37
7585 2019 April 2019 Chinese acquisitions abroad: are they different? Clemens Fuest, Felix Hugger, Samina Sultan, Jing Xing
7585 2019
April 2019
Chinese acquisitions abroad: are they different? Clemens Fuest, Felix Hugger, Samina Sultan, Jing Xing
Impressum:
CESifo Working Papers ISSN 2364-1428 (electronic version) Publisher and distributor: Munich Society for the Promotion of Economic Research - CESifo GmbH The international platform of Ludwigs-Maximilians University’s Center for Economic Studies and the ifo Institute Poschingerstr. 5, 81679 Munich, Germany Telephone +49 (0)89 2180-2740, Telefax +49 (0)89 2180-17845, email [email protected] Editor: Clemens Fuest www.cesifo-group.org/wp
An electronic version of the paper may be downloaded · from the SSRN website: www.SSRN.com · from the RePEc website: www.RePEc.org · from the CESifo website: www.CESifo-group.org/wp
CESifo Working Paper No. 7585 Category 1: Public Finance
Chinese acquisitions abroad: are they different?
Abstract In recent years Chinese acquisitions abroad have increased significantly. This paper uses a large dataset on cross-border M&A deals to investigate whether Chinese foreign acquisitions differ from acquisitions coming from other countries. We find that Chinese acquirers buy targets with lower profitability, larger size, higher debt levels, and more patents. However, private and state-owned Chinese investors differ in preferences for location in offshore financial centers, industry diversification, natural resources and technology. Chinese state-owned acquirers are similar to government-led acquirers from other countries in pursuing target firms in the resource extraction industry. Policy initiatives like the Belt and Road Initiative and Made in China 2025 influence investment patterns of Chinese state-owned acquirers but not those of private investors. Surprisingly, for acquisition prices, we find that Chinese investors pay less for firms with similar observable characteristics than investors from other countries.
JEL-Codes: G340, G380, F020.
Keywords: cross-border M&A, China, government acquirers.
Clemens Fuest
Ifo Institute – Leibniz Institute for Economic Research
at the University of Munich / Germany [email protected]
Felix Hugger Center for Economic Studies
University of Munich / Germany [email protected]
Samina Sultan Center for Economic Studies
University of Munich / Germany [email protected]
Jing Xing Antai College of Economics & Management
Shanghai Jiao Tong University, China [email protected]
2
1. Introduction
In recent years Chinese investors have increased their foreign investment activities
significantly, in particular in the form of mergers and acquisitions. In many countries,
especially in the US and in Europe, Chinese acquisitions arouse suspicion. Critics claim
that Chinese acquisitions will lead to an undesirable technology transfer to China, that
Chinese acquirers enjoy unfair advantages because they are subsidized by the Chinese
government or that the acquisitions are motivated strategically, with the objective to
gain market dominance, to increase Chinese political influence in the host countries or
to divide European countries and undermine the coordination of policies towards China
in the European Union.
At the same time there are legitimate reasons for Chinese investments abroad. For a
long time China invested the revenue from its export surplus primarily in US
government bonds. Diversifying foreign investment seems perfectly rational. For many
Chinese firms foreign acquisitions are ways to ensure access to customers or to key
suppliers, in particular raw materials. For private investors foreign acquisitions may
have the purpose to shield assets from seizure by the Chinese government. In addition,
it should not be forgotten that foreign companies have been very active as investors in
China, even if this investment is subject to a number of controversial regulations.
Nevertheless the growth in Chinese acquisitions has led several countries to tighten
regulations and restrictions on foreign acquisitions in general and those from China in
particular, and calls to restrict these investments further and to coordinate these policies
at the EU level are growing louder. This debate is mostly based on speculation and
anecdotes about the risks of these developments, but there is little systematic evidence
about the causes and effects of these investments and whether Chinese investment
differs from investment coming from other countries.
It is the objective of this paper to contribute to fill this gap by providing evidence about
the development of Chinese foreign acquisitions in the last 15 years. We focus on the
question of whether Chinese foreign acquisitions follow patterns which distinguish
them from acquisitions made by other investors, and we ask whether these patterns
might be related to Chinese economic policy strategies like the Belt and Road Initiative
(BRI) or Made in China 2025. We also investigate whether acquisition strategies of
Chinse state-owned enterprises (SOEs) differ from those of private Chinese firms and
whether Chinese investors pay more or less for their acquisitions than investors from
other countries. This sheds some light on the question of whether Chinese investors
benefit from government support, giving them ‘unfair’ advantages and distorting M&A
markets, with potentially adverse economic effects on the host countries.
Our main findings are as follows. First, a surprisingly large number of Chinese
acquisitions takes place in tax havens and offshore financial centers. This applies to
private companies but not to state-owned enterprises. A potential explanation is that
3
there are capital controls in China. Successfully bidding for firms requires the ability to
make international payments at short notice. This may require Chinese companies to
set up holding companies abroad, and tax havens and offshore financial centers may
offer the best way to do so. Second, Chinese acquirers have a stronger presence than
other investors in countries with lower per capita incomes, lower GDP growth and
smaller population sizes. Third, in terms of sectors, Chinese SOEs focus more than
other investors on infrastructure (water, energy) and on mining and commodities.
However, state-owned firms from other countries show a similar pattern. Fourth,
regarding target firm characteristics, Chinese companies tend to acquire companies
with lower profitability, larger sizes, higher levels of debt, and more patents. This
suggests that Chinese investors may have a comparative advantage in access to
financing, possibly a result of policies pursued through the state-owned Chinese
banking system. However, we uncover rich differences between private and state-
owned Chinese acquirers in terms of their preferences for target-level characteristics.
Fifth, rather surprisingly, acquisition prices paid by Chinese investors are lower than
prices paid by other firms. This questions the widespread view that government
subsidies allow Chinese investors to outbid others. Finally, we also find that the Belt
and Road Initiative as well as the Made in China 2025 strategy had an impact, but only
for Chinese SOEs, not for private investors.
2. Related literature
There is a growing literature on cross-border mergers and acquisitions. Existing studies
focus on understanding the determinants and consequences of cross-border M&As.
However, few studies have attempted to examine factors that affect Chinese overseas
M&As. One reason is that these investments have increased only recently.
Using aggregate-level data during 1985-2011, Buckley et al. (2016) examine country-
level factors that affect the location and scale of Chinese overseas M&As. They find
that the location choice of Chinese overseas M&As is affected by factors like bilateral
trade and common language. Chinese acquisitions also tend to occur in countries with
poorer rule of law and higher risk. Nonetheless, Buckley et al. (2016) do not compare
Chinese acquirers with investors from other countries. Therefore, their study does not
answer the question whether Chinese acquisitions are different.
There are several studies on determinants of Chinese outward foreign direct investment
(OFDI) (for example, Buckley et al., 2007; Lu et al. 2011; Kolstad and Wiig 2012).
These studies usually find that market size, geographic and cultural proximity, bilateral
trade, and host countries’ natural resources and political risks affect Chinese OFID.
These findings are broadly in line with the wider literature on determinants of cross-
border FDI or mergers and acquisitions. In particular, various economic, cultural,
regulatory and political factors have been shown to influence cross-border M&A
activities. These include geography, bilateral trade, relative valuation in currencies and
stock market value (Erel et al., 2012); domestic financial market development (di
4
Giovanni, 2005); accounting disclosure and accounting standards (Rossi and Volpin,
2004; Erel et al., 2012); shareholder protection and corporate governance (Rossi and
Volpin, 2004; Kim and Lu, 2013); culture differences (Ahern et al., 2015) and social
attitude (Dinc and Erel, 2013); regulatory arbitrage (Karolyi and Taboada, 2015); and
taxes (Huizinga et al., 2009).
In the literature on Chinese OFID, some studies also compare OFDI conducted by SOEs
with that by private firms (see, Duamu, 2012; Ramasamy et al., 2012; Amighini et al.,
2013; Luo et al., 2017). Some differences are found between SOEs and non-SOEs. For
example, SOEs are less concerned about political risk in the host country, less market
orientated and more resource-seeking. Even in this literature, however, few have
addressed the question whether Chinese OFDI decision is different from that by other
countries.
One important feature of Chinese acquisitions is that many acquirers are closely related
to the government. Even Chinese private acquirers may be connected to or backed by
the government in implicit ways. To the best of our knowledge, Karolyi and Liao (2017)
is the only study that explicitly compares state-backed acquirers and private acquirers.
They find that government backed acquirers are more likely than private acquirers to
come from autocratic countries with higher levels of foreign currency reserves/more
active domestic acquisitions programs, are more likely to pursue targets in countries
with richer natural resources and more potential to diversity their own industrial
structures. Compared with corporate acquirers, government acquirers are also more
likely to complete the deals. However, Karolyi and Liao (2017) find little evidence that
target-level characteristics matter differently for government acquirers.
