COMPARISON OF PREMIUMS OF CHINESE AND EUROPEAN COMPANIES IN MERGERS AND ACQUISITIONS IN EUROPE

!e goal of the study is to analyse whether Chinese Mainland companies were paying higher premiums for similar target companies in Europe in comparison with Europe-based companies during the period of 2000-2013. To determine the di"erence between premiums paid by Chinese and European companies, two samples were analysed: one represented all Chinese acquisitions in Europe which matched the set criteria, another included comparable European targets acquired by European companies. Quantitative research methods were applied to measure statistical di"erence in the premiums of the two samples. Results of the research indicate that the average premiums paid by Chinese companies are double the size of the European acquisition premiums for the similar target company in Europe. Signi#cant evidence suggests that premiums of Chinese and European companies for similar targets in Europe are not equal.


Introduction
In this global economy, companies are looking for possibilities to expand their businesses to foreign markets and one of the possible ways to do this is through mergers and acquisitions (M&A). e total value of M&A in 2013 was US$ 2215bn, with US$ 774.4bn coming from cross-border deals. In comparison with 2012, the market was slightly contracting in the USA and Europe, but growing very fast in Asia.
China as one of the booming emerging markets is showing a growing interest in cross-border M&A. e volume of Chinese outbound M&A deals rose by 40-50 times since the year 2000, and currently China is one of the leading countries in Asia by this parameter. Chinese companies' appetite is especially growing in developed countries -Europe, the USA and Australia.
Most of the research on M&A is focused on motives and value creation, thus these topics are quite well covered. However, premiums paid in M&A transactions and premium drivers seem to be still new and uncovered phenomena. Researchers found out that acquiring rms from emerging economies compared to those from developed economies have a tendency to o er higher premiums in order to acquire assets in developed countries. However, there is no unanimous opinion among researchers whether premiums, in particular paid by Chinese acquirers, can be justi ed. erefore, this paper analyses premiums paid by Chinese acquirers in M&A transactions in Europe.
e study aims to investigate whether Chinese Mainland companies in comparison with Europe-based companies were paying higher premiums for the acquisition of similar target companies in Europe during the period of 2000-2013. It reviews the previous research carried out on the topic of M&A premiums, analyses Chinese M&A in Europe, compares Chinese and European premiums paid for similar targets in Europe and the factors which might have in uenced the sizes of these premiums, as well as identi es the main M&A motives and premium drivers for Chinese companies.
Quantitative research methods are used to examine the di erence between the premiums paid by Chinese and European companies during M&A transactions in Europe. At rst, the sample of Chinese deals is determined and for each Chinese deal a comparable set of European deals is identi ed. en the comparison between premiums paid by China-based and Europe-based companies for similar targets in Europe is made. Other factors which might have in uenced the di erence in sizes of premiums paid by Chinese and European companies are identi ed and examined as well.
is paper is organised as follows: the rst section reviews the literature on the topic of M&A, in particular premiums and peculiarities of Chinese deals. Next, the methods to be applied are indicated and a description of samples is provided. In the third section, a comparison between the premiums paid by Chinese and European acquirers for similar targets in Europe is presented. e in uence of premium determinants on Chinese and European acquisition premiums are also described in the third section. In the last section, conclusions are made and directions for future research on the topic are provided.

eoretical Research on M&A and their Premiums
Companies are always looking for new possibilities to expand, and one of the ways to enter a new market is through M&A. Results from the KPMG International (2013) research revealed that executives prefer acquisition strategy among others (e.g., joint venture, green eld investment, alliance) due to the speed of entering the market and observed excellent opportunity. M&A strategy is believed to be more advantageous if the rm is seeking to enter a desired market where well-established incumbent enterprises exist and where global competitors are a empting to move (Wang, 2009). e Schoenberg classi cation model does not cover well the motives that are interrelated with more than one type of motives. Besides, external motives are hardly taken into consideration. Researchers argue that technological, economic and regulatory shocks (Mitchell & Mulherin, 1996), nancial variables (Giovanni, 2005), customers and suppliers (Ahern & Harford, 2010) a ect M&A signi cantly. erefore the authors of this paper present an extended model of classi cation of the motives for M&A (Figure 1).

