European Banks Shareholder Reactions To Announcements Of Acquisitions

This study examines announcement effects of acquisitions of listed European bidding banks. Short-term acquirers’ stock market returns in the EU-15 are considered from 2000 to 2010. We assume that the method of payment make a difference, particularly in times of crisis. Abnormal returns are defined as the cumulative average abnormal return (CAAR) where the S&P Europe 350 Financials serves as the benchmark. We found that European banks did not manage to earn abnormal returns for buy transactions over the short horizon. Furthermore, the method of payment makes no difference in this case. We conclude that in times of crisis investors dislike cash payments to finance acquisitions compared to stock-financed acquisitions.

1. Introduction

Do shareholders of bidder banks benefit from acquisitions? Although researchers came up with a plethora of investigations setting about this topic, there is no general agreement regarding the returns to bidders. For banking this issue is of particular importance because the speed of consolidation during the recent years promises to recover as further investment opportunities will come up with the removal of impediments to European cross-border banking consolidation. Next to that, after the crisis of 1980s the US banks acquired competitor banks in their attempts to resolve the thrift crisis.

Due to mergers and acquisitions the number of banks actively participating in the European market decreased significantly. Banking assets as a percentage of GDP grew from 177,2% in 1985 to 244,2% in 1997 (European Central Bank (ECB), 1999) while the number of European banks decreased from 12.670 in 1985 to 8.395 in 1999 (ECB, 2000). There are a number of reasons to believe that consolidation will increase in Europe during the coming years, according to the ECB (2008). For example, bank managers are still seeking for economies of scale and scope and opportunities for diversification. After peaking in 1999, the value of European domestic banking deals has been in decline. Interestingly, since 2003 the value of European cross-border deals has been rising year after year. We consider both domestic and European cross-border deals using a sample which contains acquisitions in the EU-15 [1] .

Considerable amount of research is being conducted in the previous decade in analyzing the announcement effects in the European banking industry (Tourani-Rad and Van Beek, 1999; Cybo-Ottone and Murgia, 2000; and Beitel and Schiereck, 2001). These papers create an overall understanding of the shareholder value creation. Several studies in the U.S. (DeLong, 2001; Houston et al., 2001) find shareholders’ reactions in the banking industry on an announcement of an acquisition as being inconclusive. Furthermore, there is a shortage of new and conclusive analysis concerning the factors having impact on the shareholders’ reactions to announcements of acquisitions around the announcement date.

Due to the enormous increase of takeover activity on the European continent described by Martynova and Renneboog (2006) new financial markets are created in order to finance the takeovers. This study examines whether methods of payment have an effect on the bidder companies’ stock prices at the announcement date of a takeover in the European banking industry.

Research conducted by Beltratti and Stulz (2009) shows that banks in the European Union performed better in times of crisis when a more solid capital structure is used. According to this study, an announcement of an acquisition including a method of payment that is sacrificing a majority of the liquidity assets could result in negative abnormal returns. Stockholders probably would not appreciate a cash payment in times of crisis.

Although the European banking industry was covered by the study performed by Beitel et al. (2003) the data of this study is outdated. This paper contributes to existing literature about announcement effects on acquisitions because no recent study is conducted on the modern European banking market. Furthermore, there is nothing available about the effects of applying a particular method of payment in the European banking sector. Lastly, the European Central Bank addresses the issues covered in this study to be relevant regarding their expectations on the M&A activities for the coming years.

The remainder of this article proceeds as follows. Section 2 details the literature and the hypotheses. The third section presents the data description and the methodology. The fourth section summarizes our findings. Finally, in the last section the conclusions of the paper will be presented.

