Even though some research dedicated to the high-tech M&A performance currently exists, authors have not compared the cases when the high-tech company is acquired by another high-tech or non-high-tech firm. Based on the literature review, the difference between these two types of companies was shown, so we assume that the M&A performance would also differ.

Hypothesis 1: The high-tech M&A performance is different depending on the high-tech or non-high-tech type of acquirer.

Considering the arguments above, would be chosen the short-term event window approach for M&A performance valuation. The following reasons made this conclusion: in the research would be considered long period and a large number of companies and this approach allow to cope with drawbacks of the information system and problem of data availability. This method allows to decrease the influence of researcher's subjectivity, the anticipation of the market players will be considered. Also, it allows eliminating the impact of some external factors not connected with the M&A deal. As the time window would be chosen 3 days' time span (-1;+1): one day before the deal - one day after the deal following by Davis and Madura (2017). Several additional windows (-3; +3), (-5; +5), (-1; +3), and (-10; +10) would be used in order to check the results obtained for the first window. Multiplication of the company's beta coefficient and the return of the corresponding index would be used for normal return determination.

2.2 Determinants of M&A performance

The results of the M&A deals appear to be contradictory (King et al., 2008), so the identification of the detailed factors explaining the performance is crucial. Following Davis and Madura (2017), Alexandris et all. (2013) we choose the factors describing the companies participating in the deal and some external factors. Thus, we introduce the original classification of the characteristics contributing to the high-tech firms M&A performance:

· Micro-level characteristics (both acquirer and target);

· Characteristics of the deal;

· Macroeconomic level characteristic.

Consideration of the factors in these three groups allow reaching the goals of the paper. Firstly, it describes the participants of the deal from different perspectives, including the current at the time of the deal financial state and growth opportunities. It also includes the deal characteristics that influence the M&A performance and consider the external factor of macroeconomic conditions that also contributed to the M&A deal activity and outcome. Secondly, these factors would be instrumental for demonstration the internal difference between the high-tech industries based on the motives of M&A considered in the previous chapter.

For each variable would be indicated the name used in the following regression analysis. The “ac” prefix indicates the relation of the variable to the acquirer and “tg” prefix - relation to the target.

Micro-level characteristics

ROE

Return on equity is one of the most common characteristics of a company's performance. Its value equals the division of the net income on the size of equity. For the paper would be used the ratio in the year before the deal. This is true for all companies' characteristics that are considering in this research. ROE would be used to characterize the operational effectiveness of the acquirer and target. In the regression would be marked as “ac_roe” and “tg_roe”.

Baker, Dutta, Saadi, and Zhu (2012) found the negative influence of the acquirer operating efficiency to the M&A deal performance. Authors separately consider the acquirer's “motivation of hubris” and goal of “empire building” and proved that in this type of deals acquirer's ROE negatively influence the M&A performance in addition to the positive impact of target's ROE. The corresponding results were obtained by Davis and Madura (2017).

According to the Baker, Dutta, Saadi, and Zhu (2012), the higher rate of return on equity indicates the higher level of operation performance, that might be a result of effective management style or business processes, special market conditions or products' specific features. Insomuch as the companies possess unique qualities that evade the common formal numerical description, the ROE would be used as a universal performance indicator that describes the company's effectiveness. Thus, the factor of both companies' operation results might have an impact on the M&A performance.

In this paper, we base the reasoning on the idea that firms are aiming to increase its value, that, in general, correspond to the calculating cumulative abnormal return as a deal performance indicator. Consequently, henceforward we assume that by conducting the M&A deal companies also targeting to the value increase, thus they are pursuing the factors that increase the M&A performance and this persecution indicates the motives of the deals conduction. Also following the Alexandridis, Mavrovitis, and Travlos (2011) we assume the market efficiency, thus the market players also have conceived the motives of the firms, the success factors in the industry and potential of the future synergy.

Taking the mentioned above into consideration, following the paper of Stennek and Verboven (2000) about the companies in the computer and electrical equipment industry is pursuing the best practices of companies' performance are aiming to acquire high-performing company, thus:

Hypothesis 2: ROE of the target has a positive influence on the M&A performance of the acquirer in the computer and electrical equipment industry.

According to the Baker, Dutta, Saadi, and Zhu (2012), for the companies intend to build an empire the acquirer ROE is negatively correlated with the M&A performance, it might be anticipated that connection for automotive and aerospace industries.

Hypothesis 3: ROE of the acquirer and ROE of the target have accordingly negative and positive influence on the M&A performance for the acquirer in automotive and aerospace industries.

R&D

Several authors have already considered (Roller, Stennek, and Verboven, 2006; Ornaghi, 2009) the possibility of the connection between the M&A and the R&D activity, but the question of correlation is still open. In the researches, there are several approaches to the inclusion of the R&D expenditures in the model. Aw, B. Y., Roberts, M. J. and Xu, D.Y. (2008) use a dummy variable to identify whether the company conducts such activity and found that the presence of the R&D in the target company influence the M&A performance. This approach has a significant disadvantage - it allows only to include the fact of conducting R&D, but do not consider the scale of it, but the finding itself is important for M&A investigation and laid the groundwork for future research.