3. Data and sample construction
We combine data from a number of sources to construct our samples. To obtain deal-
level information, we use Bureau van Dijk’s Zephyr database. Zephyr contains
information on worldwide M&A transactions. It lists information on the date of the deal,
the parties involved, the nature of the deal, methods of payment and the stake acquired.
Each deal has a unique deal number. Each target and acquirer are assigned a unique
identifier by Bureau van Dijk (BvD), which we use to link to BvD’s Orbis database. As
financial information about the targets and acquirers is limited from Zephyr, we use
Orbis to obtain financial and ownership information for both the targets and the
acquirers.
Our full sample contains information on 157,985 completed M&A deals during the
period 2002-2018.5 We exclude deals with multiple acquirers. If a firm acquires several
targets in one deal, we regard each acquirer-target pair as a single transaction. We
categorize deals into domestic and cross-border deals. To identify the location of the
5 We restrict the sample period to deals in or after 2002, as there are few Chinese deals before that year. 2002 was
also the year that China joined the WTO and therefore become more open to global investment.
5
acquirer, we use the location of the acquirer’s global ultimate owner (GUO). Frequently,
the location of the acquirer is the same as the location of its GUO, but in some cases
relying on the location of the acquirer would bias our data due to intricate ownership
structures. 6 We use the target companies’ locations directly to identify their host
countries. Using these definitions of locations for targets and acquirers, we identify
cross-border deals where the target and the acquirer are located in different countries.
We divide our full sample into three groups depending on the type of acquirer: Chinese
private acquisitions, Chinese state-owned acquisitions, and non-Chinese acquisitions.
We use the nature of the GUO of the acquirer to distinguish between Chinese SOEs and
private acquirers. We define a Chinese acquirer to be an SOE if its GUO is state-owned
or state-controlled. Our full sample contains 3,283 completed M&A deals by Chinese
acquirers, of which 1,279 are conducted by SOEs. We use this full sample for basic
descriptive results.
For analyses and estimations, we take a number of steps to further clean the full sample.
First, we focus on mergers and acquisitions where the majority of the target’s shares (at
least 50%) is purchased and exclude deals where the acquired stake is unknown. In
addition, we exclude deals in which the host country of the target is unknown. We also
exclude deals where target has negative or zero value for total assets, turnover or
employees, and if the target’s intangible fixed assets is greater than total assets in the
year one year before the deal. In our estimation sample, acquirers may be involved in
more than one deal,7 but to ensure comparability each target firm is involved in a deal
only once during our sample period. This leaves us with a total of 72,056 deals, of
which 1,168 are conducted by Chinese private investors and 732 by Chinese SOEs.8
We augment the deal and balance sheet data with country-level variables for the target
countries. We collect country-level variables from various sources including the World
Bank’s World Development and Governance Indicators, CEPII, UN Comtrade, KPMG
and EY. From the World Bank’s World Development Indicators (WDI), we obtain
general macroeconomic variables, like GDP, exchange rate, population, and resource
rents. 9 To identify tax havens, we rely on the OECD definition. 10 To measure
institutional quality, we use the World Bank’s World Governance Indicators for the rule
of law, control of corruption, political stability, and regulatory strength. We use CEPII
data for a weighted distance measure from the target country to China. The UN
Comtrade database provides us with trade volume between the target country and China.
6 For instance using the definition acquirer directly, the 2017 acquisition of the Swiss pharma company
Syngenta AG would have been classified as a Non-Chinese deal, as the direct acquirer is located in the
Netherlands. However, using the definition of the GUO, we can identify this deal as a Chinese state-
owned acquisition. 7 Again, we look at global ultimate owners of the acquirers. In total, the sample contains 20,950 global
ultimate owners of acquiring firms. 8 In the group of non-Chinese cross-border deals, most acquirers are from the US (18.55%), Great Britain
(9.7%), Germany (6.2%), and France (5.8%). 9 When information on macroeconomic variables like GDP from the WDI is missing, we use UN data
instead. 10 For a detailed list of tax haven countries see Table A3 in Appendix A.
6
Appendix A provides more details about the definitions and sources for country-level
and deal-level variables.
4. Descriptive statistics
Figure 1 shows the number and value of cross-border M&A deals by different types of
acquirers during 2002-2017, based on the full sample.11 For non-Chinese acquisitions
(Panel A), we observe a peak of transaction in both number of deals and volume around
2006-2007 and a significant drop during the 2008 financial crisis. We observe a gradual
recovery of global cross-border M&As since around 2012. These patterns are consistent
with observations made in other studies (for example, European Commission (2018)).
Thus, Panel A demonstrates that Zephyr provides a consistent picture about the global
M&A market. Panel B shows the time-series evolution of Chinese cross-border
acquisitions, which is rather different from the global trend. In particular, there was no
negative impact of the 2008 financial crisis on Chinese acquisitions. In fact, there was
a surge in the number of Chinese cross-border transactions in 2008. Over time, there is
an increasing trend for both the number and volume of Chinese overseas acquisitions.
In Panels C and D, we distinguish between Chinese private and state-owned acquirers.
These figures reveal that while there are fewer acquisitions by Chinese state-owned
acquirers, they tend to conduct larger deals compared with Chinese private acquirers.
The large spike in 2008 in terms of the number of cross-border deals (as in Panel B) is
largely driven by activities of Chinese private acquirers. For both private and
government acquirers, the total value of acquisitions rises sharply over time. However,
the rise is most prominent for acquisitions by SOEs since 2011.
Figure B1 in Appendix B shows how the number of international M&A deals with
targets located in China developed between 2002 and 2017.12 Foreign investment into
China in the form of M&A deals increased substantially from less than 50 deals in 2002
to 653 deals in 2008. From 2009 to 2017, however, the level of foreign acquisitions of
Chinese companies remained stable. The annual total deal value of foreign acquisitions
in China lies between 8 and 15 billion Euro in most years since 2005. Thus, although
investment flows into China started to rise earlier than outflows, it stagnates over the
last decade while outflows continue to rise.
The percentage of shares acquired varies significantly across deals. Figure 2 allocates
cross-border deals in our full sample by percentages of shares acquired and by acquirer
types. We observe that Chinese SOEs predominantly engage in full acquisitions (going
from 0 to 100 percent ownership). This constitutes 44.6% of all cross-border deals by
Chinese SOEs. Majority acquisitions (from 0 to more than 50 percent ownership) are
also common when Chinese SOEs are involved. In contrast, Chinese private firms tend
11 Deals are assigned to years depending on their date of completion. The year 2018 is not included in
this and similar figures as data is only available on deals until July 2018. 12 These deals are not part of the sample used in any of the other figures or regressions.
7
to increase their stakes gradually, as more than 30% of transactions involve stake
increases. Therefore, Chinese private acquirers appear to follow a more cautious
investment strategy than SOEs.
Using our estimation sample, Table 1 summarizes the number of deals, and the mean
and median deal values by acquirer types. Deal value data is available for about half of
the transactions with Chinese acquirers (private or SOE) and for about one third of the
deals with no relation to China. Table 1 confirms that Chinese SOEs are involved in
larger deals than the other acquirers, which is reflected by substantially larger mean and
median deal values. In contrast, private Chinese acquirers tend to conduct deals of
similar size to non-Chinese acquirers in terms of the median deal value.
Figure 3 looks at the distribution of cross-border deals in different geographical regions
for the three acquirers. A larger share of both Chinese and non-Chinese acquisitions
takes place in Europe. European deals amount to 66.6% of all transactions for non-
Chinese acquirers, 47.5% for Chinese government acquirers, and 38.2% for Chinese
private acquirers. Around 15-20% of all deals are located in North America for all three
types of acquirers. Differences emerge in other regions between Chinese and non-
Chinese acquirers. There are more transactions by Chinese acquirers in East Asia and
Pacific regions, as well as in Latin America and the Caribbean.
Table 2 offers a preliminary look at the distribution of deals by target countries and
acquirer types. We rank the target countries from high to low based on the number of
Chinese private acquisitions. For each target country, we provide the number of deals,
the total deal values and the corresponding percentages. The first outstanding result is
that a large percentage of Chinese private acquisitions occurs in tax haven countries. In
fact, in terms of the number of deals, Virgin Island is on top of the list for private
Chinese acquirers. The preference for tax havens is also found for Chinese SOEs, as
there are substantial activities in places like Virgin Island, Cayman Island and Bermuda.
In contrast, tax havens are less popular with non-Chinese acquirers. In Figure B2, we
plot the distribution of cross-border deals by acquirer types in selected target countries,
where Chinese acquirers’ preference for tax havens is more visible.
Table 2 also shows a geographic preference of Chinese acquirers for Asia and Pacific
countries. Based on the total value of deals, for example, a much higher share of
Chinese acquisitions happen in Australia, Japan, Malaysia, and Singapore. Nonetheless,
there is no indication that Chinese acquirers have a tendency to invest more in BRIC
countries, as their investment pattern in Brazil, Russia, and Indian is not widely
different from that of non-Chinese acquirers.
