University of Amsterdam
Faculty of Business and Economics
An empirical analysis of the effect of payment
methods, deal and firm characteristics on the stock
return of retailers who just announced a takeover
Author: C.A.H. Beijlevelt Student Number: 10763678
Supervisor: dhr. dr. Jan Lemmen Study Program: Bsc Economics & Business Specialization: Finance & Organization Date: 26th June 2018
Abstract
This paper empirically researches the effect of different payment methods, firm- and deal characteristics within mergers and acquisitions on the cumulative abnormal stock return of retailers who just announced a takeover. The expectation is that acquirers using cash as a method of payment will gain positive cumulative abnormal returns around the announcement date. Acquirers using a combination of payment methods will suffer negative cumulative abnormal returns around the announcement date. It is expected that other firm- and deal characteristics may increase or decrease these effects. To determine these effects, a unique dataset has been set with observations from January 2010 till December 2017. Eventually, 113 M&A deals have been selected. A robust regression analysis shows that retailers suffer losses in their cumulative abnormal return when they finance their deal with cash. This effect can be defended with results from papers of Firth (1979) and Dodds and Quek (1985). The cumulative abnormal returns of the retailers in this dataset have possibly been marked down because of a risk-adjustment by the market.
Statement of Originality
This document is written by Student Coen August Hendrik Beijlevelt who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are 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.
Table of Contents
Abstract 1
1. Introduction 4
2. Literature Review 7
2.1 Economic Theory 7
2.1.1 Mergers and Acquisitions 7
2.1.1.1 Efficiency Theory 8 2.1.1.2 Monopoly Theory 8 2.1.1.3 Valuation Theory 9 2.1.1.4 Neoclassical Hypothesis 9 2.1.1.5 Behavioral Hypothesis 9 2.1.1.6 Empire-building Theory 10 2.1.2 Method of Payments 10 2.1.2.1 Agency Theory 11
2.1.2.2 The Benefit of Debt Theory 11
2.1.2.3 The Signaling Theory 12
2.2 Economic Literature 12
3. Data and Methodology 17
3.1 Methodology 17
3.1.1 Event Study 17
3.1.2 Cumulative Abnormal Return 18
3.2 Data 19 3.3 Linear Regression 20 3.3.1 Regression Model 20 3.3.2 Regression Analysis 22 3.4 Descriptive Statistics 22 3.5 Hypotheses 24
3.5.1 Method of Payment Hypothesis 24
3.5.2 Industry Hypothesis 25
3.5.3 Deal Value Hypothesis 25
3.5.4 Leverage Ratio Change Hypothesis 25
4. Results 26
4.1 Table Descriptive 27
4.2 OLS Regression with Robust Standard Errors 27
4.2.1 Regression Column (1) 27
4.2.2 Regression Column (2) 28
5. Conclusion 31
6. Discussion 32
6.1 Limitations and Improvements 32
6.2 Further Research 34
Bibliography 35
Appendix 1 38
1. Introduction
In November 2010 Mckinsey & Company published a report1 about the next wave in
mergers and acquisitions (M&A) in the US retail industry. In this report they set out the expectation that after the financial crisis in 2008, the retail market will be waiting for a next M&A wave. According to the analysts of Mckinsey, each of the past three economic recessions in the US have led to waves in M&A and this time, 2010 and 2011 will be the years where the M&A business picks itself up again. Figures of Thomson Reuters and Bloomberg in the report prove this statement, because in the third quarter of 2010 the most M&A deals in two years’ time were noted.
Although, the retail sector has not suffered from the financial crisis like other sectors did, the sector still has gotten a blow from the crisis. The traditional retail did not suffer as much from the financial crisis, because people still need to consume food and beverages. However the secondary good retailers, such as technological, fashion and hard-good retailers, did suffer from the financial crisis, mainly because consumers stopped spending their disposable income on products they did not need. But underneath the surface, retailers had to face a bigger problem. The past ten years retailers have been experiencing that the traditional drivers of growth have not been satisfactory enough. Consumer spending has seen a decrease since the financial crisis, but now the economy is recovering, consumer spending still does not increase enough. A good indicator of this effect is the increase in consumer savings in the past few years. Next to that, the forecasts of analysts have not been much better. In 2010, experts of Mckinsey expected the Northern-American disposable income to remain 50 to 60 percent lower than before the crisis. In addition, the ability of retailers to pursue growth via the expansion of stores has been limited. Although, retailers have to cope with the lack of growth in consumer spending, there is one big advantage. Retailers have been cash-rich for many years now. Figures of Compustat2 prove
that the top ten retailers in the US had over $25 billion in cash in 2010.
One way for retailers to use this excess cash, was carrying out mergers and acquisitions. The number of mergers and acquisitions in the retail sector started growing again in the beginning of 2011, whilst the overall North-American M&A business started
1 Mckinsey & Company, November 2010: The next wave of M&A in retail 2 The Corporate Performance Analysis Tool of Mckinsey & Company
increasing after 2013. The value of M&A in the retail sector even started increasing much earlier, in 2009.
Although the figures suggest that mergers and acquisitions and the retail sector go hand in hand, the reality is different. Retailers have been struggling for years to experience the advantages of mergers and acquisitions. This was mainly because retail processes, which have been developing in a technological way, are hard to combine after a takeover. It could take retailers years to experience the advantages, after which they sometimes suffer from being busy sorting out the takeover instead of doing the right investments in new technological developments. Some retailers therefore prefer to pursue growth by developing their technological advantage internally. Since the decision between doing directed takeovers and growing internally is quite hard to make, it would be interesting to see the effects of mergers and acquisitions on retailers’ short-term results.
The short-term effects of M&A on retailers’ stock returns have never really been researched. Moatti et. al (2014) wrote a paper in which they compared M&A to internal growth, but they did not focus on the deal and firm characteristics within mergers and acquisitions. Therefore, it would be interesting to see if certain deal and firm characteristics can explain the variance in the stock returns of retailers.
In this paper, the short-term effects of methods of payments on the cumulative abnormal stock returns of retailers who just did a takeover deal will be researched. Next to that, there will be a regression model set up which includes deal and firm characteristics to explain more of the variance in the stock return of the acquirer.
The main question of this paper will be: What is the effect of different payment methods in mergers and acquisitions on the stock returns of the acquirer after the announcement in the retail sector in the U.S.A. from January 2010 till December 2017? To investigate these effects, three databases have been used to retrieve data on the deals (Zephyr), the stock returns (CRSP) and the deal and firm characteristics (Compustat).
