Financial Constraints and the Acquisition Premium:
International evidence
ABSTRACT
University of Groningen Faculty of Economics and Business International Financial Management
Supervisor: Dr. H. Gonenc Co-assessor: Prof dr. A. de Ridder
January 11, 2019
Name: David Markusse Student Number: S2522012
Preface
TABLE OF CONTENTS 1. INTRODUCTION ... 5 2. LITERATURE REVIEW ... 8 2.1 Financial constraints ... 8 2.2 Financial depth ... 10 2.3 Method of payment ... 11 2.4 Geographical diversification ... 13 2.5 Industrial diversification ... 14
3. DATA AND METHODOLOGY ... 16
3.1 Data ... 16
3.2 Sample selection ... 16
3.3 Sample distribution ... 17
3.4 Descriptive statistics ... 19
3.4 Methodology ... 23
3.4.1 Measurement of bidder premium ... 23
3.4.1 Firm’s financial constraints ... 23
3.4.2 Financial Depth ... 24 3.4.3 Method of payment ... 24 3.4.4 Geographical diversification ... 25 3.4.5 Industrial diversification ... 25 3.4.6 Control Variables ... 26 4. METHODOLOGY ... 28 5. EMPIRICAL RESULTS ... 30 5.1. Summary statistics ... 30
5.2 Bidder premium and financial constraints ... 31
5.3 Financial Constraints and Financial depth... 34
5.4 Financial Constraints and Method of payment ... 34
5.5 Financial constraints and Geographic Diversification ... 35
5.6 Financial Constraints and Industrial Diversification ... 35
6. CONCLUSION ... 42
7. REFERENCES ... 45
APPENDIX A ... 48
1. INTRODUCTION
Recently, the net worth of global Mergers and Acquisitions (M&A’s) has hit a new record of 3,3 trillion U.S. dollars in deal value year to September 30, 20181. With low global interest
rates, the monetary policies are aiming to stimulate corporate investments resulting in a record in corporate investment since the financial crisis in 2008. It is not only obvious that these forms of corporate investment still have a huge impact on national and global capital markets, it also makes clear that it is highly lucrative for individual firms to engage in M&A’s.
Since 1960, research has established the prevalence of enormous deal premiums with some of acquirers paying double the price for a target company (Walking and Edmister 1985; Jensen 1994). This indicates that acquirers are willing to pay large premiums but it has been displayed that their gain is negative or neutral (Eckbo, 2007). While it has been shown that target gains are on average positive. This fact illustrates that most of the acquisition gains flow to the target company (Hayward & Hambrick, 1997). Past research found that financially constrained firms act differently in terms of investment decisions, but sometimes have to make costly acquisition decisions in order to survive (Kaplan and Zingales, 1997). Therefore, a better understanding of the possible drivers of the deal premium could reduce this impediment and even relieve their financial constraints (Williamson and Yang, 2013). By empirical investigation of global acquisitions between 1990 and 2013, this thesis tries to capture the influence of country-level and deal-level variables on the bidder premium that financially constrained acquirers face.
Therefore, possible outcomes on the impact of deal level factors on the bidder premium are relevant for their internal decision-making.
One of the leading theories that could explain the occurrence of deal premiums is the theory of synergies. Synergies are the gains from acquisitions in terms of enhanced combined business capabilities, management efficiencies or knowledge and wealth transfers (Jensen, 1994). Other studies acknowledge the importance of synergies driving the deal premium, and found that they tend to be higher when synergies are apparent (Madura and Ngo, 2008). To a large extent, these studies find empirical evidence on the influence of isolated factors on the deal premium (Walking and Edmister; 1985), but not much research has been dedicated on companies’ ability to finance the acquisitions. Evidence on this topic could improve the understanding of constrained and unconstrained companies about the level of deal premium, and ultimately reduce bidder premium.
According to di Giovanni (2005) other research on corporate investment initiatives highlights the importance of the depth of the financial markets to the possibility to attract capital. In his investigation he finds that the host-country financial depth could impact the amount of M&A’s as projects with lower net-present value could be accepted. He finds evidence that companies that are active in countries with greater financial depth have a higher incidence of participating acquisitions due this smoother access to raise external capital. So the financial depth has influence on the capital attraction abilities, and so causes the firm to be less financially constrained by the external environment. Therefore, it is assumed that firm individual financial constraints interact to a high degree with the financial depth of the market. The combination of the two could bridge the internal and external factors that lead to the overall financial constraints limiting corporate decision-making.
resources and prefer stock as payment method (Almeida, Campello and Weisbach, 2004; Denis and Sibilkov, 2010; Alschwer, Sibilkov and Zaiats, 2011). But whether this is an advantage in terms of premium payable is questioned.
Geographical diversification strategies are argued to increase potential value (Roll, 1986; Moeller and Schlingemann, 2005; Harford, Humphery-Jenner and Powell, 2012) as a result of synergistic and efficiency gains but other studies argue that often these M&A’s are subject to managerial hubris and cultural clashes that decrease the firm value and therefore the willingness to pay a premium (Li, Li and Wang, 2016). Industrial diversification could be another important driver for the amount of the deal premium. Prior literature stresses that it could be either value increasing by international experience of these M&A’s according to Moeller and Schlingemann (2005) or value decreasing as it is stressed to be a result of managerial hubris, agency costs and inefficiencies (Roll, 1986; Denis, Denis and Yost, 2002). The theoretical foundations brought me to propose the following research question:
2. LITERATURE REVIEW 2.1 Financial constraints
Firms’ long- and short-term strategies are directly and indirectly affected by country’s capital market imperfections. These market imperfections increase the cost of attraction of external capital and have therefore direct influence on MNE’s investment activities (Fazzari, Hubbard and Petersen, 1988; Kaplan and Zingales, 1997). When capital market imperfections are considered as relatively large, it could be argued that those firms operating in these nations are respectively more financially constrained than others. These companies have to deal with those external circumstances and place more value on the available internal resources (Denis and Sibilkov, 2010). Also, the investors of these constrained companies place more value on the marginal dollar kept within the firm relative to those of unconstrained organizations (Faulkender and Wang 2006; Almeida, Campello, and Weisbach, 2004). The reason why the marginal dollar is more valuable for constrained companies, is a result of superior investment decicions compared to unconstrained companies. According to Denis and Sibilkov (2010), the association between value and investment is stronger for constrained firms. Constrained organizations need to be more careful in their investment decisions, as they are restricted in their internal funds (Almeida et al., 2004). As a cause of their limited resources, they only select projects with the highest return causing more value on the marginal dollar kept within the firm.
Harford (1999) finds in his study a significant influence of excess cash on the bidder premium paid during acquisitions. He finds a positive relation between a firm’s excess cash and the acquisition premium paid by the bidding company. This evidence suggests that capital abundant firms have a higher probability of over-payment during an acquisition. Vladimirov (2015) finds evidence too that firms which use debt as financing method for their acquisitions, they tend to overbid. He proposes an alternative view to the explanation of payment method to the bidder premium. He argues that when a firm is financially constrained by its external market it pays a lower acquisition premium. On the other hand, when companies are capital abundant as a consequence of easier access to competitive financing they tend to overbid. In these markets it is less costly to attract capital, leading to higher premiums when bidding for a target firm. The local capital market would financially constrain the firm in this way of argumentation. As a consequence, the firm would pay a lower premium for the target company as a matter of scarcity.
