7. CONCLUSIONS AND RECOMMANDATION
5.2 A NALYSIS FOR M ERGER I MPACTS IN THE L ONG - TERM AND THE C ONNECTION TO R ELATIVE
When analyzing long-term post-merger performance, it is to investigate the abnormal returns and cumulated abnormal returns when using market assessment and excess operating performance when using operating performance assessment for the selected companies in the long horizon (in this dissertation up to five years after mergers).
Because mean-adjusted model is used in this dissertation, the average returns for stock, ROA, and ROE in the estimation windows are used as expected return and performance to calculate abnormal returns and excess operating performance.
Deducting the mean value of the estimation windows from the actual returns and performance in post-event windows will result in the abnormal returns (AR) for stock and excess ROA (EROA) and ROE (EROE) for operating performance.
In the analysis for the long-term size impact, there are 74 samples in total, after filtering out the deals without completed financial information. Similar to the short-term analysis, the analysis was broken into five different groups, the whole sample, the first, the second, the third, and the fourth quartiles, which were arranged by relative size between combining firms from small to large in order. The analysis for post-merger performance in the long-term is presented in table 5.4 below. In table 5.3, the average abnormal returns (AAR), average cumulated abnormal returns (ACAR), average excess ROA (AEROA), and average excess ROE (AEROE) for data are presented in each panel.
Table 5.4
Long-term Post-merger Performance up to Five Years after Mergers
Quarters After Mergers Completion
1-4 5-8 9-12 13-16 17-20
Panel A: Full Sample (N=74)
AAR(%) -90.7364 -50.4809 -75.8811 -69.7480 -75.2108
(-0.5158)*** (-0.2885)*** (-0.44083)*** (-0.5254)*** (-0.5185)***
ACAR(%) -362.9456 -564.8690 -868.3932 -1147.3852 -1448.2284
(-0.5427)*** (-0.4756)*** (-0.4818)*** (-0.4997)*** (-0.5067)***
AEROA(%) -10.2171 -5.0331 -11.6574 3.9470 1.3617
(-0.4063)*** (-0.1294)** (-0.3094)*** (0.0919) (0.0541) AEROE(%) -33.1406 -20.2034 -29.5282 -8.4544 -3.5856
(-0.2964)*** (-0.2950)*** (-0.2884)*** (-0.1657)*** (-0.0951)
Panel B: The First Quartile (N=18)
AAR(%) -32.9281 -35.9485 8.4933 -8.0579 -9.1586
(-0.6596)*** (-0.5871)*** (0.1221) (-0.1298) (-0.1766)
ACAR(%) -131.7124 -275.5064 -241.5332 -273.7648 -310.3992 (-0.6588)*** (-0.6245)*** (-0.3866) (-0.3162) (-0.2889) AEROA(%) -8.8434 -13.7685 -9.3861 -3.9658 -3.2211
(-0.4426)*** (-0.5354)*** (-0.6311)*** (-0.3653)*** (-0.2788)**
AEROE(%) -13.3244 -25.8194 -18.7472 -7.0708 -5.6745
(-0.3904)*** (-0.5049)*** (-0.4761)*** (-0.2989)** (-0.2438)**
Panel C: The Second Quartile (N=19)
AAR(%) -92.2997 -88.8911 -100.372 -88.4686 -98.3296
(-0.3780)*** (-0.4616)*** (-0.4830)*** (-0.5318)*** (-0.5251)***
ACAR(%) -369.1988 -724.7632 -1126.2512 -1480.1256 -1873.4440
(-0.3838) (-0.4607)* (-0.4874)** (-0.5057)** (-0.5101)**
AEROA(%) -13.7626 -12.3092 -18.9494 -1.1520 1.0512
(-0.4660)*** (-0.4271)*** (-0.4469)*** (-0.0446) (0.0401) AEROE(%) -21.1666 -36.5945 -42.9327 -14.2344 0.8794
(-0.6546)*** (-0.4092)*** (-0.5424)*** (-0.2649)** (0.0243)
Panel D: The Third Quartile (N=18)
AAR(%) -104.9507 -48.2990 -101.8907 -77.6555 -91.1132
(-0.6736)*** (-0.3134)*** (-0.6532)*** (-0.6904)*** (-0.8472)***
ACAR(%) -419.8028 -612.9988 -1020.5616 -1331.1836 -1695.6364
(-0.8812)*** (-0.6880)*** (-0.7117)*** (-0.7309)*** (-0.7697)***
AEROA(%) -11.8852 -4.3586 -24.4002 10.018 -1.6306
(-0.8820)** (-0.0782) (-0.5522)*** (0.1354) (-0.0925)
AEROE(%) -53.6258 -27.9504 -35.9807 -22.9585 -14.4186
(-0.2673)*** (-0.5267)*** (-0.5588)** (-0.3485)*** (-0.5392)***
Panel E: The Fourth Quartile (N=19)
AAR(%) -123.7709 -29.09086 -96.4808 -94.2091 -91.6067
(-0.6931)*** (-0.12721) (-0.5079)*** (-0.6523)*** (-0.5449)***
***,**, * Statistical significance in 2-tailed tests at 1%, 5%, and 10% level, respectively.
Source: Author
For AAR, the results for the data of whole sample (panel A) show significantly negative in all the periods after mergers, and except the group of the first quartile, the similar results hold in the rest groups. The bar chart of AAR for groups from the first to the fourth quartile is shown in figure 5.3 below. When comparing AAR between each quartile group, the bar chart clearly shows that the first quartile group has the best performance as it always has the smallest negative AAR in all periods. Moreover, in the third year after merger, the AAR for the first quartile group is even positive.