Empirical analysis shows that cash financing has a negative effect on the cumulative abnormal stock return of retailers. This contradicts economic literature written by Travlos (1987) and Raad and Wu (1994) and economic theory on method of payments. On the other hand, the findings are in line with literature written by Firth (1979) and Dodds and Quek ((1985). The main argument for the negative effect of cash-financing on the cumulative
abnormal return of acquirers in retail, is the risk-adjustment of the market after a deal has been announced.
2. Literature Review
In this section we will discuss the existing theories and literature on payment methods and its effects on the stock return of the acquirer. First, we will discuss the existing economic theory. Economic theory will be split into economic theory about mergers and acquisitions and payment methods. Second, we will discuss papers written about payment methods and its effects on the cumulative abnormal stock returns of the acquirer.
2.1 Economic Theory
In this subsection we will discuss the main economic theories about mergers and acquisitions and method of payments. Since these two subjects have a lot of overlap, it will be much easier to have a look at both subjects solely to get a structured view of the economic theories.
2.1.1 Mergers and Acquisitions
Mergers and acquisitions have been a method for companies to chase growth opportunities for many years. That is why mergers and acquisitions have been a subject of research for many years as well. Therefore, researchers have discovered in the 90s that mergers and acquisitions come in waves. For many years there have been researchers, such as Blair, Lane and Schary (1991) and Brealey and Myers (1991), who had some suspicions about the wavy character of mergers and acquisitions over time. Golbe and White (1993) were the first to prove that mergers and acquisitions do indeed come in waves. Although, they stated that their research was just the beginning of developing and testing hypotheses.
Now that we know that mergers and acquisitions do come in waves, it is important to find out what could be the motives behind mergers and acquisitions. To get to know the motives behind mergers and acquisitions, we will use the paper of Trautwein (1990) which displays multiple economic theories about mergers and acquisitions. After that, we will use the paper of Mariana (2012) which displays two tested hypotheses which can be linked to theories in the papers of Trautwein (1990) and Muehlfield et al. (2012).
2.1.1.1 Efficiency Theory
First we look at the efficiency theory, which is explained in the paper of Trautwein (1990). The efficiency theory states that mergers and acquisitions mainly take place to find synergies between the acquirer and the target. The theory splits three kinds of synergies: financial, operational and managerial synergies.
Firstly, acquirers can achieve financial synergies by a decrease in the cost of capital. This can be achieved by lowering the systematic risk of the company’s investment portfolio. In addition, due to the firm’s size increase, it has bigger buying bower by which it can decrease the cost of capital.
Secondly, acquirers can achieve operational synergies by lowering the operational costs. If everything remains constant, this cost decrease will immediately affect the bottom line in a positive way. In addition, acquirers can achieve operational synergies by sharing knowledge.
Thirdly, acquirers can achieve managerial synergies by managing the target in a more effective and efficient way than the former management did.
Although the efficiency theory is well-established, it is mainly criticized on the theoretical part of the theory. Critics say that financial synergies cannot be achieved in a perfectly efficient capital market.
2.1.1.2 Monopoly Theory
The second theory is the monopoly theory, which is also explained in Trautwein (1990). This theory states that the main motive of mergers and acquisitions is achieving market power. This strategy is mainly used in conglomerate mergers, where an acquirer uses its profits from one market to acquire a target in another market. This way, acquirers can easily enter new markets and because of that diversify their operational portfolio.
Although the monopoly theory is not highly supported by the evidence, it is a theory which has been an overall accepted theory for many years. The evidence of the monopoly theory is limited, because no acquirer will admit that it thrives for a monopoly position in the market they are operating in.
2.1.1.3 Valuation Theory
The third theory which is explained in the paper of Trautwein (1990) is the valuation theory. This theory states that the biggest motive of mergers and acquisitions is that some managers have a better view on the target’s value than the market has. Managers try to take an advantage of this information asymmetry by moving rapidly.
Although this theory is quite well-established, it still gets some criticism. Critics state that in a world with perfect information, which is incorporated in the share price, it will not be possible to take a financial advantage out of information.
2.1.1.4 Neoclassical Hypothesis
The first hypothesis explained in the paper of Mariana (2012) is the neoclassical hypothesis, which could also be called the economic motivation or the disturbance theory. This hypothesis states that mergers and acquisitions are a reaction to macroeconomic events, for example technological developments or industrial overcapacity. Since macroeconomic shocks will cause an increase in the demand, companies have to scale up very quickly. One way to do this is acquiring a target.
Furthermore, Mariana (2012) also refers to Manne (1965). In his paper about mergers and the market for corporate control, he states that mergers and acquisitions are a way of relocating assets. The same applies to the relocation of capital (Jovanovic and Rousseau, 2002). After a takeover, it might be that the assets or the capital will come in the hands of firms who can use it more effectively.
Lastly, Persons and Warther (1997) state in their paper about the adoption of financial innovations that the wavy character of mergers and acquisitions might be explained by a reaction mechanism. They state that firms who see their competitor doing a takeover will be more likely to react to that by doing a takeover themselves.
2.1.1.5 Behavioral Hypothesis
The second hypothesis explained in the paper of Mariana (2012) is the behavioral hypothesis, which could also be called the managerial motivation. The behavioral hypothesis (Schleifer and Vishny, 2003) states that there is a connection between bond evaluations and the frequency of mergers and acquisitions. According to this theory, the wavy character of mergers and acquisitions can be explained by the characteristics of the players on the
market. Clearly, the behavioral theory explains the waves in M&A from a microeconomic point of view.
Schleifer and Vishny (2003) stated that managers play a huge role in the explanation of waves in mergers and acquisitions. Managers can use the valuation of their own stock as a tool to take M&A decisions. As managers see their stock price being overvalued, they might want to swap their stock for tangible assets which generate cash flows. When the stock market is in a boom, there will be a lot of shares which are overvalued. This way, Schleifer and Vishny (2003) hypothesize that there is a positive relation between capital market booms and waves in mergers and acquisitions.
2.1.1.6 The Empire-Building Theory
Another managerial theory states that managers see mergers and acquisitions as building an empire: the empire-building theory. Although this theory is very old and was already started by Berle and Means (1933), it is best explained in a paper by Trautwein (1990). This theory states that some managers who like to build an empire will maximize their own utility instead of the utility of the shareholders. To maximize their own utility, managers will close riskier deals by which the company they are managing might come in severe danger.