Finally, Figure 4 shows the distribution of cross-border M&A deals for selected
industries.13 Different patterns are shown for different types of acquirers. Notably,
13 For presentation purposes, we select industries where there are notable differences between the
different acquirer types.
8
Chinese private acquirers appear to particularly target firms in the finance and insurance
industries, which may partly explain their high regional concentration in tax havens.
Chinese SOEs, on the other hand, seem to prefer pursing targets in the manufacturing
sectors and industries related to natural resources such as mining and agriculture.
5. Are Chinese overseas acquisitions different?
The central question we attempt to address in this study is whether Chinese overseas
mergers and acquisitions are driven by different factors, compared with non-Chinese
acquisitions. To shed light on this issue, we employ the deal-level data and estimate the
following Logit regression model:
(1) Pr(𝐶𝑁𝑖,𝑗,𝑡 = 1) = 𝐹(𝛽0 + 𝛽1𝑋𝑖,𝑗,𝑡𝑇 + 𝛽2𝑍𝑖,𝑗,𝑡
𝑇𝐶 + 𝛾′𝑌𝑒𝑎𝑟 𝑑𝑢𝑚𝑚𝑖𝑒𝑠 + 𝜀𝑖,𝑗,𝑡),
where the dependent variable is a dummy indicating whether target i in country j in year
t is purchased by a Chinese acquirer. We also split the sample to differentiate between
private Chinese firms and SOEs.14 𝑋𝑖,𝑗,𝑡𝑇 is a set of target-level characteristics, and
𝑍𝑖,𝑗,𝑡𝑇𝐶 is a set of target country-level characteristics. The coefficients of interest are 𝛽1
and 𝛽2 as we want to know how target country or firm level characteristics influence
the probability of being acquired by a Chinese firm in comparison to international
investors. If a coefficient is not significant, the corresponding characteristic is either
unimportant for all investors or equally important for Chinese and international
investors. Furthermore, we include year fixed effects in all specifications to control for
general trends over time that affect all investors in the same way. In some specifications,
we also control for industry and target country fixed effects. Standard errors are robust
and clustered at the target firm level.
5.1 Effects of target country characteristics
We first examine how target country characteristics affect the probability of a target
being acquired by a Chinese firm as opposed to a non-Chinese investor. In Column 1
of Table 3, we consider a set of country-level economic indicators that are frequently
employed in the literature on determinants of outward FDI or cross-border M&As.
GDPPC is real GDP per capita in the target country; GDP growth is the annual real
GDP growth rate; Distance measures the geographical distance between China and the
target country; Population is the total population of the target country; Trade is the
bilateral trade volume between China and the target country; Inflation is the annual
inflation rate in the target country; Tax Haven is a dummy variable that equals 1 if a
target country is regarded as a tax haven; Resource is total resource rent as a ratio to the
target country’s GDP; ∆Exchange rate is the rate of appreciation or depreciation of the
host country’s currency against Chinese renminbi (RMB), and a positive value stands
14 In the regression for Chinese private companies, acquisitions by Chinese SOEs are excluded from the
sample and vice versa.
9
for appreciation of RMB. We control for year fixed effects throughout Table 3 and
report marginal effects based on the logit estimations.
Column 1 of Table 3 shows that relative to non-Chinese acquirers, Chinese acquirers
tend to conduct acquisitions in countries with lower GDP per capita, lower GDP
growth, and a smaller population.15 As expected, being closer to China and having a
larger trade volume with China both increase the probability of takeovers by Chinese
acquirers. Relative to other investors, Chinese acquirers tend to avoid inflation risk, as
the estimated marginal effect on Inflation is negative and significant. Surprisingly, in
Column 1, we do not find that resource-seeking is a more important determinant for
Chinese acquirers than for others.
Institutional quality of target countries is often thought to be an important factor
influencing cross-border M&As. To investigate whether institutional quality matters
for Chinese overseas acquisitions in the same way as for non-Chinese deals, we control
for various measures of institutional quality in Columns 2-5 of Table 3. Specifically,
we control for political stability in Column 2, regulatory quality in Column 3, rule of
law in Column 4, and control of corruption in Column 5. Throughout these columns,
however, we find little evidence that institutional quality affects decisions of Chinese
overseas acquisitions differently, since the estimated coefficients on all four indicators
are insignificant.
In Columns 6 and 7, we distinguish between Chinese private and state-owned acquirers.
While the two types of Chinese acquirers are roughly similar in other dimensions, they
are substantially different in two ways. First, relative to non-Chinese acquirers, Chinese
private acquirers are significantly more likely to conduct acquisitions in tax havens, all
else equal. In contrast, we do not find a significant difference between Chinese state-
owned acquirers and international acquirers in this dimension. Second, the two types of
Chinese acquirers are different in their preferences for natural resources. While Chines
private acquirers appear to avoid targets in resource-rich countries, Chinese state-
owned acquirers are more likely to purchase targets in such countries. Thus, resource
seeking seems a motivation for Chinese state-owned acquirers alone. In unreported
exercises, we also include the four indicators of institutional qualities and compare the
two types of Chinese acquirers with international acquirers. We continue to find that
institutional quality does not affect Chinese acquirers differently, regardless of their
ownership type.
5.2 Industry differences
Using deal-level information, we are also able to investigate whether Chinese
acquisitions are drawn to targets in specific industries. In Table 4, we include a set of
target industry dummies based on the NACE industry classification, in additional to a
15 The result is robust to using GDP instead of population as a measure for market size.
10
basic set of macroeconomic control variables.16 We use NACE industries 77 to 99 as
the reference group.17 We then report the estimated marginal effects on the industry
dummies.18
Column 1 in Table 4 shows that the probability of a takeover by Chinese firms varies
strongly across industries. For example, firms in the information and communication
industry are less likely to be targeted by Chinese acquirers compared to international
investors. In contrast, Chinese investors are keen on target firms in certain
manufacturing industries, such as manufacturing of electronics, machinery and
vehicles. Consistent with resource seeking, Chinese acquirers are also more likely to
conduct deals in mining.
In addition in columns 2 and 3, we differentiate between private and state-owned
Chinese companies, which reveals different investment patterns between the two.
Agricultural firms, for example, are more likely to be acquired by Chinese SOEs, but
the opposite is true for private Chinese firms. The same pattern holds for targets in
supply of energy, water, and gas, construction firms and companies from the mining
sector. These results are consistent with the previous finding that Chinese state-owned
acquirers are particularly attracted to natural resources and strategic sectors abroad.
Additionally a comparison of columns 2 and 3 reveals that within the manufacturing
sector the two types of Chinese acquirers display different preferences for specific
industries.
Generally speaking, we find that the investment pattern of Chinese cross-border
acquisitions exhibit some notable differences compared to international investors in
terms of targeted industries. However, there seems to be a greater distinction between
Chinese SOEs and investors elsewhere. We return to this issue in section 5.4, where we
compare Chinese SOEs to government-led acquirers from other countries.
5.3 Effects of target characteristics
Next, we consider target-level characteristics that may affect the probability of cross-
border acquisitions, again comparing Chinese and non-Chinese acquirers. We consider
the following target-level characteristics: Industry Diversity is a dummy that equals 1
if the target and the acquirer belong to different industries; Size is the natural logarithm
of total assets of the target firm; ROA is profit/loss before taxes over total assets;
Leverage is the ratio of total debt, consisting of loans and long-term debt, as to total
assets; Asset Growth is the annual growth rate of total assets; Intangibles is the
16 This includes GDP per capita, GDP growth, population, distance and bilateral trade. 17 This includes administrative and support service activities, public administration and defense,
compulsory social security, education, human health and social work activities, arts, entertainment and
recreation, and other service activities. 18 Note that Table 4 only shows a selection of industries. The following industries are dropped:
Accommodation & food service, manufacture of food, textiles, paper products and other manufacturing,
real estate, transportation & storage, and wholesale & retail.
11
percentage of intangible assets in total assets19; and Patents is the number of patents the
target company holds. All variables except Patents are measured in the year before
acquisition and are winsorized at the 1% level. We report descriptive information and
the estimation results in Tables 5 and 6. We control for the basic set of macroeconomic
variables and year fixed effects throughout different specifications. We further control
for industry and target country fixed effects in certain specifications as robustness
checks.
In Table 5, we report the sample means of these target-level characteristics for targets
acquired by different types of investors. We also conduct a t-test to formally examine
whether the sample means of target-level factors are equal between different types of
acquirers. The descriptive statistics immediately show some interesting heterogeneities.
Relative to non-Chinese investors, Chinese investors purchase larger targets (measured
by total assets) and targets with a lower return on assets. If we consider the groups of
Chinese private and Chinese SOEs separately, the only significant difference between
SOEs and non-Chinese investors is the size of the targets. However, Chinese private
investors tend to purchase targets with significantly lower ROA and more patents.