Stellner (2015) used the sample of 538 M&A deals of European companies from 2007 to 2013. In the research, author used the logarithm of absolute value of the R&D expenditures of the target and found a positive correlation of M&A performance and R&D expenses.

Another way of inclusion of this factor is the number of R&D expenditures divided by the total assets of the firm (Hatem, 2015). This approach has two positive aspects. On the one hand, it allows to consider the R&D expenditures relative to the firm size - determine the level of firms focuses on such activity. On the other hand, the relative measurements allow comparing firms with each other. However, this approach might not reflect the real state of the R&D involvedness of the company due to several reasons: firstly, the method of accounting and calculating the value of assets influence the results. Secondly, the amount of assets possessed by the companies for conducting operational activity in the different industries is varied. Thus, compatibility is illegible. Hatem (2015) conducted research based on the 136 French companies in 2007. The author found that the increase in the size of relative R&D expenditures of both companies does not significantly increase the CAR.

Another approach that was implemented by several authors (Grimpe & Hussinger, 2008, 2013; Ahuja & Katila, 2001; Sapienza, Parhankangas & Autio, 2004) is the measurement by the number of patents. The main positive feature of this approach is the relative simplicity of the data collection for developed capital markets. However, the method has several drawbacks (Shambaugh, Nunn & Portman, 2017). The patents could be the result of time-consuming researches and at the same time have no use in the operational activity of the firm. Shambaugh, Nunn, and Portman (2017) emphasize that some results of R&D activity could be saved at the know-how regime in order to elude the possibility of copying the results by competitors. Zhang, Wu, Zhang, and Lyu (2018) confirm that for high-tech industries creation and utilization of the patents is not a conventional process as it is for other industries. The new patent might include some modest alterations in the past findings and does not bring competitive advantages or be delayed for some time due to bureaucratic issues. This method is not suitable for high-tech companies. It worth to be noticed that despite the majority of the authors consider the R&D expenses level of the target company and in fewer researches this factor considered for an acquirer or both companies (Цberg and Leminen, 2017; Davis and Madura, 2017).

Coad and Rao (2008) found a positive correlation between the rate of R&D expenditures and the sales growth of tech firms. Authors emphasize the unique role of the R&D in high-tech company operations and prospects. R&D expenses connected with the future prospects of the firm, because it might become the base for product's development and creation, expansion to the new markets.

Ho, Tjahjapranata, and Yap (2006) also connected the rate of R&D expenses and the growth opportunities for the firm. Thus, considering the importance of the R&D factor and the fact that it was implemented in previous researches, it would be considered for both companies - acquirer and target. As a measurement technique would be used approach by Stellner (2015) and Davis and Madura (2017) - measurement by the natural logarithm of the rate of expenditures. In the regression this factor would be marked as “ac_rnd” and “tg_rnd”.

As we indicate in the first chapter of this paper - R&D might significantly influence the M&A deal initial motive and performance. While studying the high-tech M&A deals, one might detach two main motives connected to R&D. The first one is the acquisition of the companies with high R&D expense level in order not to conduct the R&D using acquirer's resources - in this types of deal, significant influence has R&D of the target, but not the acquirer's. According to the Goedhart, Koller, Wessels, (2010), these types of deals is pertaining to the software industry.

Hypothesis 4: R&D of the target has a positive influence on the M&A performance of the acquirer in the software industry.

The other type of deals is inhering to the medical technology and biotech, computer and electrical equipment industries. The companies in this sector aspire to obtain the critical mass of the conducted R&D in order to achieve leadership on the market. Thus, the R&D level of the target and acquirer influence M&A performance.

Hypothesis 5: R&D of the target and acquirer has a positive influence on the M&A performance of the acquirer in the medical and biotech industry.

Hypothesis 6: R&D of the target and acquirer has a positive influence on the M&A performance of the acquirer in the computer and electrical equipment industry.

Cash Flow

Powell (1997) examined the impact of cash flow as a motive to M&A. The cash flow influences the deal performance and increase the possibility for the target to be acquired for high-tech firms. The author claims that higher targets' cash flow allows acquirers to imbibe the acquisition cost that decreases the negative effect on the bidder of cash shortage because of the deal.

In the paper as an indicator of cash flow would be used cash flow from continuing operations in the financial year before the deal. Its value represents the amount of cash generated by the firm from the main business line as revenue less cost of goods sold, capital expenditures, taxation and interest paid and amortization, depreciation and with tax shield added. Besides the importance for the M&A deal, this factor is included in the paper because it represents the scale of operations of the company it might be applied for the checking of the hypotheses concerning the “scale acquisition” for some industries. In the regression would be marked as “ac_cfo” and “tg_cfo”.

