1
Post-acquisition Performance of Financially Constrained and
Unconstrained Firms
&
Differences in the Acquisition Types
Davud Nanic July 2018
Amsterdam Business School University of Amsterdam
Thesis MSc. Finance
Study track: Corporate Finance Supervisor: Dr. Tomislav Ladika
2 Statement of originality
This document is written by Davud Nanic who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3 Abstract
This thesis investigates post-acquisition performance of both, financially constrained and unconstrained US acquirers between 2005-2015. Three different methods of financing a deal were considered: cash, stock and cash-stock financing. To develop the hypotheses, this thesis makes use of The Pecking Order Theory, but it acknowledges the importance of 3 other significant theories on Capital Structure. The methodology that was used was twofold; Event Study Methodology and the OLS Regression Analysis. This thesis concluded that the rationale of The Pecking Order Theory doesn’t hold, but further research should be more focused on The Market Timing Theory. Furthermore, the results suggest that stock financing is more favourable than cash financing, and that the financially constrained acquirers tend to have higher CAAR than the full sample of firms.
4 Table of contents
Page
1. Introduction 5
2. Literature Review 8
2.1. Post-acquisition Stock Returns 8
2.2. Capital Structure 9
2.2.1. Modigliani Miller and The Trade-off Theory 10
2.2.2. Pecking Order Theory and The Market Timing Theory 12
2.3. Differences in Methods of Payment 15
2.4. Financial Constraints and Precautionary Savings 16
3. Hypotheses Development 17
4. Data and Descriptive Statistics 19
4.1. Databases & Sample 20
4.2. Acquisition Variables 21
4.3. Dependent Variable 22
4.4. Independent Variable 22
4.5. Control Variables 23
5. Methodology 26
5.1. Event Study Methodology 27
5.2. OLS Regression Analyses 30
5.2.1. Univariate Analyses 30
5.2.2. Multivariate Analyses 30
5.3. Financially Constrained Firms and Precautionary Savings 31
5.3.1. Financially Constrained Firms 31
5.3.2. Testing Hypothesis 2 32 5.3.3. Testing Hypothesis 3 32 6. Results 32 6.1. Hypothesis 1 32 6.1.1. Univariate Results 33 6.1.2. Multivariate Results 34
6.2. Hypothesis 2 and Hypothesis 3 36
6.3. Discussion 38
7. Robustness Checks 40
5 1. Introduction
Firms and various institutions undergo an acquisition when they wish to create various synergies, strategically realign or diversify their goods and services. Additionally, bidders may want to gain substantial market power, or acquire a target for financial or tax considerations. However, an acquisition is a very large step for any bidder, and hence choosing the right means of financing such an investment is a very complex predicament.
Identifying the method of financing for firms that decide to undertake a merger or an acquisition is a paramount issue in terms of keeping their capital structure stable, as well as the post M&A performance. There are two ways of funding an acquisition: by using either internal or external cash funds. In other words, the acquiring firms must determine if the deal will be financed using either cash or stock. Rappaport and Sirower (1999) argue that cash financing sends a positive signal to the market because the management is confident in the value of their stock, as well as in the value of the deal. Furthermore, Rappaport and Sirower argue that stock financing sends a negative signal to the market because it projects the idea that the management is not confident in the value of their stock. Previous research (extensively discussed in the following section) suggests that deals where cash was used as a method of financing tend to lead to a better post M&A performance than deals where stock was used.
There are four main theories that revolve around the topic at hand: The Modigliani-Miller Theorem, The Pecking Order Theory, The Market Timing Theory and the Trade-Off Theory.
The Modigliani-Miller Theorem (1958) suggests that as long as the capital markets are perfect,
the firm’s capital structure is irrelevant for the firm’s value. However, perfect capital markets don’t exist due to the presence of a number of frictions which are caused by imperfect capital markets. Hence, the way a firm finances itself matters for value. The Pecking Order Theory argues that as the asymmetric information increases, the cost of financing will also increase. There are three main sources of financing, ordered in terms of preference – internal cash, issuing debt and finally issuing equity. Each subsequent source of financing is used as the previous one has been depleted or not been available (Myers and Majluf, 1984). The Market
Timing Theory claims that the managers will perform acquisitions when they have an
information their equity is overvalued. Baker and Wurgler (2002) argue that capital structure is the cumulative consequence of previous attempts to time the market of equity. This theory states there is no pecking order in external financing; firms will issue debt when that is more favourable than issuing equity and vice versa. By trading off the benefits and costs of debt and
6 equity, the Trade-Off Theory creates an assumption there is an optimal capital structure for a given firm. A firm can choose optimal amounts of debt and equity financing by outweighing their benefits and costs – i.e. tax saving of debt vs. dead-weight implications of bankruptcy (Frank and Goyal, 2008).
Travlos and Papaioannou (1991) have researched the impact of the acquisition type on the acquirer’s stock returns on the announcement day. This research investigated the differences between the methods of payment – either a cash or a stock payment for the financially constrained acquirer. They have found that the acquirers’ abnormal stock returns are much higher when cash is used as a mean of acquisition, as opposed to using stock acquisitions which generated much lower abnormal returns. On the other hand, Suk and Sung (1997) have performed a similar study, but their results showed no differences between cash and stock acquisitions on the post-acquisition performance. Martin (1996) suggested that the acquirers tend to use stock-for-stock acquisitions as their growth opportunities are higher, as well as if their pre-acquisition stock returns increase. Shleifer and Vishny (2003) argued that stock-for-stock mergers are performed by acquirers that want to sell their stock-for-stock which is overvalued. The research question this thesis will be answering is “What are the differences in the
acquirers’ (both financially and non-financially constrained) abnormal stock returns between different types of acquisitions?”
Given the existing literature, this question is interesting to explore because such a research has not been done yet, where the method of financing is tested in relation to both, financially and non-financially constrained firms. This thesis investigates the post-acquisition stock performance of the acquiring firms. The key independent variable represents the proportion of different acquisition types acquirers can use – either cash, stock or a mix between cash and stock acquisition types. Event study methodology and the OLS Regression Analyses are used to investigate what the cumulative average abnormal returns for the acquirers’ stocks are in the event windows (i.e. the acquisition period). Moreover, the second part of the thesis investigates the differences in returns between financially constrained and unconstrained acquirers, as well as differences between financially constrained acquirers that used different methods of payment, where the KZ Index is used to measure financial constraints.
Financially constrained firms are those that do not have enough internal cash to carry out an acquisition they desire. In other words, financial constraints occur when firms have insufficient internal cash, thus preventing them from financing the desired amount of investments they
7 initially wanted to undertake (Lamont et al., 2001). Hence, they will turn to capital markets to borrow the cash needed for an acquisition. Alternatively, firms can swap their own stocks for the stocks of the target they wish to acquire.
This research relates to the existing literature as it considers the predominant four theories on capital structure, and it explores the effect of different methods of financing on the acquirer’s stock prices. For that reason, this thesis will conclude by stating which of the aforementioned theories is mostly evident by analysing the results. Furthermore, this thesis will primarily make use of the Pecking Order Theory’s rationale when deriving the hypotheses, due to its straightforward nature, as well as an influential existence in the academia.