FIG. 1. e extended Schoenberg model for classi cation of motives for M&A
Source: Compiled by the authors based on R. Schoenberg (2006) According to the surveys, strategic motives are the most important ones. KPMG International research (2013) suggests that the most popular rationales for acquisitions are consolidation and extension. Other 2 common motives are nancial e ciency and investment opportunity. However, the results of the study are based on the surveys of the management of the merging companies. erefore it is hard to measure the impact of managers' self-interest due to managers' reluctance to admit and share discrete information. Wang and Moini (2012) investigated Danish rms and obtained the results similar to KPMG International (2013); they found that two main motives are strategic ones -extension and capabilities. Financial e ciency took the third place. Both Wang and Moini (2012) and the KPMG International (2013) studies showed that managerial motives are not among the most popular ones. e acquisition motives of Chinese companies are similar to the motives of any other acquirer. Boateng, Qian and Tianl (2008) research results were consistent with KPMG International (2013) and Wang and Moini (2012) results and showed that strategic motives for Chinese M&A are the main ones encouraging M&A deals. Detailed results of the study are depicted in Table 1. Source: Boateng, Qian and Tianl (2008).
Acquirers expect to increase their market share, enter new market or gain knowhow. Financial motives, such as investment opportunity or economies of scale are also a popular explanation for M&A. However, as all studies were based on surveys of the managers, the underlying motives for transactions such as managerial hubris are hardly explored. e concept of value and price in M&A. Value and price in M&A are frequently mixed terms. According to Price Waterhouse Coopers (2013), value is the individualistic perception of the worth of a company which is di erent for di erent buyers and sellers. Fernandez (2004) distinguished six di erent types of valuation methods: balance sheet based, income statement based, mixed, cash ow discounting, value creation and option based methods. All of these methods have their advantages and disadvantages. e majority of authors stress the importance of discounted cash ow model (Damodoran, 2011;Fernandez, 2004). According to Imam, Barker and Clubb (2008), discounted cash ow model is most frequently used by buy-side nancial advisers. On the other hand, many authors admit that it is impossible to calculate the 'correct' value because in discounted cash ow approach a lot of estimations and assumptions have to be used (Damodaran, 2011). Additionally, this method is complicated, time consuming, subjective and requires inside information (Havnaer, 2012).
Relative valuation is well appraised by many authors as it represents not only the value of the target company but also the condition of the whole market (Damodoran, 2011;Fernandez, 2004). With the help of various market multiples, it is possible to compare di erent M&A deals.
Market capitalisation of a company or the price at which listed companies shares are sold on the stock exchange is the most commonly used valuation method for calculating premiums (Hayward & Hambrick 1997;Gupta & Misra, 2007;Varaiya & Ferris, 1987;Guo, Clougherty & Duso, 2013). is valuation concept is popular due to the data availability, simplicity and objectiveness. us, to be consistent with other relevant researches on acquisition premiums (Hayward & Hambrick 1997;Gupta & Misra, 2007;Varaiya & Ferris, 1987;Guo, Clougherty & Duso, 2013), further research will assume that value of a target company's share is the listed stock price prior to the acquisition announcement.
In the majority of M&A, price contains the value of the target company to the acquirer plus a premium. Price is the monetary expression of stock or cash which was used to pay for the target company (Fernandez, 2004). e price of the deal is usually paid with a premium on the price at which the target's shares are traded in the market (Emery, Finnerty & Stowe, 2004). Sometimes M&A deals are completed with a discount. Discount could alert that currently the company is overvalued and it originally had a lower value in M&A market or that management of the target company is weak (Roll, 1986). Normally, the potential acquirer compares the value (what a company is worth to the acquirer) to the current market price and if the value is below the price, the bid is abandoned (Roll, 1986).
Premiums. An acquisition premium is a ratio of the negotiated price of one target's share and the price at which the target's share is traded in the market (Laamanen, 2007;Sirower & Sahni, 2006). is premium is sometimes called a control premium in the acquisitions of publicly traded companies (RSM Bird Cameron, 2010;Komiak, 2010). However, Damodaran (2011) argues that it is a premium paid for the acquisition, but not a control premium. e researcher states that a control premium appears as the acquirer believes it could make signi cant impact on the target company. Conversely, premiums for acquisitions can be paid for many reasons, and control is just one of them. Besides, acquirers tend to overpay (Damodaran, 2011). us, agreeing with Damodaran (2011), a premium paid for acquisitions should not be referred to as a control premium.
Research revealed that a high acquisition premium could be dangerous for the acquirer because it may overweight all the synergies acquisition might create (Sirower & Sahni, 2006;Varaiya & Ferris, 1987). ere are many cases when the premium paid was too high and because of that the deal did not create additional value to its stakeholders. Sirower and Sahni (2006) stated that premiums paid for the acquisitions are negatively related with acquirers' returns and can still have an impact a er up to four years. Moeller, Schlingemann and Stulz (2005) found in their research that shareholders of the acquiring rm lost 12 cents on every dollar spent on acquisition during 1998-2001 in the United States. e accumulated total loss was about $240 billion.
Even though many studies revealed that high acquisition premiums frequently destroy value (Sirower & Sahni, 2006;Moeller, Schlingemann & Stulz, 2005), companies continue to pay high premiums. omas Reuters report indicates that the world average premium for all industries in 2012 was 31.2% and30% in 2013 ( omas Reuters, 2013).
As a result, a natural question arises as to what the major premium drivers are. A number of determinants were identi ed by the existing literature in an a empt to justify the sizes of the premiums (Figure 2). One explanation for the large premium could be that the stock on the stock exchange does not represent the real value of the company (Myers & Majluf, 1984). Having performed comprehensive due diligence analyses of the potential target, acquirers might have a deeper understanding of the real value of a target rm than the market (Laamanen, 2007). However, similarly to the acquirer who notices the target's potential for value creation, the market can also recognise the undervalued potential of the companies to be acquired. us, market prices might be adjusted with the possibility of acquisition even before the acquisition is announced (Crawford & Lechner, 1996).
Assuming that a company on the stock exchange is priced fairly, there are many other determinants for the high premiums. One of the speculative determinants for the size of the premium is believed to be managerial hubris (Hayward & Hambrick, 1997). Managers of the acquiring companies are willing to bid high to complete the deal. Research (Aktas, Bodt & Roll, 2011) proved that if the manager completed a few successful transactions in the past, he will tend to overpay for a new target company. Recent performance of the company, media praise and CEO self-con dence have a positive e ect on premiums (Hayward & Hambrick, 1997). Agarwal and Zeephongsekul (2013) state that acquirers who are risk takers pay relatively more than the risk averse acquirers. ere exists a contradictory opinion that premium is the result of either a valuation error or managerial hubris (Roll, 1986). Varaiya and Ferris (1987) noticed that overestimation of acquisition gains could cause higher premiums as well. Capron, Dussauge and Mitchell (1998) supplemented this opinion by adding that if the acquirer believes that combined resources will generate high value, the acquirer tends to be more certain about the deal. John, Liu and Ta er (2008) proved that premiums are on average 7-9% higher if both the acquirer and the target are overcon dent about potential synergies. In general, target companies' bargaining power is positively correlated with the size of the premium (Varaiya & Ferris, 1987).
Ownership of the acquirer also has an impact on the size of the premium. Guo, Clougherty and Duso (2013) research showed that state-owned companies in China are paying higher premiums in cross-border transactions in comparison with private companies.