2. Literature review

2.1 Acquisitions

In a study conducted by Huang and Walker (1987) it is stated that managerial resistance influences the announcement effects. First, abnormal returns due to announcement effects are therefore more frequently observable in pure acquisitions than in any other form of financial integration. In evaluating the abnormal returns of announcement effects this paper will purely focus on acquisitions. Second, the effects of stock price changes in the banking industry influence the possibility of being taken over, and alter the growth rate of the bank according to Levine and Zervos (1998). Therefore, managers are eager to keep returns as high as possible to prevent

2.2 Announcements effects

Several authors, including Campa and Hernando (2004), have covered the announcement effects of acquisitions. In this study, cumulative abnormal returns can be observed when the shareholders expect future synergies or wealth distributions. However, the study is inconclusive in clarifying whether the abnormal returns are positively or negatively skewed. These findings are in concordance with a study conducted by Bruner (2002) who gives an overview of different articles referring to shareholders wealth creation with respect to announcement effects. In contrast to the consensus about the stock price performance of target companies where they merely find positive abnormal returns (Kaplan and Weisbach, 1992; Bradley et al., 1988; Bhagat et al., 2005), researchers cannot reach an agreement about the announcement effects for bidders. Jensen and Ruback (1983) argue that the estimation of returns is more difficult for bidders than for target companies, because the market might fully anticipate the bidder intentions, obstructing a response to a takeover announcement. Next to this, bidders can take part in a more than one takeover at the same time.

Walker (2000) along with Mitchell and Stafford (2000) found negative returns for bidders. On the other hand, Asquith and Kim (1982) show significant abnormal returns for acquirers in the pre-announcement period. According to Jarrel and Poulsen (1989) the 1960s showed significant positive abnormal returns of 5% and of 2,2% in the 70s. However, along with other researchers they found insignificant abnormal returns in 1980s. To sum up, half of the studies present negative returns for the acquirers, whereas the other half reports zero or very small negative abnormal return.

2.3 European banking sector

According to prior research of DeLong (2003) mergers or acquisitions in the banking sector on average do not create value. Interestingly, several authors (Berger et al., 1998) have issued that announcement effects of an acquisition can be positive in the banking industry. Therefore the banking industry is a very interesting industry to study due to the at first sight reasonless inconsistency of the findings that have been reported in the last decade.

The banking industry in Europe is rapidly changing due to the credit crunch. In the study conducted by Erdinç (2009) which evaluates banks and there fragility the author concludes that fragility significantly has increased due to the credit crunch. An extensive study carried out by Chuang (2010) stresses that regulation does have impact on the banking industry and evaluates the abnormal announcement effects due to these increased regulation. In addition to these increased regulations the liquidity demands also has increased. According to the newly issued Basel Accord higher regulatory capital has to be reserved. Due to the increase in demand for regulatory capital, the method of payment could also have been altered.

Because of the inconsistent findings found in the previous research, we present the following hypothesis:

H1. Cumulative abnormal returns to the acquirer company around announcements of an acquisition in the European banking industry will be significantly different from zero.

This hypothesis will be tested against the null hypothesis that the cumulative abnormal returns of acquisitions in the European banking sector will be zero. The literature provides an inconclusive answer whether cumulative abnormal returns can be found. Therefore, an assessment of the cumulative abnormal returns of recent data could give insights that provide additional knowledge on abnormal returns.

2.4 Variables affecting abnormal returns

Knowing how the share prices of acquirer companies react to the first announcements of acquisitions, it is crucial to examine the factors having a bearing on the return of the bid. Huang and Walking (1987) as well as Shan and Wilson (2006) use the Market Model in order to define specific variables. They found variables as level of concentration, type of offer and method of payment as having an effect on the magnitude of the acquirer’s and targets’ stock reaction.

2.5 Methods of payment

According to the studies conducted by Travlos (1987) and Asquith et al. (1987) bidders experience negative abnormal returns when the acquisition is financed with stocks, and positive when the payment is made in the form of cash. There are several studies attempting to explain the choice of the payment method. In the paper of Ismail and Krause (2010) three groups of determinants of the applied methods of payment can be indentified, namely: asymmetric information, taxation and managerial control. In evaluating the method of payment and abnormal returns these factors will be taken into consideration.