This motive was indicated as one of the most important for the communication, medical equipment/biotech and software industries. The significance of this factor explained by the acquirers aiming to rapidly capture the market, as it stated for the software industry, or, as for other industries, obtain the economy of scale of operations to decrease average costs or implement operations in the new segment with the lower costs. Insomuch the acquirers are seeking for the scale of operation that helps them to strengthen on the market; we might assume the positive relation between the scale of operation for the target company and market reaction to the deal, CAR.

Hypothesis 7: Cash flow of the target has a positive influence on the M&A performance of the acquirer in the communication industry.

Hypothesis 8: Cash flow of the target has a positive influence on the M&A performance of the acquirer in the software industry.

Hypothesis 9: Cash flow of the target has a positive influence on the M&A performance of the acquirer in the medical technology and biotech industry.

Growth options

Berk, Green, and Naik (1999) presented the model of general equilibrium for assets in place and growth options. Authors focused on the research on the company's risk, that is particularly important for the high-tech companies, and found a positive correlation between the firm performance and the investment opportunities employed by it. Authors claim that the growth option determines a major part of the firm value. This idea was also developed in the Berk et al. (1999) research. Authors suppose that the firm value is the sum of the assets in place (represented by the book value of assets contributed to the current projects of the firm) value and the growth options. Authors indicate that the difference in the market capitalization value of the company and assets in place is explained by the ability of the company to transfer the R&D expenses into company's value and according to increase in intangible assets. In the regression growth options variable would be marked as “ac_go” and “tg_go”.

Growth options model was implemented several times with small alterations by a different group of authors: Andres-Alonso, Azofra-Palenzuela, and Fuente-Herrero (2006) and Bernardo, Chowdhry, and Goyal (2007). Cao, Simin, and Zhao (2008) investigated five different approaches to the growth options determination and found that the relation of the market and book value of the firm possess the highest level of significance among the others.

Another approach to this variable was introduced by Davis and Madura (2017). The novelty of the approach constituted in the method of growth option calculation. It based on the approach by Berk, Green, and Naik (1999), who found the present value of growth options and after divide the number on the amount of R&D stock. Authors confirm that this way of calculation shows the ability of the company from the market perspective to transform the investments in research and development, the future possibilities of the company, to the real market value.

Davis and Madura (2017) affirm that the companies with higher growth options are better in creating synergy as an acquirer, that is also confirmed by the result of the research, in which authors found a significant positive correlation between the acquirer growth options and M&A performance.

In the paper of Trigeorgis and Lambertides (2014) authors asserts that the higher growth options of the target lead to the lower M&A performance, due to the increase in the premiums paid by the acquirer. But while considering the acquirer's growth options, researchers indicate the positive influence of it on the M&A outcome.

Insomuch as on the M&A performance influence growth option factors for the acquirer and the target, both of them would be included into consideration. The investigating deal is referring to the high-tech, we assume that for the whole pool of deals, both variables could be significant: for the acquirer with the positive sign and for the target with a minus. However, due to the difference between the deals, industries and companies' characteristics, this assumption might be violated.

The growth options might influence to the M&A performance differently. On the one hand, targets growth option leads to the higher premium to the share price, that might be negatively perceived by the market and CAR of the acquirer decreases. On the other hand, the ability of the company to transfer the R&D investments to the increase in the market capitalization and potential synergy of these processes for target might increase the deal performance.

Considering the high-tech M&A motives, we indicated that this factor is important for the two industries: medical technology/biotech and software. In the first case, we assume the higher M&A performance for the acquirer with a bigger scale of growth options and abilities of value transfer.

Hypothesis 10: Growth options of the acquirer has a positive influence on the M&A performance of the acquirer in the medical technology and biotech industry.

In the software industry we assume that the higher premiums paid for the targets with greater growth options negatively affect the deal performance.

Hypothesis 11: Growth options of the target has a negative influence on the M&A performance of the acquirer in the software industry.

The array of variables presented above would be used in the paper in order to check the hypotheses and make the conclusions. The selection of the variables is explained by the fact that they form the different angles describe the target and acquirer, consistent with the previously made researches and also acknowledge the specific features of the high-tech companies and M&A deals.

The following two groups of variables are aiming to describe the deal and the environment of the deal, that according to the previous researches significantly influence the M&A performance.

Deal characteristics

Payment regime

The theme of connection between the abnormal returns and method of payment was investigated several times (Harford & Uysal, 2014; Uysal, 2011; Dong et al. 2006; Leeth and Borg, 2000). The fact that should be mentioned is that the results obtained by different researches are accordant.

Amihud et al. (1990) indicate that the cash payment of the deal leads to higher performance. This conclusion was supported in some researches, for example, by Davis and Madura (2017) and Alexandridis, Mavrovitis, and Travlos (2011). Alexandridis et al. (2013) found that cash deal leads to the fairer valuation of the target and according to decrease in the premiums paid in the deal. As a result, it might affect the M&A performance by increasing or decreasing in the lesser extent the CAR of the acquirer.

Thus, several authors have proved that the payment regime influence the M&A performance and this indicator should also be considered in the paper as a control variable. In the regressions this factor would be marked as “cash”.