FIG. 2. Premium determinants Source: Compiled by the authors
Besides, the origin of both acquiring and target companies has an impact on the premium. is is believed to exist due to the di erences in taxation and encouragement policies in the countries (Hope, omas & Vyas 2011;Bailey, Chung, & Kang, 1999). Hope, omas and Vyas (2011) found that acquiring rms from emerging economies in comparison with those from developed economies have a tendency to overpay in order to acquire assets in developed and therefore more stable economies. e author a ributes this to the national pride -developing countries see acquisition price as ability to make national, social, or political decisions. A few researches have been conducted to nd out whether the level of internationalisation has an impact on the premium. According to Rustige and Grote (2010), premiums of European cross-border transactions in 1985-2009 were 10 percent higher, thus companies acquiring a target from a di erent country are expected to pay more. However, most of the acknowledged researchers have focused on the acquirer's origin, such as a region or development level but have not explored the di erences in premiums paid by the acquirers from particular countries.
Also, the higher stake is purchased, the higher the premium is paid. Walkling and Edmister (1985) estimated that on average premiums are 9% higher when the bidder seeks for the majority control. Kim, Haleblianand, and Finkelstein (2011) revealed that acquirers are willing to pay higher premiums when the possibility of the acquiring company's organic growth is low. e same research also found out that premiums tend to be high if the acquirer's advisors have relatively small experience. According to Schwert (2000), hostility in the takeover has some trivial in uence on the premiums. e payment method can in uence the size of the premium as well. However, researchers do not have a unanimous opinion whether payment in cash or in the acquirer's stock has a greater impact on the size of the acquisition premium. Damodaran (2011) argues that payment in the acquirer's stock will determine larger premiums. e acquirer whose stocks are overvalued will be willing to o er a higher premium for the target in exchange for the acquirer's stock (Hajbaba & Donnelly, 2013). However, other researchers (Pinkowitz, Sturgess & Williamson, 2013;Burch, Timothy, Nanda & Silveri, 2012) state that it is the opposite, and deals have a higher premium if the payment is in cash due to the potential tax gains.
ere are other forces besides the ones dictated either by acquirers or targets that can determine the size of a premium. According to Gould (1998), the main premium determinants are M&A demand and supply, trends, similar past transactions and the industry's 'rules of thumb' . In cases when there are two or more acquirers who compete for the acquisition of one target, the winner usually pays an enormous premium (Varaiya & Ferris, 1987;Andrade, Mitchell, & Sta ord, 2001). Walkling and Edmister (1985) found out that premiums in such cases were about 30% higher. Imitation of interlock partners which have homogeneous premiums experience in uence the acquirer's willingness to pay more (Beckman & Haunschild, 2002). What is more, when relative valuation is applied, companies are frequently looking at the market multiples of the companies from the same industry (McAfee & Morley, 2010;Minjina, Dussauge & Mitchell, 2010). Hence if other companies in the industry tend to pay high premiums, it is likely that a new deal which is based on relative valuation will generate similar premiums. Besides, according to Hanemann and Rosen (2012), asset valuation uctuates with the global growth cycles and countries with higher growth perspective can o er higher premiums. All the determinants which are not in uenced directly by either acquiring or target companies can be referred to as market determinants. e above premium drivers are the major ones discussed and analysed by researchers. However, as the topic is relatively new and not much research has been conducted to explore this area, it is likely that there are still other underlying drivers for the sizes of premiums. Besides, no research has been conducted to examine the interrelation among these di erent premium drivers. is paper will mostly contribute to the research on the acquirers' origin. e ndings by Hope, omas and Vyas (2011) that emerging countries have a tendency to pay larger acquisition premiums will serve as a basis for the hypothesis about the di erence in the premiums paid by Chinese and European companies. In contrast to Hope, omas and Vyas (2011) research, this paper provides ndings on the level of particular countries and regions. e peculiarities of M&A of Chinese companies. e reasons for Chinese cross-border M&A which determine the sizes of premiums can be divided into two categories: 1) willingness and capabilities of strong Chinese companies and 2) support from the government. Peng (2012) indicates three main reasons why M&A are so popular among Chinese multinational companies. First, Chinese multinationals need a fast market entry, especially in natural resources area (Deng, 2009). To overcome branding capabilities, companies also tend to acquire existing world-class brands, such as recent acquisitions of IBM's PC division and Volvo. CEO hubris was identi ed as a third reason, as Chinese managers seek to make companies more sophisticated so that they could demand higher salaries (Peng, 2012).
In 1999, the Chinese government initiated the 'Going Global' strategy to promote Chinese investments abroad and became a country which actively supports both inward and outward FDI (Buckley, Clegg, Cross, Voss & Zheng, 2007;Guo, Clougherty & Duso, 2013). Favourable conditions such as a huge market size, low labour costs and large population let China become a dominant exporter in the world (Gao, Murray, Kotabe & Lu, 2010). Due to this, the country accumulated the largest portion of foreign reserves in the world, which accounted for US$3.4 trillion in 2012 ( e World Bank, 2013). By having such huge amount of reserves China could support their FDI activities (Luo, Xue & Han 2010). As a result, M&A became a primary mode of entering new markets by Chinese multinationals (Peng, 2012;Sauvant, Maschek & McAllister, 2009). e government promotes outward investments, which corresponds to the country's 'Going global' strategy (Buckley, Clegg, Cross, Voss & Zheng, 2007) in a few di erent formats: be er foreign exchange rates, lower interest for nancing the projects, reduced taxation etc. (Peng, 2012;Musacchio, & Flores-Macias, 2009). e main strategic industries for receiving such bene ts are natural resources (oil, gas, and minerals), services (e.g., banking, transportation, construction), and some industries involving high technologies such as computer, automobile manufacturing, and electricity power generation (Guo, Clougherty & Duso, 2013).
ose Chinese companies which get the government support could a ord a higher bid in the M&A process. However, Li, Li and Wen (2009) indicated that it is easier and more common for state-owned enterprises to get support from the government than from the private sector. e role of the government in China remains high, as most shares of Chinese publicly traded enterprises are controlled by the government (Lau, Fan, Young & Wu, 2007). Chinese state-owned enterprises (SOEs) sometimes have a 'must have it' a itude and pay higher premiums. SOEs are also motivated to achieve non-commercial objectives through M&A thus making these acquisitions of questionable value to the company (Globerman & Shapiro, 2009). According to Hanemann and Rosen (2012), acquisitions made by SOEs account for 44% of the deals by number and 79% of the deals by value of the deals. On the other hand, private Chinese companies are more rational and do not pay such large premiums (Guo, Clougherty & Duso, 2013). erefore, support from the Chinese government is one of the explanations of high premiums paid for the target companies.
China's 12 th ve year plan sets a framework for Chinese M&A deals. It seeks for its outward FDI to grow on average by 17 percent and 7 industries are named as priorities: new energy, energy conservation and environmental protection, biotechnology, new materials, new IT, high-end equipment manufacturing, clean energy vehicles (KPMG China, 2011). us, it is likely that premiums for the targets from priority industries will be larger. However, Chinese multinational enterprises have a particularly poor record of completing cross-border acquisitions that they announce (Zhang & Ebbers, 2010). In 2000-2008 only 47% of all cross-border acquisitions announced were later completed, (e.g., 67% in India) (Sun, Peng, Ren & Yan, 2011).
With the government support and Chinese companies' willingness and capabilities to acquire companies in Europe, there is an underlying possibility that premiums for similar targets in Europe will be higher if the acquirer is a China-based company rather than a Europe-based company. is would comply with the ndings of Hope, omas and Vyas (2011) that companies from emerging countries are paying larger premiums.