Moreover, prior research argues that managers from financially constrained firms are monitored in a better way by their shareholders than unconstrained firms are (Whited and Wu, 2006) From this point of view, having fewer resources to allocate for future investments means that every amount allocated should be in the right manner. The scarcity of financial resources could cause the bidding firm to be more discrete in terms of investment policies and therefore earn a higher return of their investments.
When resources are not abundant during an offer, it can be assumed that the bidder cares to an increasing level about the amount payable for the target. A financially distressed acquirer has to make valuable investment decisions as a result of the inability to accept all positive net-present-value projects (Denis and Sibilkov, 2010). Therefore it could be assumed that a financially constrained bidder will pay a lower premium. In line with this argumentation the following hypothesis is developed:
2.2 Financial depth
The possibility to obtain financial resources is according to a study of Di Giovanni (2005) determined by the financial depth of the bidding company. Financial depth is described as the size of the financial market, and captures the business environment’s ability to attract external capital. He found that MNE’s M&A activity is to a large extent influenced by the financial depth, and it has found a positive relation between financial depth and the amount cross-border investment intensity. Kandilov, Leblebicioğlu and Petkova (2017 also find evidence on the impact of source and host-country levels of financial depth and acquirers financial distress is relieved by the use of cross-border M&A’s. By means of this, they argue that there’s a substitution effect between local and source country financial depth. These studies state that financial depth is a main factor behind the decision to engage in a cross-border investment. Financial depth covers financial stabilization, economic growth and financial integration of a country (Di Giovanni, 2005; Caprio and Honohan, 2001). When the financial depth of the acquiring company’s home country is relatively big, they argue that the acquiring company has easier access to external capital.
H2:‘’The financial depth of acquiring company’s country will positively moderate the relation between acquirer’s financial constraints and the bidder premium’’.
2.3 Method of payment
Expanding on the discussion of the possible influences of bidder’s financial constraints and financial depth on the eventual acquisition premium, prior literature stresses the importance of payment method as main determinant of the bidder premium.
According to Almeida et al. (2004), capital market frictions cause firms to hold more cash reserves. By saving more cash, these firms try to safeguard the ability to finance future investment projects. Financially constrained firms also see opportunities in domestic or cross-border M&A’s, but are not able to obtain the right amount of resources to cover the deal with cash. Alschwer et al. (2011) focus in their study of 2.739 M&A’s on the payment method of constrained firms. They find significant influences on the preference of stock over cash as payment method for financially constrained firms. They also find evidence that when firms are constrained and have low (high) stock valuations, they will finance the transaction with cash (stock) in order to reduce future financing frictions. Moreover, they find that financially constrained acquirers accumulate their cash savings a year prior to the M&A. In their study they point out that acquirers are even possible to finance the deal with cash but due to uncertain financial capital markets they tend to favour stock transactions. They argue that the opportunity costs of cash are too high relative to the cost of trading stock, and thereby hold the possibility to internally fund other future investments besides large M&A’s. Other related findings state that financially distressed firms rather issue costly debt under constrained capital conditions than forego the option of financial flexibility. Related research from Graham and Harvey (2001), DeAngelo and DeAngelo (2007), and Denis and McKeon (2011) all show the effects of financing frictions on the payment method in acquisitions.
The target notices the reasoning behind the stock payment and will in his turn demand a bigger share of stock, thereby increasing the premium demanding from the acquirer. Furthermore, a transaction with stocks would mitigate the acquirers’ risk of targets’ shares over-valuation. This strategy could be seen as a contractual hedge contingent to the performance of the target company (Laamanen, 2007; Shleifer and Vishny, 2001). Further to giving a negative signal to the target company, the acquirer also signals to the target the illiquid nature of the stock offer is of importance to the formation of the new price. As the stock offer is considered as a relative less liquid offer compared to cash, prior studies found that this could increase the bidder premium (Walking and Edmister 1985). Moreover, Myers and Majluf (1984) stress that stock offers coincide with more transaction costs and legal procedures that financially harm the target. According to this, it is logical to assume that the target will demand a higher premium. From this theory, it could be deduced that stock offers will increase the bidder premium payable by the acquirer. Based on these theoretical constructs, following hypothesis is developed:
H3a: ‘’cash as payment method will positively moderate the relation between acquirer’s financial constraints and the bidder premium’’
Next to these assumed positive relationships between stock as payment method and the bidder premium, there are plausible arguments why stock as payment method could actually decrease the premium. When an acquirer pays with equity shares, the target decreases its immediate tax liability to that point in the future where it actually sells the shares (Jensen, 1994). From this point of view if the target accepts a total cash offer, it would have to pay taxes immediately. As a result of this negative side of the offer, the target company will demand a higher price increasing the bidder premium. Likewise, risks are shared between the acquirer and the target company when the transaction is made with stock. A stock payment causes the target firm to be a controlling part of the bidding firm after the transaction. This enhances the amount of trust of the acquirer in the prosperity of the acquisition and therefore becomes a proponent of the deal. By virtue of this, the target accepts a lower premium at this point in time as it is entitled to combined future profits after the acquisition (Laamanen, 2009). From this point of view the a contrasting hypothesis is formulated:
2.4 Geographical diversification
Next to the assumed effect of payment method for the amount of bidder premium paid, previous literature stresses that MNE’s geographical diversification strategies have impact on the eventual premium paid. Geographical diversification is considered as a diversification strategy that operates outside the borders of acquirer’s national country. Previous research has found evidence about the potential benefits that acquirers could attract when they involve themselves in cross-border M&A’s. These benefits can include increased corporate risk management capabilities, access to improved technologies and favourable foreign government policies (Moeller and Schlingemann, 2005).
Furthermore, having intangible assets to deploy on foreign targets could imply additional firm value expressed in synergistic and efficiency gains (Harford et al., 2011; Moeller and Schilingemann, 2005). All these benefits increase the willingness of the acquiring company to take-over the target as the future gains are promising. Prior research has not yet found whether these factors have significant influence on the bidder premium but it could be assumed observing abnormal returns in other relevant research on cross-border M&A’s. Moeller and Schlingemann (2005) have found in their work that these cross-border M&A’s create less value and significantly less abnormal returns. A higher premium payable would directly decrease the abnormal returns but that is just one of the factors of influence on the abnormal returns. So this argumentation is not entirely sufficient.
Increased agency cost and managerial hubris could cause distressing investment decisions in turn leading to elevated costs of global diversification strategies. Roll (1986) found empirical evidence that individual decision making as a consequence of managerial hubris increases the overall price paid for an acquisition in large MNE’s. From this point of view, it could be inferred that geographical diversification is associated with either high synergistic expectancies or inadequate managerial and capital allocation decisions. Both could be linked to higher premiums paid for international targets. Therefore the following hypothesis is proposed:
On the other hand, in the study of Denis and Denis (2002) is found that firms that are geographically diversified are valued at a discount. They propose that the underlying cause for this negative relationship between firm value and geographical diversification lies within the context of acquirer’s inefficient investment activities. Moreover, acquiring companies can be aware of the pitfalls of geographic diversification. They could realize in advance the potential agency problems and the disappointing synergistic gains as proposed by Roll (1986); Harford et al. (2011); Moeller and Schlingemann, (2005). Consciousness about the incidence of disappointing post-acquisition developments could lead to the demand for a lower price or even a discount. Further to organizational pitfalls when acquiring abroad, prior research (Li, Li & Wang, 2016) has shown that cultural clashes could result in an increased probability of information asymmetries during cross-border M&A’s and accumulating the post-merger costs. Experienced companies in the area of M&A’s could recognize these pitfalls and may include these facts in their decision-making, resulting in a lower premium.
