Wealth effects of abandoned takeover targets

(1)

Wealth effects of abandoned takeover

targets

Master thesis

University of Groningen

Faculty of Economics and Business

MSc Business Administration

Specialization Finance

Stefan Bos Esdoornlaan 364 9741 MB Groningen 06 – 10686627 bosstefan@hotmail.com Student number: 1654721

(2)

Abstract

(3)

1. Introduction

Mergers and acquisitions have received a lot of academic attention. Interestingly however, the share price effects of abandoned takeovers on the involved parties is nearly

unexamined. This is remarkably since there is a large number of merger and acquisitions that are started yet never finished (Pett et al, 2003).

Firms that receive a takeover offer have to make a decision whether to accept or reject the offer. While it seems to be interesting to accept the offer because of the premium in the offer, it is possible that rejecting the offer creates value. Receiving a takeover offer may signal that the assets of the target firm were not being utilized efficiently prior to the takeover attempt and that changes must be made (Dodd and Ruback, 1977 and Bradley et al, 1983). Moreover, rejecting an offer by the target firm may reveal positive information to the market because the target would only reject an offer if the underlying value of the firm is higher than the offer(Pett et al, 2003 and Bradley et al, 1983). On the opposite, when the bidding firm withdraws the offer this may reveal negative information to the market.

Therefore, Pett et al (2003) suggests further research in the market reaction based on which party abandons the offer. The difference in abnormal return for target firms in abandoned takeovers between offers abandoned by targets and offers abandoned by bidders has never been researched for the period from the announcement to after the abandonment. The research question I want to answer is therefore:

Is there a difference in share price reaction for target firms between offers abandoned by target firms and offers abandoned by bidding firms?

Managers and shareholders of target firms can make a better decision whether to accept or reject a takeover offer when the share price consequences of rejecting an offer is known. Rejection of a takeover offer by target management may communicate that the

(5)

S. Bos 2 of the takeover offer when they reject the offer. Furthermore, the probability of a new takeover offer subsequent to an abandoned offer is important to know for the decision to accept or reject an offer. When they know that the share price does not fall back after their abandonment and the probability of a new takeover offer subsequent to an abandoned offer is large, shareholders will be less inclined to accept an offer. Therefore, it is important to research whether the abnormal returns differ between abandonment initiated by bidding firms and abandonment initiated by target firms. Difference in abnormal returns between takeovers that are abandoned by targets and takeover that are abandoned by bidders has never been researched for the period from the announcement to after the abandonment.

The sample of this paper consist of target firms from the USA that have received an

abandoned takeover offer in the period 1997-2008. To measure the stock market’s response for these targets, the cumulative abnormal returns (CAR’s) will be measured during several time periods. It will be tested whether the CAR’s differ between subsamples for these periods. Furthermore, the non-parametric generalized sign test will be applied.

There are hypothesis in the literature which predict the share price of target firm at

abandoned takeovers. The synergy hypothesis states that the aggregate value of two firms is higher when they combine their resources. Therefore the share price of target firms will fall back to their pre-bid level when an offer is abandoned. Because it is synergy that causes the share price to rise and fall, it makes no difference whether the target or the bidder abandons the takeover according to the synergy hypothesis (Dodd and Ruback, 1977). The information hypothesis posits that the announcement of a takeover reveals positive information

(6)

S. Bos 3 In this paper, strong evidence is found that not only the announcement but also the

abandonment of an offer reveals positive or negative information to the market. Over the whole takeover period, from announcement to abandonment, there is a significant

(7)

S. Bos 4

2. Background literature

As mentioned previously, the consequences of takeover abandonment have been a relatively unexplored academic topic. In table 1, an overview of the results of some

important studies on takeover abandonment is given. A study by Asquith (1983) examined the cumulative abnormal returns (CAR’s) during several event time periods for bidding and target companies for successful and abandoned mergers. The share price of abandoned takeover targets rise about 12% at the announcement of the offer. However, most of this premium is already lost in the interim period. At the abandonment date, the share price is on average 5% below the pre-bid level. This decline continues and mounts up to 12% below the pre-bid level two years after the abandonment. Asquith did not divide the sample into abandonment per initiator yet posits that the decline to a below pre-bid level is caused by possible takeover candidates that are already forecasted by the market in advance of an takeover attempt and therefore there is already a takeover premium incorporated in the share price. After the abandonment of a takeover, this premium disappears and the share price declines to a level beneath the pre-bid share price.

Bradley et al (1983) examined the monthly abnormal returns in tender offers for bidding and target companies in successful and unsuccessful tender offers until five years after the announcement of the offer. Upon the month of the announcement of the offer, the share price of target firms rise about 30%. Bradley et al (1983) divided the sample into targets that received a new successful takeover offer subsequent to an abandoned offer and targets that are not taken over. The share price of targets that are taken over subsequent to an

(8)

S. Bos 5 Table 1

Overview of previous studies on abandoned takeovers.

In this table an overview is given of important studies done in the past about abandoned takeovers. The author(s), the country, study-period, type of data, sample size, event window, research period(s), the results, and the important conclusions are mentioned.

Author(s) (year)

Country Study period Data Sample Size Event window

[months] (research period)

Results Conclusion

Dodd (1980) USA 1971-77 Daily 80 [-2, +2] (announcement) [-2, +2] (abandonment) [-2, +2] (abandonment by target) +13.0% -9.75% -3.75% Permanent revaluation of 10% when target management abandons takeover.

Share price returns to pre-bid level if abandonment was not caused by target management.