Although the empire-building theory is just a theory and evidence on it is limited, there are a few researchers who have found evidence that could support the theory. For example, You et. al (1986) found that management ownership and the number of managers in the board correlate negatively with merger success. This means that the more management ownership and number of managers in the board, the less the probability a merger succeeds. Next to that, Amihud and Lev (1981) found that high management ownership is associated with conglomerate mergers, which are riskier.
2.1.2 Method of Payments
In this section we will discuss the existing economic theory on method of payments and its effects on the stock return of acquirers. Before going into the economic theory on method of payments, it is important to get an understanding of the agency theory and link this to method of payments.
2.1.2.1 Agency Theory
Agency theory, first developed by Fama (1980), provides us with the description of the relationship between the managers of the firm and its shareholders. In this case, the managers are the agents and the shareholders are the principals. Decisions about mergers and acquisitions can cause conflicts between managers and shareholders.
Jensen (1986) used agency theory to come to the agency costs of mergers and acquisitions. He states that the main issue between managers and shareholders is that shareholders might want to see the profit of the business being paid out to them, but this will negatively affect the financial resources disposable to managers. A side-effect might be that when managers reduce their resources to their disposal, they might have to go to the capital markets to obtain capital way earlier. Since financing via the capital markets causes a lot of monitoring with third parties and a possible decrease in the stock price when stock is issued, managers give the preference to use internally held capital. Next to that, managers normally have the target to let the company grow. This is good for the company, but also profitable for the managers because growth in executive compensation correlates positively with growth in sales (Murphy, 1985). After all, Jensen (1986) states that all managers look for free cash flows. But when companies have big free cash flows, the agency problem becomes bigger and bigger, because the shareholders do not want the managers to invest the free cash flows in below the cost of capital investments or waste it on organizational inefficiencies. Most of the times, managers promise to increase the dividends on the shares, but since dividends can be adjusted every single time, it is not so credible that managers will keep their promise. In advance, managers can be punished by the market when they do not keep their promise: the agency costs of free cash flows.
2.1.2.2 The Benefit of Debt Theory
To tackle the problem of high agency costs of free cash flows, Jensen (1986) has developed a theory which states that managers will keep their promise to pay out future free cash flows to the shareholders if they use debt as a method to finance the deal. Jensen explains this as follows: when issuing debt to buy back shares from current shareholders, the shareholders might take the company to the court when the company goes bankrupt. This way, shareholders can make managers keep their promise in paying out future cash flows to the shareholders instead of wasting it on bad investments. Yook (2003) adds another advantage
of debt. He states that due to a bigger possibility of bankruptcy, managers will have an incentive to work harder. On the other hand, Jensen (1986) acknowledges that debt issuance has its disadvantages, because increased leverage comes at a cost of debt, which mainly consists of bankruptcy costs.
2.1.2.3 The Signaling Theory
Next to the benefit of debt theory there is another important theory about method of payments. The signaling theory, described in a paper by Yook (2003), is a theory which states that cash is likely to be used when acquirers invest in positive net present value acquisitions. This theory makes two assumptions. The first assumption is that capital markets are not fully efficient. The second assumption is that there is information asymmetry between the managers of the firm and the market. Since this is the case, managers may use that information asymmetry advantage to signal information to the market using their method of payment. Not only can managers signal information to the market by choosing a method of payment, they can immediately change the internal capital structure as well. The theory states that managers will be more likely to issue stock when their shares and assets are overvalued. When a firm issues new shares, which is seen as a dilution of the voting power by the current shareholders, the share price will decrease due to a supply increase. This strategy is also a way to protect the shareholders against a setback once the market gets to know that the stock is overvalued.
On the other hand, when the shares and assets are undervalued, managers would want to finance with cash, because they do not want their stock to decrease any further. Offering cash in a merger or an acquisition will therefore signal to the market that the acquirer is sure that their stock is undervalued and thus has more belief that the acquisition will become a success. Furthermore, Yook (2003) quotes Myers-Majluf (1984) that when an acquirer needs outside financing, it will choose debt over equity because a stock issue is less favorable to current shareholders.
2.2 Economic Literature
Payment methods have been researched a lot in the past. Some literature focuses on the stock return of acquirers, some on the stock return of the targets and some on both the acquirer and the target. In this paper we will look at the stock return of the acquirer.
Travlos (1987) was one of the first to find hard evidence that shareholders of bidding firms who financed their takeover with stock suffered significant losses around the announcement date. On the other hand, he showed that shareholders of bidding firms who financed their takeover with cash gained normal returns around the announcement date. Next to that, Travlos (1987) proves that stock financing causes significant losses no matter what the outcome of the bid will be. These findings are consistent with the signaling theory explained in the section above.
Asquith et. al (1990) follow these findings and add to it that the relative size of the target and the bidder does influence the stock returns of the bidder. When combining a stock offer with a big relative size measure, Asquith et. al (1990) find that the effect is even more negative than with just a stock offer. They even argue that while the investment value of the takeover is positive, it is not big enough to offset the negative consequences of the combination between a stock offer and a big relative size. In addition, Asquith et. al (1990) find that it does not matter whether a bid is a tender offer or a merger offer, but it does seem that merger offers are most of the times financed with stock and tender offers are most of the times financed with cash. On average, mergers do add value according to Asquith et. al (1990).
After that, Raad and Wu (1994) studied methods of payments in combination with management equity ownership and changes in the leverage of the acquirer. They found that when firms finance their takeover with stock in combination with low management ownership of equity, the losses the shareholders had to suffer would be larger. On the other hand, stock-financing which led to a decrease in the leverage levels of the acquiring firm decreased the losses of their shareholders. They also found that when acquirers financed their takeovers with cash in combination with high management ownership and an increase in leverage, the shareholders of the acquirer would gain significant abnormal returns. Raad and Wu (1994) also chose to control for management ownership and leverage changes, by which they found that high management ownership in combination with mergers which led to an increase in the amount of leverage were associated with significant positive abnormal returns. On the other hand, high management ownership in combination with mergers which led to a decrease in leverage was associated with insignificant negative abnormal stock returns.
Chang (1998) focused his research mainly on the effect of payment methods in takeovers of privately held targets. In his paper he found results which contradicted the economic theory of positive abnormal returns in cash offers and negative abnormal return in stock offers. He found that on average bidders who financed their bid with cash, gained and lost no abnormal return. On the other hand, he found that bidders who tried to acquire privately held targets using stock as a method of payment, gained significant positive abnormal returns. Chang defends these findings with other results from the research. In the same paper he finds evidence of positive correlation between bidding firm returns in stock offers, the creation of a new large blockholder and the amount of stock issued to the target’s shareholders. This can be explained by the fact that privately held targets are highly concentrated, by which a takeover with stock creates a new large blockholder. Therefore, if a new large blockholder is created, the cumulative abnormal stock return must go up.