H4b: ‘’ Cross-border acquisitions will negatively moderate the relationship between acquirer’s financial constraints and the bidder premium’’
2.5 Industrial diversification
conclusion has been drawn that it would decrease the bidder return around the announcement due to the decrease in shareholder certainty. These theories state that the probability of managerial empire building is observable by this kind of acquisitions and that there is a higher possibility of over-payment (Denis et al., 2002). Therefore, the following hypothesis is developed:
3. DATA AND METHODOLOGY 3.1 Data
We used the data from the Thomson Reuters SDC database for completed M&A deals from the period 1990 till 2013. The subtracted data include acquirer’s financial values, transaction specific values and identity related values regrettably, only information related to target’s identification was available. All this information was necessary in order to investigate the effect of firm’s payment methods, industrial diversification and geographical diversification on the bidder premium.
M&A’s that are conducted in this period which include financial firms and public institutions (SIC Code 6000 and SIC 9000) are deleted from the dataset. These firms are intrinsically different from the other industries in terms of different capital structure. After this we retained a sample of 3052 completed domestic and cross-border M&A deals. In addition, official national financial data has been obtained from the World Bank. This data includes GDP, GDP growth, Total Stock Market Capitalization. Last, all observations are winsorized to the 1th and 99th percentile in order to alleviate the impact of outliers and reducing the skewness of the variables (Gujarati and Porter, 2009).
3.2 Sample selection
The study includes M&A completed deals from 41 countries including only listed bidders and targets. Table shows the criteria that has been applied in order to retail a full sample of 3520 completed domestic or cross-border acquisitions deals.
TABLE 1. SAMPLE CRITERIA
1.Bidder premium Deals from which bidder premium is unavailable are deleted from the sample. 2. Industry
exclusion
Deals that inhibit M&A deals from financial corporations or institutions are excluded from the sample (Kandilov et al., 2016).
3. Country identification
Transactions that involve unknown target or source countries are omitted from the sample.
4. Acquisition threshold
Acquirers need to obtain at least 50% of target’s shares in order to ensure that the M&A are ‘control deals’ (Bessler and Schneck, 2015).
5.Currency The deal contains dollar value.
3.3 Sample distribution
After inspection of table 2 (Appendix A) the following properties can be attached to the sample. When looking at the amount of deals, the majority of the deals is executed domestically in Japan with a total of 1417 deals (39%). The United States 522 deals (14%) and on the third place France with 212 completed acquisitions (6%). The sample consists out of 42 different countries.
By observing table 2 (Appendix A), the distribution of constrained and unconstrained firms can be inspected per country. The amount of constrained firms is 18% and the amount of unconstrained firms is 82%. The total amount of acquisitions with information on the deal premiums is 3588. When inspecting table 3 it is observed that in some countries2, unconstrained firms only pursue acquisitions. But the amount of acquisitions would be insufficient for building a compelling case. In Japan 93% of the total acquirers are unconstrained. The United States is that percentage lower with 60% unconstrained and in France 89% of the bidding companies are unconstrained.
Maybe it is interesting to note that Canada has a relative high percentage of acquisitions in the sample and remarkable is that 45% of the acquisitions is executed by constrained acquirers. In Poland (16) only 13% of the acquirers is unconstrained in their financial resources. In table 4 (Appendix C), the amount of acquisitions per year is observable. The year with the highest amount of acquisitions is 2007 (295) followed by 2006 (252) and 2008 (250). It is striking to see that after 2008, the beginning of the financial crisis3, the absolute amount of acquisitions
decreases while the relative amount of unconstrained acquisitions increases. This supports the theory of Almeida et al. (2004). The amount of constrained acquirers decrease and could indicate that the importance of cash reserves increases. Constrained companies only accepts investments with the highest NPV projects and otherwise save their scarce monetary resources.
In table 4 the average amount of acquisition premiums per country is observable. The percentage is the acquisition premium relative to target’s market price. The highest average
2 Countries that involve acquisitions solely from unconstrained acquirers, with the amount between brackets;
premium is of the Czech Republic with a total of 190,67, but this accounts only for 1 acquisition. Next Japan only pays 3,38% above the bidders market price with a total of 1417 acquisitions. The United States pays on average
TABLE 4. BIDDER PREMIUMS PER COUNTRY
Country N % Mean Median
Australia 200 5% 24,06 16,55 Austria 6 0% 25,58 14,59 Bahrain 1 0% 13,64 16,64 Belgium 18 0% 6,09 5,03 Brazil 25 1% 28,26 29,14 Canada 160 4% 19,55 14,20 Chile 5 0% 22,53 22,18 China 46 1% -1,20 -0,29 Colombia 5 0% 63,73 62,73 Czech Republic 1 0% 190,67 190,67 Denmark 16 0% 9,37 17,43 Finland 32 1% 21,32 19,30 France 212 6% 25,78 20,38 Germany 96 3% 20,93 15,01 Greece 12 0% 7,45 11,64 Hungary 2 0% 6,08 6,08 India 81 2% 31,63 20,40 Indonesia 8 0% 14,71 10,77 Ireland 6 0% 72,42 57,79 Israel 8 0% 20,90 22,86 Italy 41 1% 16,90 17,38 Japan 1417 39% 11,59 3,38 Luxembourg 6 0% 20,06 14,50 Malaysia 64 2% 25,06 13,14 Mexico 15 0% 18,10 2,96 Netherlands 44 1% 34,32 27,96 New Zealand 26 1% 13,06 9,13 Norway 40 1% 15,29 10,70 Philippines 10 0% 7,07 6,98 Poland 20 1% 10,15 1,22 Portugal 11 0% 14,07 6,20 Singapore 77 2% 22,06 14,00 South Africa 38 1% 25,36 17,21 Spain 25 1% 22,66 12,94 Sri Lanka 7 0% 37,87 23,53 Sweden 57 2% 31,91 20,48 Switzerland 59 2% 36,29 20,52 Thailand 38 1% 31,19 27,04 Turkey 6 0% 47,75 30,07
United Arab Emirates 3 0% 21,73 32,65
United Kingdom 172 5% 26,52 18,92
United States 522 14% 25,46 15,60
Total 3638 100% 19,18 11,63
for only 11,63% extra. This reveals that there are very high premiums that skew the distribution of the average of premium paid per country, this increases the necessity for winsorizing.
3.4 Descriptive statistics
TABLE 6. DESCRIPTIVE STATISTICS
Panel A Totals
N Mean Median Min Max St. Dev.