2.5.1 Asymmetrical information

Asymmetric information is firstly addressed by a paper of Myers and Maljuf (1984) which states that there is asymmetrical information difference between bidder and target of the value of shares of the bidder company. Moreover, they find that the choice of the payment method can be regarded as a signal that can provide information about the bidder. It has been proven that managers will finance a takeover with shares if their firm is overvalued and use cash if the company is undervalued. Cash offers mean that the assets of the firm must be undervalued and that the managers’ valuation of the synergy effect is higher than de market’s one. The finding of asymmetrical information is confirmed by several authors (Travlos (1987); Fishman (1987); Cornu and Isakov (2000); Sheifler and Vishny (2003); Rhodes-Kropf and Vismanatan (2004)). However, a study conducted by Cornett and De (1991) does not support the theory suggested by Myers and Maljuf (1984) and therefore additional investigation could be reasonable.

In the studies of Diamond (1984), Diamond (1991) and Boyd and Prescott (1986) information asymmetry is studied in banking industry. In a consistent manner the authors report that in the banking industry additional information about the method of payment in a takeover will trigger a direct movement in stock price changes. Due to the highly sensitivity of stock prices of the banking industry to additional information it could be plausible to find cumulative abnormal returns. The above review suggests that an announcement of cash-paid acquisition will generate higher announcement returns to the shareholders than announcements of acquisitions where the deal will be paid with shares. We therefore present the following hypothesis:

H2. Cumulative abnormal returns (CAAR) to the bidders around the announcement of a takeover financed with cash will be larger than CAARs around announcements of acquisitions when financed with stock.

This hypothesis will be tested against the alternative hypothesis that the method of payment could be indifferent of the announcement effects. The majority of the available literature states that a cash takeover will in general create positive abnormal returns in comparison with other financial offers due to informational asymmetries.

2.5.2 Taxation

It has also been emphasized that the taxation has an impact on the choice of the payment method (Ismail and Krause, 2010). While cash offers are considered as immediately taxable for the target’s shareholders, taxation can be postponed in case of stock offers. If this turns out to be important for the target’s shareholders, the bidder is encouraged to come up with a stock offer. Further, bidders require the compensation for an increase in taxation in the form of a higher premium. Empirical findings of studies from Wansley, Lane and Yang (1983); Harris, Frank and Mayer (1987) and Huang and Walking (1987) are in line with the taxation principle. The European banking industry is also affected by taxation policies. Although no evidence in the literature can be found referring to the method of payment and the taxation principle. Because this paper concerns a brief review, we focus only on the essentials and let the taxation issue rest.

2.5.3 Managerial control

Thirdly, managerial control is often described as a determinant in method of payment. If the bidder company is financing the takeover with purely stock the position of the current stock holders will deteriorate (Ismail and Krause, 2010). Empirical evidence exists in accordance with the findings presented above (Harris and Raviv, 1998; Faccio and Masulis, 2005). We do not investigate whether managerial control is an issue in acquisitions in the European banking industry because of the time limits of this research study.

2.6 Control variables

Below we present a brief overview of the available literature, which describes the main control variables.

2.6.1 Financial crisis

As mentioned above in the article of Erdinç (2009) the fragility of the banks has increased due to the credit crunch. The stock prices are significantly more opaque during a financial crisis according to Flannery et al. (2010). This study reveals that bank share trading exhibits strongly different features before the crisis compared to movements during the crisis. Investors will be less able to value banks properly.

Despite the fragility in the banking sector, overall announcement effects do not differ significantly in times of financial crisis suggests Amiram et al. (2010). Due to very generalistic nature of this study, we evaluate the financial crisis for the banking sector applying a control variable.

2.6.3 Increasing stake

Secondly, we consider the impact of different graduations of an increase of the stake in the target company. According to Bayne (1969) the choice between control premiums is paid when the acquirer wants more control over the target company. This is in line with the managerial control issue described in Ismail and Krause (2010), where it is stated that the current control over the bidder and target company combined will deteriorate to current stockholders when the takeover is financed with shares. According to Jensen and Ruback (1983) the control premium that is financed in takeovers also results in positive cash flow for the acquirer. By financing the control premium during a takeover according to Jensen and Ruback (1983), the acquirer will acquire newly issued projects that contain new sources of revenue. In addition to these studies, abnormal returns could be created when the control premium is financed. Therefore it becomes an interesting control variable in addition of the method of payment.