Fuller, Netter and Stegemoller (2002) researched acquirers who did five or more successful acquisition bids within three years’ time. They first researched the wealth effects of acquiring either a private target, subsidiary or a public target. They found that shareholders gain significant positive abnormal stock return when the target was a private target or a subsidiary. In addition, they found that the gain will be larger if the target is larger and when stock is used as a way of financing. When a public target was acquired, the shareholders suffered losses. In addition, they found that shareholders of acquirers suffer bigger losses when the public target is larger and when the acquirer used stock as a method of payment.
On the other side of the research spectrum, there have been researchers who found contradicting results to the results of Travlos (1987), Raad and Wu (1994) and Fuller et. al (2002).
Firstly, Firth (1979) researched the profitability of takeovers in the United Kingdom using an efficient markets theory framework. He analyzed the gains and losses of acquired and acquiring firms and concluded that there was no gain or loss associated with a takeover, just a shift of wealth. Therefore, the results back the idea that mergers and acquisitions are just for growth purposes. Firth (1979) arguments his findings in two different ways. First, he states that seemingly the market thinks the two combined companies are worth less than the two separate companies. This might be because the market thinks that the acquiring management has taken on too much work to make the combined company a success, in
which they will not succeed. Second, Firth (1979) states that the market thinks that the fees and expenses of takeover are too high and will therefore lead to less profit. In addition, Firth (1979) found that takeovers financed with cash are associated with negative cumulative abnormal stock returns. 79% of his sample of acquirers suffered risk-adjusted declines in their share prices. He therefore concludes that the market regards takeovers as very expensive and therefore marks down the stock price of the acquirer.
Secondly, Dodds and Quek (1985) researched post-merger profitability in the industrial sector during the 70s. They found a positive short-term effect of takeovers on the cumulative abnormal stock return of the acquirer. When considering the method of payment used for the takeover they found something interesting. In the month around the takeover they found negative cumulative abnormal returns for acquirers who used cash as a method of payment. 75% percent of the whole sample suffered negative cumulative abnormal return and the cumulative abnormal return was on average -1.92% for acquirers using cash. For acquirers using stock, Dodds and Quek (1985) found positive cumulative abnormal returns. The cumulative average abnormal return was 0,78%. On the other hand, just 50% of the sample experienced positive cumulative abnormal returns after the takeover, but on average the effect was positive. They motivated these findings by the fact that the acquirers are over-stretching their scarce financial resources, which the market found to risky. Next to that, he suggests that the market finds cash offers less desirable than stock offers, by which the stock prices of the acquirers using is marked down.
Since this paper is focused on the retail and wholesale sector, it is important to have a look at a paper which specifically researched the retail market and its growth opportunities. Moatti et. al (2014) disentangled two typical horizontal growth strategies: mergers and acquisitions and organic growth. They researched the effect of those two growth strategies on two indicators of firm performance: (1) firms bargaining power with respect to suppliers and customers and (2) operating efficiency arising from scale economies. For the sake of this paper, we mainly look at the outcome of the research on the M&A side. In their findings Moatti et. al (2014) empirically prove that mergers and acquisitions do have a significant positive effect on the bargaining power of a firm with respect to suppliers and customers. They even find that mergers and acquisitions do affect bargaining power in a more positive way than organic growth. On the other hand, they find that the bargaining power advantage created by M&A lasts for just two years and then
disappears. When looking at the effect of M&A on the operating efficiency, mergers and acquisitions create a disadvantage compared to organic growth. Due to the high costs of M&A and the fact that creating more operating efficiency after a takeover is way more complex than through organic growth, M&A has a negative effect on the operating efficiency. After all, Moatti et. al (2014) struggle to prove if mergers and acquisitions do have a positive or negative effect on firm performance. What they do prove, is that when measuring firm performance by return on assets or operating profit, mergers and acquisitions do worse than organic growth on the short as well as the long term.
3. Data and Methodology
3.1 Methodology
This section will explain the process of doing an event study which consists of three parts: identifying the event, measuring expected stock returns and measuring the short-term abnormal stock returns. Firstly, the start of the event study will be explained. Secondly, the way of measuring and determining stock returns will be explained. Thirdly, this section will give a look into the regression analysis done on the cumulative abnormal stock returns of the acquirer. At the end of this section the hypotheses will be summed up.
3.1.1 Event Study
This paper uses a combination of the event study methodologies of Fuller, Netter and Stegemoller (2002) and Raad and Wu (1994) to investigate the short-term effects of mergers and acquisitions. Event study is mostly used in economic and financial research where an effect of a certain event is investigated. For example, event study can be used to research stock returns, firm profitability and operating income. The event study done in this paper investigates the effect of different methods of payment, in combination with certain deal- and firm characteristics, on the short-term stock return of the acquirer. To investigate these effects, two time windows have been created: the event window and the estimation window.
The event window is the period in which the event has appeared, which in this paper will be the announcement of a merger or and acquisition. For the event window the methodology of Fuller, Netter and Stegemoller (2002) will be used. In their research to acquirers who did five or more acquisitions in three years’ time, they followed the standard event study methodology of Brown and Warner (1985). Therefore, the event window will be a five-day period (-2,2), which is exactly two trading days before and after the announcement of the takeover. In this window the cumulative abnormal returns (CARs) will be calculated, which eventually will be the dependent variable in the regression model.
The estimation window is the period in which the expected returns will be calculated. For the estimation window the event study methodology of Raad and Wu (1994) will be used. They used a 149-day period (-160,-11), which starts approximately eight
trading months before and ends two trading weeks before the announcement. Since the research of Raad and Wu (1994) is comparable to this paper, the estimation window (-160,-11) of their research will be used.
3.1.2 Cumulative Abnormal Return
The cumulative abnormal return (CAR) on an acquirers’ stock can be explained by the following. The CAR is the extra bit of return which is earned by shareholders due to the outperformance of a benchmark combined with the expected return of the stock. In this paper, Rit is the daily stock return of the acquirer and Rmt is the daily return on the S&P 500.