In that regard, as the Pecking Order Theory implies that the external financing is more costly than internal due to information asymmetry, the investors that would fund this acquisition worry it could be wasteful for reasons such as CEO empire building. Normally, the financially constrained firms hold a lot of their cash for precautionary reasons. Hence, financing an acquisition using cash is one of the ways to provide a strong signal to potential investors that the acquisition will generate high returns. The contribution this thesis adds to the area of research is relating the different methods of financing to both, financially and non-financially constrained firms to test what kind of financing is the optimal one for each of the subsamples (i.e. financially vs. non-financially constrained, whereas among the financially constrained firms, the two subsamples are the firms that used cash for acquisitions vs. those that used other methods). Hence, this thesis identifies several different subsets of firms for which cash usage should theoretically matter more for performance. There are several hypotheses developed that guided the research. Firstly, the higher the cash percentage used in the deal, the better the post-acquisition performance. Secondly, I expect to find lower CAAR for financially constrained acquirers than for non-financially constrained ones. Conversely, among the financially constrained acquirers, I expect to find higher CAAR for those that used exclusively cash than those who used other methods of payment.
The firms that will be included in the sample are all public US acquiring firms that successfully completed acquisitions between 2005 and 2015, where the minimum deal value is $5 million and where the acquirer owns 51% or more of target’s shares.
The rest of the thesis is structured in the following way. Section 2 will review the literature, trying to find the gaps that can be investigated in this research. Section 3 will develop the hypotheses necessary to test the model that will be used. Data and descriptive statistics will be
8 discussed in section 4, whereas the methodology will be explained in section 5. Section 6 will elaborate on the results, whereas the 7th section will conclude the research and offer potential suggestions for further research.
This thesis fails to reject the first null hypothesis, which was further supported with the robustness checks, because the findings suggested that as the proportion of cash financing for an acquisition increases, the firms’ CAAR will decrease. Furthermore, the results of the second hypothesis fail to reject the null hypothesis because there was limited statistical evidence that financially constrained firms have lower CAAR than financially unconstrained firms. This was supported by the robustness checks as well. Finally, the findings of the third hypothesis suggested that as the proportion of cash financing for a financially constrained firm increases, the firms’ CAAR will decrease, where the results further suggested that financially constrained firms have higher CAAR than the full sample of firms.
2. Literature Review
This research relates to the previous research by analysing the widespread theoretical framework necessary to comprehend the implications of the issue at hand, backing it up by the literature review that is oftentimes conflicting, where I am trying to find the gaps to create a sound research with unique contribution. The rest of the section is structured in the following way: Section 2.1. discusses the post-acquisition stock returns; section 2.2. discusses the main theories of capital structure; section 2.3. discusses the implications of different methods of payments and finally, section 2.4. discusses the financially constrained firms and the implications of precautionary savings. Finally, additional two sections in the Appendix discuss the motivations for undertaking M&A as well as the behaviour of stock prices of both, the acquirers and targets around the period of an M&A.
2.1. Post-acquisition stock returns
Understanding the dynamics of the stock returns after the acquisition is one of the core concepts this thesis revolves around. This underlying phenomenon is able to capture how successful the deal that took place is.
Travlos (1987) researched the effect of the method of payment on the acquirer’s stock returns. The results showed that returns are higher when cash acquisitions have been used when compared to the acquisitions financed using stock. These results were also supported by the Fuller et al. (2002), where he found that stock payments lead to negative returns, whereas the
9 cash payments lead to insignificant returns. Furthermore, Kiymaz and Baker (2008) researched various factors that could have an influence on the acquiring firms’ returns. Their results showed a negative impact on the returns when the acquiring firms have plenty of internal cash that can be used to finance the deal. On the other hand, the results also showed a positive impact on the returns when cash is used to finance the acquisition.
Travlos and Papaioannou (1991) investigated the differences in the methods of payment for financially constrained acquirors. The methods of payments used in the research were either cash or stock, and they have examined the impact of the two different method of payments’ on the stock returns of the acquiror on the announcement day of the acquisition. Despite the fact they haven’t included corporate governance levels in their analysis as the aforementioned study did, they found that the acquiror’s abnormal stock returns are greater when cash is used. A similar study conducted by Suk and Sang (1997) have found yet another different result when investigating the impact of different methods of payments on the acquiror’s post-acquisition performance. They found no differences between the different methods of payments. Furthermore, Loughran and Vijh (1997) concluded that 5-year buy and hold returns tend to be negative when stock is used to finance an acquisition, and positive when cash is used to finance an acquisition. Based on the method of payment, Rau and Vermaelen (1998) couldn’t conclude what the differences in the acquirer’s long-term performance are. However, they concluded that “glamour” acquirers tend to achieve lower returns in M&A.
Understanding the method of financing in relation to financially constrained and unconstrained firms, as well as grasping the differences in returns amongst those two subsets of firms is what is still unknown in the literature. Therefore, the contribution this thesis adds to the area of research is linking the varying financing methods to both subsets of firms to investigate what method is the most optimal for each of the subsets.
2.2. Capital Structure
This research investigates the differences in methods of payment on the post-acquisition performance of the acquiring firm. For that reason, it is important to identify what the predominant capital structure theories are and why some of them are in direct conflict, to show how achieving the optimal capital is achieved. By understanding how a certain capital structure is achieved, it will be understandable why a firm chose the certain method of payment for its acquisition – either cash, stock or a combination of the two. Therefore, this section examines the predominant capital structure theories.
10 The firm’s capital structure is the way a firm finances its assets and operations. To finance its operations, the firm can use debt, equity or different combinations of the two. This means that the firm can have a majority of an equity component, or a majority of a debt component, or an equal mix of the two. Each approach has different advantages and disadvantages. There are a number of different theories explaining capital structures, trying to determine what the relationship between the firm’s market value and financial leverage (proportion of debt in the firm’s capital structure) are. Such theories are the Trade-Off Theory, Modigliani Miller Propositions, Pecking Order Theory and Market Timing Theory.
2.2.1. Modigliani Miller Propositions and the Trade Off Theory
In the 1950s, Modigliani and Miller developed two propositions in response to the firms’ capital structures; particularly their relevance. Mainly, their view suggested that the market value of the firm will be unchanged, regardless of its changes in the levels of financial leverage. Instead, they proposed that the market value is dependent on the firm’s operating profits. Modigliani and Miller argue that the firm’s market value is also affected by its future growth prospect (i.e. the higher the growth prospects, the higher the market value of the firm). Their propositions stated that the firm’s value is neither dependent on the financing decisions, nor on the choice of capital structure. Instead, the value of the firm depends on the capability of its assets to generate value. Hence, it doesn’t matter if the assets come from internal or external capital. Moreover, Modigliani and Miller (1963) suggested that firms take on as much debt they can to benefit from tax deductibility of interest. The reason for this is to pay lower taxes than necessary, creating more value for the firm.
