3. Empirical study

In order to investigate the effects of Asian crisis, [17] used a data of nine emerging countries namely Argentina, Brazil, Chile, Indonesia, Philippines, South Korea, and Thailand between 1988 and 2002. They concluded that acquirer firms show no significant difference in abnormal returns pre and postcrisis periods. On the other hand, [18] studied the M&As in eight East Asian countries between 1997 and 2003 in order to determine market reaction to M&As during Asian crisis. Their results showed that market reaction was negative in Indonesia, Malaysia, the Philippines, South Korea, and Thailand where the bank structure was less well settled.

The results of the studies that investigate the effects of 2008 crisis are mixed. [19] utilized the M&A data in Europe between 2007 and 2010 to evaluate whether M&A differed in crisis period. They concluded that there were insignificant abnormal returns on the event date. On the other hand, abnormal returns were generated positively at the completions. However, [20] used 80 M&As in UK, USA, Canada, Germany, Japan, and France between 1999 and 2009 to determine stock returns of bidder firms. Abnormal returns precrisis and postcrisis period was not significantly different from zero. In another research, [3] examined 883 cross-border M&A deals in banking sector between 2004 and 2012. They concluded that only in M&As from emerging countries targeting developed countries, returns of the shareholders were significantly positive after the crisis. Finally, [21] gathered the M&A data for 20 emerging countries namely BRICS-T countries and Chile, Colombia, Czech Republic, Egypt, Hungary, Indonesia, Malaysia, Mexico, Morocco, Peru, Philippines, Poland, Taiwan and Thailand between 1997 and 2013. They concluded that M&As created positive abnormal returns. In addition, they found out that abnormal returns had increased after crisis for

In conclusion, results of the studies are mixed and they change according to the period.

In previous sections, it has been mentioned that there exists M&A waves. According to [22], there had been six M&A waves before the 2008 crisis, which are 1887–1907, 1919–1933, 1955–

Table 1 shows the quantity and transaction value of cross-border M&As in BRICS-T countries between 2002 and 2016 [23]. In Brazil, cross-border M&As have value of \$17 billion in 2003. It increased by 65% in 2004 and reached to \$26 billion. In 2005, there is a decrease by 58% and the value is \$15 billion. Then in 2006, there is a jump in the value and it has reached to \$74 billion. In 2007, there is a fall by 72% in value. M&As have the peak value in 2008 during the precrisis period. There is a drastic fall in value in 2009 due to crisis. In 2010, M&As have the peak value in postcrisis period. It started declining afterwards. In Russia, the value of cross-border M&As is \$35 billion in 2003, and in 2004, the value has declined by 72%. In 2005, the value has jumped to \$63 billion and during the precrisis period, M&As have the peak value in 2007. In 2008, the value has decreased by 52% and in 2009 the decrease is 43%. Then, the value has been tripled in 2010. Cross-border M&As have their peak value in 2012 in postcrisis period. In India and China, cross-border M&As have the peak value in 2007 in precrisis period and in 2010 in postcrisis period. In India, there is a jump in cross-border M&A value in 2005 (4.5 times higher than 2004), and in China, there is a high increase in 2005 as well (6.25 times higher than 2004).

2.3. Financial overview of BRICS-T countries in precrisis and postcrisis periods

1975, 1980–1989, 1992–2002, and finally 2003–2007.

134 Financial Management from an Emerging Market Perspective

target firm's stock.

In this part, first, data, methodology, and the hypothesis are explained. Then empirical results are represented.

#### 3.1. Data, methodology, and hypothesis

Our study uses daily market index returns, daily stock returns, and M&A announcement dates (event date) between January 2003–September 2008 and November 2008 and December 2013. We utilize the data from Bloomberg database for cross-border bank M&A activities in Brazil, Russia, India, China, South Africa, and Turkey. Our data consists of cross-border M&As with a transaction value over \$100 million.