Next, we formally analyze whether Chinese investors are different from other investors
by including these target-level characteristics in the Logit model, as specified by
Equation 1. Table 6 shows several interesting results. Columns 1-3 suggest that Chinese
acquirers prefer targets in industries from their own, with lower profitability, larger size,
a higher level of debt, and more patents. If a target is in a different industry from the
acquirer’s, the probability of this target being acquired by a Chinese investor increases
by 0.6-0.8%. A 10 percentage point reduction in ROA would increase the probability
of Chinese acquisition by around 0.2%. A 10 percentage point increase in target
leverage leads to around 1.3% increase in the probability of Chinese acquisition.
Interestingly, we find a positive marginal effect associated with the number of patents
the target has for Chinese private acquirers. If the number of patents the target firm has
increases by one standard deviation, this increases the probability of acquisition by
Chinese private investors by around 0.2%.20 Considering that only 2.6% of cross-
border acquisitions are made by Chinese investors in our sample, these estimated
marginal effects from ROA, leverage, and patents are rather large.21 We also find a
positive and significant marginal effect on target size, but the magnitude of the effect is
much smaller. Based on estimations in Columns 1-3, doubling the size of the target
increases the probability of Chinese acquisition by around 0.6%. Neither Asset Growth
nor Intangibles matter differently for Chinese acquirers relative to international
acquirers.
19 Intangible assets are defined as formation expenses, intangible fixed assets and goodwill arising on
consolidation. 20 The distribution of patents target firms have is highly skewed. One standard deviation equals to around
200 patents. 21 We use the Stata command firthlogit to correct for potential bias due to the small probability of
Chinese acquisitions in our sample, and the results are rather similar to logit estimation results.
12
We focus on Chinese private acquirers in Columns 4-6, and state-owned acquirers in
Columns 7-9. There, we uncover interesting and remarkable differences between SOEs
and private firms. The preference for industry diversification, as found in Columns 1-
3, is mainly driven by Chinese state-owned Chinese acquirers. In contrast, there is no
significant difference between Chinese private acquirers and others in terms of industry
diversification. Chinese state-owned acquirers favor larger targets in terms of assets.
While the estimated marginal effect on Size is also positive for Chinese private
acquirers, the effect is less robust in different specifications. The preference for highly
leveraged targets and those with patents is mainly driven by Chinese private acquirers.
These results imply that Chinese private acquirers are more likely to purchase targets
in financial distress, and that access to technology and knowledge may be a particularly
important determinant for them. Both Chinese private and state-owned acquirers prefer
targets with lower ROA, and this result is rather robust to different specifications.
To summarize, employing target-level characteristics, we find that Chinese acquirers
tend to purchase targets with a larger size, more patents, poorer performance as
measured by ROA and higher levels of debt. Chinese acquirers are also different from
international acquirers in terms of their preferences for industry diversification.
Nevertheless, there are rich heterogeneities between SOEs and private investors.
5.4 Comparison between Chinese and non-Chinese state-owned acquirers
We have so far uncovered some significant differences between Chinese and non-
Chinese acquirers. In particular, Chinese state-owned acquirers are more likely to
acquire targets from resource-rich countries, and targets with a larger size and a lower
return on assets. Chinese state-owned acquirers are also keen on industry diversification
through cross-border acquisitions. The majority of non-Chinese acquirers in our sample
are corporate acquirers. Thus, one interesting question is whether Chinese SOEs are
also different compared to state-owned acquirers from other countries.
We identify 619 non-Chinese state-owned acquirers with cross-border acquisitions,
using the ownership nature of the global ultimate owner in Orbis.22 We obtain basic
country-level characteristics for 522 non-Chinese government acquirers, which are
included in our estimations. We then run a Logit estimation where the dependent
variable equals 1 if a target is acquired by a Chinese state-owned acquirer, and 0 if it is
purchased by a non-Chinese government-led acquirer. We report the marginal effects
based on the Logit estimations in Table 7. In different columns, we use different
specifications for the year and target-country fixed effects.
In Columns 1-2, we consider target country-level characteristics that were previously
found to matter for Chinese state-owned acquirers. In Columns 3-4, we add three target
level characteristics: the indicator when the target and the acquirer are in different
22 In Orbis, the entity type of the global ultimate owner of such acquirers is labeled as “Public authority,
state, or government”.
13
industries, target size and its ROA prior to the acquisition. These three target-level
characteristics matter most for Chinese state-owned acquirers, when we compare them
to international investors previously. One caveat is that we end up with a much smaller
sample size in the last two columns, since we do not observe target-level characteristics
for the majority of government acquisitions.
While the estimated marginal effects on certain factors vary across different columns
due to changes in specifications and sample sizes, two robust results remain. That is,
Chinese state-owned acquirers are more likely to acquire larger targets and those with
poorer pre-deal financial performances, as measured by ROA. These pattern is
consistent with previous findings when we use a broader set of non-Chinese acquirers
as the control group.
Interestingly, relative to non-Chinese state owned acquirers, Chinese state-owned
acquirers no longer appear to be particularly focused on natural resources, and there is
only weak evidence in Column 3 that they are especially keen on industry
diversification. Karolyi and Liao (2017) find that government-led acquirers, in general,
are more orientated towards targets in resource rich countries, and targets with the
potential to diversity their own industry portfolio. Thus, our results show that Chinese
state-owned acquirers are no different from other government acquirers in these
dimensions.
5.5 Effects of policy initiatives
Another aspect in which Chinese acquisitions may differ from others is that they are
likely to be influenced by the strategic policy initiatives of the Chinese government.
Most notably, China announced the Belt and Road Initiative (BRI) in 2013 by the
Chinese President Xi Jinping and Made in China 2025 by Prime Minister Li Keqiang
in 2015. Do these policy initiatives have a material impact on Chinese overseas
acquisitions?
We first analyze the impact of the Belt and Road Initiative (BRI). The initial aim of BRI
is to improve trade, infrastructure and investment links between China and 65 countries
in Central, South, and South East Asia, Europe, the Middle East and North Africa.23
We use a Difference-in-differences approach to test whether the BRI changes the
regional focus of Chinese overseas acquisitions. To do so, we first construct a dummy
PostBRI, which equals 1 if the cross-border deal happened in or after 2013 and 0
otherwise. We also construct a dummy BRI, which equals 1 for 65 BRI countries
narrowly defining the outreach of the BRI initiative according to the China International
Trade Institute. Table 7 reports the result.
Table 8 shows that before 2013, Chinese acquirers were less likely to pursue targets in
BRI countries, as the estimated coefficient on BRI is negative and statistically
23 The list of BRI countries is provided in Table C1 in Appendix C.
14
significant. For Chines private acquirers, the Belt and Road Initiative fails to encourage
them to acquire targets in BRI countries since 2013, as the estimated coefficient on
BRI*PostBRI is insignificant. In contrast, the estimated coefficient on BRI*PostBRI is
positive and highly significant for Chines state-owned acquirers. These results suggest
that Belt and Road Initiative only influences state-owned acquirers.
Next, we examine the impact of another important policy initiative, Made in China
2025. This initiative defines 10 industries where the Chinese government wants
Chinese companies to become globally competitive. One way to reach that goal is
through take-overs of foreign firms in these industries. Again, we use the Difference-
in-differences estimator to investigate whether the policy has material impact on
Chinese overseas acquisitions. We construct a dummy variable CN2025 that equals 1
for industries that are related to the Made in China 2025 initiative.24 We construct
another dummy PostCN2025, which equals 1 since 2015. We then interact CN2015
with PostCN2025 in the Difference-in-differences estimations.
Table 9 reports the estimation results. There is little evidence that Chinese acquisitions
occurred more frequently in industries targeted by Made in China 2025 before 2015,
relative to non-Chinese acquisitions. Interestingly, targets in these industries become
significantly more likely to be purchased by Chinese SOEs after the policy was
introduced. Again, the policy fails to motivate Chinese private acquirers.
To summarize, we analyze the impact of government policy initiatives on Chinese
overseas acquisitions. We focus on whether government preferences manifested in its
policy guidance influences the location and industry decisions of Chinese acquirers. We
do not find that Chinese private acquirers are much affected by such policy orientations.
In contrast, state-owned acquirers appear to be adjust their investment activities in line
with the government’s intentions.
6. The deal value
In this section, we investigate whether acquisition prices are different if Chinese
investors are involved. Specifically we investigate whether Chinese acquirers pay more
or less for target companies with similar observable characteristics. There is a
widespread view that the Chinese government might support Chinese companies
investing abroad to make sure that strategic objectives like market access or technology
transfer are reached. This would suggest that Chinese companies may overpay relative
to other investors.
One challenge is that the majority of target firms in our sample is unlisted.25 As a result,
24 The industries connected to the Made in China 2025 initiative are defined relatively broadly. We match
the official announcement as closely as possibly with the industry codes provided by Eurostat. Table C2
in Appendix C provides the list. 25 Around 95% of target firms in our sample are unlisted.