However, to make their theory plausible, certain assumptions had to be made. Generally speaking, their ideas on the irrelevance of capital structure is only possible under the assumption of perfect capital markets, where the following phenomena don’t exist: taxes, transaction costs, financial distress costs, asymmetric information, agency costs. The last assumption Modigliani and Miller made was no effect of debt on the firm’s EBIT.
In that regard, the two propositions developed by Modigliani and Miller take the aforementioned assumptions into account. The first proposition revolves around the main idea that the market value of the firm is not influenced by the firm’s capital structure. This means that leveraging the firm will not increase its market value. In addition to that, this proposition suggests that the shareholders and the debt holders in firm have the equal priority when the earnings are being split between the two parties. Finally, the first proposition states that the
11 total value of the firm equals the market value of all the cash flows generated by the assets the firm has (Modigliani and Miller, 1958).
The second proposition states that the financial leverage is directly proportional to the cost of equity. If the debt component in the capital structure increases, the equity shareholders will believe there is a higher risk for the firm. Therefore, the shareholders will expect a higher return, which will subsequently increase the cost of equity. The key distinction between the two propositions is that the second one assumes that the upper hand is held by the debt shareholders when it comes on claiming the earnings. Hence, the cost of debt will decrease (Modigliani and Miller, 1958).
If any of the aforementioned assumptions are relaxed, then the capital structure indeed does matter for the firm’s value.
Trade Off Theory assumes there are optimal capital structures that can be achieved by trading off the advantages and disadvantages of debt and equity. In other words, within capital structure, there are certain advantages to leverage until the seamless capital structure is obtained. The main advantage of using debt is the tax deductibility of interest (Kim, 1978), while the main advantage of using equity is not having an obligation of repaying the acquired money. Jalilvand and Harris (1984) argued that the disadvantage of debt financing is incurring the financial distress costs, including the bankruptcy and non-bankruptcy costs (e.g. suppliers who are demanding payment terms disadvantageous to the firm). The marginal benefit of increasing debt financing decreases as debt levels increase, whereas the marginal cost of increasing debt financing increases. Hence, when a firm is choosing the levels of debt and equity financing for optimizing its capital structure, it will focus on this trade-off. In addition to this, there are two other costs a firm incurs – distress and liquidity costs. Distress cost is a cost that comes into place if the stakeholders believe the firm will not continue working, whereas the liquidity cost is a cost that the firm incurs if it liquidates the net assets, resulting in the loss of value. Another disadvantage of using debt is the agency cost that leads to the conflict between the shareholders and managers or between the shareholders and debt-holders.
The main difference between the two theories is the expected benefit from using debt financing in form of tax benefits of the interest payments. This benefit is not recognized in the Modigliani Miller Irrelevance Theory, as they assume no taxes.
Fama and French (2002) had various results regarding the Trade-Off Theory, both in favour and against its predictions. Firstly, firms with higher profitability and fewer investments have
12 greater dividend payouts. Secondly, firms that are investing more have lower market leverage. The two statements are in line with the Trade-Off Theory. However, the fact that firms with higher profitability have lower leverage shows that the prediction revolving around the firm profitability is wrong.
2.2.2. Pecking Order Theory and The Market Timing Theory
Recent studies (Mazur, 2007; Quan, 2002) have argued that there has been a shift from Modigliani Miller view of the capital structure towards the Pecking Order Theory.
The Pecking Order Theory argues that the concept of the “target capital structure” doesn’t exist. Instead, the firms will make their capital and financing decisions in the following preference order: internal cash, issuing debt, issuing equity. Each succeeding method of financing is used when the previous one had been depleted. This is due to the signalling incentives that each method of financing sends to the potential investors. Internal cash is mostly valued as it signals that the firm has enough internal funds to finance various projects without having the need to resort to debt capital markets. On the other hand, issuing equity is least favourable as it signals to the investors that the management has an information that the firm is overvalued, trying to create an unfair opportunity at the expense of investors. Myers and Shyam-Sunder (1999) found that particular data features are better explained by the Pecking Order Theory than by the Trade Off Theory.
Furthermore, Myers and Majluf (1984) claimed that there is information asymmetry between the investors and managers of the firm. This means that the managers in the firm have more inside knowledge of the firm and will act in favour the existing shareholders. Therefore, as the asymmetric information between the managers and outside potential investors increase, the cost of financing will also face an increase. Firms often pass on valuable investment opportunities if they refuse to issue stock. Hence, they prove the Pecking Order Theory (Myers and Majluf, 1984).
Fama and French (2002) tested the predictions of the Pecking Order Theory and the Trade Off Theory regarding Dividends and Debt. Their results showed that firms with fewer investments and firms that are more profitable are paying higher dividends. In addition to that, they found that firms that are more profitable are also less levered, proving the Pecking Order Theory. On the other hand, firms with higher investment activities have less market leverage, proving the Trade Off Theory, as well as a more complex form of the Pecking Order Theory. Furthermore, lower long-term dividend payments are associated with firms that have higher investment
13 activities, but dividends don’t differ to accommodate short-term differences in investment. Short-term differences in earnings and investment is, as the Pecking Order Theory suggests, mainly absorbed by debt.
The solution Myers and Majluf offer to the information asymmetry problem is for firms to finance their investments using retained earnings instead of issuing new security. The implication this has is that issuing equity will become more expensive as the asymmetric information increases. Therefore, in firms where information asymmetric is large, issuing debt should be used to avoid selling securities which would be under-priced. However, Goyal and Frank (2003) argued that the Pecking Order Theory isn’t consistent where it should be, mainly for smaller firms where asymmetric information is an important issue.
A firm’s stock price decreases when the news of decreasing capital structure events get out. These news may include events such as an initial public offerings. On the other hand, news of increasing capital structure events are received well by the market because in these events, various prestigious financial institutions often become insiders that monitor the performance of the firm. Moreover, as the managers, as well as insider shareholders have better inside information than the potential outside investors, the use of equity is likely to be limited. The reason for this is to keep the control of the firm in tact (Hutchinson, 1995). Therefore, as Myers and Majluf already established, outside investors can only rely on various noisy signals that the firm gives out (i.e. level of capital structure).
Gaud et al. (2005) found that the transaction costs play a crucial role in the capital structure decisions. Namely, obtaining external financing (i.e. issuing debt and issuing equity) has much higher transaction costs than obtaining internal financing that do not generate any transaction costs. This conclusion is very consistent with the Pecking Order Theory. Furthermore, Zeidan, Galil and Shapir (2018) found that private firms’ owners in Brazil abide to the Pecking Order Theory – despite the presence of subsidized loans, the choice of capital structure indicates that the owners follow the Pecking Order Theory. The authors also show that the status of being a financially constrained firm is irrelevant on their findings.
A theory that is in direct conflict with the implications of the Pecking Order Theory is the Market Timing Theory. According to Baker and Wurgler (2002), capital structure is the cumulative consequence of various attempts the management had to time the market. This means there is no pecking order in external financing, but simply a matter of a better opportunity at a given point in time. Firms will issue debt when that is more favourable than
14 issuing equity and vice versa. In other words, acquiring financing for M&A investments depends on market fluctuations. For example, equity will be issued when the firm is overvalued and debt will be issued when the firm is undervalued. Therefore, management will acquire financing from the external market fluctuations when the timing is favourable. The conclusion that was reached was that the newly updated capital structure will be strongly related to the firm’s own historical market values.