The S&P500 is the index of the 500 biggest companies in the United States of America, which will be used as an indicator of the return on the market. In this case, beta (βi) is an
indicator of how much the stock of the acquirer will move when the S&P 500 index moves up or down with 1%. Using these indicators, the abnormal return can be calculated as follows:
ARit= Rit - αi -βiRmt (1)
As stated in the estimation period section, the expected return of the stock should be calculated according to the return earned in the estimation period. To calculate the expected return, the event study commands in Stata have been used. Using these commands, Stata sets up a simple market model and calculates the expected returns according to that market model. In the calculation of the abnormal returns, the market model can be indicated by the calculation of the expected return in the estimation period:
E(Rit)=αi +βiRmt (2)
Here it can be seen that the expected stock return of the acquirer consists of a constant and a certain reaction to the market movement. As indicated by formula (1), the expected return will be subtracted from the stock return in the event window, which will eventually give you the abnormal return.
To calculate the cumulative abnormal return, the abnormal returns on the stock of the acquirer should be summed. This can be done in Stata as well by using the command to sum the abnormal returns.
The formula for the calculation of the cumulative abnormal return is as follows:
CAR(t1, t2) = ∑ARit (3)
In this case t1 and t2 indicate the first and last day of measuring the abnormal returns in the event window. The cumulative abnormal return shows how much excess return has been earned by the shareholders in the period around the announcement of the merger.
To calculate how much cumulative abnormal return has been earned by the shareholders in the total dataset, the cumulative average abnormal return (CAAR) can be calculated. The CAAR is the average of all the cumulative abnormal returns in the dataset. In this case, it is an indicator how much return an investor can gain on average in the retail and wholesale sector when a merger or an acquisition takes place.
The formula to calculate the CAAR is as follows:
CAAR = (1/N) *∑CAR(t1, t2) (4)
3.2 Data
The data for the empirical research has been retrieved from different databases. Firstly, the Zephyr database has been used to gather a set of M&A deals. In Zephyr it is possible to select different types of conditions which the M&A deals must fulfill. First of all, the deals must all be mergers or acquisitions. Different forms of deals, such as a management buy-outs or buy-ins, have not been considered. In addition, all acquirers must be listed at an American stock exchange. Since listed companies have the obligation to share financial information, it is easy to gather this financial information in databases.
The next condition states that the events should have taken place between the first of January 2010 and the thirty-first of December 2017. In addition, targets must be listed or unlisted. Moreover, the acquirers should operate in either the retail or the wholesale sector. Targets do not have this constraint, since this paper also controls for the industry the target is in. This will be done according to the industry measure, which measures whether the target and acquirer are in the same sector or not.
The last condition is the minimum and maximum deal value. The minimum deal value is stated on fifty million US dollars and the maximum deal value is stated on one billion US dollars.
After all conditions have been met, the total size of the dataset is 142 M&A deals. Since Zephyr just supplies the M&A deals, a different database had to be used for the stock returns of the acquirer.
To gather all the stock data of the acquirers, the CRSP database has been used. To get stock returns out of CRSP a text file had to be set up with all company Cusip codes. This way, CRSP can recognize which company is which and whether the database has stock data of these companies. Due to the lack of stock data of some acquirers, the dataset had to be reduced to 116 M&A deals. Using Stata during the event study the M&A deals have been linked to the stock data in the estimation and event window. After the stock data had been retrieved from CRSP, data had to be gathered from Compustat.
The Compustat database can be used to retrieve financial information of the acquirers, such as debt and equity book values. With these values the leverage ratios have been calculated. The debt and equity values retrieved from Compustat were the book values the fiscal quarter before and after the event. With these values it was possible to calculate the percentage change of the leverage ratios. Due to some negative values in the equity book values, three M&A deals had to be subtracted from the dataset. Finally, 113 M&A deals were left, which all have a value for the variables of the model.
3.3 Linear Regression
In this section the regression model will be explained. The intuition behind the regression model has already been explained in the literature review, because all variables in the regression model are based on earlier research. Later in this section, the regression analysis done in Stata will be explained.
3.3.1 Regression Model
Finally, when all data has been retrieved and put into Stata, it will be possible to do a regression analysis on the stock returns of all acquirers. The most important question here is whether the variance in the independent variables can explain the variance in the dependent variable: the cumulative abnormal stock return of the acquirer. The independent variables chosen in this model mainly come from the independent variables used in the papers of Travlos (1987), Fuller et. al (2002), Raad and Wu (1994) and Asquith et. al (1990). To do the regression analysis, a regression model had to be set up.
The regression model is as follows:
CAR=β0+β1*Cash+β2*Combi+β3*Industry+β4*LN(DealValue)+β5*LeverageRatioChange+εi
Since this paper mainly focuses on the effect of different payment methods on the cumulative abnormal stock return of the acquirer, the variables which represent these payment methods will be placed first into the model.
The first independent variable cash is a dummy variable which will have value 1 if cash is used as the method of payment and 0 if another method of payment is used.
The second independent variable combi is a dummy variable as well, which will have value 1 if the acquirer uses a combi of payment methods which consists of the following options: stock, cash, deferred payment and earn-out. The combi variable will have value 0 if just one payment method is being used by the acquirer.
The third independent variable is an industry indicator, which is a dummy variable as well. It indicates whether an acquirer acquires a target which is in the same industry as him or not. This is important to know, because cross-industry acquisitions can be riskier, since the acquirer might enter a new business area. The variable has value 1 if the target was in the same industry as the acquirer. When the target is not in the same industry as the acquirer, the variable will have value 0.
The fourth variable is the natural logarithm of the announced merger value in US dollars. It will give an indication on whether the takeover is a big or small takeover. Since the minimum and maximum deal value do lie quite far from each other, we used the natural logarithm to bring all the observations closer to each other.
The fifth variable is the LeverageRatioChange, which will give an indication on whether the acquirer has increased its leverage ratio during the takeover. The leverage ratio is equal to the book value of the debt divided by the book value of the equity. Since the market values of the debt were not available, the book values for both the debt and the equity had to be used. This variable could be worthy, because most cash takeovers will in the end be financed with the issuance of debt. When an acquirer issues more debt, its leverage ratio will increase if the equity book values remain the same. It therefore might explain something extra in the variance of the stock return of the acquirer. This last variable concludes the regression model. All details about the variables can be found in tables 3 up till 7 in Appendix 1.
3.3.2 Regression Analysis
After the regression model has been set up, the regression analysis can be done in Stata. Linear regression analysis can be done in Stata in many ways. The most common used regression analysis is the Ordinary Least Squares (OLS) regression analysis. OLS regression regresses the dependent variable on the independent variables in such a way that the sum of squares of the residuals of the dependent variable and the independent variables is as small as possible. The OLS regression analysis is the best estimator when some conditions are fulfilled: (1) consistency because regressors are exogenous, (2) unbiased because the error terms are homoscedastic and serially uncorrelated and (3) error terms are normally distributed. The OLS regression analysis can easily be done in Stata by selecting the linear regression button using the standard errors which follow the OLS methodology.