15
we do not observe the total number of shares and the market value of equity of the target
firms. It is therefore impossible to calculate price per share before the transaction and
the take-over premium as it is usually done in the literature. Alternatively, we first
calculate 𝑃𝑟𝑖𝑐𝑒𝑖,𝑗,𝑡, which is the amount the acquirer paid for 1% of equity of target firm
i in country j in year t.26 We then estimate Equation 2 as below:
(2) 𝑃𝑟𝑖𝑐𝑒𝑖,𝑗,𝑡 = 𝛼 + 𝛽1𝐶𝑁𝑖,𝑗,𝑡 + 𝛽2𝐸𝑞𝑢𝑖𝑡𝑦𝑖,𝑗,𝑡 + 𝛽3𝑅𝑂𝐴𝑖,𝑗,𝑡 + 𝛽4𝐹𝑢𝑙𝑙 𝐴𝐶𝑖,𝑗,𝑡 + 𝛽5𝐴𝑛𝑦 𝑃𝑎𝑡𝑒𝑛𝑡𝑖,𝑗,𝑡
+ 𝛾𝑍′𝑖,𝑗,𝑡 + 𝑡𝑖𝑚𝑒 𝑒𝑓𝑓𝑒𝑐𝑡𝑠 + 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐹𝐸 + 𝑇𝑎𝑟𝑔𝑒𝑡 𝑐𝑜𝑢𝑛𝑡𝑟𝑦 𝐹𝐸 + 𝜀𝑖,𝑗,𝑡
In Equation 2, CN is a dummy that equals 1 if the acquirer is a Chinese firm. To
differentiate between Chinese private and state-owned acquirers, we instead include in
some specifications a dummy CNpriv that equals 1 if the acquirer is a Chinese private
firm, and CNSOE is a dummy that equals 1 if the acquirer is a Chinese state-owned or
state-controlled firm. ROA and Book equity are the average value of return on assets
and book value of equity over the three years prior to the deal.27 Full AC is a dummy
variable indicating whether 100% of the target were acquired. Any Patent is a dummy
indicating whether the target firm holds any patent. We include a set of country-level
variables, 𝑍′𝑖,𝑗,𝑡, as further controls. We control for industry and target-country fixed
effects, and allow for different time fixed effects in different specifications.
Table 10 shows the estimation results based on Equation 2. As expected, larger book
equity or return on assets increases the payment for the target for all types of acquirers.
This result is robust throughout different columns in Table 10. We find positive point
estimates on Full AC and Any Patent in most specifications, but these estimates are not
always significant. Most importantly, controlling for these observable characteristics,
we find a tendency of underpayment by Chinese acquirers relative to non-Chinese
investors (Columns 1, 5, and 7).
When we distinguish between private and state-owned/controlled Chinese firms, we
find that the pattern of underpayment is driven by private Chinese firms—the estimated
coefficient on CNpriv is negative and highly significant in Columns 2, 4, 6 and 8. This
means that private Chinese firms pay less than other international investors for targets
with similar characteristics. While the coefficient for state-owned Chinese firms is also
negative, it is not statistically significant.
These results question the theory that Chinese systematically investors outbid others,
driven by hidden government support. There are different possible explanations for the
negative price effect. First, it could be that Chinese (private) investors pursue smarter
investment strategies or are better negotiators. Of course, this raises the question of why
they should be smarter than others. Second, our analysis of the target characteristics has
26 This is to account for the fact that not all acquirers in our sample bought 100% of the target firm. 27 We control for equity value instead of total assets because acquirers purchase the equity of the target
firm, which is different from asset acquisition. However, our result is robust to controlling for total assets
instead of equity.
16
shown that Chinese investors focus on companies with higher debt, potentially because
they have easier access to financing through the state owned Chinese banking system.
This advantage could translate into lower acquisition prices as other investors tend to
avoid highly indebted targets. Yet another possible explanation is that Chinese investors
put more emphasis on observable balance sheet indicators, as opposed to less visible
but relevant indicators like the business model or future prospects of the target. This
would imply that the lower prices are no real advantage for the acquirers. Finally, one
cannot exclude the possibility that some non-Chinese acquirers may avoid competing
too aggressively with Chinese investors in takeover bids because they are concerned
that this might have a negative impact on their own operations in China.28 This would
indeed imply that Chinese foreign investment policies would be unfair. Of course, such
a theory would have to be supported by more evidence to be credible.
7. Conclusions
Chinese overseas acquisitions have increased significantly in the last decade, and this
increase gives rise to concerns in particular in Europe and the US about potential
strategic objectives the Chinese government might pursue through foreign acquisitions.
In this study, we examine whether Chinese acquisitions differ from acquisitions of other
international investors. We use a comprehensive dataset for cross-border acquisitions,
and analyze how characteristics of the host countries and targets affects decisions of
Chinese acquirers. Our results indicate that Chinese acquirers are more attracted to
targets located in countries with lower GDP per capita, lower GDP growth and smaller
population. Chinese acquirers prefer targets with larger sizes, poorer profitability,
higher levels of debt and more patents. They also prefer targets located in industries
different from their own. However, we find that private and state-owned Chinese
acquirers have distinct features in their preferences for using tax havens, natural
resources, industry diversifications and patents. While some of these differences
persists between Chinese and non-Chinese acquirers, others reflect different motives
for corporate and government-owned acquirers.
We also show that government policies affect decisions of Chinese acquirers. We find
strong evidence that government initiatives like the Belt and Road Initiatives and Made
in China 2025 have a significant impact on the investment pattern of Chinese SOEs, in
terms of their regional and industry focus. However, these policies fail to have the same
influence on Chinese corporate acquirers.
Finally, we analyze acquisition prices, motivated by the widespread view that
government support might enable Chinese companies to outbid other investors, which
would suggest that Chinese acquirers pay higher prices. Surprisingly we find that prices
paid by Chinese acquirers for firms with similar characteristics, as far as they are
28 In principle governments from all countries could try to pursue this strategy. But in other countries the
separation between private companies and governments is clearer, and a certain size and market power
would be needed, which countries other than China and the US do not have.
17
directly observable in financial data, are actually lower. This might be related to the
focus of Chinese investors on firms with higher debt levels, or to a different weight
international investors put on observable characteristics like balance sheet items and
less directly observable characteristics like the business model or future prospects of
the target firms.
What are the implications of our results for the debate about changing policies of host
countries towards Chinese acquisitions? Our analysis does confirm that Chinese
investment is different from investment coming from other countries. It is influenced
by strategic initiatives of China like the Belt and Road Initiative and Made in China
2025. In addition, the observation that Chinese firms focus on targets with higher debt
and low profitability suggests that Chinese firms might have easier access to finance
than other investors, giving them a competitive advantage. But whether this beneficial
or harmful for the host countries is an open question. To shed more light on this it is
necessary to investigate, among other things, how firms develop after they have been
taken over by Chinese investors. We intend to investigate post-acquisition
performances for Chinese acquisitions in future research.
18
References
Ahern, K. R., Daminelli, D., & Fracassi, C. (2015). “Lost in translation? The effect of cultural
values on mergers around the world,” Journal of Financial Economics, 117 (1), 165–189.
Amighini, A., Rabellotti, R., and Sanfilippo, M. (2013). “Do Chinese state-owned and private
enterprises differ in their internationalisation strategies?” China Economic Review, 27, 312–
325.
Buckley, P. J., Clegg, L. J., Cross, A. R., Liu, X., Voss, H., and Zheng, P. (2007). “The
determinants of Chinese outward foreign direct investment,” Journal of International Business
Studies, 38(4), 499–518.
Buckley et al. (2016). “The Institutional Influence on the Location Strategies of Multinational
Enterprises from Emerging Economies: Evidence from China’s Cross-border Mergers and
Acquisitions,” Management and Organization Review, 12(3): 425–448.
Conrad, B., Ives, J., Meissner, M., Wübbeke, J., Zenglein, M. J. (2016). “Made in China 2025
- The making of a high-tech superpower and consequences for industrial countries,” Mercator
Institute for China Studies.
Cosentino, B., Dunmore, D., Ellis, S., Preti, A., Ranghetti, D., Routaboul, C. (2018). “Research
for TRAN Committee: The new Silk Route - opportunities and challenges for EU transport,”
Directorate-General for Internal Policies, European Parliament’s Committee on Transport and
Tourism.
di Giovanni, Julian. (2005). “What drives capital flows? The case of cross-border M &A
activity and financial deepening,” Journal of International Economics, 65(1): 127-149.
Duanmu, J. L. (2012). “Firm heterogeneity and location choice of Chinese multinational
enterprises,” Journal of World Business, 47, 64–72.