Acquirer Div. Payer 3520 0.83 1 0 1 0.38
Firm Level Size 3520 16.89 16.79 9.78 22.68 3.07
Leverage 3520 0.38 0.38 0 0.93 0.24
Market to Book 3520 2.62 1.69 0.08 20.89 3.10 Tobin's Q 3520 1.69 1.27 0.55 9.81 1.39
ROA 3520 0.08 0.06 -0.20 0.31 0.07
Cash to Assets 3520 0.14 0.11 0.00 0.63 0.13
Deal level Bidder Premium 3520 18.06 10.51 -74.86 190.67 38.04
Method of Payment 3520 0.56 1 0 1 0.50
Geographical Div. 3520 0.73 1 0 1 0.44
Industrial Div. 3520 0.51 1 0 1 0.50
Attitude 3520 0.81 1 0 1 0.39
Deal value 3520 3.34 3.15 0.10 8.32 3.07
Country level Mcap/GDP 3520 4.41 4.41 1.88 5.72 0.46
Note: all definitions of variables are observable in Appendix B, table 5. All variables are winsorized at the 1th and 99th percentile except for dummy variables. Further, the total amount of observations included in the
regressions are lower than stated the for variables in tables 2 to 4 due to missing values at certain variables.
The correlations observed from the correlation matrix show that there is a substantial correlation between the acquirers ROA and whether it pays dividends to its shareholders (0.28). Theoretically, it could be reasoned why this might be observable. When a company profitable, often measured by ROA it has a higher incidence of paying out dividend to its shareholders as there’s more profit from firm activities.
TABLE 7. PEARSON CORRELATION MATRIX
Note: All definitions of variables are observable in Appendix B, Table 5. The full sample consists out of 3502 acquisitions between 1990 and 2013. The denoted star (*) is the significance level 0.01
Another high correlation is observable between Mcap/GDP and Tobin’s Q (0.23). This correlation is probably observable as both variables are shaped by the market capitalization. Tobin’s Q is a variable that contains it and Mcap/GDP by the total national value of stock market capitalizing5. Next, a remarkable high negative correlation is observed between Size and Mcap/GDP (0.29). Size is measured in terms of book value, so no obvious explanation could be inferred from this. A negative correlation is found between Tobin’s Q and Leverage (-0.22). It could be argued that when a firm has a large debt outstanding, the market value of the company is lower. The highest negative correlation is observed between Leverage and Cash to Assets (-0.41). This is in accordance with the pecking order of finance theory6, so when firms have more leverage they tend to have less cash reserves.
Lastly, a notable significant correlation is observed between Cash to A. and Tobin’s Q (0.31). A logical assumption behind this correlation could be that firms are increasingly valued by the market when they have more cash on hand respecting their their assets. After all, having more cash-reserves as a firm could be a safeguard for a save investment. To conclude, no striking theoretically impossible correlations are observed, and there are no correlations that exceed the threshold of 0.70. Therefore no variables need to be deleted from the analysis.
5 See appendix B, Table 5 for detailed definitions of the variables.
6 This theory implies that firms will burn first through their cash reserves, then issue debt and at last issue equity
3.4 Methodology
3.4.1 Measurement of bidder premium
In this description of the variable used for bidder premium the definition of Officer (2007) is applied, ‘’the aggregate amount of each form of payment (cash, equity, debt, etc.) offered to target shareholders 4 weeks prior to the acquisition’’ subtracted by the market value of the target firm as a percentage of the market value of the target. We use the data 1 month prior to the announcement date of the acquisition, in order to correct for the potential of increase in premium because of a run-up effect. The values that are in line with the variable could be either positive or negative. Positive values mean that the acquirer pays a premium for the acquisition and negative values mean that the acquisition is offered at a discount. This study uses the following definition to describe the variable:
Bidder Premiu𝑚𝑖 =
𝑷𝟎− 𝑺𝑴𝟏𝑴 𝑺𝑴𝟏𝑴
Whereby DP1m is the deal premium that that is calculated by subtracting the market value (𝑆𝑀1𝑀) of the offer price (𝑃0) divided by the market value of the target. The variable is winsorized at the 1st
and 99th percentile in order to correct for outliers that potentially skew the data (Gujarati and Porter, 2009).
3.4.1 Firm’s financial constraints
We identify financially constrained acquirers using a dummy variable whether the acquiring company pays dividend. It is argued that constrained companies do not have the cash reserves to pay dividends to their shareholders. Whether the acquiring firm pays dividend is therefore a good proxy for the categorization to determine if acquiring firm is financially constrained or not. This is widely used to fully measure level of financial constraints or frequently used an important subset of it in an index. (Kaplan and Zingales, 1997; Whited and Wu; 2006 Hadlock and pierce, 2010). In order to use this variable in the data analysis a dummy variable is created to check whether the acquiring firm allows dividends during the entire year that acquisition is conducted (Lament et al., 2001; Alschwer
it does not (constrained). The following notation is used to denote the dummy variable financial constraints:
𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑡𝑠 = 0 𝑫𝒊𝒗𝒊𝒅𝒆𝒏𝒅 𝑷𝒂𝒚𝒆𝒓𝒊
3.4.2 Financial Depth
In order to examine the effect on bidder premiums paid the host country financial depth as mentioned earlier is considered, therefore two widely accepted proxies are used to capture this variable. Kandilov et al. (2017) make use of country’s total stock market capitalization as a percentage of the GDP calculated 7. The variable will be used to capture the depth of the host countries’ financial equity markets from 1990 until 2013. Improvements in financial depth over these years will cause an increase in capital raising abilities for those located in that particular host country. The following variables are created:
FINANCIAL DEPT𝐻𝐼 = (
𝑺𝑴𝒄 𝑮𝑫𝑷)𝒊
Where FINANCIAL DEPT𝐻𝐼 is the size acquirers’ financial markets. This is calculated by dividing its total stock market capitalization by the GDP.
3.4.3 Method of payment
A dummy variable is created for the payment method for the use of regressions. When an M&A is conducted with cash the dummy variable becomes 1, when a payment is conducted with stock the dummy turns 0. A minority of M&A’s is conducted with fragmented payment methods, meaning that for a part of it is paid for with cash and a part with stock. As this study aims to clearly distinguish between the influences cash and stock on the dependent variable, it needs to set a restriction. When an M&A is conducted for 51% or more with a specific payment method, then it will mean that the variable will get the value that is assigned to that category. Data descriptions show that the amount of acquisitions conducted solely with cash is 1989 (55%). Furthermore, the number of acquisitions conducted totally with stock was 502 (14%), and mixed or unknown payments resulted in a total of 1148 (31%). But as the threshold of 51% is used to denote whether the payment is cash or stock, the
distribution of cash and stock payments is different. The following notation is used to describe the variable method of payment:
𝑀𝐸𝑇𝐻𝑂𝐷 𝑂𝐹 𝑃𝐴𝑌𝑀𝐸𝑁𝑇𝐼 = 1 𝐂𝐀𝐒𝐇𝐏𝐀𝐘𝐌𝐄𝐍𝐓,𝒊
3.4.4 Geographical diversification
For assessing whether an M&A is conducted across geographical borders, a dummy variable is created. When a geographical border has not been crossed, the dummy variable turns to 1. In other words, the dummy variable is 1 when it concerns a domestic M&A and it turns 0 when it is a cross- border acquisition. The amount of cross-border acquisitions that were conducted is relatively large. The total amount of cross-border deals is 2575 (71%), and the amount of domestic deals was 1064 (29%). The variable:
𝐺𝐸𝑂𝐺𝑅𝐴𝑃𝐻𝐼𝐶𝐴𝐿 𝐷𝐼𝑉.𝐼 = 0 𝐃𝐎𝐌𝐄𝐒𝐓𝐈𝑪𝒊
The above-mentioned notation is used to denote geographical diversification dummy variable
3.4.5 Industrial diversification
The assessment of industrial diversification is likewise geographical diversification. The dummy variable is created to check whether the M&A is focused on the same industry or that actual diversification strategies are in place. The variable turns 1 when it involves an M&A of the same industry and it concerns industrial diversification when the dummy variable turns 0. The following notation is used to describe the variable:
3.4.6 Control Variables
Acquirers Tobin’s Q is used as a control variable. Moeller, Schlingemann and Stulz (2004) report that a high Tobin’s Q is a signal for the overvaluation of bidder’s stock. This could result in an increase of the total price a target demands for the acquisition. As this is a possible proxy for overvaluation, the variable is included in the model. The Tobin’s Q is calculated by dividing the total market value of the companies assets by the total replacement costs of the companies assets (book-value).