Research Methodology
e quantitative survey was performed by using two samples: the rst represented Chinese acquirers, the second involved European acquirers. Figure 3 indicates the summarised research scheme. e research is focused on 2 issues. First, it compares Chinese acquisition premiums with the premiums paid by European acquirers and identi es whether there is a statistically signi cant di erence in these premiums. Second, it investigates other premium determinants and their in uence on the di erence between the premiums paid by Chinese and European acquirers.
Data collection. Data for analysis was extracted from Mergemarket (a globally acknowledged database on M&A). e sample of Chinese deals was based on 7 criteria: one for identi cation of the acquirer, two for identi cation of the target and four for the identi cation of the deals (Figure 4).
First of all, only those acquirers which are Chinese Mainland companies were included in the sample, i.e. Hong Kong, Macao and Taiwan were not a subject of the research. Targets for these acquirers had to be only Europe-based companies which were listed on the stock exchange (at least during the period of acquisition). e time frame of 14 years (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013) was set so as to exclude earlier deals which might have been completed under slightly di erent conditions without the Only the deals the value of which was publicly announced and only the ones in which more than 10% shares had been acquired are included in the sample. Due to the fact that only 47% of Chinese deals which are announced get completed (Zhang & Ebbers, 2010), only the deals that had both announcement and completion dates were included in the sample.
In total there were 117 Chinese deals in Europe with announced acquisition value from 01/01/2000 to 01/01/2014. A er application of 7 criteria, 17 deals were recognised as eligible for further analysis and constitute the sample.
Due to a relatively small sample size (N=17) and the fact that each Chinese acquisition was unique and hardly comparable, each acquisition is analysed separately. erefore a comparable set of deals for each Chinese acquisition was obtained by applying the same 7 criteria for European acquirers, which had to be a Europe-based company and its target company had to operate in the same sub-industry. In total 17 datasets covering 139 European deals were obtained.
Comparison of European and Chinese deals. Premiums in the database were computed using formula 1: Share price at time t is a price in stock exchange 1 or 30 days before the announcement day. Two acquisition premiums are used: 1 day prior to the announcement day -to diminish the risk of taking into account events which were not related with the acquisition but in uenced uctuations of the stock price (Kim, Haleblian & Finkelstein, 2011, Jordan & Wort, 2009, and 30 days -because this pre-acquisition stock price is more representative as it is not in uenced by possible information leakage about potential acquisition (Reuer, Tong & Wu, 2012). e 30 days period is widely used by many other researchers (Guo, Clougherty & Duso, 2013;Hayward & Hambrick, 1997;Kim, Haleblian & Finkelstein, 2011). e research is composed of two parts. e rst part seeks to identify whether premiums paid by Chinese acquirers are higher than European acquisitions premiums for the similar type of a company in Europe and is focused on comparison of 17 Chinese deals with 17 datasets of comparable European deals thus representing a purely comparative research. e second part helps to identify whether there are any other factors besides regional ones which in uence the size of the premiums. It complements the results gained in the rst part and also tests whether premiums paid by Chinese acquirers are larger than those paid by European acquirers for similar targets in Europe. In the second part, 17 datasets were aggregated into one sample and 17 Chinese deals were compared with 139 European deals, therefore, comparability of deals was partially neglected as each Chinese deal was represented by a di erent number of comparable deals.
If the hypothesis of sample distribution normality was rejected, lambda ( ) values (square root, logarithm, square) were applied to transform non-normal data into normally distributed data. If the transformation did not succeed to obtain normally distributed data, only non-parametric tests were applied.

Comparison of premiums based on comparability of targets.
In order to compare whether premiums paid by Chinese acquirers are higher or lower than the mean of the premiums of comparable acquisitions by European companies, at rst all 17 Chinese deals were compared with 17 datasets (based on comparability of targets). A er that deeper analysis of the cases was executed. Table 2 summarises the methods and tools applied. e null hypothesis in both parts is the following: H 0 : Premiums paid by Chinese and European companies for similar targets in Europe were equal.
To test whether there is a signi cant di erence between the premiums paid by European and Chinese acquirers for similar targets in Europe, the average premium paid by Europe-based companies of the representing dataset was subtracted from the premium paid by the Chinese acquirer. Binomial test is used to measure signi cance of frequency of cases and independent samples t-test to measure whether there is a signi cant di erence between sizes of premiums. e number of cases with a statistically signi cant di erence between the premiums of European and Chinese acquirers was identi ed by conducting one sample t-tests. e cases with signi cant di erence where divided into two groups. One group consisted of cases where a higher premium was paid by European acquirers and the second one was comprised of Chinese acquirers; a binomial test was conducted to conclude whether there is a signi cant di erence between premiums paid by Chinese and European companies. Chinese acquirers were found to pay larger premiums.