Relatedness

The relatedness of the companies conducted the M&A was initially brought under consideration by Jaffe (1986). Hussinger (2010) found that the lower technological distance between the companies, the higher M&A performance. The author indicates it based on the idea that companies in similar industry are better to create synergy, thus the market considers the deal as a perspective. This idea was also proved by Flanagan and O'Shaughnessy (2003) who claims that industry relatedness between target and acquirer leads to better M&A performance due to easier integration and better synergy. Kohers and Kohers (2001) also offer confirmation of these findings on the high-tech firms sample. In the regressions this factor would be marked as “relatedness”.

Deal attitude

Another issue that is sometimes applied by authors while studying M&A performance is the deal attitude: friendly or hostile. Powell (1997) indicates that during the hostile deals acquirers pursue the opportunity of targets stocks price growth, so the premiums paid during the deal accordingly increase and the M&A performance for the acquirer decrease. These findings also proved by Alexandridis, Mavrovitis, and Travlos (2011). Authors of both researches indicate that this is not typical for friendly acquisition. Also, researches assert that the potential synergy for the friendly acquisitions is higher, that appears to the better market reaction and accordingly better M&A performance. In the regressions this factor would be marked as “attitude”.

Relative size

Madura and Ngo (2008) and Jansen, Sanning, and Stuart (2013) indicate the connection of the M&A performance and the size of the deal. For the relative size determination authors use the division of the target's book value of assets to the acquirer's book value of assets. Researches assert that the relatively large deals have a significant negative influence on the acquirer's M&A performance and this effect decrease with the reduction of the relative size of the deal. The significance of this factor and its negative influence on the high-tech M&A performance was also highlighted by Davis and Madura (2017). In the regressions this factor would be marked as “rel_size”.

Macroeconomic factor

Market conditions

According to the Dittmar and Dittmar (2008), the M&A performance ceteris paribus varies due to the market conditions. Due to the economic expansion investors tend to value companies shares higher and tend to estimate the company's prospects more positively. Authors found that M&A activity is positively related to the change in the gross domestic product. Harford (2005) also confirm this finding and demonstrate that the economic shocks, that could be controlled in the research by GDP change influence the willingness of the companies to acquire other market players. In the regressions this factor would be marked as “gdp_usa100”.

Considering all the previously mentioned variables, the M&A performance measurement method, the base regression that would be used for hypotheses testing is the following:



In conclusion, in this chapter were developed main hypotheses (summarized in the Table 7 in supplement) that would be checked by the regression analysis on the empirical data in the next part of the research. The methodology of the M&A performance determination were described. The array of indicators that potentially specifically influence the high-tech M&A performance and explain the impact of the acquirer's motives for the particular industries was selected: R&D expenses, growth options, return on equity and cash flows from continuing operations for the acquirer and the target. Several control variables that, according to the previous findings, also influence the deal outcome was chosen. They are divided into two groups. First includes the deal characteristics: method of payments, the relatedness of the companies, relative size, and deal attitude. The second group - is the control factor of the common market activity - economic expansion.

Chapter 3. Methodology and results

3.1 Sample description and methodology

The sample used for the study was obtained from the Capital IQ system by means of market screening instruments. As a benchmark for the developed capital markets was used the USA market for the following reasons:

· The homogeneity of the market is higher than for the European Union, in which are included most of the other developed countries. The homogeneity of the US market could be explained by the strong influence of the common national legislation on the state's legal frame;

· According to the Hofstede and McCrae (2004), the differences within one country in company's corporate culture is drastically lower, thus on the M&A performance this factor influence in the lesser extent and allow to focus the research on the industry specificity;

· Companies under consideration operate in similar conditions, thus the impact of the external factors resemble and relatively easy to indicate and accommodate;

· The market is active and the amount of M&A is high, that being combined with the strict financial reporting procedure allow obtaining high-quality data with uniform indexes and greater sample size.

In order to implement all the positive aspects mentioned above we would consider deals than the acquirer and the target both US incorporated companies.

In each observation, the target company belongs to the high-tech industries. Following the paper of Bowen, Davis, and Matsumoto (2005) we would indicate the high-tech industry by the SIC (Standard Industrial Classification) codes:

High-tech industry	SIC Codes
Software	737
Medical technologies (incl. drugs)	283, 382, 873
Communications	366, 481, 489
Computer equipment	357
Electrical equipment	360 - 365, 367
Automotive and aerospace	37

Sources: Bowen, Davis, and Matsumoto (2005), author's analysis

Authors conducting the researches dedicated to the high-tech companies have not indicated the common way of such industries determination. However, approaches have insignificant divergence and the method presented above includes the codes that is the majority of the authors choice as high-tech industries (Bowen, Davis, and Matsumoto, 2005; Clem, Cowen and Jeffrey, 2004; Dowdell and Press, 2004). Companies were determined by the main SIC code of their operation activity. Other companies which are not operating in these industries in the paper called non-high-tech.