3. Data and methodology

3.1 Data

The hypotheses described in the previous section haven been tested using data from 2000/01/01 to 2010/11/1. Each acquisition in our sample satisfies the following selection criteria: a) both the bidder and target companies are banks; b) bidder and target companies are from EU-15 countries; and c) the acquirer is listed. These criteria lead to 608 observations.

We excluded delisted and unlisted companies for which no ISIN numbers were available. The Zephyr database of Bureau van Dijk is used to find the announcement dates. For this study it is required to set the rumor date equal to the announcement date, in order to have a dataset for which no information leakage can have occurred before the announcement date. This leaves us with 155 observations.

From this sample we considered those observations where the acquirer have share price data available in the Datastream database. The final sample consists of a panel of firms with a total of 63 observations corresponding to 39 firms.

3.2 Sample description

Table one present the characteristics of the acquisitions of the European banking industry. In panel A the origin of the bidder companies who announced an acquisition within the European banking industry are summarized. The most announcements of acquisitions are from the United Kingdom (44,4%). Italian banks are second with 30,2% of the announcements of the acquisitions. Furthermore, it must be noticed that the database has no observations of Austria, Finland, France, Greece, Ireland, Luxembourg, Netherlands and Portugal. Mainly, these observations are lacking due to incomplete stock returns or missing ISIN numbers. Furthermore, in Panel B the announcement years are described. In the years 2000, 2003 and 2008 the most announcements where made. Most of the acquisitions announced are imposed to finance with shares. Moreover, in panel C the increase of relatively stake in the target company is being analyzed. Most of the acquisitions are acquisitions that involve a complete takeover which are mostly financed with cash. Finally, acquisitions announced during the financial crisis consist of 19% of the sample. The way of financing during the financial crisis is inconclusive. 50% of these announcements are financed with cash and the other 50% with shares and other payment forms.

Table 1: Sample characteristics

Number of acquisitions Cash Shares and other*

Panel A: Acquirer country

Belgium 3 3 0 Denmark 2 2 0

Germany 2 0 2

Italy 19 16 3

Spain 8 0 8 Sweden 1 1 0

United Kingdom 28 15 13

Panel B: Announcement year

2000 10 6 4

2001 6 2 4

2002 6 6 0

2003 8 4 4

2004 4 3 1

2005 7 5 2

2006 4 2 2

2007 4 2 2

2008 8 4 4

2009 3 2 1

2010 3 1 2

Panel C: Increase of stake in the target company

0 Withdrawn 1 1 0 1 0-25% stake increase 11 6 5

2 25-50% stake increase 12 7 5

3 50-75% stake increase 5 4 1

4 75-100% stake increase 34 19 15

Panel D: Financial crisis

Acquisition in the financial crisis** 12 6 6

Acquisition outside the financial crisis 51 31 20

* shares and other including deferred payment, debt assumed and loan notes

** An announcement during the financial crisis is considered to be within 1-1-2007 and 31-12-2008

3.3 Empirical Methodology: event study

The traditional event study methodology is used to measure the impact of the announcement of an acquisition on the security’s return; cumulative average abnormal returns are examined. We focus here only on short term cumulative changes in stock prices of the bidder company [2] . Furthermore, we measure the abnormal returns to the bidders around the event day. The event day is the trading day and is defined as t=0. If an acquisition is announced on a non-trading day, we consider the next trading day as announcement date. According to MacKinlay (1997) it is advisable to include the days around the event day in the event window to check if they also have abnormal returns. This is recommended because the company’s returns may be affected by information that becomes available to the market prior to the event day. We assume an event window τ = 5.