Shleifer and Vishny (2003) investigated the misevaluations in the stock market of the merging firms. They concluded that companies whose equity is overvalued are better suited to undergo acquisitions of other companies, as well as grow and survive in the competitive markets. On the other hand, companies that are undervalued will become the takeover targets of the former ones. Selvarajah and Ursel researched what type of financing occurs in different periods and reached a conclusion that during periods of a high merger activity, there will be an increase of debt financing. These findings are consistent with the Pecking Order Theory, but not with the Market Timing Theory (Selvarajah and Ursel, 2012). Moreover, Harford focused on future leverage targets of the firm that will perform a merger in the article “Do firms have leverage targets?”. Hartford’s article pointed towards the fact that the financing method depends on the firm’s capital structure prior to the acquisition and the firm will choose the financing method to reach a desired future capital structure (Harford et al., 2009).
Equity market timing involves issuing stocks when the prices are high and repurchasing them when the prices are low. This means exploiting the fluctuations in the cost of equity. This theory works only in imperfect and inefficient markets. Hence, the Market Timing Theory is also in direct conflict with the Modigliani Miller Propositions that assume that the markets are efficient and integrated. Since in their assumptions the costs of different capitals don’t differ, there could be no gain from timing the market and switching between debt and equity. However, when the markets are inefficient, the cost of equity differs from other costs of capital. Another influential article, The Method of Payment in Corporate Acquisitions (Martin, 1996), identifies that firms which are primarily using equity financing have six common characteristics – high growth, high investment prospects, low cash, target’s & acquirer’s q-ratios are high, tender offers and presence of blockholders and institutional shareholders.
15 2.3. Differences in Methods of Payments
The different methods of payments this thesis considers are cash, stock and a combination of the two. Therefore, it is crucial to explore the characteristics of each, as well as the implications they entail. This section examines what the differences between each method of payment. Rappaport and Sirower (1999) identify a framework and two straightforward guidelines that managers of both parties involved in an M&A should follow when either paying for a deal or accepting it. Firstly, the acquiring firm must determine if the deal will be financed using either cash or stock. Secondly, if the managers of the acquiring firm opt for issuing stocks, then they have to determine if they will offer a fixed number of shares or a fixed value of shares. Rappaport and Sirower argue that offering cash entails a strong signalling message to the markets because the management team is confident in the value of their own stock, but also confident in the value of the deal proposed. However, this type of an acquisition places all the possible risks, but also rewards on the acquiring firm. On the other hand, Rappaport and Sirower claim that offering stock entails an unfavourable signalling message to the markets because it shows a lack of confidence in the acquiror’s stock value. Therefore, in essence, the acquiror proposes to share the recently merged/acquired firm with the stockholders of the firm that has been taken over. If the acquiring firm offers a fixed number of shares to finance the acquisition, then this act will only reinforce the unfavourable signalling impression. This is because the stockholders of the selling firm will be required to partake in the risk that the acquiror’s stock value will decrease before the deal actually goes through. On the other hand, if the acquiring firm offers a fixed value of shares to finance the acquisition, then the signalling message this act sends to the markers will be perceived as a more confident one. This is because the acquiring firm will assume all the risk there is.
Yang, Guariglia and Guo (2017) examined the extent to which the acquiring firms’ liquidity impacts method of payment, acquisition decisions and post-merger performance. Firms that are rich with cash are more prone to be acquirors (especially if such firms are inclined to tunnelling), whereas acquirors with high growth opportunities are less prone to acquire their targets using cash. Such an effect is more pronounced for acquirors that are financially constrained because they the opportunity costs of holding onto more cash are more severe. Lastly, using cash as a method of payment leads to under-performance in short and long-term, as opposed to using stock acquisitions. Therefore, cash-rich firms with poor corporate
16 governance system, where tunnelling is taking place, are motivated to undertake acquisitions using cash. Otherwise, it is more favourable to undertake acquisitions using stock.
2.4. Financial constraints and precautionary savings
Firms are labelled as financially constrained if they cannot finance the desired amount of opportunities using their internal funds. As it was previously established, such firms can turn to the external capital markets to acquire funds necessary for acquisitions. Namely, by issuing debt and/or equity (Lamont et al., 2001). In the “Modigliani and Miller world”, where various assumptions have been made, the financing constraints don’t present an issue because they argued that the cost of financing is the same if the firm uses internal or external funds. These types of financing are the “same” because any firm can get the required cash at the same cost as their own cost of capital. However, this doesn’t hold when the aforementioned assumptions are relaxed (Modigliani and Miller, 1958), because various frictions are present. Some of the financial frictions that cause the external financing to be more expensive than internal financing are agency problems, tax regulations and transaction costs (Jensen and Meckling, 1976). In addition to that, Myers and Majluf (1984) also argued that financial distress costs and asymmetric information issues are also financial frictions that cause the cost of external financing to be larger than internal financing.
As the Pecking Order Theory states, the choice of financing is determined on the costs ascribed to each of the three methods of financing, where the financially constrained firms cannot acquire the mostly desirable type of financing. Such constraints might leave an effect on the firm’s method of payment used in the acquisition. Lamont et al. (2001) argue that financially constrained firms will face a number of shocks and will have to deal with varying returns. The information asymmetry in the market allows the acquirer to finance the acquisition using stocks if the firm is overvalued or with cash if the firm is undervalued. Either method has a different signal of the acquirer’s value of shares because it is assumed that the management has better information about the firm than the outside investors.
These factors are crucial for this research as I am investigating the post-acquisition differences in returns between financially constrained and unconstrained acquirers, as well as the differences in returns among the financially constrained acquirers that used different methods of payment.
17 In addition to that, the precautionary savings are savings that occur if firms face an uncertainty about their future income. The precautionary motive to postpone the consumption into the future and save in the current period exists because there is a lack of completeness of insurance markets. This means that firms aren’t able to insure against the unfavourable state the economy will face in the future. Firms that engage in precautionary savings anticipate their earnings will decrease if this unfavourable state is realized. Hence, to retain a smooth path of consumption and to avoid unfavourable outcomes of future income fluctuations, firms will consume less in the current period, and invest the savings in the future, when they anticipate that the income will be lower. Firms often use the proceeds of the precautionary savings to invest into a project that will achieve economic growth and generate fixed capital (Skinner, 1987). Precautionary savings are crucial to consider in this research because financially constrained firms hold large amount of their cash for precautionary reasons. Therefore, an implication of financing an acquisition using cash is one of the ways for the financially constrained acquirers to provide a strong signal that the acquisition will generate high returns.
