.0218* -0.0054 0.0227 (0.73) (1.58) (1.01) (1.87) (-0.24) (0.82) 0.0592 0.108 0.0566 0.112 0.263** 0.284** (1.03) (1.44) (0.98) (1.49) (2.50) (2.17) 0.35 -0.218 0.341 -0.231 0.401 0.326 (1.62) (-0.73) (1.63) (-0.79) (1.00) (0.63)
No Yes No Yes No Yes
No Yes No Yes No Yes
2,129 2,129 2,129 2,129 626 626
0.0089 0.0208 0.0103 0.0239 0.0327 0.0433
Acquirer Controlling Shareholder
Year Fixed effects
Table 5. Probit regressions on Diversification 2005-2014
This table looks at the likelihood of whether the deal is based on sector diversification. The higher this binary variable is, the more likely the focus is on a diversified M&A. Columns 1-4 gives the results of the whole sample (2,129 obs) and columns 5-8 gives the results of only cross-border deals (626 obs). To prevent omitted variable bias for the method of payment, the dummy variable 'other' is removed. Year fixed effects are the dummy variable for each year and Firm Fixed effects are the percentage initial stake of the acquirer in the target and the dummy variable listed acquirer. There is no R-squared presented but a pseudo R-squared because a probit model does not measure the fit with a likelikhood function (Stock and Watson,2012, p.440). Robust standard errors are used to take into account for better estimations. Because this is a probit model, the coefficients are shown in a cumulative standard normal distribution function. The reported *, ** and *** give the significance of the variable at 10%, 5% and 1%, respectively.
Domestic Deal
Dependent variable: Sector diversified M&A
Debt Payment
All deals included Only cross-border deals included
Firm Fixed effects
N
Pseudo R-squared Upmarket M&A
Constant Log Deal Value
Cash Payment
Share payment
Investor Protection Index
24
6. Robustness checks
This robustness section will check what variables affect the takeover price. In the theory we assume that emerging market firms cannot fully utilize the value of a cross-border deal. The test will look what factors will affect the price for emerging market firms. This test does not look at the upmarket M&A, but it looks at emerging countries itself.
Table 6 describes the effect of the on the deal value. The variable diversification is negative and significant related to the deal value. In cross-border deals, when there is a diversified deal, the effect of this variable is 35.7% lower. This can be explained by the fact that non-diversified deals will give firms direct synergies, for example in the cost structure and in the productivity (Bertrand & Betschinger, 2001). This is even the case in table 6 in the domestic deals as in the cross-border deals. In the case with all deals, the domestic deals are negative related to the deal value. Deals in the same country are having less value. A possible explanation is that firms from the same country can perhaps have more information about the other party. So, the optimal price can be easier found with this advantage.
When the acquirer is an emerging market firm, the deal value is lower. In the complete deal sample, the effect is -21.4% and in cross-border deal -52.1%. Possible reason for this result is that emerging market firms do not acquire very large targets, because of low M&A experience and capabilities (Bertrand & Betschinger, 2001). The deal value is higher in the presence of a controlling shareholder. Lane et al. (1998) found that controlled firms are monitored by the block holder and less involved in unrelated M&As. In this case are the cross-border deals not significant. So with this assumption of Lane et al. (1998) there can be conclude that better monitored firms do higher acquisitions. The methods of payment are positive en significant related to the deal value. The GDP growth in the acquiring country is not significant and negative related to the deal value. Countries with a better investor protection do lower valued acquisitions. In cross-border deals firms with better protected shareholders do lower valued acquisitions as well.
25 (1) (2) (3) (4) -0.412*** -0.366*** -0.363** -0.357** (-4.52) (-4.00) (-2.31) (-2.29) -0.811*** -0.660*** (-8.35) (-6.67) -0.293** -0.214 -0.474* -0.521* (-2.05) (-1.36) (-1.72) (-1.81) 0.206** 0.445*** -0.191 0.155 (2.34) (3.71) (-1.26) (0.82) 0.451*** 0.478*** 0.658*** 0.635** (2.60) (2.79) (2.69) (2.55) 1.228*** 1.203*** 1.200*** 1.231*** (5.24) (5.28) (3.61) (3.71) 0.672*** 0.639*** 0.669** 0.714** (3.63) (3.53) (2.07) (2.20) 0.0391** -0.0242 0.0215 -0.0241 (2.39) (-1.18) (0.71) (-0.64) -0.0423** -0.178*** -0.0663** -0.191*** (-2.33) (-4.69) (-2.24) (4.00) 11.74*** 12.39*** 11.95*** 12.21*** (54.36) (33.41) (40.04) (23.69) No Yes No Yes No Yes No Yes 2,129 2,219 626 626 0.0618 0.0940 0.0388 0.0868 Diversified Deal
Table 6 OLS regression on Ln Deal Value
Year Fixed effects Firm Fixed effects
N Domestic Deal Cash Payment Debt Payment Share payment Emerging Market Firm GDP Growth acquirer
Dependent variable: Ln Deal Value
This table looks at the effect on deal value. The first two colums describe all deals. The columns 3 and 4 describe only the cross-border deals. Under the coefficients is the t-value presented. The Year fixed effect is the dummy variable for each year and firm fixed effects are the percentage initial stake of the acquirer in the target and the dummy variable listed acquirer. To prevent omitted variable bias for the method of payment, the dummy variable other payment is omitted. The R-square is the measure of fit for each OLS regression. Robust standard errors are used to take into account for better estimations. The reported *, ** and *** give the significance of the variable at 10%, 5% and 1%, respectively.