Comparison of premiums, partially neglecting comparability of targets
To prove that premiums paid by Chinese and European acquirers are di erent while taking into account other possible factors, the set of 17 Chinese deals is compared with 139 European deals. e second null hypothesis to be tested was the following: 2H 0 : sizes of premiums paid by Chinese and European companies were not signi cantly in uenced by other than regional factors. e tests on whether there is a signi cant di erence between premiums in the two samples, and whether the premiums paid by Chinese companies and European companies were correlated with potential premium determinants (the deal announcement date, stake in the target company, the deal value, the target country development level and payment method) were conducted (Table 3). Additional 2 variables were tested only for the European deals sample: the nature of the deal (domestic or cross-border) and the bidder's country development level (developed, frontier or emerging). e Kruskal Wallis Test was used to test whether the sector has an impact on acquisition premiums and partial correlation was used to test the correlation between the bidder's region (China or Europe) and paid premiums while controlling for other 5 variables. It

Empirical ndings on dissimilarity between the premiums of Chinese and European companies
e results of the research conducted by using the two methods described suggest that premiums paid by Chinese and European companies for similar targets in Europe were di erent. During 2000-2013, the average acquisition premium paid by Chinese acquirers was approximately two times higher than the average premium paid by European companies.
Description of the sample of Chinese deals. Most of the target companies which were acquired by Chinese companies are based in Western Europe ( Figure 5). is complies with the ndings of Hanemann and Rosen (2012) that more than 90% of all Chinese acquisitions were completed in Western Europe.  (Figure 6). is corresponds to other research (Zhang & Ebbers, 2010;Hanemann & Rosen, 2012) where these sectors are indicated as the strategic ones for Chinese outbound investments. In general, most of Chinese deals had positive acquisition premiums (Figure 8). In 14 out of 17 deals, 30 days acquisition premiums were between 0% and 100% and in 2 deals the premium was more than double the stock price. e size of 30 days premiums varied between -32.2% and 175%. e mean of the 1 day acquisition premium (36.6%) was lower in comparison with the 30 days mean (42.2%).
is is consistent with other research (RSM Bird Cameron, 2010), which showed that due to the potential information leakages, markets anticipated the acquisition

Deal value M USD
Numer of deals before its announcement date. More than 90% of deals were completed with positive acquisition premiums. Table 4 indicates that there is no evidence of signi cant di erence between 1 and 30 days Chinese acquisition premiums. Description of the sample of European deals. e European sample (obtained by pooling all 17 datasets of European deals) contains 139 deals with speci ed 1 day premium and 138 deals with 30 days premium (the di erence in the number of the deals in two samples appeared due to the fact that one target company had not been listed on the stock exchange 30 days before the deal announcement date). Consistently with the Chinese sample, most of the European targets were located in the developed European countries (87%) (Figure 9).

FIG. 8. Distribution of 1 day and 30 days Chinese premiums
Source: compiled by the authors 1 day premium 30 day premium

Frequency Frequency
Most of the acquirers in the European deals sample were located in the developed European countries as well ( Figure 10). e most active acquiring country is the United Kingdom (33 bids). Bidders from Italy and France completed 15 acquisitions each, while bidders from Germany, Greece and Norway closed 10 deals each.  In the European sample, most of the deals could be referred to as domestic rather than cross-border deals. 92 deals were completed by bidders and acquirers which are from the same country and 47 deals can be named as cross-border deals. Most of domestic deals were completed in the United Kingdom (N=25) and Italy (N=10).
Proportions of sectors of target companies in the sample of European deals were slightly di erent from the Chinese deals sample (Figure 11). Energy (N=43) and transportation (N=30) sectors were the main sectors of target companies in the European sample. Mining, which in the sample of Chinese deals was the largest target sector, ranked third (N=22) in the sample of European deals sample. Automation was a dominant sector (18 deals). Analysis of the pooled European deals sample showed that similarly to Chinese deals, most of the European acquisitions were completed with a positive premium. For 70% of the deals premiums paid by European companies were between 0% and 50% ( Figure 13). However, in contrast to the sample of Chinese deals, the European deals sample is more skewed, due to a few deals which were accomplished with extremely high premiums. e largest premium paid by European acquirers was 358%, while the largest Chinese acquisition premium was only 150%. A positive extreme was observed in the 1 day premium. e comparison of the means of premiums complies with the results obtained from the analysis of the Chinese deals, which showed that the 1 day premium is lower than the 30 days premium, and is in line with RSM Bird Cameron (2010) research.

FIG. 13. Distribution of 1 and 30 days European premiums
Source: Compiled by the authors 1 day premium 30 day premium