Next to that, a regression analysis with robust standard errors can be done. A robust regression will be used when the dataset has outliers which can compromise the validity of the model. Robust regression is less dependent on all conditions which must be fulfilled when using the OLS regression analysis. OLS regression analysis might give misleading results when the conditions are not met properly, this way OLS regression analysis will not work as well as it should work.
Since the dataset used in this paper is quite spread out, we used robust standard errors. The main reason for this is to prevent heteroscedastic errors which compromise the validity of the model. In addition, using robust standard errors will cause the standard errors of the variables to be smaller. Eventually, this will give us more significant variables which will be able to explain more of the variance in the dependent variable.
3.4 Descriptive Statistics
All the descriptive statistics of the variables discussed in section 3.3.1 can be found in table 3, which is placed in Appendix 1. The cumulative abnormal return has a mean of 2,04%, which is quite normal for average cumulative abnormal returns. Especially, when we compare this to the findings of Travlos (1987) and Raad and Wu (1994) who found average cumulative abnormal returns within the event window around the 2%-mark as well. If we then have a look at the standard deviation, we can see that the standard deviation is 6,74%. This tells us that the cumulative abnormal returns are lying quite far from each other within the dataset, which can also be noticed when we look at the minimum and maximum value
of cumulative abnormal return. These are respectively -24,35% and 35,59%. From this we might conclude that some acquirers do much better than others.
The dependent variable cash is a dummy variable, so the mean and standard deviation do not say anything. What we can conclude from table 4 is that most of the acquirers choose to finance their acquisition with cash: 79 cash takeovers against 34 combi takeovers.
The dependent variable combi is also a dummy variable, so we cannot conclude anything out of the mean and standard deviation. In table 5 we can see that most acquirers choose an alternative payment method instead of choosing a combination of payment methods: 30 combi takeovers against 83 alternative method of payments.
As the variable industry is a dummy variable as well, we cannot conclude anything out of table 3. Table 6 tells us that in this dataset, the acquirers mainly choose to acquire targets which are in the same industry as them: 97 same industry targets against 16 different industry targets. This might tell us that big players in the retail and wholesale market acquire smaller players to fight competition and therefore increase their market share.
The variable change in leverage ratio has a mean of 15,20% and a standard deviation of 46,38%, which shows us that the leverage changes do differ a lot amongst the acquirers. This can also be seen in the minimum and maximum values of leverage ratio change. Since these are already percentages, there is no need to take the natural logarithm to decrease the differences.
Finally, we can see in table 3 that the natural logarithm of the deal value has a mean of 19,12 and a standard deviation of 0,81. The minimum and maximum values are 17,73 and 20,64. The minimum and maximum value of the actual deal values are $50 million and $922 million. In table 7 we can see that the deal values are quite equally distributed, with most deals lying between $100 and $250 million.
Table 1: Correlation Matrix
CAR Cash Combi Industry LnDealValue Leverage
CAR 1 Cash -0.1829 1 Combi 0.1489 -0.9164 1 Industry -0.0295 0.0103 0.0142 1 LnDealValue -0.0661 -0.0743 0.0743 -0.1610 1 Leverage 0.0550 -0.2165 0.2731 0.0103 0.0054 1
In Table 1 above we can see the correlation matrix. It is important that the variables do not correlate too much with each other, because if they do, they will reduce each other’s explanatory power. This phenomenon is called multicollinearity and we want to prevent this. In the correlation matrix we can see that almost all correlations are far under the 0,7-mark, which is the mark where multicollinearity appears. The only two variables who correlate a lot with each other are cash and combi, but that is logical. This is the case, because the cash and combi variables are dummy variables, so if one of the two has value one the other variable must have value 0. This way they affect each other a lot, but this should not be a problem, because it cannot be prevented.
3.5 Hypotheses
In this section we will look at the hypotheses of this paper. For every single independent variable we will state a hypothesis. The idea is that every single independent variable should have their own effect on the dependent variable.
3.5.1 Method of Payment Hypothesis
The null hypothesis will be that shareholders do not gain or lose any cumulative abnormal stock return around the announcement date. The alternative hypothesis for cash-financing will be that shareholders of acquirers using cash as a method of payment will gain positive cumulative abnormal return.
The null hypothesis will be that shareholders do not gain or lose any abnormal stock return around the announcement date. The alternative hypothesis for combi-financing will be that shareholders of acquirers using a combination of payment methods will suffer negative losses in their cumulative abnormal return.
3.5.2 Industry Hypothesis
The null hypothesis will be that shareholders do not gain or lose any cumulative abnormal stock return around the announcement date. Based on the paper of Fuller et al. (2002), the first alternative hypothesis will be that shareholders of an acquirer will suffer a negative cumulative abnormal return when the target is in the same industry as the acquirer provided that the target is listed. The second alternative hypothesis is: when a target is unlisted, the shareholder of an acquirer will gain a positive cumulative abnormal return when the target is in the same industry as the acquirer.
3.5.3 Deal Value Hypothesis
The null hypothesis will be that shareholders do not gain or lose any cumulative abnormal stock return around the announcement date of a merger or an acquisition. Based on economic theory, the alternative hypothesis will be that shareholders will suffer a negative cumulative abnormal return as the deal value increases.
3.5.4 Leverage Ratio Change Hypothesis
The null hypothesis will be that shareholders do not gain or lose any abnormal stock return around the announcement date of a merger or an acquisition. Based on the paper of Raad and Wu (1994), the first alternative hypothesis will be that shareholders will gain a positive cumulative abnormal return if the acquirer pays in cash and increases its leverage ratio. The second alternative hypothesis will be that shareholders will suffer a negative cumulative abnormal return provided that the acquirer pays with stock and decreases its leverage ratio.
4. Results
In this section the results of the regression analysis will be explained. Firstly, the results of the regression with just the two payment methods as independent variables and the cumulative abnormal return as the dependent variable will be explained. Secondly, the results of the final regression with all independent variables included will be explained.