Dinc, I. S., and Erel, I. (2013). “Economic nationalism in mergers and acquisitions,” Journal of
Finance,
Erel, Isil, Liao, Rose C., and Weisbach, Michael S. (2012). “Determinants of cross-border
mergers and acquisitions,” Journal of Finance, 67 (3): 1045–1082.
European Commission (2018). “Commission Staff Working Document on the Movement of
Capital and the Freedom of Payments,” available at
https://ec.europa.eu/info/sites/info/files/2018-capital-market-monitoring analysis.pdf
19
Huizinga, Harry P., and Johannes Voget. (2009). “International Taxation and the Direction and
Volume of Cross Border M&As,” Journal of Finance, 64(3): 1217-1249.
Karolyi and Liao. (2017). “State capitalism‘s global reach: Evidence from foreign acquisitions
by state-owned companies,” Journal of Corporate Finance, 42: 367–391.
Karolyi and Taboada. (2015). “Regulatory arbitrage and cross‐border bank acquisitions,”
Journal of Finance.
Kim, E. H., and Lu, Y. (2013). “Corporate governance reforms around the world and cross-
border acquisitions,” Journal of Corporate Finance, 22, 236– 253.
Kolstad, I., and Wiig, A. (2012). “What determines Chinese outward FDI?” Journal of World
Business, 47(1), 26–34.
Lu, J., Liu, X., and Wang, H. (2011). “Motives for outward FDI of Chinese private firms: Firm
resources, industry dynamics, and government policies,” Management and Organizational
Review, 7(2), 223–248.
Luo, L., Qi, Z, and Hubbard, P. (2017). Not looking for trouble: Understanding large-scale
Chinese overseas investment by sector and ownership,” China Economic Review, 46: 142-164.
OECD (2000). “Towards Global Tax Co-operation – Report to the 2000 Ministerial Council
Meeting and Recommendations by the Committee on Fiscal Affairs.”
Ramasamy, B., Yeung, M., and Laforet, S. (2012). “China’s outward foreign direct investment:
Location choice and firm ownership,” Journal of World Business, 47 (1), 17–25.
Rossi, S., Volpin, P.F. (2004). “Cross-country determinants of mergers and acquisitions,”
Journal of Financial Economics, 74 (2), 277–304.
20
Figure 1: Number and Value of Deals by Deal Category (full sample)
Notes: This figure shows the development of the number and value of deals over the sample period 2002-2017.
We thereby differentiate between different deal categories depending on the nature of the acquirer: Non-Chinese
acquirers (Panel A), Chinese acquirers (Panel B). We furthermore decompose Chinese acquirers into private firms
(Panel C) and state-owned firms (Panel D). The number of deals is reported in the right hand scale and the value
of deals, which is measured in current billion Euros, is reported in the left hand side.
21
Figure 2: Type of Deals by % shares acquired (full sample)
Notes: This figure shows the share of different type of deals for the three deal categories of interest. We
thereby differentiate deals by the percent of shares acquired. Full means that 100% of the target firm
were acquired. Majority means that at least 50% but less than 100% were acquired. Minority means that
less than 50% were acquired. Stake increased means that the acquirer already owned a share of the target
firm, but increased this share. Unknown means that we do not know how much of the target firm was
acquired.
Figure 3: Distribution of cross-border M&As by target regions
Notes: This figure shows the distribution of cross-border M&As by different target regions. The category
“Other” includes countries in Central Asia, Sub-Saharan Africa, Middle East & North Africa, and South
Asia.
22
Figure 4: Distribution of cross-boder M&As by target industries
Notes: This figure shows the distribution of cross-border M&As for a selection of target industries. It is
based on the NACE industry classification.
23
Table 1: Summary Statistics by acquirer types based on the estimation sample
Acquirer type
Number of deals Mean deal
value
(in €million)
Median deal value
(in €million) All
With deal
value
CN private 1,168 577 159.0 20.0
CN SOE 732 391 394.3 54.6
Non-CN 70,156 21,038 263.8 23.1
Total 72,056 22,006 263.4 23.0
Notes: This table shows summary statistics by acquirer types based on the sample used for estimations.
For number of deals we differentiate whether the deal value is known or not.
Table 2: Number of deals and deal value by target countries and acquirer types
Number of deals Total deal value (in €million)
By Target
Country
Non-Chinese
firms
Private Chinese
firms
SOEs Non-Chinese firms Private Chinese
firms
SOEs
Count Percent Count Percent Count Percent Value Percent Value Percent Value Percent
Virgin Isl. 553 0.79% 223 19.09% 54 7.38% 21,147 0.38% 12,246 13.35% 8,168 5.3%
US 9,885 14.09% 138 11.82% 90 12.30% 2,000,000 35.35% 25,382 27.68% 4,814 3.12%
Gr. Britain 10,105 14.40% 104 8.90% 61 8.33% 890,037 16.04% 4,666 5.09% 18,766 12.17% Germany 4,897 6.98% 84 7.19% 87 11.89% 197,819 3.56% 1,499 1.64% 2,922 1.9%
Cayman Isl. 271 0.39% 76 6.51% 24 3.28% 44,624 0.8% 4,383 4.78% 9,935 6.44%
Singapore 682 0.97% 47 4.02% 24 3.28% 35,446 0.64% 3,992 4.35% 6,067 3.94% Australia 2,118 3.02% 46 3.94% 46 6.28% 151,910 2.74% 4,718 5.14% 14,996 9.73%
France 3,032 4.32% 34 2.91% 19 2.60% 174,697 3.15% 115 0.13% 2,247 1.46%
Italy 1,720 2.45% 23 1.97% 22 3.01% 79,307 1.43% 2,229 2.43% 355 0.23%
Japan 280 0.40% 23 1.97% 8 1.09% 24,064 0.43% 1,789 1.95% 567 0.37%
Netherlands 3,234 4.61% 23 1.97% 27 3.69% 294,542 5.31% 3,148 3.43% 1,641 1.07%
Spain 3,144 4.48% 22 1.88% 41 5.60% 105,120 1.89% 2,942 3.21% 2,103 1.36% Malaysia 475 0.68% 21 1.80% 19 2.60% 6,989 0.13% 2,542 2.77% 489 0.32%
Bermuda 142 0.20% 20 1.71% 12 1.64% 55,982 1.01% 7,972 8.69% 3712 2.41%
Canada 2,602 3.71% 16 1.37% 23 3.14% 237,480 4.28% 358 0.39% 17,450 11.32% Belgium 1,482 2.11% 14 1.20% 5 0.68% 73,004 1.32% 228 0.25% 1,825 1.18%
India 239 0.34% 13 1.11% 6 0.82% 8,802 0.16% 867 0.95% 13 0.01%
Switzerland 1,296 1.85% 12 1.03% 8 1.09% 108,013 1.95% 4,364 4.76% 38,090 24.71%
Russia 1,727 2.46% 11 0.94% 4 0.55% 98,120 1.77% 76 0.08% 2,735 1.77%
Brazil 1,005 1.43% 8 0.68% 15 2.05% 60,706 1.09% 940 1.03% 2,002 1.3%
RoW 21,267 30.32% 210 17.99% 137 18.7% 919,871 16.57% 7,250 7.9% 15,259 9.89%
World 70,156 100% 1,168 100% 732 100% 5,587,688 100% 91,713 100% 154,167 100%
Notes: This table shows the number of deals and total deal value by target countries and acquirer type.
Total deal value is reported in current €million.
24
Table 3: Target country characteristics and probability of Chinese acquisitions
Probability of being All Chinese acquirers CN Private CN state-owned
acquired by
GDPPC -0.009*** -0.008*** -0.007*** -0.008*** -0.009*** -0.005*** -0.004***
(0.001) (0.001) (0.002) (0.002) (0.002) (0.001) (0.001)
GDP growth -0.001* -0.001* -0.000 -0.001* -0.001* -0.000 -0.001***
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Distance -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.001*** -0.000*
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Population -0.009*** -0.009*** -0.009*** -0.009*** -0.009*** -0.005*** -0.004***
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Trade 0.013*** 0.014*** 0.013*** 0.013*** 0.013*** 0.008*** 0.005***
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Inflation -0.001** -0.001** -0.001** -0.001** -0.001** -0.000* -0.000
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Tax haven 0.016*** 0.015*** 0.014*** 0.015*** 0.016*** 0.014*** 0.000
(0.005) (0.005) (0.005) (0.005) (0.005) (0.003) (0.005)
Resource -0.010 -0.014 -0.020 -0.018 -0.010 -0.039*** 0.021**
(0.018) (0.019) (0.020) (0.020) (0.019) (0.015) (0.010)
∆Exchange rate -0.003 -0.004 -0.002 -0.003 -0.003 0.012** -0.014***
(0.009) (0.009) (0.009) (0.009) (0.009) (0.006) (0.004)
Political stability
-0.002
(0.002)
Regulatory quality
-0.003
(0.002)
Rule of law
-0.002
(0.002)
Control of Corruption
-0.000
(0.001)
Year FE Yes Yes Yes Yes Yes Yes Yes
No. of observations 63,085 63,085 63,085 63,085 63,085 62,536 62,373
Notes: In this table, we consider how target country-level economic and institutional characteristics affect
the likelihood of Chinese cross-border acquisition. We report the marginal effects from Logit estimations.