In order to test the sample for abnormal returns, we have to determine the normal (benchmark) return for the particular acquisition that is linked to the announcement. The market model as outlined by MacKinlay (1997) is used to estimate the normal returns in an estimation window of 100 trading days (starting 110 days before the announcement date). The event period itself is not included in the estimation period in order to prevent the event from being a factor in the normal performance model parameter estimates. These are estimated from the coefficients of the market equation, given by [3] :



where Riτ is the return of the bidder’s security at time τ (event i). τ is 110 days until 11 days before the event i. αi is the coefficient and the intercept term for the regression equation related to the acquirer’s security. Coefficient βi covers the systematic risk of the acquirer’s security. Market return is calculated by using the return on the index of Standard and Poor’s Europe 350 Financials [4] . Return data is based on dividends reinvested and obtained from Datastream. Lastly, εit is the error term. We used OLS-regressions to calculate the coefficients of equation 2 for all acquirer’s securities in the estimation window.

The coefficients we found in the forgoing paragraph are used to predict the returns in the event window. The abnormal return is the actual return of the security over the event window minus the normal return of the firm over the event window. For firm i and event date τ the abnormal return is:


The abnormal returns for event i are estimated for timed period τ, where τ is based on the event day and the same number of days before and after the event day. τ is set equal to 5 given that short-term reactions around the announcement date are considered. In order to make statistical inferences about the data the abnormal returns observations need to be aggregated through time and across securities. The abnormal returns based on estimation period from day 11 through day 110 are averaged across N acquiring to produce AAR, which are cumulated to form CAAR. CAAR is calculated for day 0, two days prior to the announcement day 0, and two days after the announcement day.

The variance of AAR is the sum of the variances of the abnormal returns divided by the square of the number of events after adjusting for the loss of degrees of freedom ((N-1)2). The cumulative average abnormal returns (CAAR) are estimated by aggregating over the event window τ = T1 to τ = T2. The variance of CAAR is found by summing the variances of the abnormal returns over the event window. We test for each τ if the CAAR differs from zero for the banking industry using the following t-statistic:


3.5 Cross-section analysis

We apply a multiple cross-section regression analysis, as presented below.

CAARi = a1 + a2MOPi + a3FCi + a4INCRSTi + εi,

where CAARi is the cumulative average abnormal returns which is measured for τ = 5 for the bidder in acquisition announcement i. MOPi refers to the dummy variable of the method of payment which equals 1 if the payment form is cash, and 0 if the payment form is shares. FCi is the dummy variable that equals 1 if the announcements has been made in the time interval of 2007/01/01 and 2008/12/31. Otherwise, it equals 0. Lastly, the dummy variable INCRSTi contains four classes (0-25%, 25-50%, 50-75% and 75-100%) of the increase of stakes. We inserted four dummy variables while excluding the constant term to avoid multicollinearity errors.

4. Results

CAARs of bidder banks

Table 2 shows a high T-statistic which implies that the cumulative average abnormal returns (CAARs) for the estimated alphas and betas (equation 1) in an event window of 5 days for acquisitions of banks in Europe are not significantly different from zero. Therefore, the first hypotheses (H1) that cumulative abnormal returns to the acquirer company around announcements of an acquisition in the European banking industry will be significantly different from zero can be rejected.

Table 2: Descriptive statistics and t-test: cumulative average abnormal returns


for bidder company’s in the European banking industry

Event window 5 days*

Panel A CAAR (63 observations)

Mean 0,000

Maximum 0,001

Minimum -0,001

Standard deviation** 0,001

T-statistic (mean of zero) 0,995

* In the event window of 5 days the alfa and beta are estimated using OLS in the market model

** The standard deviation are rounded and therefore it could be seen as high

Method of payment

Table 3 reports that the second hypotheses (H2) can be rejected, regarding the high T-statistics for both cash payment and shares and the other payment forms. This means that we do not find significant differences between announcements of acquisitions including cash payments and announcements of acquisitions including payments with stock. Therefore, cash payments do not result in higher abnormal returns compared to payment types.