Event study is employed for the analysis. Event studies aim to determine whether there are abnormal returns around the date an event is announced to the market. Abnormal returns are the returns that are less or more than normal returns when the related event is announced. These returns are usually related with the performance of the market index returns [24, 25]. The event is the M&A announcement date. There could be different event windows, which include the announcement date. In this study, we investigate abnormal returns for different event window lengths:


We choose market model in order to estimate market α<sup>4</sup> and β<sup>5</sup> over a prediction period, which is 128 days prior to and 9 days prior to event date, that is, (�128, �9). The market model is as follows:

$$R\_{it} = \alpha\_i + \beta\_j R\_{mt} + \varepsilon\_{it} \tag{1}$$

Rit stands for the return of the ith security at time t and Rmt denotes the return of the market at time t.

Then, abnormal return (AR) is calculated using predicted αiand β<sup>i</sup> :

$$AR\_{it} = R\_{it} - \widehat{\alpha}\_i - \widehat{\beta}\_i R\_{mt} \tag{2}$$

ARit represents abnormal return for the ith bank at time t and Rit is the actual return on bank i. Later, average aggregate abnormal return (AAR) is calculated:

$$AAR\_t = \frac{\mathfrak{I}}{N} \sum\_{i=1}^{N} AR\_{it} \tag{3}$$

After that, by adding daily abnormal returns up, cumulative abnormal returns are obtained:

$$
\mathbb{C}AR\_{i(T\_1 - T\_2)} = \sum\_{t=T\_1}^{T\_2} AR\_{it} \tag{4}
$$

Here, CARi is the cumulative abnormal return for bank i over the event window T<sup>1</sup> and T2. Finally, average aggregate cumulative abnormal return is calculated (AACAR):

$$\mathsf{ACAR}(T\_1, T\_2) = \frac{1}{N} \sum\_{i=1}^{N} \mathsf{CAR}\_i(T\_1, T\_2) \tag{5}$$

and ð Þ T1; T<sup>2</sup> are (0, +1), (�1, +1), and (�2, +2).

5 Responsiveness of a security to the market return.

<sup>4</sup> Return on security when the expected return on market is zero.

#### 3.2. Empirical results

Event study is employed for the analysis. Event studies aim to determine whether there are abnormal returns around the date an event is announced to the market. Abnormal returns are the returns that are less or more than normal returns when the related event is announced. These returns are usually related with the performance of the market index returns [24, 25]. The event is the M&A announcement date. There could be different event windows, which include the announcement date. In this study, we investigate abnormal returns for different

We choose market model in order to estimate market α<sup>4</sup> and β<sup>5</sup> over a prediction period, which is 128 days prior to and 9 days prior to event date, that is, (�128, �9). The market model is as follows:

Rit stands for the return of the ith security at time t and Rmt denotes the return of the market at

ARit <sup>¼</sup> Rit � <sup>α</sup>b<sup>i</sup> � <sup>β</sup>

AARt <sup>¼</sup> <sup>1</sup> N X N

ARit represents abnormal return for the ith bank at time t and Rit is the actual return on bank i.

After that, by adding daily abnormal returns up, cumulative abnormal returns are obtained:

CARiðT1�T2<sup>Þ</sup> <sup>¼</sup> <sup>X</sup>

Here, CARi is the cumulative abnormal return for bank i over the event window T<sup>1</sup> and T2.

1 N X N

i¼1

Finally, average aggregate cumulative abnormal return is calculated (AACAR):

ACARð Þ¼ T1; T<sup>2</sup>

and ð Þ T1; T<sup>2</sup> are (0, +1), (�1, +1), and (�2, +2).

Responsiveness of a security to the market return.

Return on security when the expected return on market is zero.

i¼1

T2

t¼T<sup>1</sup>

b i

Rmt þ eit (1)

Rmt (2)

ARit (3)

ARit (4)

CARið Þ T1; T<sup>2</sup> (5)

:

Rit ¼ α<sup>i</sup> þ β<sup>i</sup>

• Two days before and two days after the event date (�2, +2)

Then, abnormal return (AR) is calculated using predicted αiand β<sup>i</sup>

Later, average aggregate abnormal return (AAR) is calculated:

• A day before and a day after the event date (�1, +1) • The event day and a day after the event date (0, +1)

136 Financial Management from an Emerging Market Perspective

event window lengths:

time t.