-140
Bar Chart of AAR for the Groups from the First to the Fourth Quartile
Source: Author
The analysis result of ACAR is tightly connected to AAR, due to the reason that ACAR is the cumulated result of AAR in different periods. Because AAR for the whole samples is significantly negative, the cumulated results of AAR are significantly negative in all periods as well. Therefore, ACAR for the group of the whole sample increase gradually year after year, from -362.9456 of the first year to -1448.2284 of the fifth year cumulatively. When dividing the whole sample into four equal groups in accordance with relative size between combining firms, most ACAR
for each group is still significantly negative. Figure 5.4 further analyzes the results of ACAR for different groups. It can be observed that comparing to the rest groups, the ACAR for the first-quartile group keep increasing its negative ACAR with a relatively smaller volume each year. On the other hands, other groups show very large increasing in negative ACAR each year.
-2000 -1500 -1000 -500 0
1 2 3 4 5
Year After Mergers
ACAR% 1sr Quartie
2nd Quartile 3rd Quartile 4th Quartile
Figure 5.4
Bar Chart of ACAR for the Groups from the First to the Fourth Quartile
Source: Author
For AEROA, the results show significantly negative in the first three years (1-12 quarters after mergers) for the group of whole samples shown in panel A. From the fourth year to the fifth year, the mean performance turned out to be positive, but not significantly different from zero. The similar results roughly hold in research groups of quartiles first, second, and third, showing that the AEROA started to become less negative comparing to the earlier years after mergers no matter significantly or not.
However, the results for the fourth quartile shows different picture. The AEROA for the first year is the same as the previous groups, but it becomes positive right starting from the second year after mergers, though not significantly. After the fourth year, the
AEROA for the fourth quartile becomes significantly positive for the rest two years.
The bar chart of AEROA for the groups from the first to the fourth quartiles is shown in figure 5.5 as below.
Bar Chart of AEROA for the Groups from the First to the Fourth Quartile
Source: Author
For AEROE, the results for full samples show significantly negative for the first four years (1-16 quarter after mergers), but it became less negative as years went by.
Extending the investigation to four different quartiles, it shows that the AEROE for the first year is all significantly negative for all groups, and the results for group of first quartile had the smallest AEROE. After the second year, the AEROE for those four groups became better and better corresponding to the timeline. What worth mentioned is to look at the group of the fourth quartile. It can be observed specifically for the fourth quartiles that, different from the formal three groups of the first to the third quartiles, though not significantly, AEROE for the second, fourth, and the fifth year after mergers were positive. The bar chart of AEROE for the groups from the
first to the fourth quartiles is shown in figure 5.6 as below.
-60 -50 -40 -30 -20 -10 0 10
1 2 3 4 5
Year After Mergers
AEROE %
1st Quartile 2nd Quartile 3rd Quartile 4th Quartile
Figure 5.6
Bar Chart of AEROA for the Groups from the First to the Fourth Quartile
Source: Author
5.3 Summary and Conclusion
This chapter presents the analysis results of post-merger performance for the short-term (for the first quarter after mergers) and the long-term (up to five years after mergers) For both short-term and long-term analyses, in addition to analyzing the whole samples, the analysis was broken into four quartile groups according to the relative size of combining firms.
In the short-term size impact analysis, there are 58 samples in total. In the percentage change analysis, it was found that for both market and operating performance assessments, percentage changes are all significantly negative. Among the four
quartile groups categorized in order by absolute sizes of the acquiring companies, the percentage change of ROE for the first quartile is found to get more negative for the larger companies. On the other hands, when investigation is done from the relative-size perspective, the percentage change of ROE in the first quarter is especially found to get more negative with decreasing the size difference between the combining companies. The results of multivariate regression did not show much significant relationship between percentage changes in performance and those independent variables. Yet, two significant findings are shown: the positive association between percentage change in stock returns for the first quarter after mergers and relative size of combining firms; the positive association between percentage change and stock returns in the first quarter after mergers and absolute size of the acquiring firms. The unclear results might be due to the large standard deviation and to an unsystematic relation between the dependent and the independent variables.
In the long-horizon event studies for post-merger performance, 74 samples are analyzed in total. Most AAR in the post-event windows is significantly negative, but AAR for the first quartile group was always smaller than other groups. ACAR is associated with the results of AAR, where only the first quartile group keeps relatively lower negative ACAR comparing to the rest groups. AEROA was significantly negative in the first three years after mergers for whole groups and became less negative afterwards AEROA for the fourth quartile group started to become positive, though not significantly, from the second year after mergers. The analysis results for AEROE are similar to AEROA. Most AEROAs are significantly negative during the post-event windows but for the fourth quartile groups. Although not significantly different from zero, most AEROE after the second year turned out to be positive.
6. D ISCUSSION OF F INDINGS
Chapter 6 focuses on the investigative results for the dissertation. The discussion is based on the regression analysis and event study results presented in the previous chapter. There are three sections in this chapter. The first section of 6.1 discusses the post-merger performance in the short-term and its connection to the relative firm size of combining firms in M&A deals. Relating the analysis in the previous chapter, the arguments in Chapter 3 are also to be discussed. The second section of 6.2 discusses the post-merger performance in the long-term and its connection to the relative firm size of combining firms in M&A deals. The summary of the chapter as a whole and a brief conclusion will be presented in the last section, 6.3.