Pett et al

(2003)

USA 1982-92 Daily 84 [-1/2, +1/2] (announcement) : [-1/2, +1/2] (abandonment) :

+17.0% -4.0%

The value of the target after the abandonment is higher than pre-announcement value.

Dodd and Ruback (1977)

USA 1958-76 Monthly 36 [0] (announcement) : [+1, +12] (announcement) :

+ 19.0% -3.3%

Target shareholders realize a significant capital gain regardless of the outcome of the offer.

Bradley et al (1983)

USA 1963-80 Monthly 112 [0] (announcement) [+1, +24] (announcement)

+29.4% -27.5%

Share price falls back to pre-bid level within two years.

(9)

S. Bos 6 Author(s)

(year)

Country Study period Data Sample Size Event window

[months] (research period)

Results Conclusion

Bradley (1980) USA 1962-77 Daily 97 [-1/10, +1/10] (announcement) (announcement till abandonment)

+ 29.0% +16.0%

Rejecting a tender offer is a value-maximizing decision for the target shareholders.

Asquith (1983) USA 1962-75 Daily 91 [-24, -1] (announcement) [0] (announcement) (interim period) [0] (abandonment) [+1, +12] (abandonment) -10.5% +11.7% -8.1% -6.4% -8.7%

Share price falls back below the pre-bid level after the

abandonment

(10)

S. Bos 7 Dodd (1980) examined the stock price reaction around the announcement date and around the abandonment date for successful and abandoned mergers. The sample of abandoned merger proposals are split into two subsamples: merger proposals that are abandoned by the target firm and merger proposals that are abandoned because of another reason (for example by the bidding firm or by market authorities). In the two days around the

announcement, target firms earn an abnormal return of about 14%. After the announcement of the offer, the returns of the subsamples differ. Shareholders of

abandoned targets earn a significant abnormal return of -3.75% around abandonment date when the proposal is abandoned by the target firm and a significant abnormal return of -9.75% when the proposal is not abandoned by the target firm. Dodd (1980) shows that the market reacts less negative to an abandonment initiated by the target firm than from abandonment not initiated by the target firm. According to Dodd (1980) there is a large positive revaluation of 10.25% of target shares when the takeover is abandoned by the target firm and a small positive revaluation of 4.25% when the abandonment is not initiated by the target firm. However, Dodd (1980) measured the abnormal return only around the announcement and around the abandonment of the offer and did not measure the

abnormal return in the interim period between the announcement and abandonment of the offer. The studies of Asquith (1983) and Bradley et al (1983) have shown the importance of measuring the abnormal return in this period to measure the total effect of takeover

abandonment for target firms. In the interim period, the abnormal returns for target firms is negative (Asquith, 1983 and Bradley et al, 1983). Therefore, the return for target firms over the total takeover period is probably less than the 10.25% and 4.25% given and possibly negative when the abandonment was not initiated by the target firm.

Concerning the announcement return, I expect, following prior research to find positive abnormal returns. Therefore, hypothesis 1 becomes:

(11)

S. Bos 8 This paper the abnormal returns of target firms involved in abandoned takeovers will be researched in different time periods. For the other time periods, I point out the following hypotheses:

Hypothesis 2: The interim period between the announcement and the abandonment of a takeover offer, will be associated with negative abnormal returns for target firms.

Hypothesis 3: Abandonment of a merger or acquisition will be associated with negative abnormal returns for the target firms around the event day.

Hypothesis 4: In the year subsequent the abandonment date of a merger or acquisition there will be no abnormal returns for target firms.

Hypothesis 5: For target shares at abandoned takeovers there will be no permanent revaluation over the total takeover period, from announcement until abandonment of the takeover offer.

I expect the abnormal returns around the abandonment date to differ between

abandonment by target firms and abandonment by bidding firms. Therefore, hypothesis 6 becomes:

Hypothesis 6: Around the abandonment date of a takeover, the abnormal return of target firms where the target firm abandons the takeover is higher than the abnormal return of target firms where the offer is abandoned by the bidding firm.

(12)

S. Bos 9 takeovers. However, the two hypotheses have contradictory predictions for the returns of abandoned takeovers.

According to the synergy hypothesis, the aggregate value of two firms is higher when they combine their resources because this creates synergy (Bradley et al, 1983). Therefore, the share price of the target firm will rise at the announcement of a takeover offer. In the interim period the share price will fall because of uncertainty about the probability of a successful takeover. When the takeover abandons, the share price of the target will gradually fall back to a pre-bid level because synergy can only be accomplished when the resources of the companies are combined (Dodd and Ruback, 1977). According to the synergy hypothesis the share prices of target companies can drop to below pre-bid level because there is already a premium incorporated in the share price for possible takeover candidates (Asquith, 1982). After the abandonment of a takeover, this premium disappears and the share price declines to a level beneath the pre-bid share price. Because it is synergy that causes the share price to rise and fall, it makes no difference whether the target or the bidder abandons the takeover according to the synergy hypothesis. The conclusions of the studies of Bradley et al (1983) and Asquith (1982) are in line with the synergy hypothesis because according to these studies the premium at the announcement proves that the two firms would be worth more if they combine than as separate firms. After the abandonment this premium disappears and the share price falls back to or below its pre-bid level.

(13)

S. Bos 10 wealth will increase if the inefficiency is eliminated. The wealth of the target shareholders will increase irrespective of the outcome of the offer unless there are permanent barriers to eliminate the inefficiency (Dodd and Ruback, 1977). The share price of target firms can decline in the interim period and at the abandonment date according to the information hypothesis because of uncertainty whether the inefficiency can be eliminated (Dodd, 1980 and Dodd and Ruback, 1977) . However, this decline will not eliminate the total gain from the announcement of the offer.