4

5

This section introduces the empirical results. First, abnormal returns for the entire period are shown without separating the before/after crisis periods in Section 3.2.1. Then, abnormal returns for pre / post crisis periods are given in Sections 3.2.2 and 3.2.3, respectively.

#### 3.2.1. Aggregate daily abnormal returns

This part introduces the aggregate results, which means that abnormal returns of the M&A activities are included to the analysis without considering pre- and post-crisis periods. A total of 36 banks with M&A transaction values over \$100 million are taken into consideration.

Table 2 shows average aggregate daily abnormal returns two days before after the event date. The AARs before and on the announcement date are negative and significant at 5% while AARs are positive and significant at 5% significance level. The AARs increase through the event window. In other words, the AAR two days before the event day is �0.043, it is �0.040 on the day before the event day, and it is larger but still negative on the event day. One day after the event day, the AAR reaches to the largest value. There are excess returns on the M&As. On the second day, AAR decreases again.

Table 3 shows aggregate CARs for the related event window. In 5-day event window (�2, +2), CAR is �0.042, and it is statistically significant at 5%. Then, in 3-day event window (�1, +1), CAR increases to �0.018, and this value is statistically significant at 5%. Finally, in 2-day event window, CAR increases to �0.007, and it is statistically significant at 5%.


Table 2. AARs for the Related Event Associated with M&A Activities.


1 Time period that includes several days prior and after the event.

Table 3. CARs for the Related Event Windows Associated with M&A Activities.

Table 4 shows the distribution of 5-day CARs. The results show that there are negative abnormal returns in Brazil and Russia, while there are positive cumulative abnormal returns in China, India, South Africa, and Turkey. The results are significant at 5% significance level. In terms of 5 day CARs, Brazil has the lowest CAR among other countries, and it is followed by Russia. Although, there is positive CAR in India, South Africa, Turkey, and China have more CAR than India. CARs in South Africa and Turkey are very close. China has the largest CAR among these countries in 5-day event window.

Table 5 shows the distribution of 3-day CARs. Brazil, Russia, and India have negative CARs while China, South Africa, and Turkey have positive CARs. The results are significant at 5% significance level. In 3-day event window, Brazil has the least CAR among other countries and it is followed by India and Russia. While India has slightly positive CAR in 5-day event window, it has negative CAR in 3-day event window. South Africa and Turkey have positive CAR in 3-day event window as well as the 5-day event window. Turkey has the largest CAR among other countries in 3-day event window.

Table 6 shows the distribution of 2-day CARs. Brazil, Russia, and India have negative CARs while China, South Africa, and Turkey have positive CARs. The results are significant at 5% significance level. In 2-day event window, Brazil has the least CAR and it is followed by Russia and India. Turkey has the largest CAR among other countries in 2-day event window as well as 2-day event window.

Table 7 shows the distribution of mean CARs. On an average, Brazil, India, and Russia have negative cumulative abnormal returns and China, South Africa, and Turkey have positive


Table 4. Distribution of 5-day CARs (�2, +2) in BRICS-T countries.


Table 5. Distribution of 3-day CARs (�1, +1) in BRICS-T countries.

cumulative abnormal returns between 2003 and 2013 for banking industry. The results are significant at 5% significance level.

#### 3.2.2. Daily abnormal returns in precrisis period

Table 4 shows the distribution of 5-day CARs. The results show that there are negative abnormal returns in Brazil and Russia, while there are positive cumulative abnormal returns in China, India, South Africa, and Turkey. The results are significant at 5% significance level. In terms of 5 day CARs, Brazil has the lowest CAR among other countries, and it is followed by Russia. Although, there is positive CAR in India, South Africa, Turkey, and China have more CAR than India. CARs in South Africa and Turkey are very close. China has the largest CAR among these