Brumberg (1956) concluded that current period savings were perceived significant to close the wedge between income in the current period and the highest previously earned income. Therefore, it was argued that precautionary savings were a hedge against a firm’s fluctuations in income. Furthermore, Aiyagari (1994) suggested that lower fluctuations of firm’s earnings resulted in a lower precautionary saving rate. Similarly, as the fluctuations of firm’s earnings began increasing, the saving rate doubled. Bearing in mind the business cycles of firms, Challe and Ragot (2010) argued during times of recessions, the motive for precautionary savings is embraced, causing aggregate consumption to decrease and aggregate saving to increase. This creates the domino effect in the economy, further worsening the impacts of macroeconomic recessions.
This thesis only deals with financial constraints where financial constraints are measured using the KZ Index. Hence, financial distress, economic distress and bankruptcy risk are not considered here, despite the fact these issues are highly related to one another.
3. Hypotheses Development
This section of the thesis makes use of the Pecking Order Theory’s rationale for its straightforward nature, as well as an influential existence in the academia.
Given that there is information asymmetry between managers and shareholders, the Pecking Order Theory is plausible because as managers know more about the inherent value of the firm,
18 they will finance their acquisitions accordingly. Specifically, using internal cash or issuing debt to acquire the cash necessary for the acquisition sends a signal that the firm is profitable and/or undervalued, and that the management is convinced the acquisition is profitable. On the other hand, using the stocks for acquisition sends a signal of a lack of confidence in the management, and it would lead the potential shareholders to believe the firm is overvalued.
This thesis explores the perception of (potential) firm’s investors’ decision between perceiving the firm as undervalued, and thus investing more in it, or overvalued, and investing less in it. I believe that the shareholders will attribute more value to the undervalued firms (i.e. firms that use cash for acquisitions, no matter what the source is). Otherwise, they are consciously giving the firm free money. If the results of the thesis aren’t supporting the hypotheses, this thesis will conclude that the Pecking Order Theory is not applicable on the given sample of firms and will try to deduce what the detrimental characteristic of firms that causes this disparity is.
I hypothesize that all firms’ post-acquisition performance will increase, but also that the firms that have used cash to finance acquisitions will have higher abnormal returns than the firms that have used stocks around the period of their acquisitions. In addition to that, the performance of the firms that have used both, cash and stock as means of acquisition will be lower than cash acquisitions but higher than stock acquisitions. This will depend on the ratio of cash to stock in that type of the acquisition – i.e. the higher the cash percentage, the better
the post-acquisition performance.
Hypothesis 1:
H0: as the proportion of cash financing for an acquisition increases, there will be no change or decrease in firms’ CAAR
H1: as the proportion of cash financing for an acquisition increases, the firms’ CAAR will increase
In response to the financial constraints and precautionary savings, another two hypotheses have been created. Precautionary savings imply that financially constrained firms hold more cash so they are able to fund their acquisitions in times of lower cash holdings. Therefore, financial constraints affect investment because financially constrained firms will not be able to acquire all the targets they desire, but only the most optimal ones. Hence, the signal that is being sent is related to the Pecking Order Theory because constrained acquirers will prefer to use cash as it signals the project they have been saving up for is the most optimal and signals the management believes the project will be profitable. In other words, the Pecking Order Theory
19 implies that external financing is more costly than internal due to the information asymmetry between the managers and investors. Therefore, the investors are worried the acquisition could be wasteful if the acquirer tries to obtain the money externally, as they may anticipate empire building by the CEO. Hence, financing an acquisition using cash sends a strong signal to outside investors that the acquisition will generate high returns.
On the other hand, financially constrained firms will underperform due to adverse financing circumstances, largely due to higher external costs of financing, which will be observed in their stock price.
In that regard, I expect to find lower CAAR for financially constrained acquirers than for non-financially constrained ones. Conversely, among the non-financially constrained acquirers, I expect to find higher CAAR for those that used exclusively cash than those who used other methods of payment.
Hypothesis 2:
H0: financially constrained firms have equal or higher CAAR than financially unconstrained firms
H1: financially constrained firms have lower CAAR than financially unconstrained firms
Hypothesis 3:
H0: among financially constrained firms, an increase in cash financing leads to unchanged or
decreased CAAR
H1: among financially constrained firms, an increase in cash financing leads to increased CAAR, where financially constrained firms have higher CAAR than the full sample of firms
4. Data and Descriptive Statistics
Section 4. discusses the databases used to generate and form the sample used in the thesis, as well as the disadvantages of them. In addition to that, this section discusses how the sample was formed and the key variables used. The rest of the section is structured in the following way: 4.1 discusses the databases used and the formation of the sample; 4.2 discusses the acquisition variables; 4.3 discusses the dependent variable; 4.4 discusses the independent variable and 4.5 discusses the control variables.
20 4.1. Databases & Sample
SDC Platinum, nowadays referred to as Thomson ONE, is the database that allows access to various company information, as well as event and industry information offered by different Thomson Reuters sources. The database includes information to the Deals Module (e.g. Mergers and Acquisitions), Equity Module (e.g. Initial Public Offerings), Private Equity Module, Share Ownership Module and the basic financials.
Thomson ONE has been used to compile the initial sample consisting of acquisitions’ information. This sample is made up of the US acquiring firms that have undergone an Initial Public Offering (i.e. US public firms) and have successfully completed designated acquisitions in the period between 1.1.2005 and 1.1.2015. This sample only includes deals whose value exceeded $5 million. Furthermore, this sample excludes all firms in the financial (SIC industry codes 6000-6999) and services (SIC industry codes 4000-4999) industries. In addition to that, only the acquirers that have owned more than 51% of target’s shares after the acquisitions are taken into consideration. There have been four different Deal Types considered in this thesis. These include Disclosed Value M&A, Acquisitions of Remaining Interest, Tender Offers and Exchange Offers. Disclosed Value M&A was included as this thesis deals only with transparent and known information. Therefore, the Undisclosed Value M&A would distort the results as it wouldn’t lead to a representative sample. Acquisitions of Remaining Interest represent a subsequent acquisition of the previous acquisition where less than 100% of the target was acquired. For this reason, it has been included in this thesis. A Tender Offer is one of the most usual types of acquisitions where the acquiror offers to buy shares (using cash) from the target’s shareholders. Acquirors usually add a premium when making Tender Offers. Lastly, Exchange Offers are a form of Tender Offers, where the Method of Payment is acquiror’s shares instead of cash. Finally, this thesis makes use of seven different Acquisition Techniques: Exchange Offer, Open Market Purchase, Tender Offer, Stock Swap, Financial Acquiror, Acquiror as an Investor Group and an Institutional Buyout. Financial Acquiror is the acquiror who performs acquisitions for the sole purpose of profiting from purchasing undervalued targets.
The resulting output of the aforementioned sample characteristics provides the information of the Method of Payments – either cash, stock, a mix between cash and stock, other means of payment and unknown types of payment.
Only the information regarding takeovers by means of cash, stock and/or a mix of the two is considered; any other type is dropped from the sample (i.e. other and unknown). In that regard,
21 the key independent variable named “Cash-to-Stock Proportion” is constructed that represents the proportion of the cash to stock used in the takeover. Section 4.4 discusses this in more detail.