The announcement of the abandonment of the offer also reveals new information to the market (Pett et al, 2003). This information is less negative when the target firm abandons the offer compared to when the bidding firm abandons the offer (Pett et al, 2003). The target firm will only abandon the offer if they think the underlying value of the firm is higher than the offer or expect a higher offer in the future (Pett et al, 2003, Bradley, 1980, and Bradley et al, 1983). According to Bradley (1980) target managers may be acting in the interest of their shareholders by opposing an outstanding offer. The abnormal

announcement return for targets of abandoned takeover offers is 29%. After the abandonment, the abnormal return is advanced to 45% when the offer is abandoned because of opposition by the target management. On the opposite, when the bidding firm abandons the offer this may reveal negative information to the market because a bidding firm will only abandon the offer when they have new information about the underlying value of the target firm or when there is a barrier to eliminate the inefficiency. The results of Bradley (1980) and Dodd (1980) are in line with the information hypothesis since target shares are positively revaluedirrespective of the outcome of the offer.

To answer the research question, I will research whether the abnormal returns for target firms differ between abandonments caused by target firms and abandonments caused by bidding firms. Hypotheses 7 and 8 therefore becomes:

(14)

S. Bos 11 Hypothesis 8: Over the whole takeover period, from announcement to abandonment, there is no abnormal return for target sharers when the abandonment is caused by the bidding firm.

In general, the share price of the target firms involved in abandoned takeovers will rise at the announcement of the offer and decline around the abandonment of the offer. The hypotheses in the literature differ in their forecast of the time and the magnitude of the decline. Table 2gives an overview of the predicted wealth effects in different time periods of target companies involved in abandoned takeover offers according to the synergy

hypothesis and to the information hypothesis.

Table 2

Overview of the predicted wealth effects in different time periods of target companies involved in abandoned takeover offers according to the synergy and the information

hypotheses.

A distinction has been made between an abandonment initiated by the bidding firm and abandonment initiated by the target firm.

Time period Synergy

Hypothesis Information Hypothesis Abandonment by bidder Abandonment by target Abandonment by bidder Abandonment by target Announcement of offer Positive Positive Positive Positive

Interim period Negative Negative Zero / negative Zero / negative

Abandonment Negative Negative Negative Negative

Post-abandonment period Negative Negative No impact No impact Abnormal return over total

period Zero / negative Zero/negative Negative Positive

If the information is positive when the target abandons the offer and negative when the bidder abandons the offer, this would also mean that the probability of a new successful offer subsequent to an abandoned offer is larger for targets that abandoned the offer compared to targets where the bidder abandoned the offer. Therefore, it will be interesting to research whether the probability of a takeover differs between the subsamples.

(15)

(16)

S. Bos 13

3. Methodology

This section will discuss the methodology used in this paper. Section 3.1 will discuss the event study methodology. Section 3.2 describes the structure of this study. The computation of the abnormal return and the choice of the market index will be described in Section 3.3. Last, the non-parametric generalized sign test will be explained in section 3.4.

3.1 Event study

To examine the returns of target companies at abandoned takeover offers, the event study methodology is used. An event study examines the abnormal company stock return at or around a certain date, the event date, to measure the impact of an event on the value of a firm (MacKinlay, 1997). The event study methodology uses an estimation window, an event window and a post-event window. However, this study is different from most other event studies in finance because it has 2 events, the announcement of a takeover offer and the abandonment of the takeover offer. Therefore, this study has two event windows. Furthermore, the cumulative abnormal return between the announcement and

abandonment is measured. The efficient market hypothesis states that security prices adjust quickly to fully reflect new information (Brown and Warner, 1985). Therefore, the abnormal performance at the event date is a measure of the wealth impact of that event on the shareholders of the company.

3.2 Structure of this study

The aim of this study is to examine whether the abnormal returns of target firms differ between abandoned takeovers that are initiated by target firms and abandoned takeovers that are initiated by bidding firms . To evaluate the stock market’s response to an

(17)

S. Bos 14 Figure 1: Timeline of this study

Figure 1 gives a timeline of this study. This study is different from other studies in finance in that it has 2 events, the announcement and the abandonment of a takeover offer. Furthermore, the CAR

between the events is measured.

The estimation period

The estimation period is the period prior to the announcement of the takeover offer and is used to calculate the expected return. Most other literature on the effects of abandoned takeovers do not use an estimation window. The literature that did use an estimation window, used monthly data. Measurement of expected return important in this study because the abnormal returns are measured over long horizons. Therefore, I choose for a relatively long estimation period of 250 days.

The announcement period

The event date is the day of the announcement of the takeover offer and is defined as day 0. The event window is some days around the event and is used to examine if there is a share price reaction around the announcement of the takeover. Setting the right event window is important for an event study. The event window has to be long enough to cover the entire effect of the event on the share price. For this study the event period can also not be too long. For some takeover offers, there will be uncertainty about the outcome of the offer soon after the announcement. Setting a too long event window will result in a too low announcement return. Furthermore, setting a long event window will exclude more firms because the announcement and abandonment window will overlap each other. Therefore, an [-2, 2] event window is used in this paper. This is shorter than other studies on the effects

of abandoned takeovers (see Table _{). However, based on an analysis of the daily abnormal} returns around the events, this event window is long enough to capture the effect of an announcement or abandonment. The event window is excluded from the estimation period

Estimation period Post-abandon

period Interim period Announcement of takeover offer Abandon date 0 T1 T0 0

(Interim period) (Event window 2) (Post-event window) (estimation window) (Event window 1)

T5 T4

(18)

S. Bos 15 to prevent influence of the event on the parameters of the market model (MacKinlay, 1997). As a result, the estimation period is set as [-252, -3].