With regard to hostile versus friendly take-overs, according to the study of Walkling and Edmister (1985) there is a significant influence of the attitude of the acquisition and the bidder premium. They report that when a take-over is considered as hostile the management could defend themselves in several ways which results into an increased total price, therefore also increasing the premium. A dummy variable is created in order to control for potential increase in premium.
Deal value is measured in millions of dollars. This variable is included for control as it is stressed that a higher deal value increases the acquisition premium in an absolute manner. According to Madura Viale and Ngo (2012), the absolute amount of the deal value in dollars has a positive influence on the premium paid by the acquirer.
Acquirer size is measured in book value. It has been shown that relative premiums decrease by the size of the bidding or target organization. The decision to use the book value is to diminish the impact of market fluctuations on the size measurement of the company that could be influenced by the measurement in different points of time (Hayward and Hambrick, 1997).
Next, acquirer cash to assets is included as a control variable. Prior research has found significant effects of free cash flow (FCF) on the bidder premium (Madura Viale and Ngo, 2012). The abundance of cash reserves indicates less financial distress, and causes the acquirer to offer a higher price for the acquisition. The amount of cash is taken as ratio of the bidder’s total asset, to see for every acquirer the amount of cash reserves relative to its operational size. This ratio gives an improved indication on the level of cash that could be spent on such investments, and is free from market fluctuations (Madura Viale and Ngo, 2012).
4. METHODOLOGY
In order to examine the relation of the independent variables financial constraints, and the firm-level moderators method of payment, geographical and industrial diversification and the country level moderator financial depth on dependent variable the bidder premium, the following regressions are estimated: (1) 𝐵𝑖𝑑𝑑𝑒𝑟 𝑃𝑟𝑒𝑚𝑖𝑢𝑚𝑖,𝑡 = 𝛼𝑡+ 𝛽1 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖+ 𝛽2 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖+ 𝑌𝑒𝑎𝑟𝑡+ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑡+ 𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝑐+ 𝜀𝑖,𝑡 (5)𝐵𝑖𝑑𝑑𝑒𝑟 𝑃𝑟𝑒𝑚𝑖𝑢𝑚𝑖,𝑡 = 𝛼𝑡+ 𝛽1 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖+ 𝛽2 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑑 𝑑𝑒𝑝𝑡ℎ𝑖+ 𝛽3𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟 𝑋𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑝𝑡ℎ𝑖 + 𝛽4 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖+ 𝑌𝑒𝑎𝑟𝑡+ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑡+ 𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝑐+ 𝜀𝑖,𝑡 (6)𝐵𝑖𝑑𝑑𝑒𝑟 𝑃𝑟𝑒𝑚𝑖𝑢𝑚𝑖,𝑡 = 𝛼𝑡+ 𝛽1 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖+ 𝛽2 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑝𝑡ℎ𝑖+ 𝛽3𝑀𝑒𝑡ℎ𝑜𝑑 𝑜𝑓 𝑃𝑎𝑦𝑚𝑒𝑛𝑡𝑖 + 𝛽4𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟 𝑋 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑝𝑡ℎ𝑖+ 𝛽5 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖 𝑋 𝑀𝑒𝑡ℎ𝑜𝑑 𝑜𝑓 𝑃𝑎𝑦𝑚𝑒𝑛𝑡𝑖+ 𝛽6 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖+ 𝑌𝑒𝑎𝑟𝑡 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑡+ 𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝑐+ 𝜀𝑖,𝑡 (7)𝐵𝑖𝑑𝑑𝑒𝑟 𝑃𝑟𝑒𝑚𝑖𝑢𝑚𝑖,𝑡 = 𝛼𝑡+ 𝛽1 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖+ 𝛽2 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑝𝑡ℎ𝑖+ 𝛽3𝑀𝑒𝑡ℎ𝑜𝑑 𝑜𝑓 𝑃𝑎𝑦𝑚𝑒𝑛𝑡𝑖+ 𝛽4𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑖 + 𝛽5 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖 𝑋 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑝𝑡ℎ𝑖+ 𝛽6 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖 𝑋 𝑀𝑒𝑡ℎ𝑜𝑑 𝑜𝑓 𝑃𝑎𝑦𝑚𝑒𝑛𝑡𝑖 + 𝛽7𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖𝑋 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑖+ 𝛽8 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖+ 𝑌𝑒𝑎𝑟𝑡+ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑡+ 𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝑐+ 𝜀𝑖,𝑡 (8) 𝐵𝑖𝑑𝑑𝑒𝑟 𝑃𝑟𝑒𝑚𝑖𝑢𝑚𝑖,𝑡 = 𝛼𝑡+ 𝛽1 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑑𝑒𝑟𝑖+ 𝛽2 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑝𝑡ℎ𝑖+ 𝛽3𝑀𝑒𝑡ℎ𝑜𝑑 𝑜𝑓 𝑃𝑎𝑦𝑚𝑒𝑛𝑡𝑖+ 𝛽4𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑖 + 𝛽5 𝐺𝑒𝑜𝑔𝑟𝑎𝑝ℎ𝑖𝑐𝑎𝑙 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑖+ 𝛽6 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑑𝑒𝑟 𝑋𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑝𝑡ℎ𝑖 + 𝛽7 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖 𝑋 𝑀𝑒𝑡ℎ𝑜𝑑 𝑜𝑓 𝑃𝑎𝑦𝑚𝑒𝑛𝑡𝑖+ 𝛽8𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖𝑋 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑖𝑎𝑙 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑖 + 𝛽9𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑 𝑃𝑎𝑦𝑒𝑟𝑖𝑋 𝐺𝑒𝑜𝑔𝑟𝑎𝑝ℎ𝑖𝑐𝑎𝑙 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑖+ 𝛽10 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖+ 𝑌𝑒𝑎𝑟𝑡+ 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑡+ 𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝑐+ 𝜀𝑖,𝑡
diversification and Geographic diversification are dummy variables that equal 1 when the deal concerns respectively cash payments, domestic acquisitions and acquisitions within the same industry.