Sample

The sample consists of 327 deal that was conducted in the USA between the companies incorporated in the USA in the period between 01.01.2000 and 28.02.2019. All the companies in the sample are public. Target companies sample consists of only high-tech companies, that is vital in order to research the high-tech M&A deals performance. Acquirers in the sample belong to the high-tech and non-high-tech industries. The deal implicates the acquisition of more than 50% of the shares of the target of the merger of the companies. All the financial data are taken as the last financial result of the company in the year before the deal. The structure of the acquirers: 86% of the companies are related to the high-tech industries, 14% - to the non-high-tech industries.



Picture 1. The distribution of the acquirers by industries

Source: author's analysis

Among the high-tech industries, most of the deals were conducted by the acquirers from medical technology and biotech (27%), computer and electrical equipment (21%) and software (28%) industries. The number of deals in the communications and automotive and aerospace industries is lower.

Another important factor of the deal in terms of industry specificity is the relatedness of the field of operating activities between the target and acquirer. The relatedness is determined by the sic codes of the companies' activity. If the first two digits are similar firms are considered as related and not related if otherwise. In the sample, only 70% of the deals were described as “related” that indicates the presence of cross-industrial deals. The superiority of the high-tech acquirers might be explained by the fact that this types of companies actively participate in the M&A deals and natural conservatism and unwillingness of the majority of the firms to enter the unrelated industry.

The majority of the deal is friendly, the share of the hostile transaction account for approximately 2%. As the payment regime the dominated method is the mix of shares and cash. The only cash paid 21% of the whole pool of the deals, and approximately 90% of them paid 50% and more percent of the value by cash.

Table 2. Descriptive statistics on targets and acquirers

	Targets	Acquirers
		High-Tech	Non-High-Tech
	Q1	Median	Q3	Q1	Median	Q3	Q1	Median	Q3
ROE	2.7%	10.6%	32.9%	6.1%	25.4%	46.9%	4.1%	19.3%	45.2%
Growth	1.1	8.1	56.2	0.89	7.9	64.8	0.3	1.1	10.6
CFO (mln usd)	0.9	16.8	189.4	7.8	216	3770.4	8.5	18.8	433.9
RnD(mln usd)	2.3	9.4	54	2.9	24.4	180.4	3.0	8.2	84.1

Source: author's calculations

Table 2 represents the descriptive statistic of the sample. The values of cash flow from continuing operations and RnD expenses are taken as absolute numbers, however, further for the regression analysis would be taken the logarithm of them. The low growth, CFO, and RnD values could be explained by the absence of the values in some observations in the sample.

As it might be seen on the descriptive statistic the high-tech acquirers who obtain high-tech targets possess a higher return on equity that indicates the higher efficiency of the operational activity conducted by such firms. However, among non-high-tech acquirers incredibly efficient companies (with ROE of 47% and higher) occupy a quarter of the sample. It might indicate that better performance firms operating in non-high-tech industries are expanding their operations in the high-tech field. Targets in the sample show the relatively low operating efficiency, but while considering it is necessary to take into account the specificity of these types of firms. The higher amount of the intangible assets that being amortized and accordingly included in the cost of goods sold or services provided strongly influence the performance result that is calculated as return on equity. This is also fair for the high-tech acquirers which perform in general weaker than companies in the non-high-tech sphere.

The most drastically distinct feature of these companies is growth options. The lowest level of this factor show companies from the non-high-tech industries that supports the idea of the significant difference of the high-tech companies and the necessity of its separate investigation as it stated in the first chapter. Comparing the two other groups, it is necessary to mention that the first quartile and the median indicators are higher for the high-tech targets. This demonstrates the prospect of such companies more effectively transfer the advantage of the conducted R&D into the company market capitalization. High-tech acquirers are not as capable in value creating from the R&D as high-tech targets but excel the non-high-tech firms. The reason for the not as good value transfer might serve the fact that these companies are noticeably bigger than the targets.

The cash flow from operations indicator represents the scale of operation of the considered company. The smallest values pertain to the target companies, however there are array of the companies in the last quartile which scale of operations is comparable to the one of the acquirers, that indicates the possibility to prove the hypotheses about acquiring the scale of operations for communication, medical/biotech and software industries.

The interesting feature is the small difference in R&D expenses amount for high-tech and non-high-tech acquirers. Companies in the high-tech sphere are usually conducting the researches and invest heavily in them, that is the distinctive feature of the industry. As it might be seen from the descriptive statistic, the companies that are acquiring the high-tech company are as much as interested in R&D as high-tech ones. While considering relatively to the scale of operation of the company, we might compose the profile of the non-high-tech company acquiring the high-tech firm. The most important feature of it would be the high intensity of R&D, that makes these companies are similar to the high-tech. The amount of R&D expenditure of the targets is not as high in absolute measurement while being compared to the acquirers, but relatively to the scale of operation, it constituted almost a quarter of the cash flow and additionally indicated the vital character of this activity to the high-tech company.