The interim period

The interim period is the period from the announcement until the abandonment of the takeover offer. The abnormal return is measured from three days after the announcement until two days before the abandon date. The time between the announcement of the takeover offer and the abandonment of the offer differs for every event. Hence, it is not possible to directly compare the CAR’s of firms in this period because of the difference in time period. Table 3 gives an overview of the time (in trading days) between the

announcement date and the abandonment date for the total sample and for the subsamples.

Table 3

Time in trading days between announcement and abandonment of the takeover offers

In this table an overview of the number of trading days between the announcement of the takeover offer and the abandonment of the offer for the sample of abandoned takeovers is given. The subsamples are given separately. The last column gives the total number of events in the sample.

(19)

S. Bos 16 Measuring abnormal return over long-horizons involves some problems that do not arise or are less important with short horizons. The first problem is that all models for measuring expected return show problems describing average returns. This problem is known as the bad-model problem, this problem is unavoidable (Fama, 1997). As a result, any found abnormal returns is a combination of the error in the model for measuring expected return and the return from the event. Bad-model problems are more serious when measuring and testing returns over long horizons because the standard error changes only with the square root of the error wherewith the mean of the CAR changes because of this bad-model problem (Fama, 1997).

Another problem of measuring abnormal returns over long horizons can be cross-sectional dependence in the sample. Cross-sectional dependence increases with the length of the test period and with sample size (Fama, 1998). I choose not to correct the data for

cross-sectional dependence because the procedures for correcting the data are mostly used for monthly data, like the calendar time portfolio approach (Fama, 1998), and for measuring abnormal returns over periods of more than one year. Instead, I choose to add a non-parametric test to reduce the problems with measuring abnormal returns over long horizons. Nonparametric procedures appear to have fewer potential problems, and

conclusions based on these procedures seem less likely to be due to misspecification of the test (Kothari and Warner, 1997).

The abandonment period

The abandonment of the takeover offer is the second event of this study. Like with the announcement of the takeover the CAR is calculated from two days before the

abandonment until two days after the abandonment.

The post-abandonment period

(20)

S. Bos 17 abandonment. Therefore, there can still be a premium in the share price because the market can anticipate on a new takeover offer. This premium will resolve over time. The

post-abandonment period starts the third day following the abandon date of the takeover offer and the abnormal returns are examined until one year after the abandonment.

3.3 Measurement of abnormal return

The daily return on a firm share and the market index are calculated as the continuously compounded returns:

= 100% × ln

− 1 1

where is the return of a firm i on day t, and is the price of the firm share on day t.

To calculated the return by continuously compounded returns is normal for finance studies. The advantages of this method in this study is that the continuously compounded returns are time-additive (Brooks ,2002, pp. 6-8). This is important for this study since returns are calculated for periods longer than one day.

The abnormal return of a company is measured by the difference between the actual return and the expected return of the company share. The expected return is based on a

benchmark. Brown and Warner (1980, 1985) and MacKinlay (1997) discuss three statistical models to calculate the expected returns; the mean adjusted model, the market adjusted model, and the market model. Although simpler models, like the constant mean return model and the market adjusted return model, often yields similar results to more

(21)

S. Bos 18 The market model is a one factor model. Other statistical models use multiple factors to reduce the variance of the abnormal return. An example is the three factor model by Fama and French (1992). Fama and French added two factors, market capitalization and book-to-price ratio, to the market model. However, the gains from additional factors for event studies are limited (MacKinlay, 1997). For this reason, this paper will not use a multifactor model.

The abnormal return of security i is the difference between the actual return of the security and the expected return estimated with the market model. For any security i the expected return is:

= α + ß R+ 2

where

= -252,...,-3 (days relative to the announcement), and

E _, = 0 and var _, = !_"$_#

And where is the period t return on securityi, _% is the period t return on the market index, _, is the time period disturbance term for security & with a mean of zero and variance !_"$_#.

The abnormal return (AR) of security i is measured by:

(22)

S. Bos 19 Where AR is the abnormal return of securityi at day t and α and ß are the OLS values from the estimation period.

Abnormal returns can be aggregated through time and across securities. A difficulty in this study is that the length of the interim period varies across firms. To solve this difficulty, I start with aggregate through time by calculate a cumulative abnormal return (CAR) for each firm i:

CAR = * AR 4 ,

-.

Where n is the number of days of that period. For the interim period, the average number of days between the announcement and the abandonment of the offer (in that subsample) is used.

Next, these CAR’s are averaged to calculate the average cumulative abnormal return for firms in the period:

CAR

////// = 1_{N * CAR} 1 -.

5

where N is the number of firms.

T-test

(23)

S. Bos 20 calculated. The cross-sectional method conducts a t-test by dividing the average event-period CAR////// by its contemporaneous cross-sectional standard error:

= CAR//////

σCAR////// √n⁄ 6

where σ(CAR//////) is calculated as:

σ CAR////// = 6 1

N − 1 *CAR & − CAR//////$

1 -.

7

To determine if there is a significant difference between the mean abnormal returns for certain sub-groups, the following t-test will be conducted:

= 8.− 8$ 9!.$

:. + !:$$$

8

where the subscripts 1 and 2 stand for the different sub-groups, n is the number of observations in a particular sub-group, and σ is the standard deviation of the sub-group.