Table 5 shows the distribution of 3-day CARs. Brazil, Russia, and India have negative CARs while China, South Africa, and Turkey have positive CARs. The results are significant at 5% significance level. In 3-day event window, Brazil has the least CAR among other countries and it is followed by India and Russia. While India has slightly positive CAR in 5-day event window, it has negative CAR in 3-day event window. South Africa and Turkey have positive CAR in 3-day event window as well as the 5-day event window. Turkey has the largest CAR among other

Table 6 shows the distribution of 2-day CARs. Brazil, Russia, and India have negative CARs while China, South Africa, and Turkey have positive CARs. The results are significant at 5% significance level. In 2-day event window, Brazil has the least CAR and it is followed by Russia and India. Turkey has the largest CAR among other countries in 2-day event window as well

Table 7 shows the distribution of mean CARs. On an average, Brazil, India, and Russia have negative cumulative abnormal returns and China, South Africa, and Turkey have positive

Name of the country CAR (�2, +2) (%)

Name of the country CAR (�1, +1) (%)

Brazil �0.129 China 0.007 India �0.012 Russia �0.011 South Africa 0.012 Turkey 0.026

Brazil �0.287 China 0.017 India 0.001 Russia �0.006 South Africa 0.013 Turkey 0.014

Table 4. Distribution of 5-day CARs (�2, +2) in BRICS-T countries.

Table 5. Distribution of 3-day CARs (�1, +1) in BRICS-T countries.

countries in 5-day event window.

138 Financial Management from an Emerging Market Perspective

countries in 3-day event window.

as 2-day event window.

This section introduces the abnormal return analysis results of M&As in banking sector during the pre-crisis period, that is, between September 2003 and November 2008. In this manner, 22 banks M&As with a M&A transaction value more than \$100 million have been investigated.

Table 8 shows AARs for pre-crisis period. There are negative AARs before and on the event date. However, there are positive abnormal returns after the announcement date. The results


Table 6. Distribution of 2-day CARs (0, +1) in BRICS-T countries.


Table 7. Distribution of mean CARs in BRICS-T countries.


Table 8. AARs for the Related Event Windows Related to M&A Activities Before Crisis (2003-2008/9).

are significant at 5% significance level. Two days before the announcement day AAR is �0.057 and 1 day before the event day it increases to �0.051. On the announcement day, AAR increases to �0.037 and 1 day after the event day, it turns to positive and has its peak value. In other words, the AARs follow an increasing path until 1 day after the announcement day. Two days after the announcement date, it decreases but it is still positive. This figure is very similar to the aggregate case in Section 3.2.1.

Table 9 shows CARs for 5-day, 3-day, and 2-day event windows. There are negative cumulative abnormal returns in precrisis period. The values tend to increase as the event window gets narrower to the event date. The results are significant at 5% significance level.

Table 10 shows 5-day CARs in BRICS-T countries. In Brazil, India, and Russia, CARs are negative and significant at 5% level. In China, South Africa, and Turkey, CARs are positive and significant at 5% level. Brazil has the least CAR and it is followed by India and Russia in 5-day event window. China has the largest CAR in 5-day event window in precrisis period and it is followed by South Africa and Turkey. Note that Turkey had the largest CAR in the aggregate case.

Table 11 shows the distribution of 3-day cumulative abnormal returns in BRICS-T countries in precrisis period. In Brazil, India, and Russia, CARs are negative and significant at 5% level. In China, South Africa, and Turkey, cumulative abnormal returns are positive and significant at 5% level. Brazil has the least CAR in 3-day event window during the precrisis period. However, the CAR value has increased with respect to 5-day event window. Russia and India follow Brazil and their CAR values have decreased compared to 5-day event window. China still has the largest CAR but the values have decreased in 3-day event window. CARs in South Africa and Turkey have increased in 3-day event window


Table 9. CARs for the Related Event Windows in Response to M&A Activities Before Crisis (2003-2008/9).


Table 10. Distribution of 5-day CARs (�2, +2) in BRICS-T countries before crises (2003–2008/2009).