Investor Protection Index
R-squared
All deals Cross-Border Deals
Acquirer Controlling Shareholder
26
7. Summary and conclusion
This paper analyzed the effect of an upmarket M&A in two ways. First, the Cumulative Abnormal Return for the shareholders of acquiring firm after a M&A announcement during 2005-2014. In the sample of 29 upmarket M&As are the CARs in the one-day and three-days event small negative and small positive. But in the twenty-days and thirty-days windows are the CARs 4.57% and 5.22% and statistically significant. The results of this test were not predicted by the related literature. Andrade et al. (2012) predicted a negative CAR for the [-1, +1] and the [-20, +20] event periods. Others estimated also a negative return because of agency problems, method of payment and incomplete monitoring of the managers (Jensen, 1986; Rau& Vermaelen, 1995; Duchin & Schmidt, 2013). Shareholders react negative on an industry diversified upmarket M&A. For the twenty-day window the effect is -4.29% and for the thirty-day window the effect is -7.45% and for both significant. This is in line with the results of the Free Cash-flow theory of Jensen (1986). He stated that an industry diversified M&A value destructive is. Also, Bertrand and Betschinger (2011) found that diversified deals miss the potential advantages of the cost and production synergies.
Second, the probability of an upmarket M&A and a diversified deal is calculated in a worldwide sample of 2005-2014 with 2,219 observations. This is measured by a probit model. The likelihood of an upmarket M&A is higher when payment method is by cash or debt instead of shares. This is exactly what Jensen (1986) and Rossi and Volpin (2004) founded. The likelihood of a diversified deal is not negatively affected in the presence of a controlling shareholder as Lane et al. (1998) predicted. In a cross-border based sample is this founding significant. These shareholders have a monitoring function of the managers to prevent any agency problems. Reason for this is because managers could act in a selfish way (Duchin & Schmidt, 2013).
To answer the research question ‘Which factors affect the return and probability of an
upmarket M&A focused on industrial diversification?’ there can be concluded that the choice for an
industry diversified M&A not an optimal choice is, especially in an upmarket M&A. This is in line with the declining probability of an upmarket M&A if there is a diversification factor. Cash or debt is the optimal choice to finance a upmarket M&A and the presence of a controlling shareholder will increase the CAR and surprisingly the probability of an upmarket M&A.
The limitations of this approach are that some financial factors are not included in the used models. For example Rau and Vermaelen (1995) and Erel et al. (2012) used the book-to-market factor as firm variable for the acquiring and the target firm and the deal value compared to the EBITDA. Also this model used one year-fixed factor in the model. For example the GDP growth can be different per year because of a financial crisis or a recession. A regression with different lags is perhaps an option to get more reliable estimators. In this case the year before the M&A
27 protection index. This index is the average of different measures and some of them could be
interpreted as itself. For example the separation of the conflict of interest index and the shareholders governance index can be made (Djankov et al., 2004). With this separation you will get more internal and external estimations about the governance regulations of firms.
With these results, an emerging market firms that consider acquiring a developed market firms could better use cash or debt as payment to acquire the other firm. Reason for this is that Cumulative Abnormal Returns for their own shares will be higher. But considering Linn and Switzer (2001) this choice can only be made if there is private information about potential synergies. So the conclusion that cash or debt bids are superior to stock bids cannot per se be made. Other
implications of this research are the large number of missing and incomplete data, especially for the upmarket acquirers. Lot of financial data cannot be found. Possible reasons for this are that the accessible databases do not provide this data. There should be more observations than a sample of 29 upmarket M&As to calculate the Cumulative Abnormal Return. A possible solution is to add Foreign Direct Investments (FDI) to the sample. Reason for this is that the total value of cross-border FDI deals from emerging markets was 53% in 2013 and 49% in 2007 (World Investment Report 2014, 2014, p. 44). But FDI have difficulties because of the presence of inter-company loans and some countries do not provide extensive information about FDI flows (Erel et al., 2012). This will give possible biased estimations.