M&A performance determination

For M&A performance determination is used Cumulative abnormal return. As a real return for every acquirer's share were calculated the percentage difference for each day relative to the previous trading day. As a normal return was taken the index approach. The value of normal return equals the index percent changes for the same time interval, multiplied by the beta of the analyzed company's shares. For high-tech acquirers were taken NASDAQ (National Association of Securities Dealers Automated Quotation) index because it specialized specifically on the high-tech companies. For the non-high-tech acquirers were implemented NYSE Composite index and the similar methodology was used.

Table 3. Cumulative abnormal returns by acquirer type *P-values are significant at 10% level.

**P-values are significant at 5% level.

***P-values are significant at 1% level.

Event - window	Acquirer type
	High-Tech	Non-high-tech
	CAR	St. error	CAR	St. error
(-1;+1)	6.27%*	0.03759	1.233%*	0.0073
(-1;+3)	5.75%**	0.02742	1.12%*	0.00675
(-3;+3)	5.61%*	0.03121	0.89%	0.00685
(-5;+5)	5.15%*	0.02701	0.78%	0.00709
(-10;+10)	4.42%*	0.02426	0.25%	0.00226

Source: author's calculations

In Table 3 presented the results obtained by the acquirer during the M&A deal. M&A performance of the high-tech acquirer is significantly higher and increase with the event-window enlargement. For the non-high-tech acquirers, the results obtained are significantly different. The M&A performance is hovering around zero, but significantly positive. Thus, the market is more positively reacts to high-tech mergers when the acquirer is a high-tech firm. This outcome might be a consequence of the following reasons: firstly, according to the Alexandridis, Mavrovitis, and Travlos, 2011 and Eliasson, Hansson, and Lindvert, 2017) M&A deals with the companies from the different industries ceteris paribus perform worse than the transaction between related in the industry specificity. Secondly, the measurement of the M&A performance of these companies is conducted based on the different market indexes, that is a necessary assumption due to company's specificity.

Therefore, the first hypothesis of the difference between the M&A performance of the high-tech and non-high-tech acquirers are proved due to the significant distinction between the cumulative abnormal returns of the deal obtained by the firms, that represented in Table 3 above.

Table 4. M&A performance by the acquirer's industry *P-values are significant at 10% level.

**P-values are significant at 5% level.

***P-values are significant at 1% level.

	CAR(-1;+1)	CAR(-1;+3)	CAR(-3;+3)	CAR(-5;+5)	CAR(-10;+10)
Medical technology	1.56%***	5.02%***	5.60%**	2.57%**	5.45%**
Computer equipment and electrical equipment	4.42%**	5.27%*	3.19%	5.15%	9.55%*
Automotive, aerospace	5.04%*	3.38%*	3.70%*	7.80%	3.97%*
Communications	1.68%***	1.54%*	4.09%	3.82%*	4.40%*
Software	3.80%***	5.32%**	5.71%**	4.67%**	4.06%**

Source: author's analysis

Within the group of high-tech companies, M&A deals are viewed by the market differently. Industries which constituted the high-tech sphere are distinct, thus the performance of the deals is also dissimilar.

As evidently revealed by the table, the results are positive and quite consistent within the industries. Acquirers from the automotive/aerospace and computer/electrical equipment industries are in general perform better in M&A than other companies. A little lower is M&A performance for the companies in the software industry.

We calculated M&A performance measurements for 3, 5, 7, 11, and 21 day's event-windows as presented in the Table 4. This analysis was conducted in order to check the results and prove that the choice of the particular short event window will not lead to the significant results biases. On the result, when the long event-window is considered, have influence the market perturbations that could not be connected directly to the company. Also, the biasness could cause other announcements and circumstances surrounding the deal. In the following research the regression analysis would be made based on the CAR with 3 day's event window (1 day before the deal - 1 day after the deal).

Methodology for variables

In order to smooth and accommodate the non-linear influence of the factors, instead of the absolute empirical values is used the logarithm of these numbers. This approach is applied in the paper for the following variables: cash flow from continuing operations and R&D expenses and growth options. ROE and the relative size would be used in the decimal order. Cash payment would be represented as a dummy variable, equals to 1 if the deal was paid only by cash and 0 otherwise. Deal attitude is characterizing by the dummy variable, that constituted 1 if the deal was friendly and 0 if it was hostile. Relatedness of the companies also described by the dummy variable, that account for 1 in case of related companies participate in the deal and 0 otherwise. Relatedness is determined based on the main SIC code of the companies' operational activity. If the first two digits of the relevant codes are equal, then the dummy equals one, 0 - otherwise.

In order to accommodate the macroeconomic factor external for both firms, the GDP variable would be included in the regression. This factor represents the increase in economic activity of the USA, that, consequently influence the M&A activity. The variable was created in the following way: the amount of GDP (Gross Domestic Product) of the US in 2000 was taken as a basis, so the variable for this year is equal to 1. For the following years variable is represented by the division of the GDP of the considered year to the basis. For the deals in 2019 was taken the forecast of the Congressional Budget Office.

In this section we describe the sample that would be used in the following research, the method that is was obtained and the M&A performance calculation, the method of variables values application and the common methodology of the research.