Z-test

To test whether the probability of a takeover subsequent to an abandonment differs between the subsamples, the following z-statistic is calculated (Keller, 2005):

< = => − =. >$ 9=̂ 1 − =̂ @ 1_:

.+ 1:$A

9

(24)

S. Bos 21 pooled proportion of both subgroups, :_. is the sample size of the first subgroup, and :_$ is the sample size of the second subgroup.

Choice of market index

For computing the abnormal return, a proper market index has to be found as a benchmark. Most important, because of the scope of this research, the index should contain USA shares only. In my opinion, a broad value-weighted index containing USA shares is the most proper benchmark for this study. I have chosen for the S&P 500 index. The S&P 500 index is a value weighted index containing the 500 largest USA firms.

3.4 Non-parametric test

In this paper, the abnormal returns are measured over long-horizons. An advantage of non-parametric tests is that in comparison with the non-parametric tests, they are likely to reduce misspecification of measuring long horizon abnormal returns (Kothari and Warner, 1997). Another advantage of non-parametric tests is that it does not make assumptions concerning the distribution of returns. Common non-parametric tests are the sign test, the rank-test, and the generalized sign test. A disadvantage of the sign test is the requirement of

symmetrical distributed abnormal returns while the distribution of long-run abnormal stock returns is positively skewed (Barber and Lyon, 1997). The rank test and the generalized sign test do not require symmetry in cross-sectional abnormal returns (Corrado, 1989, Cowan, 1992). Moreover, the generalized sign test is less sensitive to increases in the length of the event window compared to the rank test (Cowan, 1992). Because of these reasons, this study uses the generalized sign test. The generalized sign test compares the proportion of positive abnormal returns in the event window to the proportion of positive abnormal returns at the 250 day estimation window (Cowan, 1992).

=̂ = _{C *}1 ₂₅₀1 D -. * E, $FG -. 10

(25)

S. Bos 22

E, = H 1 if 8_{0 otherwise}, > 0 11S

The z-value of the generalized sign test statistic is:

TU = V − C=̂

(26)

S. Bos 23

4. Data

This section discusses the data used in this paper. Section 4.1 will discuss the data sources and the construction of the data sample. Section 4.2 will describe the characteristics of the data.

4.1 Data sources and sample construction

The empirical tests of this paper are based on the returns of target firms involved in an abandoned takeover offer process in the period 1997-2009 in the USA. The sample is divided into two subsamples according to whether the offer was abandoned by the bidding firm or by the target firm.

The daily share prices and the daily price of the market index for computing abnormal returns are obtained from Datastream using International Securities Identifying Numbers (ISIN). The returns are adjusted for dividend payments and stock splits. Because a large part of the firms are currently delisted and have no ISIN number any more, the share prices of those firms are obtained manually from Datastream. For collecting data, the Zephyr

(27)

S. Bos 24 Table 4

Data criteria Variable Criteria

Type of offer - Withdrawn

Country - USA

Deal type - Mergers

- Acquisitions

Time period - Announcement date after January 1997 - The abandonment must lie before June 2008

- The announcement date and the abandon date of the offer have to be known

- The interim period between the announcement and the abandonment must be at least 5 days.

Percentage of stake - Initial stake is maximum 5% - Bid is aimed at 100% ownership

Quotation - Quoted target

- The share price of each target firm needs to be available from one year before the takeover offer

Deal value - Minimum of 100 million USA Dollar

(28)

S. Bos 25 Two event dates are used in this paper; the ‘announcement date’ and the ‘abandonment date’. The announcement date is the date when the announcement of the takeover offer was made. The abandonment date is the day that the takeover offer is cancelled. The announcement dates are given by Zephyr. The abandonment date is not always given by Zephyr. In a some cases, Zephyr reported the abandonment date in the deal comments. In case Zephyr did not report an abandonment date, the date of the bidding company’s press release of the abandonment was used. The press release is obtained from the website of the bidding firm. In case the announcement and the abandonment occurred within 5 days, the event is deleted from the sample because it is not possible to measure the impact of the announcement or the abandonment of the offer.

In case more than one bidding firms compete for one target, the target is included only one time to avoid double influence of that firm. This study takes the day of the first takeover offer as announcement date because this announcement has the largest impact on the share price of the target firm. The day the offer of the last firm is abandoned is taken as the

abandonment date. If a bidder makes multiple attempts to acquire the same target, the day of the first announcement is used as the announcement date and the day of the

abandonment of the last offer is used as the abandonment date.

4.2 Data characteristics

The final sample of this study consist of 160 abandoned takeover targets. The reason for the small sample size is that it was not directly possible to search for abandoned takeovers in Zephyr. It was possible to search for withdrawn offers but most of these withdrawn offers were followed by successful offers. Therefore, most of the takeovers on the Zephyr list turned out to be successful takeovers. Although the sample size is small, it is larger than other studies on abandoned takeovers (table 1).

(29)

S. Bos 26 great chance that the target will eventually be acquired. Furthermore, the probability that the target is acquired in the year subsequent an abandoned takeover offer is larger when the target company rejects the offer than when the bidder withdraws the offer. In section 5.4 it will be tested whether this difference is significant.

Table 5

Target firms acquired in the year subsequent an abandonment per initiator of abandonment

Percentages are given in parentheses

Initiator of abandonment Target firms acquired within one year subsequent an abandoned offer

Target firms not acquired within one

(30)

S. Bos 27

5. Results

This section discusses the results of the event study. The results will be discussed per period: Section 5.1 discusses the abnormal return around the announcement of a takeover offer for the whole sample of target firms and the subsamples separately. Next, section 5.2 discusses the abnormal return between the announcement and abandonment, section 5.3 around the abandonment date, and section 5.4 will discuss the abnormal return in the year subsequent to a takeover. In section 5.5 the abnormal return over the total takeover period will be discussed.