Table 12 shows the distribution of CARs in two-day event window. Brazil, India, and Russia have negative abnormal returns while China, South Africa, and Turkey have positive abnormal returns. The results are significant at 5% significance level. Brazil has the least CAR in 2-day event window. This value of CAR in 2-day event window is larger than the value of CAR in 3 day event window. CAR for Russia in 2-day event window is less than CAR in 3-day event window and CAR for India in 3-day event window is larger than CAR in 2-day event window. China still has the largest CAR in 2-day event window and the value has increased compared to the 3-day event window. CAR in South Africa has increased while CAR in Turkey has decreased in 2-day event window with respect to 3-day event window.

are significant at 5% significance level. Two days before the announcement day AAR is �0.057 and 1 day before the event day it increases to �0.051. On the announcement day, AAR increases to �0.037 and 1 day after the event day, it turns to positive and has its peak value. In other words, the AARs follow an increasing path until 1 day after the announcement day. Two days after the announcement date, it decreases but it is still positive. This figure is very

Table 9 shows CARs for 5-day, 3-day, and 2-day event windows. There are negative cumulative abnormal returns in precrisis period. The values tend to increase as the event window gets

Table 10 shows 5-day CARs in BRICS-T countries. In Brazil, India, and Russia, CARs are negative and significant at 5% level. In China, South Africa, and Turkey, CARs are positive and significant at 5% level. Brazil has the least CAR and it is followed by India and Russia in 5-day event window. China has the largest CAR in 5-day event window in precrisis period and it is followed by South Africa and Turkey. Note that Turkey had the largest CAR in the

Table 11 shows the distribution of 3-day cumulative abnormal returns in BRICS-T countries in precrisis period. In Brazil, India, and Russia, CARs are negative and significant at 5% level. In China, South Africa, and Turkey, cumulative abnormal returns are positive and significant at 5% level. Brazil has the least CAR in 3-day event window during the precrisis period. However, the CAR value has increased with respect to 5-day event window. Russia and India follow Brazil and their CAR values have decreased compared to 5-day event window. China still has the largest CAR but the values have decreased in 3-day event window. CARs in South

narrower to the event date. The results are significant at 5% significance level.

similar to the aggregate case in Section 3.2.1.

140 Financial Management from an Emerging Market Perspective

Africa and Turkey have increased in 3-day event window

�2, +2 �0.053 �1, +1 �0.019 0, +1 �0.006

Event window Average CAR (%)

Name of the country CAR (�2, +2) (%)

Table 10. Distribution of 5-day CARs (�2, +2) in BRICS-T countries before crises (2003–2008/2009).

Brazil �0.368 China 0.064 India �0.025 Russia �0.039 South Africa 0.041 Turkey 0,009

Table 9. CARs for the Related Event Windows in Response to M&A Activities Before Crisis (2003-2008/9).

aggregate case.

Table 13 shows the distribution of mean CARs in BRICS-T countries for precrisis period. Accordingly, Brazil, India, and Russia generates negative and statistically significant abnormal


Table 11. Distribution of 3-day CARs (�1, +1) in BRICS-T countries before crises (2003–2008/2009).


Table 12. Distribution of 2-day CARs (0, +1) in BRICS-T countries before crises (2003–2008/2009).


Table 13. Distribution of mean CARs in BRICS-T countries before crises (2003–2008/2009).

returns while China, South Africa, and Turkey obtains positive cumulative abnormal returns between September 2003 and September 2008 for banking industry for M&A transactions with a value more than \$100 million. Brazil has the least mean CAR and it is followed by Russia and India while China has the largest mean CAR and South Africa and Turkey follow it.

#### 3.2.3. Daily abnormal returns in postcrisis period

This section introduces abnormal returns in M&As in the banking industry during the postcrisis period, that is, between November 2008 and December 2013. In this manner, 14 bank M&As with a M&A transaction value more than \$100 million have been investigated.

Table 14 shows the AARs for the related event window in postcrisis period. Two days before the announcement date, AAR is positive; 1 day before the announcement date AARs is negative; and on the event date, AAR is positive. One-day and 2-day after the event date, AARs are slightly negative. The results are significant at 5% significance level.