In the above displayed models (1,5,6,7 and 8), are the used models for hypothesis testing. In model 1 the effect of financial constraints is tested on the bidder premium. In model 5 the moderating effect of cash payment is tested. In model 6, the moderating effect of industrial diversification is tested. Furthermore, model 7 and model test respectively the moderating effects of Industrial and Geographical diversification on the bidder premium. 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑖 are the control variables that are included in the model (Tobin’s Q, Attitude, Deal Value, Size, Cash-to-assets). Also, 𝑌𝑒𝑎𝑟𝑡, 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑡 and 𝐶𝑜𝑢𝑛𝑡𝑟𝑦𝑐 describe the year-, industry- and country-fixed-effects that are included in each model.
To control for potentially biased results the Gauss-Markov assumptions have to be met in order to reduce the risk of biased results. By inclusion of a constant term, the zero mean assumption of the standard error will be fulfilled. To check for potential multicollinearity the VIF8 is inspected. The
5. EMPIRICAL RESULTS 5.1. Summary statistics
In table 8 the difference of constrained and unconstrained acquirers can be inspected on firm-, deal- and country-level factors. The means and medians are compared between the groups to check for significant differences between the two. When firm level factors are inspected, the means and medians for the variables; Size, Tobin’s Q, ROA, and Cash to Assets are significantly different between the groups on 0.01 significance level. The higher means for unconstrained firms of the variables Size and ROA are supported by our theory that unconstrained firms are more profitable and bigger in size (Erickson and Whited, 2000; Almeida, Campello and Weisbach, 2004). Next, unconstrained firms have a slightly higher leverage than constrained firms with a median difference of 4% on a 0.05 confidence level. This also supports our notion that unconstrained firms have higher leverage due to their easier access to external capital. At deal level, certain highly significant values directly point out structural differences between constrained and unconstrained acquirers. These variables are Method of payment, Industrial diversification and Deal Value. Their means and medians are significant on a 0.01 level. Unconstrained acquirers slightly prefer cash as payment method over stock payment, which is in accordance with our expectations of unconstrained firms being cash abundant and therefore preferring that as payment method. The mean and median are higher for industrial diversification for constrained firms, meaning that constrained acquirers prefer diversifying into other industries as it could relieve their financial constraints according to research of Williamson and Yang (2013)9. It is also noted that unconstrained companies slightly prefer to acquire firms from different countries. Cross-border acquisitions are seen as a riskier investment and could be seen as an argument that supports this significant (0.10) difference. When inspecting the country level financial depth difference, I see a negative difference for Mcap/GDP. It is argued that the financial depth causes firms being less constrained in terms of financial possibilities. So the difference between means and medians of both variables should be positive for unconstrained acquirers. But these are not tested on their significance as they are country level factors that could not be tested on differences by a single firm-level factor (financial constraints).
9 Williamson and Yang (2013) found in their research that acquirers are able to relieve their financial constraints by
TABLE 8. SUMMARY STATISTICS
Note: All definitions of variables are observable in Appendix B, Table 5. The full sample consists out of 3052 acquisitions between 1990 and 2013. The reported significance levels are denoted with respectively *=0.01, **=0.05 and ***=0.10. Mean and median comparisons are based respectively a t-test and a Wilcoxon signed-rank test.
5.2 Bidder premium and financial constraints
In table 9 the outcomes of the OLS regression are observable. In Model 1 impact of being financially constrained is tested on the bidder premium. The control variables that were relevant according to prior theoretical arguments10 are included. In this model country, year and industry fixed effects are implemented in order to control these influences on the bidder premium. When we observe the coefficient of Dividend Payer in model 1 we see that is positive and significant on a 0.05 level. This means that there is a positive influence between being unconstrained and the bidder premium. This supports our theory. In this research I argue that due to an improved availability of financing possibilities it would increase the bidder premium (Roll, 1986; Jensen 1986; Harford, 1999; Vladimirov; 2015). Furthermore, we see that the coefficient for attitude is positive and significant on a 0.10 level in model 1 and on a 0.05 level for the models 2 to 4. This is remarkable as it shows that when the deal concerns a friendly acquisition, the deal premium is higher then when it would be a hostile over (Walking and Edmister, 1985). A possible argument why friendly
overs would have higher deal premiums is due to their longer negotiation processes preceding these kinds acquisitions, that eventually increase the premium. It is noted that Deal Value is significant to a 0.01 level for all of the models. Meaning that a higher transactional value causes the premium11 relatively to increase. It is observed that might be that greater transactional values occur during acquisitions of greater size. Acquisitions of greater organizational size causes more complexity and could increase agency-costs when these companies are put together. The target might demand a higher price as a result of this (Roll, 1986; Denis et al., 2002). Increased bargaining capacity would be another argument. A rather strange observation is found when analysing the variable ROA. The coefficient is for models 1 to 4 negative and significant to a 0.10 for the models 1, 2 and 4 and significant to a 0.05 level for model 3.
When inspecting models 2 till 4, we observe direct relations with payment method and the acquisition premium. The coefficients are positive and significant on a 0.01 level. Which means that cash payment directly increases the deal premium which is in congruence with related literature on this topic (Meyers and Majluf, 1984; Walking and Edmister; 1984). Furthermore, the impact of industrially diversification strategies on the bidder premium is positive and significant in model 4 with to a 0.10 level. Which means that it increases the deal premium what is in accordance with our theoretical stated theoretical arguments. Next, when inspecting the financial depth variable in model 4 we see a positive coefficient of Mcap/GDP on the deal premium. This is in accordance with our theory, that when the financial depth of a bidder’s country is larger the deal premium would be higher. Unfortunately this is not confirmed empirically by our results.
TABLE 9. REGRESSION COEFFICIENTS
Model 1 Model 2 Model 3 Model 4
Measure of Financial Constraints
Dividend Payer 5.120** 5.012** 5.020** 4.619** [2.322] [2.320] [2.299] [2.343] Financial Depth Mcap/GDP 3.270 [3.466] Deal variables Method of Payment 4.097*** 3.878*** 4.100*** [1.394] [1.392] [1.405] Industrial Diversification 2.433 2.745* [1.485] [1.499] Geographical Diversification -6.756*** -6.478*** [1.994] [2.034] Control variables Leverage 0.366 0.157 0.437 0.819 [3.776] [3.764] [3.764] [3.820] Tobin’s Q 1.038 1.057 0.950 0.704 [0.699] [0.702] [0.708] [0.714] ROA -24.290* -25.881* -27.522** -26.553* [13.460] [13.498] [13.463] [13.663] Attitude 3.332* 3.743** 3.678** 4.437** [1.761] [1.764] [1.750] [1.747] Deal value 2.129*** 2.217*** 2.124*** 2.195*** [0.404] [0.402] [0.403] [0.406] Cash to Assets -1.050 -1.268 -0.274 0.567 [7.222] [7.240] [7.180] [7.326] Size -0.208 -0.200 -0.465 -1.150 [2.499] [2.482] [2.481] [2.489] Country ID -0.083 -0.086 -0.118 -0.115 [0.081] [0.081] [0.081] [0.082] (Constant) -2.245 -2.861 5.945 -8.352 [13.303] [13.011] [13.326] [18.166]
Country fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Industry fixed effects Yes Yes Yes Yes
adjusted R2 0.122 0.124 0.129 0.131
Observations 3520 3520 3520 3417
5.3 Financial Constraints and Financial depth
In model 5 (table 10) the interaction between a firm being financially constrained and the financial depth of its market is tested in the bidder premium. When I take a look at the moderating variables I may directly conclude that there is not a significant relationship between the interaction of Dividend Payer and Financial Constraints to the bidder premium. My argument on the enforcing relationship between being financially constrained and financial depth causing the premium to be positively moderated is not valid for this sample. Next, the direct influence of the Mcap/GDP on the premium is present and significant to a 0.10 level. This is in agreement with our argument that the financial depth has a positive influence on the bidder premium as there is more incidence in M&A activity12 causing the demand to acquire to increase and therefore potentially increase the premium. But nothing can be said with certainty because we do not observe this variable to be significant in the models 6 to 8. To conclude financial depth of the acquiring companies country doesn’t moderate the relationship between financial constraints and the bidder premium and evidence on the influence of financial depth on the bidder premium stays largely unidentified.