Since the aim of this thesis is to find the Cumulative Abnormal Returns around the period of successful acquisitions, an event study has been performed using Eventus (section 5. discusses methodology in more detail). This platform allows the users to perform an event study by making use of the data from any pre-extracted source or from CRSP database. This thesis uses the Fama French Daily Event Study, where Fama French factors are used as benchmark. In addition to that, this thesis uses the PERMNO security identifier. This security identifier, together with the dates of each acquisition from the sample generated the Cumulative Abnormal Returns.
However, a major limitation of Thomson ONE Database is that it doesn’t allow the user to download the PERMNO security identifier, as the only identifier it contains is the CUSIP. Therefore, one has to turn to other databases to acquire PERMNO security identifiers for each firm used in the sample, as well as the other crucial firm identifiers including PERMCO and GVKEY. Within Wharton Research Data Services, the following two databases were used to acquire the necessary firm and securities identifiers: CRSP and CRSP/COMPUSTAT Merged. CRSP has been used to download nine-digit CUSIPs, PERMNOs and PERMCOs, whereas CRSP/Compustat Merged has been used to download the GVKEYs.
4.2. Acquisition variables
Understanding the acquisition variables is crucial for further understanding of the results of the thesis, as they provide us with a general understanding of how the takeovers performed, on a general level. These variables were obtained from the Thomson One database, once the sample was created.
Table 1. Descriptive Statistics of acquisition variables
(1) (2) (3) (3) (4) (5)
VARIABLES #Obs Mean Median Std. Dev. p5 p95
Owned After Transaction (%) 812 99.16 100 4.001 95.59 100
Value of Transactions ($ Millions) 812 769.8 154.94 1,691 9.441 3,618
Table 1. reports the descriptive statistics for the acquisition variables used in the formation of the Acquisitions sample from Thomson ONE Database. Those two variables are the Percentage Owned After Transaction and
22
the Value of Transaction (in millions of dollars). For each of the two variables, the reader can find the number of observations, the mean, standard deviation, the 5th and the 95th percentiles, in each respective column.
Percentage Owned After Transaction reported 812 observations, where the mean percentage was 99.16% and the median was 100%. These infer that the majority of takeovers entail an acquisition very close to 100%, meaning that the acquirers tend to purchase the entire firm. In addition to that, the standard deviation was 4.001%, and the 5th and the 95th percentiles are 95.59% and 100%, respectively, further augmenting the idea that firms tend to purchase entire firms.
Value of Transaction also reported 812 observations, where the mean value was $769.8 Million and the median was $154.940 Million. The standard deviation of the values is 1,691. The two latter indicators point towards the fact that there was a very broad array of different amounts of transactions taking place. To further strengthen that argument, the 5th percentile was $9.441 Million, whereas the 95th percentile was $3,618 Million.
4.3. Dependent variable
The dependent variable this thesis used was the cumulative abnormal returns. Essentially, this thesis generates three different CAARs for each of the given event windows (discussed in more detail in section 5.). The reason cumulative abnormal returns throughout the event window was chosen to be a dependent variable it to observe the differences between cumulative average abnormal returns across three different event windows, once the multivariate regressions have been completed. Positive cumulative abnormal returns infer favourable value implications, whereas negative cumulative abnormal returns infer unfavourable value implications. 4.4. Independent variable
The key independent variable this thesis used was generated using the output provided by Thomson One Database. This output was in form of cash and stock percentages used by each firm to finance their acquisitions. The independent variable that was generated takes the values between 0 and 1, where 0 represents acquisitions that were financed using 0% cash and 100% stock, whereas 1 represents acquisitions that were financed using 100% cash and 0% stock. The values in between 0 and 1 represent the proportion of cash to stock that was used to finance the acquisition, meaning the higher the number, the higher the ratio of cash used in the acquisition. Hence, this variable was named “Cash-to-Stock Proportion”.
23 4.5. Control variables
The control variables that will be discussed in this subsection are both, firm and deal specific variables, and they are: firm size, leverage, ROA, Debt HHI index, relative firm size – equity and relative firm size – assets.
The first few variables were computed using the natural logarithm. We do this to normalize the data, as it can be mis scaled. Hence, we scale it using the natural logarithm, so the variables can be interpretable.
Firm characteristics Firm size
The size of the firm is often measured by taking the natural logarithm of firm’s total assets. This thesis reports the firm’s size by taking the natural logarithm of the firm’s book value of assets at the end of the fiscal year prior to the year which the acquisition occurred in. Schwert (1995) argues that the likelihood of a takeover has a negative relationship with the firm size, meaning that larger firms aren’t likely subjects of takeovers. Therefore, firm size could also be perceived as a defence against takeovers. Further on, Dong et al. (2006) concluded that the shareholder wealth is positively affected by takeover defence – i.e. firm size. This thesis expects to find the firm size to be negatively related to the CAAR, meaning that an increase in firm size will result in a decrease in CAAR.
Cash holdings
This thesis measures the cash holdings by taking the natural logarithm of cash holdings at the end of the fiscal year prior to the year which the acquisition occurred in. As theory suggests, an increase in cash holdings is associated with firms being financially constrained, as they are generating precautionary savings that will be consumed when the period of lower cash holdings takes place. However, there could be various different reasons why a firm holds more or less cash.
Leverage
Acquirer’s leverage is calculated by dividing the total debt by the total assets, where the total debt is the sum of current liabilities and the long-term debt. Rajan and Zingales (1995) concluded that larger firms have more leverage, claiming that larger firms rarely fail and are often more diversified. Therefore, firm size can be an inverse proxy for the probability of
24 default. Further on, with an increase in the acquirer’s debt, the creditors will more closely monitor the management. Hence, the management of firms with higher leverage have an incentive to enhance the firm performance. In addition to this, Lang, Stulz and Walkling (1991) found that bad acquisitions aren’t an often occurrence in firms that have a high leverage because they cannot afford to spend the little cash they have on unfavourable takeovers. This thesis expects to find the negative relationship between leverage and CAAR, meaning that the CAAR will decrease when leverage increases.
Return on assets
The return on assets, ROA, is computed by taking the ratio of the acquirer’s net income by acquirer’s total assets. ROA is used to control for the acquirers’ performance in acquisitions. Morck, Schleifer and Vishny (1990) concluded that firms with higher levels of ROA have a tendency to to undertake better decisions when performing takeovers. Barber and Lyon (1996) argued that ROA is amongst the best measure of firm performance. This thesis expects to find the ROA to be negatively related to the CAAR, meaning that an increase in ROA will result in a decrease in CAAR.
Deal characteristics
Relative firm size – equity
The relative size of the firm is often measured by dividing the value of the deal by the acquirer’s equity market capitalization at the end of the fiscal year prior to the year the acquisition occurred in. Asquith, Bruner and Mullins (1983) concluded that the acquirer abnormal returns increase as their relative size increases. Therefore, such firm dynamics is consistent with value-maximizing behaviour by the acquiring firms’ management. This thesis expects to find the relative firm size to be negatively related to the CAAR, meaning that an increase in relative firm size will result in a decrease in CAAR.