Table 15 shows CARs for the related event windows in postcrisis period. In 5-day event window, CAR is positive and in 3-day and 2-day event windows cumulative abnormal returns are negative. Average CAR has the largest value in 5-day event window, the least value in 3-day event window. The results are significant at 5% significance level.

Table 16 shows distribution of the 5-day CARs in BRICS-T countries in postcrisis period. Brazil has the least CAR among other countries. South Africa and China follow Brazil. Note that South Africa had positive CAR in precrisis period. Russia has the largest CAR and it is followed by


Table 14. AAR for the Related Event Windows in Response to M&A Activities After Crisis (2008/11-2013).


Table 15. CARs for the Related Event Windows in Response to M&A Activities After Crisis (2008/11-2013).

India and Turkey. Another remarkable points are that India had negative CAR in precrisis period and China had the largest positive CAR in precrisis period

returns while China, South Africa, and Turkey obtains positive cumulative abnormal returns between September 2003 and September 2008 for banking industry for M&A transactions with a value more than \$100 million. Brazil has the least mean CAR and it is followed by Russia and India while China has the largest mean CAR and South Africa and Turkey

This section introduces abnormal returns in M&As in the banking industry during the postcrisis period, that is, between November 2008 and December 2013. In this manner, 14 bank M&As with a M&A transaction value more than \$100 million have been

Table 14 shows the AARs for the related event window in postcrisis period. Two days before the announcement date, AAR is positive; 1 day before the announcement date AARs is negative; and on the event date, AAR is positive. One-day and 2-day after the event date,

Table 15 shows CARs for the related event windows in postcrisis period. In 5-day event window, CAR is positive and in 3-day and 2-day event windows cumulative abnormal returns are negative. Average CAR has the largest value in 5-day event window, the least value in 3-day

Table 16 shows distribution of the 5-day CARs in BRICS-T countries in postcrisis period. Brazil has the least CAR among other countries. South Africa and China follow Brazil. Note that South Africa had positive CAR in precrisis period. Russia has the largest CAR and it is followed by

Table 14. AAR for the Related Event Windows in Response to M&A Activities After Crisis (2008/11-2013).

Table 15. CARs for the Related Event Windows in Response to M&A Activities After Crisis (2008/11-2013).

AARs are slightly negative. The results are significant at 5% significance level.

event window. The results are significant at 5% significance level.

Event day Average abnormal returns (%)

Event window Average CAR (%)

�2, +2 0.002 �1, +1 �0.004 0, +1 �0.001

�2 0.002 �1 �0.002 0 0.001 1 �0.001 2 �0.001

follow it.

investigated.

3.2.3. Daily abnormal returns in postcrisis period

142 Financial Management from an Emerging Market Perspective

Table 17 shows the distribution of 3-day CARs in BRICS-T countries in postcrisis period. South Africa now has the least CAR among other countries in 3-day event window. CAR value in Brazil does not change compared to the 5-day CAR but the CAR value in South Africa has decreased. CAR values in China and India have also decreased while CAR in Turkey has increased. Russia has the largest CAR among other countries and the value has increased.

Table 18 shows the 2-day CARs in BRICS-T countries during postcrisis period. The figure is similar to the 3-day CAR case. South Africa has the least CAR and its value has not changed.


Table 16. Distribution of 5-day CARs (�2, +2) in BRICS-T countries after crises (2008/2011–2013).


Table 17. Distribution of three-day CARs (�1, +1) in BRICS-T countries after crises (2008/2011–2013).


Table 18. Distribution of 2-day CARs (0, +1) in BRICS-T countries after crises (2008/2011–2013).


Table 19. Distribution of mean CARs in BRICS-T countries after crises (2008/2011–2013).

The CAR value in China has increased slightly. CARs in India have decreased while the CARs in Russia and Turkey have increased.

Table 19 shows the distribution of mean CARs in BRICS-T countries in postcrisis period for M&A transactions with a value more than \$100 million. Brazil, China, and South Africa have negative and statistically significant mean CARs while India, Russia, and Turkey have positive and statistically significant mean CARs. South Africa has the least mean CAR and Brazil and China follow it. Russia has the largest CAR and Turkey and India follow Russia.