Table 3: Descriptive statistics and t-tests: bidder company’s cumulative

average abnormal returns divided into two groups cash and shares (and others)

Event window 5 days*

Panel A cash (37 observations)

Mean 0,000

Maximum 0,001

Minimum -0,001

Standard deviation** 0,001

T-statistic 0,940

Panel B shares and other (26 observations)

Mean 0,000

Maximum 0,001

Minimum -0,001

Standard deviation** 0,001

T-statistic 0,933

* In the event window of 5 days the alfa and beta are estimated using OLS in the market model

** The standard deviation are rounded and therefore it could be seen as high

Control variables

We expected to find negative returns to announcements with cash payments made during the financial crisis (2007 and 2008). Table 4 (panel A) presents the results of the cross-sectional regression analysis, where we find that announcements of an acquisition including cash payments made during the recent financial crisis result in a significantly negative development of CAAR.

Panel B shows that only an increase of the stake from 75 to 100% paid with cash results in an increase of CAAR. The other classes of increase are not significant.

Table 4: Test on cumulative abnormal returns in the Europeans stock market the method of payment and the control variables

Cash Shares and other

N Coefficient p-value N Coefficient p-value

Panel A (financial crisis)

Acquisition in crisis 6 -0,001 0,050 ** 6 -0,000 0,729

Acquisitions outside crisis 31 0,000 0,419 20 0,000 0,909

Panel B (Increase of stake

in the target company)

0 Withdrawn 1 0,000 0,964 0 - - 1 0-25% stake increase 6 0,00 0,934 5 -0,000 0,441

2 25-50% stake increase 7 0,000 0,478 5 0,000 0,351

3 50-75% stake increase 4 -0,000 0,347 1 0,000 0,520

4 75-100% stake increase 19 0,001 0,097 *** 15 0,000 0,797

* Alpha and beta are estimated applying the market model

** The effect of an acquisition in cash during the financial crisis has a significant negative effect on CAAR at a 10 significance level

*** The effect of acquisition in cash when there is a stake increase of 75-100% has a significance positive effect on CAAR at a 10% significance level

5. Conclusions

This study examines the short-term reactions of shareholders of European banking companies that announce acquisitions of banks, which are also located in Europe (EU-15). Research conducted in the past reports inconclusive results regarding abnormal returns for bidder companies in general. Although, specific investigation in the banking sector have not frequently been reported.

The method of payment applied makes no difference in obtaining abnormal returns for the shareholders of the bidder bank. However, we did find evidence for suggesting that it matters which method of payment is applied in times of crisis. Shareholders do not appreciate cash payment in liquidity absorbing conditions. Furthermore, it does make sense to conclude that a 75-100% stake increase should be performed using cash. An acquirer that finances the acquisition completely with cash will reveal increasing cash flow opportunities in the future. This is inline with findings in the research of Jensen and Ruback (1983).

The conclusion that bidder returns are not influenced by the method of payment is in contrast with the results found by existing studies that dig into acquisitions of non-banks, which report significantly higher bidder returns in case of cash payments than from stock-financed acquisitions (Asquith et al., 1987; Travlos, 1987).

5.1 Limitations

Firstly, the restrictions implemented to limit this research and secondly the availability of data resulted in a small sample. This has impact on the possibilities of making significant explanations. It is possible that we could find more significant results if we had more announcements of acquisition available for the European banking market. Furthermore, it must be mentioned that the normality assumptions for applying event studies are violated.

5.2 Recommendations

First, according to the study conducted by Beltratti and Stulz (2009), an announcement of an acquisition including a method of payment that is sacrificing a majority of the liquidity assets could result in negative abnormal returns. Stockholders probably would not appreciate a cash payment in times of crisis. We did not find evidence for this suggestion due to the limitations of the sample. However, we believe this would be an interesting issue to investigate in future research projects.

Second, PricewaterhouseCoopers (2006) stresses the importance of considering mergers and acquisitions of banks in Eastern Europe, carried out by bidder banks located in the EU-15. Many banks have sought to grow beyond their home markets by merging with or acquiring stakes in banks in neighbouring countries that may have faced similar domestic market conditions. The so called ‘emerging European countries’ as Poland, Czech Republic, Russian Federation, and Ukraine are attractive by offering more growth opportunities compared to home countries, and the regional expansion and therefore diversification opportunities.