Standard errors are robust and clustered at the target firm level. * p < 0.10, ** p < 0.05, *** p < 0.01.
25
Table 4: Target industries and probability of Chinese acquisitions
Probability of being (1) (2) (3)
acquired by All Chinese acquirers CN Private CN state-owned
Agriculture 0.010 -0.009* 0.019**
(0.009) (0.005) (0.008)
Construction 0.001 -0.003 0.005*
(0.004) (0.003) (0.003)
Energy, water, gas. 0.004 -0.003 0.006**
(0.005) (0.004) (0.003)
Finance & Insurance 0.004 -0.001 0.005***
(0.003) (0.002) (0.002)
Information & Comm. -0.008*** -0.007*** -0.001
(0.002) (0.002) (0.001)
M. chem/oil, pharma -0.002 -0.007** 0.005**
(0.003) (0.003) (0.002)
M. elec & machinery 0.023*** 0.007** 0.017***
(0.004) (0.003) (0.003)
M. metal products 0.011** -0.006* 0.017***
(0.005) (0.003) (0.004)
M. vehicles 0.048*** 0.017** 0.033***
(0.010) (0.008) (0.008)
Mining 0.016*** -0.007** 0.025***
(0.006) (0.003) (0.005)
Prof./scientific/techn. activ -0.004 -0.008*** 0.003**
(0.003) (0.002) (0.001)
Macro Controls Yes Yes Yes
Year FE Yes Yes Yes
No. of observations 62,353 61,723 61,373
Notes: In this table, we consider how target industries affect the likelihood of Chinese cross-border
acquisition. Classification of industries is based on NACE industry classification. NACE industries 77
to 99 are used as the reference group. Table 4 shows a selection of industries. The following industries
are dropped: Accommodation & food service, manufacture of food, textiles, paper products and other
manufacturing, real estate, transportation & storage, and wholesale & retail. We report the marginal
effects from Logit estimations based on Equation 1. Standard errors are robust and clustered at the target
firm level. * p < 0.10, ** p < 0.05, *** p < 0.01.
26
Table 5: Target-level characteristics by acquirer types
Variable Non-
Chinese
All CN T-test of equal means (p-
value)
CN
SOE
T-test of equal means (p-
value)
CN
private
T-test of equal means (p-
value)
Total assets 101,189 702,026 0.0000 858,103 0.000 581,966 0.000
Leverage 0.191 0.256 0.7779 0.252 0.862 0.260 0.824
ROA 0.003 -0.045 0.0317 0.002 0.982 -0.080 0.005
Intangibles 0 .050 0.056 0.3705 0.055 0.570 0.056 0.483
Asset growth 14.258 4.942 0.8462 1.003 0.853 8.142 0.925
Patents 4.927 22.357 0.0003 8.927 0.568 30.819 0.000
Notes: In this table, we report the sample means of target-level characteristics, including size, leverage, return on assets (ROA), intangibility, asset growth, and number of
patents. We report the sample means of each variable for targets acquired by non-Chinese, all Chinese, Chinese state-owned, and Chinese private investors, separately. We also
test null hypothesis that the sample means of each variable are equal between targets acquired by non-Chinese and all Chinese investors, the null hypothesis that the sample
means of each variable are equal between targets acquired by non-Chinese and Chinese state-owned investors, and the null hypothesis of equal sample means between targets
acquired by non-Chinese and Chinese private investors. We report the p-values from the associated t-tests. For definitions of target-level characteristics, see Table A2.
27
Table 6: Targets’ financial characteristics and probability of Chinese acquisitions
Pro(All Chinese acquirers) Pro(CN Private) Pro(CN state-owned)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Industry Diversity 0.008** 0.006 0.007* 0.002 0.001 0.001 0.006** 0.006* 0.006*
(0.003) (0.004) (0.004) (0.002) (0.003) (0.003) (0.003) (0.003) (0.003)
Size 0.007*** 0.006*** 0.006*** 0.002*** 0.001* 0.001 0.006*** 0.005*** 0.005***
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
ROA -0.018*** -0.020*** -0.020*** -0.006* -0.008* -0.009* -0.013*** -0.014*** -0.013***
(0.004) (0.005) (0.005) (0.004) (0.004) (0.005) (0.003) (0.004) (0.004)
Leverage 0.007 0.014** 0.013** 0.006 0.009** 0.007 0.001 0.006 0.008
(0.005) (0.006) (0.006) (0.004) (0.005) (0.005) (0.004) (0.005) (0.005)
Asset Growth -0.004 -0.004 -0.004 -0.002 -0.002 -0.001 -0.003 -0.002 -0.002
(0.003) (0.004) (0.004) (0.003) (0.003) (0.003) (0.002) (0.003) (0.003)
Intangibles -0.005 -0.001 0.001 -0.003 0.001 0.005 -0.002 -0.003 -0.003
(0.013) (0.015) (0.015) (0.009) (0.011) (0.011) (0.010) (0.013) (0.014)
Patents 0.001** 0.001*** 0.001*** 0.000*** 0.001*** 0.001*** 0.000 -0.000 -0.000
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.001) (0.001) (0.001)
Macro controls Yes Yes Yes Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes Yes Yes Yes
Target country FE Yes Yes Yes Yes Yes Yes
Target industry FE Yes Yes Yes
No. of observations 8,786 7,509 7,509 8,459 6,918 6,787 8,410 6,947 6,849
Notes: In this table, we consider how targets’ financial characteristics affect the likelihood of Chinese cross-border acquisition. We report the marginal effects from Logit
estimations based on Equation 1. Standard errors are robust and clustered at the target firm level. * p < 0.10, ** p < 0.05, *** p < 0.01.
28
Table 7: Comparison between Chinese and non-Chinese state-owned acquirers
Probability of being
acquired by Chinese state-owned (1) (2) (3) (4)
GDPPC -0.050 1.724** -0.063 1.365
(0.033) (0.714) (0.053) (1.300)
GDP growth -0.010 -0.030** -0.008 -0.021
(0.008) (0.013) (0.017) (0.026)
Distance -0.008 -2.592 -0.017 0.902
(0.007) (2.043) (0.015) (4.129)
Population -0.052* -0.023 -0.040 -1.346
(0.027) (1.940) (0.053) (4.638)
Trade 0.107*** -0.050 0.068 0.017
(0.027) (0.218) (0.051) (0.501)
Resource 0.027 0.941 -1.303 -11.382
(0.480) (2.669) (1.027) (12.552)
∆Exchange rate -0.554** -0.670** -0.412 -0.382
(0.261) (0.294) (0.460) (0.564)
Industry Diversity 0.116** 0.054
(0.055) (0.060)
Size 0.063*** 0.064***
(0.013) (0.015)
ROA -0.264** -0.301***
(0.106) (0.112)
Year FE Yes Yes Yes Yes
Target country FE Yes Yes
No. of observations 928 828 271 233
Notes: In this table, we compare Chinese and non-Chinese state-owned acquirers. The dependent variable
is a dummy that equals 1 if a target is purchased by a Chinese state-owned acquirer, and 0 if it is purchased
by a state-owned acquirer from other countries. We report the marginal effects from Logit estimations.
Standard errors are robust and clustered at the target firm level. * p < 0.10, ** p < 0.05, *** p < 0.01.
29
Table 8: The impact of The Belt and Road Initiative
(1) (2) (3)
All CN CN Private CN state-owned
PostBRI -0.115 -0.182 0.103
(0.254) (0.327) (0.392)
BRI -0.223* -0.0118 -0.539**
(0.132) (0.163) (0.220)
BRI×PostBRI 0.0320 -0.181 0.386*
(0.142) (0.182) (0.230)
Macro controls Yes Yes Yes
Year FEs Yes Yes Yes
No. of observations 69,269 68,574 68,186
Notes: In this table, we analyze the impact of the BRI on Chinese cross-border acquisitions. We report
the point estimates from Logit estimations. PostBRI is a dummy that equals to 1 if the deal took place in
or after 2013. BRI is a dummy variable that equals to 1 if the target country is one of the 65 BRI countries
(see Table C1 for the list of countries). Standard errors are robust and clustered at the target firm level. *
p < 0.10, ** p < 0.05, *** p < 0.01.