5.4 Financial Constraints and Method of payment
Model 6 tests hypothesis 3a and 3b. These hypotheses are tested between for eventual negative and the positive moderation effects of payment method to the relationship of financial constraints and the bidder premium.
In models 2 to 4, the direct effect of cash as payment method on the bidder premium is tested. I observe for the three models a positive effect. For these models I could say with 99% certainty that cash has a positive direct influence on the bidder premium. Model 6 does not indicate a direct effect of cash on the bidder premium. The moderating effect of payment method and financial constraints is also not significant on any level. The coefficient of the interaction term is positive in favour of hypothesis 3a but nothing can be said due to insignificant results. When looking at the adjusted R squared the number slightly rises meaning that the explanatory power of the model increases to a small degree. Finally, both hypothesis 3a and 3b are not supported by our results.
12 According to the research of Di Giovanni (2005) the depth of the financial markets causes more M&A activity causing
5.5 Financial constraints and Geographic Diversification
Looking at the direct effects of geographical diversification and the bidder premium in models 3 and 4, we observe negative and highly significant coefficients. The results of these models are significant to a 0.01 level. This indicates that when a firm domestically acquires, it pays a lower premium. This finding is in agreement with Moeller and Schlingemann, (2005)13 When the moderating effects
between financial constraints and industrial diversification is tested in model 7, no generalizing conclusions may be derived from this. We observe a negative but insignificant coefficient for this interaction effect. When a company is not constrained and domestically acquires, the premium will be lower. This is in accordance with our hypothesis 4a, when a company is constrained and engages in a cross-border acquisition, it will pay a higher premium. Unfortunately this interaction effect is not statistically significant.
5.6 Financial Constraints and Industrial Diversification
After inspection of model 3 and 4, the direct effect of industrial diversification on the bidder premium is tested. For both models the coefficient is positive and significant to a 0.10 level in model 4. This indicates that the premium is larger for domestically executed acquisitions. When the direct effect is inspected in model 8, the coefficient increases to a high degree meaning a stronger relationship but still insignificant. As I observed all three models for direct effects, the majority is insignificant. Therefore we cannot infer any valid conclusions from this.
In pursuing to check for moderating effects of industrial diversification on the main relationship whereby nothing can be said with certainty. The coefficient is not significant on any level. The coefficient is negative, meaning that when a company is unconstrained and acquires another company from a different industry, it pays a lower premium. This partly supports our hypothesis 5 as it positively moderates the relation between financial constraints and industrially diversified acquisitions. Following the arguments of Roll, (1986) and Denis et al. (2002), industrial diversification increases agency related costs and gives rise to organizational inefficiencies. Organizations are aware of these common pitfalls and extra costs and take this into consideration when making an offer for a company from a different industry.
TABLE 10. REGRESSION COEFFICIENTS
Model 5 Model 6 Model 7 Model 8
Measure of Financial Constraints
Dividend Payer 26.417 4.293 9.171** 6.621* [19.391] [-3.087] [-2.875] [-3.408] Financial depth Mcap/GDP 8.386* 3.762 3.614 4.130 [4.889] [3.548] [3.462] [3.556] Deal variables Method of Payment 0.359 [3.738] Geographical Div. -3.470 [4.153] Industrial Div. 4.594 [3.765] Moderating variables
Dividend payer X Mcap/GDP -2.588 -4.615
Dividend Payer X Method of Payment 4.789
[4.031]
Financial Constraints X Geographical Div. -4.370
[4.586]
Financial Constraints X Industrial Div. -1.801
[4.108] Control variables Leverage 0.974 0.533 0.516 1.011 [3.842] [3.820] [3.849] [3.844] Tobin's Q 0.733 0.819 0.680 0.782 [0.706] [0.708] [0.709] [0.711] ROA -22.173 -24.974* -24.820* -23.625* [13.585] [13.639] [13.605] [13.661] Attitude 4.062** 4.490** 3.961** 4.142** [1.754] [1.761] [1.744] [1.756] Deal value 2.188*** 2.285*** 2.157*** 2.136*** [0.407] [0.405] [0.406] [0.410] Cash to assets -0.572 -0.303 0.693 -0.460 [7.356] [7.385] [7.340] [7.333] Size -0.996 -1.061 -1.355 -0.899 [2.514] [2.489] [2.513] [2.499] Country ID -0.088 -0.084 -0.108 -0.085 [0.082] [0.082] [0.083] [0.082] (Constant) -36.935 -16.319 -9.993 -20.731 [22.779] [18.699] [18.645] [18.526]
Country fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Industry fixed effects Yes Yes Yes Yes
adjusted R2 0.124 0.126 0.128 0.125
Observations 3417 3417 3417 3417
5.7 Financial Constraints, Financial Depth, Method of Payment, Geographical Diversification and Industrial Diversification
By inspecting models 1 to 4, the main dependent variable dividend has a significant (0.05) and positive relationship to the bidder premium. From this we may conclude that when a company is financially constrained it pays a lower premium. When we incrementally add method of payment, geographical industrial diversification to the model we observe a small increase in explanatory power14 of the overall model. The highest is in the last model being 0.131. The moderating effects of financial depth, payment method, geographical and industrial diversification are not significant in any of the models. The adjusted R squared is lower for models 4 to 8 compared to models 1 to 4, meaning that the latter models have less explanatory power than the first. Moderating effects between financial constraints and any of the other independent variables are not present. when I look at models 1 to 8, the control variable Deal Value seems significant to a 0.01 level and positive for all of the models, implicating that the level a higher transactional value causes the relative deal premium to increase as well. I also observe positive coefficients for attitude at model 1 to 8, with significance levels of 0.05 except for model 1 (0.10). This shows us that a friendly attitude causes the premium to increase. Last, the control variable ROA, which is generally a proxy for profitability, causes the deal premium to decrease. In other words, a better profitability causes the premium to go down.