Future research can extend this model by using other factors that were not included in this research. This research is mainly focused on factors that are close to the firm itself. Factors as the currency effect, law and accounting standards could be used as Erel et al. (2012) and Rossi and Volpin (2004) did. Also in this research the acquiring factors are used. Perhaps adding more target factors will increase the measure of fit of each model. The robustness check is added to see if the upmarket M&As are done with a lower deal value. Also factors as controlling shareholders and diversification are related to this size literature in terms of deal value. For future research table 6 can be extended. Another possible direction is to measure how successful an upmarket M&A is, in terms of
profitability of the firm itself. This can be done with the Tobin’s Q, the measure for firm performance. In this way the choices of emerging market firms between domestic or cross-border deals and
diversified or non-diversified deals can be evaluated.
Last recommendation for future research in the Cumulative Abnormal Return is also estimate all cross-border deals instead of only the upmarket M&As. In this case will emerging market firms be compared with other non-emerging market firms.
28
8. References
Andrade, G., Mitchell, M., & Stafford, E. (2001). New evidence and perspectives on mergers. Journal
Of Economic Perspectives, 15(2), 103-120.
Athreye, S., & Kapur, S. (2009). Introduction: The internationalization of Chinese and Indian firms trends, motivations and strategy. Industrial and Corporate Change, dtp007.
Bertrand, O., & Betschinger, M. (2011). Performance of domestic and cross-border acquisitions: Empirical evidence from Russian acquirers. Journal of Comparative
Economics, 40(3), 413-437.
Bodie, Z., A. Kane & A.J. Marcus (2011). Investments and Portfolio Management. New York: The McGraw-Hill Companies.
Chari, A., Ouimet, P. P., & Tesar, L. L. (2009). The value of control in emerging markets. Review of
Financial Studies, hhp090.
Chung, K. H., & Pruitt, S. W. (1994). A simple approximation of Tobin's q. Financial
management, 70-74.
Djankov, S., La Porta, R., Lopez-de-Silanes, F., & Shleifer, A. (2008). The law and economics of self- dealing. Journal of financial economics, 88(3), 430-465.
Duchin, R., & Schmidt, B. (2013). Riding the merger wave: Uncertainty, reduced monitoring, and bad acquisitions. Journal of Financial Economics,107(1), 69-88.
Dyck, A., & Zingales, L. (2004). Private benefits of control: An international comparison. The Journal
of Finance, 59(2), 537-600.
Erel, I., Liao, R. C., & Weisbach, M. S. (2012). Determinants of cross‐border mergers and acquisitions. The Journal of Finance, 67(3), 1045-1082.
Huyghebaert, N., & Luypaert, M. (2010). Antecedents of growth through mergers and acquisitions: Empirical results from Belgium. Journal of Business Research, 63(4), 392-403.
International Monetary Fund (2005). World Economic Outlook Database, April 2015
(https://www.imf.org/external/pubs/ft/weo/2015/01/weodata/download.aspx), June 2015. Jensen, M. C. (1986). Agency cost of free cash flow, corporate finance, and takeovers. Corporate
Finance, and Takeovers. American Economic Review,76(2).
Lane, P. J., Cannella, A. A., & Lubatkin, M. H. (1998). Agency problems as antecedents to unrelated mergers and diversification: Amihud and Lev reconsidered. Strategic Management
Journal, 19(6), 555-578.
Linn, S., & Switzer, J. (2001). Are cash acquisitions associated with better postcombination operating performance than stock acquisitions? Journal Of Banking & Finance, 25(6), 1113-1138. Martynova, M., & Renneboog, L. (2008a). A century of corporate takeovers: What have we learned
and where do we stand?. Journal of Banking & Finance,32(10), 2148-2177.
Martynova, M., & Renneboog, L. (2008b). Spillover of corporate governance standards in cross- border mergers and acquisitions. Journal of Corporate Finance, 14(3), 200-223.
Mishkin, F.,2009. The Economics of Money, Banking and Financial Markets. Pearson (US), Ninth Edition.
Mitchell, M. L., & Netter, J. M. (1994). The role of financial economics in securities fraud cases: Applications at the Securities and Exchange Commission. The Business Lawyer, 545-590. Rau, P. R., & Vermaelen, T. (1998). Glamour, value and the post-acquisition performance of
acquiring firms. Journal of financial economics, 49(2), 223-253.
Rossi, S., & Volpin, P. F. (2004). Cross-country determinants of mergers and acquisitions. Journal of
Financial Economics, 74(2), 277-304.