3.2 Regression analysis and results

In this section, we present the results of empirical data regression analysis based on the methodology described in the previous section. As a result of the analysis would be checked the hypotheses made in Chapter 2.

As depended variable in every regression is used M&A performance. As long as we do not observe the drastic difference between the results obtained by the calculation of the M&A performance on distinct event windows, and as it made by Rheґaume and Bhabra (2008), one event-window would be used as the benchmark. Following Davis and Madura (2017), we apply the (-1; +1) 3-days event window. As a method of regression analysis would be used ordinary least squares (OLS) with robust standard errors in order to obtain consistent results under heteroscedasticity

The first regression allows to compare the factors that influence the M&A performance for high-tech and non-high-tech acquirers.

Table 5. Regression analysis by the type of acquirer *P-values are significant at 10% level.

**P-values are significant at 5% level.

***P-values are significant at 1% level.

VARIABLES	High-tech	Non-High-Tech
	CAR (-1;+1)	CAR (-1;+1)
ac_roe	-0.000145	-0.00211**
	(0.000641)	(0.000936)
ac_go	0.00797**	0.00221
	(0.00395)	(0.00432)
ac_cf	-0.000273	0.0111*
	(0.00313)	(0.00659)
ac_rnd	0.00787	0.00618*
	(0.00522)	(0.00307)
tg_roe	1.25e-05	0.0159*
	(2.46e-05)	(0.00912)
tg_go	-0.00266	0.00318
	(0.00582)	(0.00689)
tg_cf	-0.00322	-0.00219
	(0.00335)	(0.00464)
tg_rnd	0.00891***	0.0169**
	(0.00326)	(0.00704)
all_cash	-0.00574	0.00954*
	(0.00619)	(0.00486)
attitude	0.00556	0.0351**
	(0.0163)	(0.0170)
rel_size	-0.00176*	-0.00814***
	(0.000931)	(0.00245)
actg_related	0.000655
	(0.00738)
gdp_usa100	0.1028	0.0667*
	(0.0890)	(0.0365)
Constant	-0.0346	-0.144**
	(0.0665)	(0.0707)
Observations	281	46
R-squared	0.0262	0.563

Source: author's calculations

As revealed by Table 5 above, on the deal performance for the high-tech firms influence significantly positive the amount of research and development expenses of the target company. The negative effect has a relative size of the target to the acquirer. Competitiveness of the deal does not have a significant impact on the deal results, despite the fact that this was anticipated, based on the previous researches. The reason why this factor does not have significant impact might be in the relatively small amount of the hostile deals in the sample and the impact of the factor is not fully reflected. According to the literature review, we anticipate that R&D of the acquirer would be a significant factor, that was not proved by the regression. Acquirer's growth options significant and has positive sign.

The results of the second regression where the acquirer in the non-high-tech company indicates that different factors impact M&A performance. First of them, that as against the previous case return on equity of the target is positive and significant. It indicates that for this type of acquirer is important the effectiveness of the target company management. This fact might be explained by the lack or insufficient level of expertise in high-tech company management possessed by the acquirer, so in order to create synergy the target company should initially be high-performing.

Negative significant variable for ROE of the acquirer and positive significant target's ROE combined indicate the importance of empire building motive in high-tech M&A for non-high-tech acquirers, as it was also shown in the literature review for automotive and aerospace industry.

As it was mentioned before, non-high-tech companies that are considered in the sample partly possess the features of the high-tech firms - higher level of R&D expenses. In this type of deal R&D expenses of target and acquirer influence the prospect synergy and provide the acquirer company with better market response in terms of share price. Also, in the second regression significant variable is the GDP. It indicates the higher dependency of the M&A deals performance from the economic environment when the acquirer is a non-high-tech company.

Obtained regressions were tested on the overall significance. According to the results of the F-test (Fisher test) commonly factors are not equal to zero. R square statistics also indicate that regressions possess sufficient quality and explanatory power for the model based on the empirical data. Since the data is empirical and the results of the observations are attributed to the one specific company we might assume the presence of the multicollinearity, despite the fact that the variables were selected in the way to eliminate intercrossing the parameters. Thus, the quality of the obtained models is sufficient to make conclusions. As it revealed by the regressions above, the M&A performance of the acquirer is depended from the different factors, which varies due to the industry specificity of the acquirer.

The next step of the research is to examine the difference between the high-tech industries. To accomplish that we conduct the regression analysis similar to the previous, but additionally, we include the dummy variables that indicate the industry of the acquirer. To consider every industry independently we create regression, which content five dummy variables for particular industry: medical technology and biotech, communication, software, computer and electrical equipment, automotive and aerospace.

As it revealed by the Table 8 in the supplement of the research, the dummy variables that are significant and indicate that the industry-specificity is presented in the medical technology and biotech, software, and computer equipment and electrical equipment industries. Unfortunately, due to the insignificance of other variables, in the paper, we would not further consider the two other industries: automotive and aerospace and communication. The insignificance of the factors could be explained by the low number of the deals in these industries in the sample and relatively low M&A activity in this spheres of the economy.