Empirical evidence on the di erences in premiums
e results of the analysis of consolidated premiums and cases are presented to show whether premiums paid by Chinese acquirers were larger than those paid by European acquirers.
Analysis of consolidated premiums showed that in 62% of the cases the premiums paid by Chinese acquirers were higher than the mean of representative premiums paid by European acquirers (Table 5). Chinese acquirers paid more both in 1 and 30 days acquisition premiums more frequently. In 8 cases, both premiums paid by Chinese acquirers were higher than the premiums paid by European acquirers and only in 4 cases both premiums paid by European acquirers were higher. Table 6 summarises the results of the binomial test to measure whether the number of cases where Chinese paid higher premiums in comparison with European companies is su cient to draw a conclusion that Chinese in general were paying higher premiums. Even though the observed probability for paying larger acquisition premiums is higher when the acquirer is a Chinese company (65% and 59% for 1 and 30 days premiums respectively), it provides insu cient evidence that all Chinese deals are more frequently completed with higher premiums. e binomial test provides results on the frequency of cases where the premium is higher/lower neglecting the size of the premiums.  Analysis of consolidated premiums showed that the means of Chinese and European acquirers' premiums di er by nearly 50%. e means of the premiums paid by Chinese acquirers are 36.5% and 42.2% for 1 and 30 days premiums respectively, while the means of European acquirers' premiums are 18.7% and 23.1%, but the results from independent samples t-test have indicated that there is no signi cant di erence between the means of two groups even when taking into account the sizes of the premiums (Table 7).
Nevertheless, the one-sample Kolmogorov-Smirnov test showed that with 95% con dence, the levels of both 1 day and 30 days premiums are distributed normally; the independent samples t-test and the binomial test are not statistically reliable due to the very small size of the samples, which varied between 6 and 17 companies. Due to the fact that the di erence between the means is nearly 50% and the fact that tests were conducted for very small samples, it is very likely that, with a larger sample, statistically signi cant evidence that premiums are larger will appear. However, this could only be possible to con rm by either diminishing the set of criteria for the Chinese sample (e.g., including Hong Kong deals or deals with acquisition of less than 10% stake in the target company) or waiting for more Chinese deals to appear on the market in the future. Although the results showed that Chinese acquirers were paying two times higher premiums in comparison with European acquirers for similar targets in Europe, it was not possible to prove that this di erence in premiums is statistically signi cant due to the small sample size.
Analysis of cases was used to analyse separately all the 17 cases in order to nd out whether there is a statistically signi cant di erence between the Chinese acquirer and the representative sample of premiums paid by European acquirers. One-sample Kolmogorov-Smirnov test showed that in 15 cases, distribution of each European dataset was normal, 2 Chinese deals had only 1 representative European deal. e analysis of cases showed that Chinese acquirers completed the deals with signi cantly higher premiums than the acquiring European companies. One-sample t-tests presented in Table 8 have revealed that in 9 cases out of 15 there is statistically signi cant evidence that the mean of premiums paid by European companies is not equal to the premium paid by the Chinese acquirer for a similar target in Europe.

Source: Compiled by the authors
ere is signi cant evidence that 1 day premium was higher in 6 deals and lower in 1 Chinese case. e 30 days premium was signi cantly higher in 7 cases and lower in 1 Chinese deal. is proves that in more than a half of cases Chinese paid higher premiums than the average premium paid for a similar target company by European acquirers. e binomial test (Table 9) was conducted for only those cases where the di erence between premiums was statistically signi cant. In 86% of the cases Chinese were paying a higher 1 day premium and in 88% of cases a higher 30 days premium. At 90% con dence level, there is su cient evidence that when there is a signi cant di erence between the premiums paid by Chinese and European acquirers, Chinese are paying higher premiums. is binomial test has 2 shortcomings. First of all, it only measures frequency of cases, neglecting the size of the premium. Second, the sample size is very small and because of that it is very di cult to test any hypothesis. Supposedly, if the proportion of cases remains the same and the sample size increases, binomial test would show evidence that if there is a signi cant di erence between the premiums paid by Chinese and European acquirers, Chinese acquirers would be the ones paying higher premiums.
Summarising the results, we can state that Chinese acquirers have paid higher premiums in comparison with European acquirers for similar targets in Europe. Also, Chinese acquirers were paying larger premiums more frequently. However, the small sample size was the major reason why this did not always prove to be statistically signi cant.

Comparison of premiums partially neglecting comparability of targets
A er pooling 139 European deals in one sample and comparing it with the sample of 17 Chinese deals (partially neglecting comparability), the results indicate that Chinese acquirers were on average paying a premium which is double the premium paid by European acquirers. 17 Chinese acquirers on average paid 36.5% 1 day premiums and 42.2% 30 days premiums in comparison with 139 European acquirers who paid 18.4% 1 day and 21.1% 30 days premiums (Table 10).
Normality tests of samples indicate that the data from the European deals sample is not distributed normally. e main reasons why the one-sample Kolmogorov-Smirnov test rejected the hypothesis of normal distribution is skewness of the data. A empts to transform data into normally distributed data using lambda (λ) functions did not succeed. As parametric tests required normality of distribution, only non-parametric Wald-Wolfowitz and Median tests which are equivalent to the parametric independent sample t-test were conducted while analysing the pooled European deals sample. e Wald-Wolfowitz and Median tests (Table 10) also show that premiums paid by Chinese acquirers are signi cantly larger than premiums paid by European buyers. e results of the pooled sample are slightly di erent from the ones obtained during the rst part of the research where the di erence in the means between the premiums of the deals of Chinese and European companies was not always signi cant. is di erence might be due to enlarged number of deals in the sample, also because of the Median test which only measures the median, not the mean. However, both research methods indicated that the premiums paid by Chinese companies are twice as large as the average premium paid by a European acquirer. e only di erence was that the rst research method failed to prove that this di erence in premiums is always statistically signi cant.