Table 9: The impact of Made in China 2025
(1) (2) (3)
All CN CN Private CN state-owned
PostCN2025 0.0116 -0.100 0.343
(0.299) (0.374) (0.484)
CN2025 -0.0166 0.0960 -0.185
(0.0868) (0.107) (0.148)
CN2025×PostCN2025 0.0815 -0.218 0.402*
(0.143) (0.198) (0.214)
Macro controls Yes Yes Yes
Year FEs Yes Yes Yes
No. of observations 62,353 61,723 61,373
Notes: In this table, we analyze the impact of the Made in China 2025 policy on Chinese cross-border
acquisitions. We report the point estimates from Logit estimations. PostCN2025 is a dummy that equals
to 1 if the deal took place in or after 2015. CN2025 is a dummy variable that equals to 1 if the target
belongs to the industries defined in the Made in China 2025. Standard errors are robust and clustered at
the target firm level.* p < 0.10, ** p < 0.05, *** p < 0.01.
30
Table 10: Pricing of the targets by Chinese acquirers
Dep. variable: 𝑃𝑟𝑖𝑐𝑒𝑖,𝑗,𝑡 (1) (2) (3) (4) (5) (6) (7) (8)
CN -2427.696* -2229.151 -2940.868** -3398.903*
(1420.333) (1528.258) (1480.273) (1790.727)
CNpriv -4765.002** -4374.736** -5098.502** -5078.082*
(2007.446) (1989.946) (2381.328) (2994.417)
CNSOE -447.000 -426.125 -1120.374 -1932.354
(1534.659) (1820.065) (1291.260) (1699.657)
Equity 0.016*** 0.016*** 0.016*** 0.016*** 0.016*** 0.016*** 0.015*** 0.015***
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004) (0.003) (0.003)
ROA 2382.042*** 2390.066*** 1794.686*** 1799.143*** 2679.112*** 2690.157*** 3841.581*** 3868.283***
(753.757) (753.519) (645.184) (645.058) (912.243) (911.779) (1263.931) (1267.184)
Full AC 401.124 384.606 765.643* 738.924* -20.386 -34.732 172.758 149.425
(402.105) (401.802) (413.887) (413.569) (424.533) (423.511) (791.654) (792.945)
Any Patent 672.084 663.561 962.042 954.343 1069.188* 1058.839* 1618.138* 1617.997*
(617.958) (616.569) (677.999) (676.635) (559.936) (559.289) (972.187) (971.200)
Macro controls Yes Yes Yes Yes Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes Yes Yes Yes Yes
Target country FE Yes Yes Yes Yes Yes Yes Yes Yes
Year FE Yes Yes
Industry-year FE Yes Yes
Target country-year FE Yes Yes
Target country-industry-year FE Yes Yes
No. of observations 5315 5315 5307 5307 5288 5288 3485 3485
Notes: In this Table we analyze whether the pricing of targets by Chinese acquirers is different from non-Chinese investors. The dependent variable 𝑃𝑟𝑖𝑐𝑒𝑖,𝑗,𝑡 is what the acquirer paid for 1% of the share of the target
firm (in thousand €). CN is a dummy that equals 1 if the acquirer is a Chinese firm. CNpriv is a dummy that equals 1 if the acquirer is a Chinese private firm, and CNSOE is a dummy that equals 1 if the acquirer is a Chinese
state-owned or state-controlled firm. ROA and Equity are the average value of return on assets and book value of equity over the three years prior to the deal. Full AC is a dummy variable indicating whether 100% of the
target were acquired. Any Patent is a dummy indicating whether the target firm holds any patent. Standard errors are robust and clustered at the target firm level. * p < 0.10, ** p < 0.05, *** p < 0.01.
31
Appendix A: Variable definition and summary statistics
Table A1: Country-level control variables
Variable Description Source Obs Mean Std. Dev.
GDP PC GDP per capita (USD) WDI 68,342 39,166 18,484
GDP growth GDP growth rate (%) WDI 68,363 2.25 2.58
Distance Population weighted distance
to China (1000 km)
CEPII 71,783 8.87 2.42
Population No. of inhabitants (millions) WDI 69,656 93.37 153.57
Trade Export and import in goods
with China (billions, USD)
UN
Comtrade
69,543 92.33 142.03
Inflation Annual inflation of consumer
prices (%)
WDI 68,306 2.49 3.03
Tax haven Dummy=1 if the host country
is defined as a tax haven
according to the OECD
OECD 72,056 0 .0276 0 .1639
Resource Share of resource rents in GDP WDI 63,653 0.0212 .0203
∆Exchange rate Annual growth rate of target
country currency relative to
Chinese Yuan
WDI and
own
calculations
66,687 0.0211 0.2590
Political stability Measure for political stability
on a scale from -2.5 to 2.5.
WGI 68,918 0 .5381 0.6089
Regulatory quality Measure for regulatory quality
on a scale from -2.5 to 2.5.
WGI 68,894 1.2838 0.6331
Rule of law Measure for rule of law on a
scale from -2.5 to 2.5.
WGI 68,917 1.2867 0.750
Control of
corruption
Measure for control of
corruption on a scale from -2.5
to 2.5.
WGI 68,896 1.2884 0.8798
Notes: In this table, we report the definitions and data sources for the set of country-level variables
employed in our estimations. We also provide the number of observations, the sample average and
standard deviations for each variable.
32
Table A2: Firm-level control variables
Variable Definition Source Obs Mean Std.Dev.
Industry Diversity Dummy=1 if the target and the
acquirer are in different
industries
Orbis 58,385 0.5429 0.4982
Size Natural logarithm of total
assets of the target firm
Orbis 21,999 8.51 2.12
ROA (Profit/loss before taxes)/Total
assets
Orbis 21,907 0.0267 0.3393
Book equity Total assets-(loans+long-term
debt)
Orbis and
own
calculation
23,589 71019.67 806934.4
Patents Number of patents the target
firm holds
Orbis 71,525 5.39 204.33
Any patent Dummy variable indicating
whether the target firm holds
any patent
Orbis 71,525 0.133 0.3393
Leverage (Short-term loan+long term
debt)/Total assets
Orbis 18,591 0.3133 6.91
Asset Growth Annual growth rate of total
assets
Orbis 23,783 14.04 1114.21
Intangibles Intangible fixed assets/Total
assets
Orbis 20,550 0.0504 0.14
Notes: In this table, we report the definitions and data sources for the set of target-level variables
employed in our estimations. We also provide the number of observations, the sample average and
standard deviations for each variable. Size, Leverage, ROA and Asset Growth are winsorized at the 1%
level. Book equity is the average value of the three years prior to the deal.
33
Table A3: The list of tax haven countries
Andorra Gibraltar Netherlands Antilles
Anguilla Grenada Niue
Antigua and Barbuda Guernsey Panama (English)
Aruba Isle of Man Samoa
The Bahama Jersey San Marino
Bahrain Liberia Seychelles
Bermuda Liechtenstein St. Lucia
Belize Malta St. Kitts & Nevis
British Virgin Islands Marshall Islands St. Vincent and the Grenadines
Cayman Islands Mauritius Turks & Caicos Islands
Cook Islands Monaco US Virgin Islands
Cyprus Montserrat Vanuatu
Dominica Nauru
Notes: In this table, we provide the list of countries we regard as tax havens in estimations. The list is
based on OECD (2000).
34
Appendix B: Additional descriptive figures
Figure B1: Number and value of foreign acquisition of Chinese firms
Notes: This figure shows the development of the number and the value of foreign acquisitions of Chinese
firm during the sample period 2002-2017. The number of deals is reported on the right hand scale. The
value of deal, in current €billion, is reported on the left hand scale.
Figure B2: Distribution of cross-border acquisitions in selected countries
Notes: This figure shows the distribution of cross-border acquisition for a selection of target countries
across the different deal categories. GB stands for Great Britain, US for the United States, DE for
Germany, FR for France, BR for Brazil, VG fir British Virgin Islands, JP for Japan and KY for the Cayman
Islands.
35
Appendix C: Countries and industries covered by policy initiatives
Table C1: BRI countries
Region Countries
East Asia China, Mongolia
Southeast Asia Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines,
Singapore, Thailand, Timor-Leste, Vietnam
South Asia Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan,
Sri Lanka
Central Asia Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan, Uzbekistan
Middle East and
North Africa
Bahrain, Egypt, Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Oman,
Qatar, Saudi Arabia, Palestine, Syria, United Arab Emirates, Yemen
Europe Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina,
Bulgaria, Croatia, Czech Republic, Estonia, Former Yugoslav
Republic of Macedonia (FYROM), Georgia, Hungary, Latvia,
Lithuania, Moldova, Montenegro, Poland, Romania, Russia, Serbia,
Slovakia, Slovenia, Turkey, Ukraine
Notes: In this table, we report the list of 65 countries initially covered by the BRI initiatives. Source:
China International Trade Institute (Cosentino et al., 2018).
Table C2: Industries targeted by Made in China 2025
New generation information technology
High-end computerised machines and robots
Space and aviation
Maritime equipment and high-tech ships
Advanced railway transportation equipment
New energy and energy-saving vehicles
Energy equipment
Agricultural machines
New materials
Biopharma and high-tech medical devices
Notes: In this table, we list industries targeted by the Made in China 2025 initiatives. Source: Conrad et
al. (2016).