Table 2: Regression output
Table 2: results empirical analysis Cumulative Abnormal Return
Independent Variables (1) (2) Cash -0.042** (0.019) -0.044** (0.020) Combi -0.018 (0.026) -0.015 (0.029) Industry -0.010 (0.033) Ln (DealValue) -0.012 (0.007)
Leverage Ratio Change 0.006
(0.011) Constant 0.055*** (0.019) 0.293** (0.148) N 113 113 R-squared 0.036 0.0650 F-value 3.01* 1.86 Df 2, 110 5, 101
Robust standard errors are parentheses below the estimated coefficients. Significance is indicated as follows: *** if p<0,01, ** if p<0,05 and * if p<0,10. N is the total observations. R-squared indicates the percentage of the variance in the dependent variable explained by the variance in the independent variable(s). F-values indicate the significance of the model,
4.1 Table Descriptive
The results of the regression analysis are displayed in table 2. On the left-hand side the independent variables are shown. The dependent variable, the cumulative abnormal return, is placed above the two regressions done. In column (1) the first regression can be seen. In this regression, only the variables cash and combi are included. In this way, we can distinguish the effect of payment methods in isolation from the other independent variables and included in the final model with the other independent variables. These independent variables are then regressed against the dependent variable: cumulative abnormal return. In column (2) the final regression can be seen. In this regression, all the independent variables are regressed against the dependent variable.
4.2 OLS Regression with Robust Standard Errors 4.2.1 Regression Column (1)
In the first column of table 2 the variables cash and combi are regressed against the dependent variable: cumulative abnormal return. The model is significant at a 10% level.
According to the regression results, cash-financing has a negative impact on the cumulative abnormal return. The coefficient of cash has a value of -0,042 with a standard error of 0.019. The coefficient is significant at the 5%-level. From this we can conclude that the cumulative abnormal return will decrease when an acquirer uses cash as a method of payment. This contradicts the alternative hypothesis, which states that acquirers using cash as a method of payment will gain positive abnormal returns. Although this might seem strange, other researchers have found the same result for cash acquisitions. Firth (1979) found significant negative cumulative abnormal returns after a takeover bid when the acquirer proposed a cash offer. According to Firth (1979) this is due to risk-adjustment by the market. Like Firth (1979), Dodds and Quek (1985) also found negative cumulative abnormal returns after cash offers. They also concluded that this could be the case due to a mark-down of the market, but they added that this would be the case if the market found that the acquirers were dealing too risky with their financial sources. This could be a reason why in this paper a negative effect of cash-financing has been found. Possibly, the market found the takeovers done in this dataset too risky to pursue. If this is the case, it is in line
with existing literature of Firth (1979) and Dodds and Wuek (1985) that cash-financing has a negative effect on the cumulative abnormal stock return of the acquirer.
The coefficient of the combi variable has a value of -0,018 with a standard error of 0,026, but is insignificant. The direction of the coefficient is in line with the alternative hypothesis that picking a combination of payment methods does have a negative impact on the cumulative abnormal return of the acquirer, but this cannot be concluded with 1, 5 or 10% significance. It could be the case that due to the multi-collinearity between the cash and combi variables, the combi variable loses explanatory power and thus has a lower significance.
4.2.2 Regression Column (2)
The second column of table 2 shows the total model in which the payment methods and the other independent variables are regressed on the cumulative abnormal return of the acquirer. This model is not significant at a 1,5 or 10% level.
The cash variable has a coefficient of -0,044, which is more negative than in the first regression, and has a standard deviation of 0,020. It is significant at the 5% level. This tells us that when an acquirer chooses cash as the method of payment, his cumulative abnormal return will decrease by 0,044. As stated in the first column explanation, this finding is in line with the findings of Firth (1979) and Dodds and Quek (1985). It might be that in a high-competitive market as the retail and wholesale market, the market adjusts for the risk of a takeover much faster than in other markets. Because the effect is negative, we cannot reject the null hypothesis.
The combi variable has a coefficient of -0,019 and a standard deviation of 0,029. Although it is not significant, the direction of the coefficient is in line with the literature stating that using a combination of payment methods will have a negative effect on the cumulative abnormal stock return of the acquirer. The insignificance might be caused by the multicollinearity, because the combi variable correlates heavily with the cash variable. On the other hand, if this would be the case, then the cash variable should lose explanatory power as well. It might also be that the disparity in the data causes this variable to be insignificant. Because the combi variable is not significant at 1,5 or 10%, we cannot reject the null hypothesis.
The industry variable has a coefficient of -0,010 with a standard deviation of 0,033, it is not significant at any level. This might be the case because the industry measure is not evenly distributed. There are way more acquirers who chose to acquire a target which is in the same industry. Therefore, the variance might be too big, by which it loses explanatory power. The direction of the coefficient does match the findings of Fuller et. al (2002) partly. Their findings suggest that when an acquirer and a target are in the same industry and the target is either listed or a subsidiary, the cumulative abnormal return will be affected negatively. On the other hand, they find that a takeover of a private target in the same industry sorts in a positive abnormal return for the acquirer. Since the industry variable is not significant at any level, we cannot reject the null hypothesis.
The deal value variable, which is the natural logarithm of the actual deal value, has a coefficient of -0,012 and a standard deviation of 0,007. It is almost significant at a 10% level with a p-value of 0,108. The direction of the coefficient is in line with economic theory, which states that a higher deal value will sort in a lower cumulative abnormal return. As the natural logarithm has a positive relation to the amount under the log, the outcome will increase when the amount under the log increases. Therefore, if the deal value increases, the cumulative abnormal return will decrease. This is quite logical, because a bigger deal value requires more financing, which might require more risk. If a stock gets riskier, less investors will buy the stock if one assumes that investors are risk-averse. Therefore, more investors will supply the stock instead of demanding it, by which the stock price will drop. Since the natural logarithmic deal value variable is not significant at 1,5 or 10%, we cannot reject the null hypothesis.
The leverage ratio change variable, which indicates whether an acquirer increases its leverage ratio before a takeover, has a coefficient of 0,006 and a standard deviation of 0,011. This variable is not significant, but if we look at the direction of the coefficient we can say that it matches the literature. As Raad and Wu (1994) found in their paper: an increase in the leverage ratio is positively related to the cumulative abnormal return provided that the acquirer pays in cash. If the acquirer chooses stock or a combination as the method of payment, a leverage decrease will make the negative cumulative abnormal return worse according to Raad and Wu (1994). The effect of the coefficient will be positive on the cumulative abnormal return, but this cannot be said with 1,5 or 10% significance. It might be that this variable has data which lie too far out of each other, by which its variance is too
high and it loses explanation power. Next to that, it could be possible that the data sample is just too small for the leverage ratio change variable to be significant. It has been tried to winsorize the data, but this did not sort in a better or significant effect. Since the leverage ratio change variable is not significant, we cannot reject the null hypothesis.