5.8 Robustness
First of all, the term ‘financially constrained’ is rather extensive. Many researchers use different kinds of proxies to capture financial constraints. The use of Size, Tobin’s Q, Market-to-book and acquirer Cash are not uncommon proxies of financial constraints (Madura viale and Ngo, 2012; Hayward and Hambrick, 1997; Lang and Walking 1991). Following the argument of Kaplan and Zingales (1997), when companies face financial constraints there is a wedge between the costs of internal and external sources of finance. Recent research from Mielcarz, Osiichuk and Behr (2018), argue that net working capital (NWC) 15 is a more accurate proxy of financial constraints because it
does not cover growth prospects like cash flow proxies of financial constraints do. They argue that NWC is a more precise proxy for financial constraints as it covers available internal resources of the
14 Explanatory power is measured by the adjusted R squared. The adjusted R squared measures the Goodness-of-Fit
company. It is argued that a high level of NWC is a proxy for being financially unconstrained as it indicates that there are sufficient amount of internal resources. An advantage of using NWC over Dividend Payer for financial constraints is that the firm does not make the decision itself. To clarify, in this study I assume that companies that do not pay dividends to their shareholders are financially constrained but in reality there might be additional motivations than saving resources to fill the gaps of capital market frictions exclusively. Fact is that firms themselves chose whether or not to pay dividends, and thereby have influence to our categorization of being financially constrained. This isn’t the case when NWC is used.
After inspection of table 11, I observe slight differences. The significance level of NWC as main predictor is 0.10 in models 9 to 11, and the 0.05 at model 12. No changes in direction of coefficients are observed for all of the models. The significance levels and direction of coefficients for Method of payment and Geographical Diversification are similar. In this instance, the correctness of hypothesis 1, stating that being financially constrained reduces the bidder premium, is correct and robust. Moreover, when models 13 to 16 are examined more differences are raised between the original models of moderators. The measure for financial constraints is not significant in model 15 but is highly significant in model 16. Al of the deal-level variables, Payment method, Geographical Div. and Industrial Diversification, are now significant to respectively a 0.01, 0.01 and 0.05 level. These findings highlight the importance of the variables to determination of the bidder premium. No changes in the direction of these coefficients are found. By analysing the results of this regression it can be said with 99% certainty that the moderation effects of geographic and industrial diversification are in place.
Geographic Diversification negatively moderates the relation between financial constraints and the bidder premium, which is in accordance with our hypothesis 4a. When a company is constrained and engages in a cross-border acquisition, it will pay a higher premium.
lower premium. This contradicts our last hypothesis. Also, the R squared a bit higher for models 13 to 16 meaning that they have more explanatory power.
TABLE 11. REGRESSION COEFFICIENTS
Model 9 Model 10 Model 11 Model 12
Measure of Financial Constraints
NWC 13.287* 12.812* 12.961* 14.383** [7.263] [7.261] [7.235] [7.253] Financial depth Mcap/GDP 3.097 [3.491] Deal variables Method of payment 4.024*** 3.812*** 4.014*** [1.392] [1.390] [1.404] Industrial diversification 2.563* 2.892* [1.490] [1.500] Geographical Diversification -6.669*** -6.389*** [2.003] [2.042] Control variables Leverage 3.213 2.900 3.241 3.966 [4.176] [4.164] [4.167] [4.231] Tobin's Q 1.114 1.131 1.026 0.790 [0.696] [0.699] [0.706] [0.711] ROA -26.335* -27.846** -29.483** -28.814** [13.434] [13.475] [13.445] [13.638] Attitude 3.369* 3.775** 3.710** 4.465** [1.773] [1.776] [1.761] [1.758] Deal value 2.163*** 2.249*** 2.157*** 2.229*** [0.406] [0.404] [0.405] [0.408] Cash to assets 1.179 0.859 1.862 2.876 [7.485] [7.505] [7.451] [7.593] Size 0.194 0.181 -0.058 -0.730 [2.479] [2.464] [2.464] [2.480] (Constant) -3.661 -4.124 4.520 -9.309 [13.160] [12.884] [13.185] [18.100]
Country fixed effects Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes
Industry fixed effects Yes Yes Yes Yes
adjusted R2 0.123 0.126 0.131 0.133
Observations 3502 3502 3502 3399
I may hereby conclude that initial findings relating to the relevance of financial constraints to bidder premium are robust. The additional regressions add to the relevance of being financially constrained and diversifying geographically or industrially.
Model 13 Model 14 Model 15 Model 16
Measure of Financial Constraints
NWC -45.463 13.595 2.941 50.842*** [50.506] [9.034] [8.916] [12.101] Financial depth Mcap/GDP 4.092 3.671 3.956 3.723 [3.570] [3.565] [3.565] [3.507] Deal variables Method of Payment 4.173*** [1.415] Geographical Div. -6.890*** [2.024] Industrial Div. 3.341** [1.499] Moderating variables NWC X M.cap/GDP 13.737 [11.414] NWC X Method of Payment 0.903 [10.696] NWC X Geographical Div. -49.283*** [13.273] NWC X Industrial Div. 24.340** [11.472] Control variables Leverage 3.697 3.495 3.850 2.868 [4.246] [4.239] [4.222] [4.236] Tobin's Q 5.132** 4.892** 5.015** 5.370** [2.381] [2.385] [2.375] [2.374] ROA 0.922 0.895 0.885 0.713 [0.702] [0.703] [0.694] [0.696] Attitude -25.661* -27.143** -26.671** -28.869** [13.665] [13.653] [13.551] [13.480] Deal value 4.147** 4.512** 4.119** 3.960** [1.774] [1.772] [1.773] [1.750] Cash to assets 2.219*** 2.323*** 2.164*** 2.224*** [0.409] [0.407] [0.412] [0.405] Size 2.295 1.705 1.141 2.480 [7.667] [7.651] [7.599] [7.583] Country ID -0.424 -0.580 -0.406 -0.440 [2.488] [2.482] [2.483] [2.491] (Constant) -22.177 -19.352 -20.857 -9.606 [19.001] [18.349] [18.587] [18.663]
Country fixed effects Yes YES YES YES
Year fixed effects Yes YES YES YES
Industry fixed effects Yes YES YES YES
adjusted R2 0.125 0.127 0.128 0.134
Observations 3399 3399 3399 3399
6. CONCLUSION
Financial constraints have been prominently present in the business world since the work of Fazzari et al. (1988). In this early work on capital market frictions, they pointed out the importance of internal sources of finance, access to equity or debt and the functioning of credit markets to firm related investments. Research by Kaplan and Zingales (1997) it is stated that the wedge between internal and external sources of capital causes constrained companies to save costly internal resources for their best possible investment decisions. These company specific financing frictions have influence on the financial management, resulting in discreet firm behaviour in terms of investment behaviour (Denis and Sibilkov, 2010). This argument is grounded in our research as we find significant evidence on this topic. The bidder premium is lower for constrained firms than for unconstrained firms. This finding could be backed by the argument that financial unconstrained firms are less well monitored on their financial decisions caused by low financial distress, while constrained firms are on the other hand more cautious in terms of capital spending (Roll 1986; Harford, 1999).
Mentioned by Fazzari et al., external financial factors could be of great influence on firm’s internal financial decisions. This theory is supported and extended by di Giovanni (2005). He finds evidence on the impact of that financial depth of acquirers’ home country of origin causes an increased cross-border M&A propensity. During this study it is argued that it interacts with firms financial constraints because financial deepening causes the costs of external depth to decrease. These decrease costs of attracting external capital relieve capital financial constraints on firm level by relieving the financial constraints this would cause the financial constraints and thereby impacting the relation to the bidder premium positively. However in none of the models this fact is proven. A possible reason for this may be that solely the size of the equity markets is insufficient to capture this wide sense of financial depth.