Table 6. Regression analysis for the industries *P-values are significant at 10% level.

**P-values are significant at 5% level.

***P-values are significant at 1% level.

VARIABLES	Medical technology	Software	Computer and electrical equipment
	CAR (-1;+1)	CAR (-1;+1)	CAR (-1;+1)
ac_roe	-0.000379	0.000270	-0.000865
	(0.000647)	(0.000293)	(0.00166)
ac_go	0.00704**	0.00286***	0.00255
	(0.00352)	(0.000817)	(0.00318)
ac_cf	0.00496*	-0.00300***	-0.00583
	(0.00259)	(0.000929)	(0.00504)
ac_rnd	0.00604**	-0.000965	0.0155**
	(0.00252)	(0.00157)	(0.00663)
tg_roe	4.04e-05*	9.99e-05	0.0121**
	(2.12e-05)	(0.000312)	(0.00559)
tg_go	0.00455	-0.00387**	0.0109**
	(0.00293)	(0.00152)	(0.00436)
tg_cf	0.00540**	0.00086*	-0.00402
	(0.00230)	(0.00119)	(0.00420)
tg_rnd	0.00534**	0.00319**	0.0106**
	(0.00222)	(0.00127)	(0.00469)
all_cash	0.00605**	0.00276*	0.0164**
	(0.00294)	(0.00148)	(0.00681)
rel_size	-0.00176	-0.000688**	-0.00706*
	(0.00113)	(0.000324)	(0.00403)
actg_related	0.00771*	0.00644**	0.000600
	(0.00423)	(0.00254)	(0.0155)
gdp_usa100	0.0543	-0.00579	-0.0133
	(0.0437)	(0.0157)	(0.0434)
attitude		-0.00151	0.00444
		(0.00241)	(0.00717)
Constant	-0.0432	0.0647***	0.0448
	(0.0520)	(0.0178)	(0.0677)
Observations	88	92	69
R-squared	0.451	0.436	0.372

Source: author's calculations

In Table 7 presented results of the three regressions for the industries. The first one is regression for the acquirers operating in the medical technology industry, that includes biotech, drugs and medical equipment development and manufacturing and other associated activities. According to the table above we conclude that on the M&A performance have impact research and development expenses of both sides of the deal. Hypotheses 5 implicate the joint influence of the factors. To prove it we conduct the likelihood ratio test for nested models. We check the joint insignificance of the variables - target's and acquirer's R&D expenses. The result is presented on Picture 2 in supplement. The test's statistic is higher than the critical value, thus the initial specification of the model is correct and target's and acquirer's R&D expenses have positive joint influence on M&A performance of the acquirer in medical and biotech industry. Thus R&D synergy creation and willingness of the acquirer company to obtain a new level of R&D scope in order to take advantage from the R&D economy of scale as a motive of M&A deal is valid and important for the medical technology industry.

As it was shown in the literature review, authors tend to suppose that acquirers in the medical industry are motivated by the acquisition of scale of operation not only in R&D sphere. We observe the positive significant influence of the target's scale of operations on the M&A performance, the variable for the acquirer is not significant. Thus, by means of the M&A companies are intending to capture new markets and customers and not to increase the efficiency of operations by the economy of scale.

Also, an important factor is tending to be the growth option of the acquirer. In the regression it is significant and has positive sign. It means that the higher ability of the company to transform the R&D expenses into the market firm's value, the more positive market players evaluate the combined company's potential. Significance of this factor proves the motive of acquisition with aim of more efficient realization of the target's R&D potential to the market capitalization in medical and biotech industry.

The second regression in Table 7 represents the analysis of the M&A deals when the acquirer in the software industry. For software acquirers M&A performance the important factor is the amount of target's R&D expenses. R&D could be implemented in the new or existing products and services and might contribute to the acquirers' market position improvements hereafter. This conclusion also explains the negative significant influence of the target's growth options. Companies in the software industry are in general conduct less fundamental research than, for example, firms in medical industry. Thus, the results of R&D more frequently transform into the market products, increase the sales, scale of operation and, as a result, rise to market capitalization of the company. This conclusion is applicable to considered sample, the share of the deals with target and acquirer from the software sphere constituted 85% of all deals in this industry. Thus, software acquirers are motivated to obtain the company with high amount of R&D expenses which potential was not fully realized in market operation.

Another important motive of M&A in software industry is scale accumulation. The increase in scale of operation through M&A allows to obtain the benefits from economy of scale and to enter new market or get access to new clusters of customers. Due to importance of this factor market reacts to this deals positively and target's free cash flow positively influence the M&A performance of the target.

Last regression describes the factors that have an impact on the M&A performance of the acquires in the computer and electrical equipment industry. The significance of the variables legitimates the approach when we combined these industries due to the similar reason for the companies to conduct the deals. The first significant factor is the return on equity of the target that makes this deals relatively similar to the ones where the acquirer is non-high-tech firm.