5. Conclusion
In this paper the effect of different payment methods in combination with deal and firm characteristics on the cumulative abnormal stock return of acquiring retailers has been investigated. To research the effect of financing a deal with cash or a combination of different payment methods, a dataset from Zephyr has been used. After linking the deals from Zephyr to the stock returns found in CRSP, an empirical analysis has been done with just the payment methods as independent variables. The main finding of this empirical analysis is that cash-financing can have a negative effect on the cumulative abnormal stock return of the acquirer. This finding is significant at a 5%-level. Using a combination of payment methods does not have a significant effect on the cumulative abnormal return of the acquirer.
After retrieving more deal and firm data, it was possible to set up a second regression analysis. In the second multiple regression analysis, the effect of payment methods, an industry measure, the deal value and the leverage ratio change on the cumulative abnormal return of the acquirer has been researched. The variable cash is significant at a 5%-level. It still has a negative relation to the cumulative abnormal return of the acquirer.
These findings are in line with the findings of Firth (1979) and Dodds and Quek (1985), who also found negative effects of cash-financing on the stock return of the acquirer. Firth (1979) defended his findings by stating that the market adjusted the stock prices of the acquirers for the risk of the acquisition. Dodds and Quek (1985) supported the finding of Firth (1979) adding that the market would risk-adjust the stock price when the market thought the acquirer was dealing too risky with its financial resources. It might be that in this paper, where the retail and wholesale sector is used, this is the case as well. It could be plausible that the market adjusts for a certain kind of risk, because the North-American retail and wholesale sector is highly competitive. As such, investing in different technologies or investing in other retailers, is perceived as a “high risk” activity. As stated in the interview with Miel Janssen, retailers experience a period of turmoil and anxiety following a takeover. Since turmoil and anxiety can have negative effects in the long-term, it is possible that the stock market incorporates this effect immediately into the stock price. In the discussion section, we will discuss this matter further.
6. Discussion
In this section we will discuss the limitations of this paper, suggest some improvements which could be implemented and sum up some subjects which would need more research.
6.1 Limitations and Improvements
This paper has its limitations. Firstly, the amount of observations is a bit low. In this paper 113 observations have been used to construct the regression model. When doing an empirical regression analysis, it is key that the amount of observations is minimally 100 observations. This to prevent a lack of explanatory power in the variables, because a small sample cannot explain phenomena of a big population.
In this paper we used acquirers who operate in the retail and wholesale sector, but the findings cannot say anything about other sectors. Next to that, it is not possible to state that the findings in this paper represent the total retail and wholesale sector, because we only studied North-American retailers. The possible lack of observations can be one of the reasons that the regression model is not as significant as one might want. This lack of observations is mainly caused by the sector itself. When searching for deals in the retail and wholesale sector, the maximum amount of deals found was 142 deals. It could be that in other sectors the maximum amount will be much higher. This could also be explained following the paper of Moatti et. al (2014), who found that creating internal growth was way more profitable in the long-term than doing an acquisition. Therefore, retailers will decide to do a takeover less often than in other sectors might be the case. The data of the Institute of Mergers Acquisitions and Alliances3 shows us that the retail sector is not really a
sector which does a lot of M&A compared to other sectors, such as the high tech or the financial sector. Since 1985 the retail sector has had almost 5% of all merger and acquisition deals. Compared to the financial sector, with 12% one of the highest, the retail sector does little M&A deals.
The lack of observations could also be due to the time period in which this paper has measured its variables. In eight years (2010-2017) 142 deals have been found, which is not that much. On the other hand, this period has been chosen on purpose. For the sake of this paper it had to be prevented that variables were measured during the financial crisis, because this might affect the explanatory power in a negative way, since the American
economy was still recovering back then. A longer period could have benefited the amount observations, but on the other hand it could have confused the regression model and its variables.
The lack of observations could also be due to the databases used. The Zephyr database has a lot of M&A deals, but does not have all of them. In addition, the CRSP database did not have all the stock data of the acquirers selected in Zephyr. Next to that, the Compustat database which has been used to calculate the leverage ratios did not have all the data as well. Probably, when the amount of observations would have been greater, the variables would be way more significant by which the model would have had better explanatory power. Therefore, if the data is available, further research should gather much more observations than this paper.
Secondly, the way of calculating the leverage ratio could be done differently if the data would be available. In this paper, the book values of both equity and debt have been used, but it would have been much better if the market values of the debt and the equity could have been used. Especially, when it is possible to measure the market values on a daily basis, we can construct a way better model. That way, we could have linked an increase of debt to an M&A deal in a much better way. Now we estimated the change in the leverage ratio by picking fiscal quarter book values from before and after the takeover. Further research, if the data is available, might want to use the market measures to get a better result. An alternative way to calculate the actual values of debt is to calculate the present values of the book values of the debt issued. In this paper it has not been possible to apply this method, because the interest rates of debt issuance are not known. We could have used the risk-free rate, but since the interest rates on treasury bonds is near zero, this would not have had an effect on the present values.
Thirdly, this paper uses a high level of significance (10 percent) for the measurement of significance of the first regression model. The problem with picking a higher level of significance is that the probability of a type I error gets bigger. The probability of a type I error is described as the probability of rejecting a null hypothesis while it is in fact true. When the significance level is higher, we will reject the null hypothesis earlier, so the probability of rejecting a true null hypothesis will be bigger. To reduce the probability of a type I error, we should measure with a lower level of significance. To then get a significant model, we will need more observations.
6.2 Further Research
Lastly, more research is needed to investigate the advantages mergers and acquisitions might have on the retail sector. Since this is a sector which is highly competitive, has low margins and is exposed to a lot of development, more research is needed whether mergers and acquisitions are a good and profitable way to pursue growth. Moatti et. al (2014) proved in their paper that M&A is just a short-term success story, but it would be interesting to research whether and in what ways M&A could also be a long-term success story for retailers.
Next to that, more research will be needed to investigate in what ways mergers and acquisitions can be made a better fit for the retail sector. For example, the retail sector is now trying to develop an omnichannel shopping experience. It could be attractive to research the effects of retailers acquiring tech companies who have the assets to deliver such a shopping experience on their firm performance. It may be questioned whether tech companies can easily be integrated in the retail sector and what development the retail sector should make to adjust to these technological improvements.
In addition, research has to be done on whether the findings of this paper contain in a bigger sample, because the findings in this paper contradict the economic theory and some economic literature. It might be that the retail sector is a sector which experiences different effects of deal and firm characteristics within M&A. Next to that, if the findings in this paper still contain in a bigger sample, it would be interesting if further research could investigate why retailers are still using method of payments within their takeovers which are not profitable at the short-term.