5.1 Cumulative abnormal return at announcement date

To test whether the announcement of a merger or acquisition offer is associated with positive abnormal returns for target firms in the US the cumulative abnormal return is measured and tested around the event date. Table 6_{presents the cumulative abnormal} return around the announcement of a takeover offer. As expected, there is a significant positive cumulative abnormal return of 22.8% around the announcement for target firms. The positive results in this sample corresponds to values reported by other studies in this area of research. Bradley et al (1983) and Bradley (1980) show larger abnormal returns for targets of abandoned takeovers of 30% and 29% respectively. On the opposite, studies of Asquith (1983) and Dodd (1980) show smaller abnormal returns of 12% and 13%

(31)

S. Bos 28 Table 6

Cumulative abnormal return for target firms around the announcement date In this table the cumulative abnormal returns around the announcement date are shown for the whole sample and for the sub samples of abandonments driven by targets and bidders. All abnormal

returns are percentages. T-values are given in the next column. In the last column the Z-value of the generalized sign test are given. *** Denotes statistical significance at the 1% level; ** denotes

statistical significance at the 5% level, and * denotes statistical significance at the 10% level.

No. of firms Cumulative abnormal return (%) T-test Generalized sign test

All abandoned offers 160 22.799 2.482*** 12.062***

Offers rejected by target 93 23.991 2.386*** 9.037***

Offers withdrawn by bidder 63 20.623 2.640*** 7.454***

Difference between subsamples 3.368 1.001 -

5.2 Cumulative abnormal return between announcement and abandonment

Table 7_{presents the cumulative abnormal returns for target firms in the interim period} between the announcement and the abandonment of a takeover offer. There is a significant abnormal return of -10,71% in the interim period for the sample of target firms. This result implies that hypothesis 2, that there is a negative abnormal return in the interim period, cannot be rejected. The negative abnormal return in the interim period correspond to the results of the studies of Asquith (1983) and Bradley et al (1983). When comparing the results of the subsamples there is a remarkable difference. I expected that the difference in

abnormal return between the subsamples would occur around the abandonment date. However, targets of takeover offers that are abandoned by bidding firms lose all of the takeover premium in the interim period while targets that abandon the takeover offer themselves show a small negative abnormal return that is not significant. The relatively low t-values in this period is a result of the longer period of measurement. The negative

(32)

non-S. Bos 29 parametric generalized sign test. The results of the generalized sign test show that the

proportion of target firms with a negative cumulative abnormal return in the interim period for offers that are abandoned by bidders is significant larger than normal. The proportion of firms with a positive cumulative abnormal return in the interim period is not significant different from normal for offers that are abandoned by target firms.

Table 7

Cumulative abnormal return for target firms between announcement and abandonment In this table the cumulative abnormal returns in the interim period between announcement and abandonment are shown for the whole sample and for the sub samples of abandonments driven by targets and bidders. All abnormal returns are percentages. Difference between sub-samples is given

in the last row. T-values are given in the third column. In the last column the Z-value of the generalized sign test are given. *** Denotes statistical significance at the 1% level; ** denotes

statistical significance at the 5% level, and * denotes statistical significance at the 10% level.

All abandoned offers 160 -10.709 -2.821*** -2.197**

Offers rejected by target 93 -4.084 -0.836 0.724

Offers withdrawn by bidder 63 -20.289 -4.741*** -3.403***

Difference between subsamples 16.146 2.194** -

5.3 Cumulative abnormal return at abandonment date

Table 8_{presents the cumulative abnormal return for target firms around the abandonment} date of a takeover offer. There is a insignificant negative cumulative abnormal return of -5.46% around the abandonment date. This result implies that hypothesis 3, that the

(33)

S. Bos 30 cumulative abnormal return around the abandonment date is significant larger than normal. The results of the subsamples show that the negative cumulative abnormal return around the abandonment date is caused by offers that were abandoned by the bidding firm. The cumulative abnormal return of target firms for offers that are abandoned by the bidding firm is significantly lower than the abnormal return of target firms of offers that are abandoned by the target firm. This result implies that hypothesis 6, that the abnormal returns of target firms of offer abandoned by the target is higher than the abnormal return of target firms for offers abandoned by the bidding firm, cannot be rejected. Furthermore, this result

corresponds to the study of Dodd (1980) who found a higher abandonment return for targets that abandoned the takeover offer compared to targets that did not abandon the offer.

Table 8

Cumulative abnormal return for target firms around the abandonment date In this table the cumulative abnormal returns for the period around the abandonment date are

shown for the whole sample and for the sub samples of abandonments driven by targets and bidders. All abnormal returns are percentages. Difference between sub-samples is given in the last row. T-values are given in the third column. In the last column the Z-value of the generalized sign test

are given. *** Denotes statistical significance at the 1% level; ** denotes statistical significance at the 5% level, and * denotes statistical significance at the 10% level.

All abandoned offers 160 -5.457 -0.693 -2.831***

Offers rejected by target 93 0.765 0.124 0.308

Offers withdrawn by bidder 63 -14.718 -1.731** -4.666***

Difference between subsamples 15.483 4.704*** -

5.4 Cumulative abnormal return in the year subsequent an abandoned takeover

(34)

S. Bos 31 hypothesis 4, that the abnormal return in the year subsequent to the abandonment is

negative, is rejected. The insignificant result in this period is in contrast to the studies of Asquith (1983), Bradley et al (1983), and Dodd and Ruback (1977) which found negative abnormal returns in the years after the abandonment. The results for the two subsamples and the non-parametric generalized sign test are insignificant too. It can be concluded that no abnormal returns in the period after the abandonment can be attributed to the events.