Relative firm size – assets
Similarly, the relative firm size can also be measured by dividing the deal value by the natural logarithm of the acquirer’s assets at the end of the fiscal year prior to the year the acquisition took place in. This thesis expects to find the negative relation between the relative firm size (using assets) with the CAAR, meaning that the CAAR will decrease when the relative firm size increases.
25 Fixed effects
Year-fixed effects
The year-fixed effects capture the impact of aggregate time series movements. It is paramount to control for year-fixed effects because otherwise, the regressions attain the impact of aggregate trends that are unrelated with causal relationships. Therefore, we don’t want aggregate trends influencing regressions, so we include the year-fixed effects.
Industry-fixed effects
This thesis also adds the industry-fixed effects to the multivariate analyses by taking into account the 2-digit Acquirer SIC industry codes.
Table 2. Descriptive statistics for key variables
N Mean Median Std. Dev. Pct. 5 Pct. 95
CAR (-50, 50) 812 -0.046 -0.022 0.287 -0.554 0.359 CAR (-30, 30) 812 -0.026 -0.011 0.217 -0.432 0.322 CAR (-3, 3) Cash-stock proportion Firm Size Cash holdings Leverage Return on assets
Relative firm size – equity Relative firm size – assets
812 812 806 801 806 806 775 806 0.003 0.571 6.923 4.556 0.207 -0.001 0.323 91.050 -0.0004 0.702 6.818 4.477 0.152 0.041 0.112 23.629 0.077 0.441 2.109 2.173 0.219 0.192 0.521 186.137 -0.115 0 3.447 1.086 0 -0.389 0.007 1.963 0.129 1 10.609 8.322 0.644 0.168 1.319 392.431
Table 2. reports the summary statistics for the CAR (-50, 50), CAR (-30, 30), CAR (-3, 3), Cash-to-stock, firm size, leverage, ROA, Debt HHI index, relative firm size – equity, relative firm size – assets and year-fixed effects. For each of the variables, the reader can find the number of observations, the mean, median, standard deviation, the 5th and the 95th percentiles, in each respective column.
CAR (-50, 50) reports the Cumulative Average Abnormal Return of -0.046, where the median and standard deviation are -0.022 and 0.287, respectively. CAR (-30, 30) reports the CAAR of -0.026, where the standard deviation and median are 0.217 and -0.011. The third and final CAR (-3, 3) reports the CAAR to be 0.003, where the standard deviation and median are 0.077 and -0.0004, respectively. From the three different CARs windows, it can be observed that as the event window decreases, the CAAR increases.
26 Table 3. Matrix of correlations
Variables (1) CAR ( -50, 50) (2) CAR (-30, 30) (3) CAR (-3, 3) (4) Cash-stock (5) Firm Size (6) Cash Hold. (7) Lev. (8) ROA (9) Rel. size eq. (10) Rel. size as. CAR (-50, 50) 1.000 CAR (-30, 30) 0.784 1.000 CAR (-3, 3) 0.272 0.401 1.000 Cash-stock Proportion 0.040 0.014 -0.093 1.000 Firm Size 0.064 0.002 -0.026 0.374 1.000 Cash Holdings 0.073 0.030 -0.019 0.316 0.825 1.000 Leverage 0.018 -0.024 0.035 0.072 0.198 -0.025 1.000 Return on Assets 0.030 0.023 -0.010 0.331 0.488 0.376 0.014 1.000
Relative firm size – equity -0.018 -0.043 0.106 -0.362 -0.272 -0.301 0.064 -0.292 1.000
Relative firm size – assets 0.037 0.007 0.022 -0.064 0.408 0.360 0.044 0.163 0.283 1.000 Table 3. reports the matrix of the correlations between all of the variables, where each variable was correlated
once with one another variable.
It is significant to report these values as we can observe the relationships between the variables in isolation of any other factors. The dependent variable for the largest event window, (-50, 50), has a negative relationship with the Relative firm size - equity. For the second event window, (-30, 30), the dependent variable has a negative relationship with leverage and the Relative firm size - equity. Finally, the dependent variable for the shortest event window, (-3, 3), has a negative relationship with the key independent variable Cash-stock proportion, Firm size, Cash Holdings and the Return on Assets. From these observations, it can be inferred that the most precise dependent variable is the one from the shortest event window, as it most accurately presented the negative relationships between the CAAR and the independent and control variables. This thesis expects to find negative relationship between the CAARs and the included independent and control variables, as the theory argues that the acquirer’s post acquisition stock prices decrease.
5. Methodology
This section discusses the two methodologies used in this thesis to examine the impact of an acquisition on the acquirer’s share price with the primary purpose to test the outlined hypotheses. The methodologies used in this thesis are the Event Study Methodology and the Ordinary Least Square (OLS) Regression Model. The Event Study Methodology looks at the impacts of various corporate events on the share prices of the parties involved. In this case, the thesis looks at the Cumulative Abnormal and Cumulative Average Abnormal Returns of a number of different acquirers. On the other hand, the OLS Regressions test if acquisitions with higher cash ratios generate higher acquirer returns. Furthermore, the OLS regressions also test
27 if the financially constrained firms have lower CAAR than non-financially constrained firms, as well as if the financially constrained firms have higher CAAR if the method of payment was cash. Detailed descriptions of both methodologies are further elaborated in the following two subsections.
5.1 Event Study Methodology
The consequences of various corporate events are instantly evident in the stock prices of the firm involved in that event. For that reason, it is important to test the aforementioned hypotheses to assess the impact of the acquisition on the stock price of the acquirer. In Econometrics, the model that is used to test the given data is an Event Study Methodology. The key assumption considered in this thesis is that markets are efficient. Such an assumption provides the foundation for a successful Event Study (Fama, E. F., 1998) because it implies that the relevant financial and economic information that is accessible by market participants is fused with the stock prices as soon as the event takes place. This assumption is commonly referred to as the Efficient Market Hypothesis.
Event Study Methodology essentially involves obtaining abnormal returns of the events (i.e. acquisitions) by fine-tuning for the returns that result from the movements of the stock prices, which are in fact the market perceptions of the events being studied.
Event Study Methodology can be performed using several different models: Market Model, Constant Mean Return Model, Market Adjusted Return Model, Capital Asset Pricing Model (i.e. CAPM) and the Three Factor Model. These models differ in the alternating Estimation Windows and the market return benchmarks. This thesis uses the Market Model for its straightforward nature, as well its widespread use across academia. Several critical stages are required to carry out a successful Event Study. Firstly, the relevant events have to be identified that will undergo the Event Study Analysis. This thesis deals with all acquisitions carried out by the US public firms in the period of 2005-2015 where the deal value was larger than $5 million and the acquirer had more than 51% ownership of the acquired target.
Secondly, the Market Indices this thesis uses are CRSP Value Weighted, whereas the Benchmark Options are the Fama French Two Step and the Momentum Factor. The reason why Value Weighted Market Indices were chosen is because it accurately depicts the reality as the whole market is Value Weighted. On the other hand, Fama French is the ordinary “three factor” model, where the Momentum Factor was the fourth factor added to the analysis.
28 Thirdly, each firms’ normal returns are to be calculated. This is achieved by establishing an Estimation Period that occurs prior to the event taking place, where two parameters are estimated.