e e ect of other variables
Further research aims to identify whether there are any other factors which have in uenced the di erence between the size of the premiums paid by Chinese and European companies.
As the data from the Chinese deals sample is distributed normally, a parametric Pearson correlation test was conducted to test whether the size of the premium is correlated with a year of the deal announcement, the stake in the target company or the deal value. It showed that none of the tested variables had a signi cant relationship with the premium paid by the Chinese. Correlation coe cients of all 3 variables were close to zero (Table 11). However, the sample size of 17 Chinese deals might be too small to reach reliable results. us, there is no evidence that Chinese companies' premiums were not in uenced by other than regional factors.
Other 4 variables were not tested for correlation with the premiums paid by Chinese acquirers because all the 17 Chinese deals in the sample included targets from developed European countries, the payment method was cash, and the nature of the deal and the bidder's country development level was also the same.
As the data of the European deals was not distributed normally, non-parametric Spearman's rho correlation was applied to test this relationship. e correlation test between the premiums paid by European companies and 7 variables showed that the premium size is correlated with 3 variables (Table 12): the stake in the target company, the deal announcement date and the payment method for the 30 days premium. European companies' premiums do not depend on the deal value and this complies with the Guo, Clougherty and Duso (2013) research as they have not found the relationship either between the premiums or the transaction value.
In addition, there is no relationship between premiums paid and the fact whether the bidder and the target was of the same origin or they were registered in di erent European countries (Table 12). However, this contradicts Rustige and Grote's (2010) research that Europeans usually pay approximately 10% more in cross-border acquisitions than in domestic ones. is contradiction might be a result of a few reasons: rst, in this research, only those cross-border acquisitions which were made within Europe were included in the sample. Rustige and Grote's (2010) research covered not only Europe. Also, it could be due to sectoral di erences. e last reason could be the time frame. In this sample the time period is between 2004 and 2013 (to be consistent with the Chinese deals sample), while Rustige and Grote (2010) analysed deals which occurred in 1985 e positive relationship was observed between the premiums paid by European acquirers and the deal announcement date (Table 12), the more recent was the deal, the higher was the premium. However, this is not consistent with Mergermarket's (2013) review on premiums of all M&A in Europe. According to the statistics, premiums were growing between 2004 and 2008 and then were quite stable in 2008-2013. Statistics covered all acquisitions which took place in Europe not specifying the bidder country or region. Besides, all industries were included in the sample. ese are the two reasons explaining the slight di erence in the relationship of premiums and the deal announcement dates between this research and Mergermarket (2013) results.   Source: Compiled by the authors e same positive relationship was observed between the premiums of European companies and the purchased stake in the target company: the larger stake was purchased, the higher premium was paid. is complies with Walkling and Edmister (1985) who proved that the stake purchased in the target company has a positive in uence on the acquisition premium. It could be associated with the bene ts of owning the majority control and capability of making real changes within the acquired company. Besides, shareholders of the target company are expecting additional premium for abandoning their majority control.
Payment method was found to have a relationship with European acquisition premiums as well (Table 13). Positive relationship indicated that higher premiums were paid for the deals with payment in equity rather than in cash. is complies with Damodaran (2011) and Hajbaba and Donnelly (2013), who argue that payment in acquirers' stock will determine larger premiums. However, it does not comply with opinions of other researchers (Pinkowitz, Sturgess & Williamson, 2013;Burch, Nanda & Silveri, 2012) who stated that deals have a higher premium if the payment is in cash due to the potential tax gains. Besides, payment method has a relationship with the stake purchased in the target company. Deals with a larger value tend to be paid in the acquirer's stock. Also, the relationship between the payment method and the nature of the deal (domestic or cross-border) was observed. More domestic deals were completed by paying for the target company in equity, while payment in cash was a more popular payment method during the cross-border deals. is corresponds to the Chinese sample as all the 17 cross-border deals were paid in cash.

Source: Compiled by the authors
Premiums paid by both Chinese and European companies were not signi cantly in uenced by the sector in which the target company operated (Table 13). To test the di erence between premiums in 8 sectors, a non-parametric Kruskal Wallis test was conducted, which indicated that the highest premiums by European acquirers were paid for targets which operate in leisure, consumer-retail, and industrial automation sectors. Chinese acquirers paid highest premiums for targets which operate in consumer (other), leisure, transportation and mining sectors. However, signi cant di erence between the means was neither observed in Chinese nor in European deals. e Kruskal Wallis test provides nearly the same results.
ere is no evidence that Chinese acquirers are paying higher premiums for the priority industries such as energy, mining, transportation and automation (Guo, Clougherty, Duso, 2013). Predominance of the deals in energy and mining sectors indicates that even though it is hard to prove that premiums in these sectors were signi cantly higher, there is evidence that Chinese companies were acquiring a signi cantly larger portion of companies from priority industries.
ere is an underlying possibility that Chinese premiums were higher than European premiums due to other than the regional facts. To test the e ect of other variables on the correlation between the region (China or Europe) and the size of premiums, Spearman's partial correlation was used (Table 14). ose variables which are relevant to both Chinese and European premiums -the deal announcement date, the stake in the target company, the deal value, the target sector, the target country development level and the payment method -were controlled in a row. Two other variables, the development level of the bidder's country (emerging, frontier or developed) and the deal nature (cross-border or domestic) were rejected due to the fact that Chinese acquisitions are only cross-border and the bidder's country development level is emerging for all deals. e stake in the target company, the deal value, the target sector, the payment method and the target country development level did not have a signi cant positive in uence on the di erence between Chinese and European acquirers' premiums. e results in Table 14 show that when all the 5 variables were controlled, the correlation was even stronger than without controlling these variables. is means that there is su cient evidence that a er elimination of any e ect of these 5 variables, the size of the premiums and the bidder's region will still correlate. On the other hand, the deal announcement date seems to have slight in uence on the di erence between the premiums paid by Chinese and European companies. While controlling the deal announcement date, the correlation coe cient between the region and size of the premiums diminished from 1.59 to 1.51 for the 1 day premium and from 1.6 to 1.52 for the 30 days premium. Without controlling any variables, the p-value was 0.047, thus, even a slight decrease in the correlation coe cient would have caused a loss in signi cance. is loss in signi cance could have been caused by slight inconsistencies between the deal announcement dates in the Chinese and European deals samples.
ere were a di erent number of deals in each of the 17 European datasets and thus a er pooling all datasets into one sample, slight dates inconsistencies might have occurred. However, a er analysis of two samples, no signi cant di erence in announcement dates was observed. Most of the deals in Chinese and European samples occurred in 2011 and 2012. Another European deals peak was in 2008, which is not consistent with the sample of Chinese deals. erefore, the time frame had a slight in uence on correlation between the size of the premium and the bidder's region.
If all the variables are controlled at the same time, the correlation between the bidder's region and the size of the premiums remains signi cant. e last row in Table  14 indicates that while controlling for 6 variables -the stake in the target company, the deal value, the target sector, the payment method and the target country development level, which increased the correlation coe cient, and the deal announcement date, which decreased correlation coe cient -the nal correlation coe cient increased.
us, there is insu cient evidence that correlation between the bidder's region and the size of the premiums was signi cantly in uenced by these 6 variables. e bidder's origin is a reasonable explanation for the di erence in the premiums paid by Chinese and European companies. erefore, the previous ndings gained through the Wald-Wolfowitz and Median tests that there is a signi cant di erence between the medians of the Chinese and European deals should be retained. e results of this research suggest that the statistically signi cant di erence between the premiums of Chinese and European companies for similar targets in Europe does exist.
Chinese acquirers' willingness to pay more in comparison with the premiums of European acquirers for the targets in Europe could be explained by 3 main reasons: Chinese companies' underlying motivational factors, support from Chinese government and China's growth.
According to Peng (2012), most Chinese companies complete mergers as they need a fast market entry and ability to increase branding capabilities. Also, national pride and managerial hubris are other major reasons for M&A as Chinese managers of sophisticated companies could demand higher salaries (Peng, 2012). Due to the underlying need for acquisitions and strong nancial capabilities, Chinese companies are paying high premiums to acquire target companies in Europe and in other developed countries. What is more, about half of all Chinese cross-border deals were completed by state owned enterprises (Hanemann & Rosen, 2012). According to Guo, Clougherty and Duso (2013), state owned companies are frequently trying to reach non-commercial objectives during the M&A and thus are paying signi cantly larger premiums in comparison with Chinese private companies.
Besides, Chinese government is actively promoting outbound investment through its 'Going Global' Strategy. A Chinese company which is willing to acquire a foreign company could expect such bene ts as be er foreign exchange rates, lower interest for nancing the projects or reduced taxation (Peng, 2012;Musacchio, Flores-Macias, 2009). is support is likely to be one of the reasons why Chinese companies could a ord higher bids in the M&A process.
According to Hanemann and Rosen (2012), asset valuation uctuates with the global growth cycles, and countries with higher growth perspective can o er higher premiums. As China is currently one of the fastest growing economies in the word, it has stronger position in bidding for the targets. As a result, M&A became a primary mode of entering new markets by Chinese multinationals (Peng, 2012;Sauvant, Maschek & McAllister, 2009). us, these three reasons are believed to be the major ones in uencing the signi cant di erence in Chinese and European companies' acquisition premiums.