Table 9

Cumulative abnormal return for target firms in the year subsequent to the abandonment In this table the cumulative abnormal returns in the year subsequent to the abandonment are shown

for the whole sample and for the sub-samples of abandonments initiated by targets and bidders. All abnormal returns are percentages. Difference between sub-samples is given in the last row. T-values are given in the third column. In the last column the Z-value of the generalized sign test are given. *** Denotes statistical significance at the 1% level; ** denotes statistical significance at the 5% level,

and * denotes statistical significance at the 10% level.

No. of

firms

Cumulative abnormal return (%)

T-test Generalized sign test

All abandoned offers 160 -2.306 -0.454 0.179

Offers rejected by target 93 -4.041 -1.178 0.516

Offers withdrawn by bidder 63 -1.830 -0.275 -0.121

Difference between subsamples 2.211 -0.212 -

A z-test is used to test whether the difference between the proportion of target firms that is acquired after the offer is abandoned by the target is significant higher than the proportion of targets that is acquired after an offer is abandoned by the bidder. In table 5 an overview is given of the number of firm that are taken in the year subsequent to the abandonment. According to the z-test, the difference between the proportions is significant on a 1% confidence level with a z-value of 3.886. Hypothesis 9, that the probability of a new

(35)

S. Bos 32 5.5 Cumulative abnormal return of total takeover period and discussion of results

Table 10 presents the cumulative abnormal returns for target firms over the whole takeover period, from announcement till abandonment. There is a significant positive revaluation of 20.7% of target sharers when the target firm abandoned the offer. This result implies that hypothesis 7, that there is a positive revaluation when the target firm abandons the

takeover offer, cannot be rejected. This corresponds to the result of Dodd (1980) who found a positive abnormal return when the target abandoned the offer as well. Furthermore, there is a significant negative revaluation of -14.4% of target shares when the bidding firm

abandoned the offer. This result implies that hypothesis 8, that there is a negative

revaluation when the bidding firm abandons the takeover offer, cannot be rejected. For the total sample of target firms there is a insignificant positive revaluation of 6.6%. This result implies that hypothesis 5, that there is no revaluation of target shares over the total

(36)

S. Bos 33 Table 10

Cumulative abnormal return for target firms of abandoned takeovers over period from announcement till abandonment

In this table the cumulative abnormal returns for the total takeover period are shown for the whole sample and for the sub samples of abandonments driven by targets and bidders. All abnormal returns are percentages. Difference between sub-samples is given in the last row. T-values are given in the third column. In the last column the Z-value of the generalized sign test are given. *** Denotes

statistical significance at the 1% level; ** denotes statistical significance at the 5% level, and * denotes statistical significance at the 10% level.

All abandoned offers 160 6.6 1.116 3.955***

Offers rejected by target 93 20.672 3.568*** 6.344***

Offers withdrawn by bidder 63 -14.384 -2.774*** -3.196***

Difference between subsamples 35.056 3.854*** -

(37)

S. Bos 34 Figure 2 gives an overview of the cumulative abnormal returns of the total sample and the two sub-samples.

Figure 2

Cumulative abnormal return of target firms that receive an abandoned takeover offer

In this figure an overview of the cumulative abnormal returns for the period from announcement till abandonment of the takeover offer is given for the total sample, and for the subsamples of offer that are abandoned by targets and by bidders.

% -20 -15 -10 -5 0 5 10 15 20 25 30 Announcement date

Interim period Abandonment date All targets

Offers abandoned by target

(38)

S. Bos 35

6. Conclusions and

suggestions for further research

This study examines the shareholder wealth effects of target firms involved in abandoned takeovers. The aim of this study is to research whether the wealth effects for target shareholders differ between when the target abandons a takeover offer compared to the wealth effects when the bidder abandons the offer. Therefore, the sample of 160 target firms, originated in the USA that received an abandoned takeover offer in the period 1997-2008, is divided into these two subsamples. The event study methodology in addition with the generalized sign test is used to examine the wealth effects of target shareholders. The wealth effects is researched in four distinct periods: the announcement of the offer, the interim period between the announcement and the abandonment of the offer, the abandonment, and the year subsequent the abandonment.

The announcement of a takeover offer is accompanied by significant positive abnormal returns. Most of the premium of the announcement of an offer is already lost before the abandonment of the offer. It can be concluded that it is important for studies on abandoned takeovers to measure the abnormal return in the interim period to fully measure the effects of takeover abandonment. The abandonment of the offer is has a negative effect for

shareholders of target firms when the bidding firm abandons the offer. The results of the non-parametric generalized sign test show that the proportion of targets with a negative abnormal return around the abandonment date is higher than normal for the total sample of abandoned targets. However, the negative abnormal return around the abandonment date for the total sample is not significant. Also the abnormal returns in the year subsequent an abandonment are not significant different from zero. It can therefore be concluded that the market is efficient, once the offer is abandoned there is no share price reaction that is related to the abandonment.

(39)

S. Bos 36 by the bidding firm. The positive revaluation when the target abandons the offer is slightly smaller than the premium in the offer. However, this study shows that most of the targets (63% in this sample) are being acquired within one year subsequent an abandoned takeover offer. Targets would only accept a new offer if it is higher than the abandoned offer. My conclusion from this research is therefore that rejecting an offer can be interesting for target shareholders and management. Shareholders do not have to be afraid to lose all of the premium from the announcement by rejecting the offer.