Sehgal, S., Banerjee, S., & Deisting, F. (2012). The impact of M&A announcement and financing strategy on stock returns: Evidence from BRICKS markets. International Journal of
Economics and Finance, 4(11), p76.
Stock, J.H., M.M. Watson (2012). Introduction to Econometrics. London: Pearson, Third Edition. Stucchi, T. (2012). Emerging market firms’ acquisitions in advanced markets: Matching strategy with
29 resource-, institution-and industry-based antecedents. European Management Journal,
30(3), 278-289.
S&P Dow Jones Indices LLC. (2014) Country Classification Methodology. McGraw Hill Financial . S&P Down Jones (2015) Emerging Markets Indices Fact Sheet. McGraw Hill Financial.
World Investment Report 2014, Investing In The SDGs: An Action Plan, UNCTAD, United
30
9. Appendix Appendix 1
Emerging Markets
Brazil Indonesia South Africa
Chile Malaysia Taiwan
China Mexico Thailand
Colombia Morocco Turkey
Czech Republic Peru United Arab Emirates Egypt Philippines Greece Poland Hungary Qatar India Russian Federation Developed Markets
Australia Ireland Singapore
Austria Israel Spain
Belgium Italy Sweden
Canada Japan Switzerland
Denmark Luxembourg U.K.
Finland New Zealand U.S.
France Netherlands
Germany Norway
Hong Kong Portugal
Sources:
S&P Dow Jones Indices LLC. (2014) Country Classification Methodology. McGraw Hill Financial S&P Down Jones (2015) Emerging Markets Indices Fact Sheet. McGraw Hill Financial
31
Appendix 2
Source:
S&P Dow Jones Indices LLC. (2014) Country Classification Methodology. McGraw Hill Financial
Frontier Emerging Developed
Initial Eligibility Criteria
Req Req Req Req Req Req Req Req Req Req Req Req Req Req Req= required criterion
GDP Criteria
GDP (PPP) per capita of greater than $15,000
A minimum of two Req
A minimum of three
Req Domestic market cap of over $15 bn
Settlement period of T+3 or better Sovereign Debt rating of BB+ or above Non-occurrence of hyperinflation
No significant foreign ownership restrictions Freely-traded foreign currency
Summary of S&P Dow Jones Indices Country Classification requirements
S&P Country Classification Criteria
Domestic market cap of over $2.5 bn Domestic turnover value of over 1 $ bn Exchange development ratio of over 5%
32
Appendix 3
Acquiror country N % Target country N %
Aus 8 1.3% Arg 1 0.2% Aut 1 0.2% Aus 24 3.8% Bel 12 1.9% Aut 3 0.5% Bul 2 0.3% Ba r 1 0.2% Ca n 46 7.3% Bel 10 1.6% Chi 7 1.1% Ber 4 0.6% Col 1 0.2% Bra 4 0.6% Cro 1 0.2% Bul 1 0.2% Den 3 0.5% Ca n 36 5.8% Fi n 7 1.1% Ca y 9 1.4% Fra 45 7.2% Chi l 4 0.6% Ger 37 5.9% Chi 10 1.6% Gre 4 0.6% Cyp 3 0.5% HK 6 1.0% Den 7 1.1% Hun 1 0.2% Egy 4 0.6% Ice 2 0.3% Fi n 3 0.5% Indi 8 1.3% Fra 28 4.5% Indo 1 0.2% Ger 30 4.8% Ire 14 2.2% Gi b 1 0.2% Is r 2 0.3% HK 4 0.6% Ita 18 2.9% Ice 1 0.2% Ja p 53 8.5% Indi 1 0.2% Ken 1 0.2% Indo 2 0.3% Kor 3 0.5% Ire 6 1.0% Kuw 1 0.2% Is r 11 1.8% Lux 2 0.3% Ita 8 1.3% Ma l a 4 0.6% Ja m 1 0.2% Ma l t 1 0.2% Ja p 3 0.5% Mex 4 0.6% Jor 1 0.2% Net 22 3.5% Kor 2 0.3% New 4 0.6% La t 1 0.2% Nor 7 1.1% Li t 1 0.2% Pa p 2 0.3% Lux 2 0.3% Phi 4 0.6% Ma c 3 0.5% Pol 4 0.6% Ma l a 3 0.5% Por 2 0.3% Ma u 1 0.2% Qa t 2 0.3% Na m 1 0.2% Rus 6 1.0% Net 22 3.5% Sa u 1 0.2% New 2 0.3% Si n 10 1.6% Ni g 2 0.3% Sl o 1 0.2% Nor 8 1.3% Sou 4 0.6% Pa p 1 0.2% Spa 19 3.0% Pol 3 0.5%
Tota l cros s -border Tota l cros s -border