Conclusions and suggestions
e primary goal of this study was aimed at identifying whether Chinese Mainland companies were paying higher premiums for similar target companies in Europe during the period of 2000-2013 in comparison with Europe-based companies and whether it can be explained by other than regional factors. To determine the di erence between the premiums paid by Chinese and European companies, two samples were analysed -one representing all Chinese acquisitions in Europe which matched the set of criteria, another sample representing comparable European targets acquired by European companies. e analysis of literature revealed that premiums paid by Chinese acquirers might be higher than the premiums paid by European acquirers. e size of the premiums might be in uenced by various deal-related and market-related factors such as managerial hubris, expected gains, origin and ownership of the acquirer, payment method, market trends, demand and supply etc. e ndings of other researchers revealed that companies from emerging countries tend to pay larger acquisition premiums when acquiring companies in developed countries and that this is particularly popular among Chinese companies when entering foreign markets due to the government support, companies' motivational factors and expected high economic growth in China.
Two research methods employed provided similar results and showed that Chinese companies were paying larger premiums compared to European companies' premiums paid for similar targets in Europe. e average Chinese premium was double the size of the average European companies' premium paid for similar targets in Europe. In cases when there was a signi cant di erence in premiums among comparable Chinese and European deals, there was about 90% probability that a higher premium was paid by a Chinese acquirer.
When partially neglecting comparability of deals, the signi cant correlation between the sizes of premiums and the bidder's region was observed. e results proved that there is a signi cant di erence in the medians of Chinese and European premiums. When taking all the comparable deals, pooling them into one sample and controlling for other variables which might in uence the size of premium, signi cant di erence was found between the premiums paid by Chinese and European companies.
Even though the research which was aimed at comparison of 17 Chinese deals with 17 comparable datasets failed to nd a signi cant di erence in premiums, the comparison of 17 deals with a pooled European sample of 139 deals indicated signi cant di erence in premiums. e inconsistency in the results of two studies was mainly caused by the sample size.
When comparability was partially neglected due to the fact that in the pooled sample there was a di erent number of European deals representing a particular Chinese deal and partial correlation helped to control for premiums determinants (sector, deal announcement date, payment method, target country's development level, stake in a target company and deal value), correlation between the sizes of premiums and the bidder's region was present even when controlling for all the named variables.
erefore it could be concluded that Chinese companies were paying higher premiums for similar targets in Europe during 2000-2013 in comparison with European acquirers and this large di erence in premiums was not signi cantly in uenced by other than regional factors.
is study contributed to the ideas of other researchers and provided new insights on the topic of premium determinants. It complemented the rather outdated research of Walkling and Edmister (1985) and showed that sizes of premiums were in uenced by the stake purchased in the target company. What it more, the research backed up Damodaran's (2011) theory that premiums are higher when payment method is in equity, in contradiction to Pinkowitz, Sturgess and Williamson (2013) and Burch, Nanda and Silveri (2012) ndings. Also, it was found out that domestic deals tend to be completed by paying for the target company in equity, while payment in cash was a more popular payment method during the cross-border deals. Research provided ndings that deals with larger value tend to be paid in equity, but there is no signi cant relationship between the size of the premium and the deal value. What is more, results of the research provided insu cient evidence that there is a di erence between European domestic and European cross-border acquisition premiums. e general trend shows that the number of Chinese acquisitions in Europe increases year by year. e European Union Chamber of Commerce in China (2013) also forecasts a potential jump in the acquisitions. With a larger sample it would be possible to conduct a broader research and examine the sources and reasons for such di erence further.