The results of the total sample of targets are in line with the synergy hypothesis. There is no significant revaluation of target shares. However, there is a significant difference in return of target firms between abandonments initiated by targets and abandonments initiated by bidders. There is a significant revaluation of 20.7% of target sharers when the abandonment is initiated by the target firm and a significant revaluation of -14.4% of target sharers when the abandonment is initiated by the bidding firm. These results show that the party which abandons the takeover offer reveals information to the market. This is in line with the information hypothesis. According to the synergy hypothesis there would be no difference between the results of the subsamples.

Suggestions for further research

Abandoned takeovers have received relatively little academic attention. In this thesis the wealth effects of abandoned takeovers are explored for firms in the USA. Perhaps the effects of takeover abandonment is different for firms in the USA compared to countries/regions with another corporate governance system.

In this thesis a no distinction has been made between abandonment by target management and by target shareholders. To further research the consequences of takeover abandonment it would be interesting to make this difference.

(40)

(41)

S. Bos 38

References

Asquith, P., 1983, “Merger bids, uncertainty, and stockholder returns”, Journal of Financial Economics, Vol. 11, pp. 51-83.

Barber, B.M. and J.D. Lyon, 1997, “Detecting long-run abnormal stock returns: The empirical power and specification of test statistics”, Journal of Financial Economics, Vol. 43, pp. 341-372.

Boehmer, E., Musumeci, J., and A.B. Poulsen, 1991, “Event-Study Methodology under conditions of Event-Induced Variance”, Journal of Financial Economics, Vol. 30(2), pp. 253-272.

Bradley, M., 1980, “Interfirm Tender offers and the market of Corporate Control”, Journal of Business, Vol. 53, pp. 345-376.

Bradley, M., A. Desai and E.H. Kim, 1983, “The rationale behind interfirm tender offers: Information or Synergy?”, Journal of Financial Economics, Vol. 11, pp. 183-206.

Brooks, C., 2002, “Introductory econometrics for finance”, Cambridge University Press: Cambridge

Brown, S.J. and J.B. Warner, 1980, “Measuring security price performance”, Journal of Financial Economics, Vol. 8, pp. 205-258.

Brown, S.J. and J.B. Warner, 1985, “Using daily stock returns”, Journal of Financial Economics, Vol. 14, No. 1, pp. 3-31.

(42)

S. Bos 39 Dodd, P., 1980, “Merger proposals, management discretion and stockholder wealth”, Journal of Financial Economics, Vol. 8, pp. 105-137.

Dodd, P. and R. Ruback, 1977, “Tender offers and stockholder returns”, Journal of Financial Economics, Vol. 5, pp. 351-373.

Fabozzi, F.J., M.G. Ferri, T.D. Fabozzi and Tucker, J., 1988, “A Note on unsuccessful Tender Offers and Stockholder Returns”, The Journal of Finance, Vol. 43, No. 5. pp. 1275-1283.

Fama, E.F., 1997, “Market efficiency, long-term returns, and behavioral finance”, Journal of Financial Economics, Vol. 49, pp. 283-306.

Holl, P. and D. Kyriazis, 1996, “The Determinants of outcome in UK Takeover bids”, International Journal of the Economics of Business, Vol. 3, No. 2, pp. 165-184.

Jensen, M.C. and R. Ruback, 1983, “The Market for Corporate Control: The Scientific Evidence”, Journal of Financial Economics, Vol. 11, pp. 5-50.

Keller, G., 2005, “Statistics for management and economics”, 7th edition, Thomson Brooks/Cole.

Kothari, S.P. and J.B. Warner, 1997, “Measuring long-horizon security price performance”, Journal of Financial Economics, Vol. 43, pp. 301-339.

MacKinlay, A., 1997, “Event studies in economics and finance”, Journal of Economic Literature, Vol. 35, pp. 13-39.

Mitchell, M.L. and E. Stafford, 2000, “Managerial Decisions and Long-Term Stock Price Performance”, Journal of Business, Vol. 73 (3), pp. 287-329.

(43)

S. Bos 40 Taffler, R.J. and P. Holl, 1991, “Abandoned Mergers and the Market for Corporate Control”, Managerial and Decision Economics, Vol. 12, No. 4, pp. 271-280.

(44)

S. Bos 41

Appendix A: Summary statistics

Table A.1

Average alpha and beta of the companies in the sample

Alpha Beta 0,0005 0,7066

Table A.2

Number of events per sample year

Year Announcement date Abandonment date

1997 1 1 1998 2 0 1999 4 0 2000 11 8 2001 7 13 2002 13 8 2003 19 24 2004 23 17 2005 13 15 2006 22 21 2007 19 20 2008 26 33 Total ₁₆₀ ₁₆₀

(45)

S. Bos 42

Appendix B: relationship between number of days between announcement

and abandonment and cumulative abnormal returns

Table B.1:

OLS between the CAR in the interim period and the number of days between the announcement and the abandonment of a takeover offer

Dependent Variable: CAR Method: Least Squares Included observations: 160

Coefficient Std. Error t-Statistic Prob.

C -0.009887 0.044016 -0.224622 0.8226 DAYS -0.001021 0.000352 -2.904636 0.0042 R-squared 0.050691 Adjusted R-squared 0.044683 F-statistic 8.436912 Prob(F-statistic) 0.004204 Table B.2:

OLS between the CAR on the abandon date and the number of days between the announcement and the abandonment of a takeover offer

Dependent Variable: CAR Method: Least Squares Included observations: 160

Coefficient Std. Error t-Statistic Prob.

Wealth effects of abandoned takeover targets