𝑁𝑅𝑖𝑡 = 𝛼̂𝑖 + 𝛽̂𝑖 ∗ 𝑅𝑚𝑡 (1)
Where NRit represents the normal return on security i in period t, Rmt represents the market return in period t, and αi and βi represent unknown parameters that are to be estimated for each security i. This thesis uses tree different Estimation Periods. The first one, the “End Before Event Date (EST)” was chosen to be -31 days. This represents the final day of the estimation period in the analysis. The second one, “Minimum Estimation Length (MINESTN)” was chosen to be 90 days. This represents the least number of days that the acquirer’s stock can be traded during the estimation window in order for the acquirer to still be included in the chosen sample. The third one, “Maximum Estimation Length (ESTLEN)” was chosen to be 500 days, as the maximum estimation should provide enough information for a thorough analysis. In all three cases, the Estimation Method was OLS.
Fourthly, this thesis uses three different Event Windows, where, according to the Efficient Market Hypothesis, the effects of information disclosure are captured in the stock prices of the firms in the sample. The abnormal returns over each Event Window for every firm in the sample are calculated using the following formula:
𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡− 𝑁𝑅𝑖𝑡 = 𝑅𝑖𝑡− (𝛼̂𝑖+ 𝛽̂𝑖∗ 𝑅𝑚𝑡) (2)
Where: 𝐴𝑅𝑖𝑡 represents the abnormal returns on a security i in period t and 𝑅𝑖𝑡 represents the actual returns on a security i in period t. Average abnormal return is calculated by computing the mean value of the abnormal returns of each firm for every day of the corresponding event window. The abnormal returns for each firm are aggregated under the assumption that for all the firms in the sample, the abnormal returns’ distribution is independently and normally distributed. Finding the mean value of the abnormal returns reduces any noise effect of an individual outlier in the sample. Abnormal performances are usually spotted by large deviations from zero in the average abnormal returns. The following formula is used to compute the average abnormal returns in period t:
𝐴𝑅 ̅̅̅̅𝑡 = 𝐴𝐴𝑅𝑡 = 1 𝑁∑ 𝐴𝑅𝑖𝑡 𝑁 𝑖=1 (3)
29 However, an Event Study Methodology considers the performance of the stock around the period of the event – i.e. the Event Window. This is computed using the cumulative abnormal returns, where we aggregate the abnormal returns throughout the entire event window, and the formula for calculating is the following:
𝐶𝐴𝑅𝑖(𝑡1,𝑡2) = 𝐴𝑅𝑖𝑡1+. . . +𝐴𝑅𝑖𝑡2 = ∑ 𝐴𝑅𝑖𝑡 𝑡2
𝑡=𝑡1
(4)
Where CAR represents the Cumulative Abnormal Returns on a security i over the Event Window.
The estimation periods this thesis uses are the following: [-50, 50], [-30, 30] and [-3, 3]. Three different Event Windows have been chosen to investigate the Cumulative Average Abnormal Returns over different time periods. Essentially, all three of them capture the effect of the acquisition on the price of the acquiror, by taking into account the stock price movements around the period of the acquisitions. The reason why three different ones are used in the analysis is to observe the differences across different periods used, after including the control variables. Another important reason to use different periods is to observe whether there is information leakage, i.e. returns increasing before the announcement. Cumulative abnormal returns with a positive sign would imply that the market reacted well to the acquisition, whereas a negative sign would imply that the market didn’t react well to the acquisition.
Cumulative Average Abnormal Returns have been computed for each of the three event windows by finding the mean value of all firms’ Cumulative Abnormal Returns in that given event window. The formula to calculate Cumulative Average Abnormal Returns is:
𝐶𝐴𝐴𝑅 = 1
𝑁∑ 𝐶𝐴𝑅𝑖𝑡 𝑁
𝑖=1
(5)
Similarly to the CAR inference, a CAAR with a positive sign would imply a positive market reaction, whereas a CAAR with a negative sign would imply a negative market reaction. Finally, this thesis uses one statistical test to assure the event study methodology is reliable. The test used in this thesis is the Patell Test (Patell, 1976) which is able to point out, at a certain level of significance, if the cumulative abnormal returns are significantly different from zero. Furthermore, Eventus Software was used in this thesis to carry out the Event Study Methodology. Such a software allows the users to perform reliable and valid studies.
30 This test gives the Cumulative Abnormal Returns for each separate acquisition in the sample. 5.2 OLS Regression Analyses
To reach the conclusion and the outcome of all three hypotheses, the OLS Regressions test if higher acquirer returns are generated by firms with higher cash ratios as a method of payment in a takeover and if financially constrained firms generate lower returns than non-financially constrained firms.
5.2.1 Univariate Analysis
Daily average abnormal returns of the firms in the sample are compared on day 0 to perform the univariate analysis. Upon completion, a number of control variables have been added to perform the multivariate analysis which tested the validity and reliability (i.e. achieving identical results by using different methodologies) of the results that were obtained using the univariate analysis. Essentially, the multivariate analysis measures the impact of the independent variable, cash-stock percentage, on the dependent variable, the abnormal returns of the firms when the acquisition takes place.
5.2.2 Multivariate Analysis
The regression model developed in this thesis investigates the relationship between the two key variables: the generated independent variable, cash-stock percentage, and the computed abnormal returns. This model regresses the dependent variable, the abnormal returns, on the independent variable, cash-stock percentage, as well as on the selected control variables. The control variables include firm size, cash holdings, leverage, ROA, Debt HHI index, relative firm size – equity and relative firm size – assets. The usage of these control variables diminishes the omitted variable bias.
The regression model for each event window, where all the control variables have been added, will be computed in the following way:
Regression for the Event Windows (-50, 50); (-30, 30); (-3, 3)
𝐶𝐴𝐴𝑅𝑖 = 𝛼𝑖 + 𝛽1∗ 𝐶𝑎𝑠ℎ𝑠𝑡𝑜𝑐𝑘𝑖𝑡+ 𝛽2∗ 𝐹𝑖𝑟𝑚 𝑠𝑖𝑧𝑒𝑖𝑡+ 𝛽3∗ 𝐶𝑎𝑠ℎ 𝐻𝑜𝑙𝑑𝑖𝑛𝑔𝑠𝑖𝑡 + 𝛽4 ∗ 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖𝑡+ 𝛽5∗ 𝑅𝑂𝐴𝑖𝑡+ 𝛽6∗ 𝑅𝑒𝑙. 𝑠𝑖𝑧𝑒 𝑒𝑞𝑢𝑖𝑡𝑦𝑖𝑡+ 𝛽7
∗ 𝑅𝑒𝑙. 𝑠𝑖𝑧𝑒 𝑎𝑠𝑠𝑒𝑡𝑠𝑖𝑡+ (𝑌𝑒𝑎𝑟 𝐹𝐸𝑖𝑡) + (𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸𝑖𝑡) + 𝜀𝑖𝑡 (6)
To ensure the validity and reliability of results, this thesis used StataSE 15 (64-bit) programme to run the aforementioned regressions.
