State Capitalism’s Global Reach:

Evidence from Foreign Acquisitions by State-owned Companies

By

G. Andrew Karolyi and Rose C. Liao1

February 2013

Keywords: Cross-border acquisitions; government-controlled corporations; sovereign wealth funds.

JEL Classification codes: G3, F3.

1 Professor of Finance and Global Business and Alumni Chair in Asset Management, Johnson Graduate School of Management,

Cornell University and Assistant Professor of Finance, Rutgers Business School, Rutgers University, respectively. This project

was initiated while Karolyi was a professor and Liao was a Ph.D. student at the Fisher College of Business at Ohio State

University. Karolyi thanks Ohio State University’s Dice Center for Financial Economics for financial support. The authors

thank Yiorgos Allayannis, Ilona Babenko, Warren Bailey, Christos Cabolis, Willie Chan, Paul Choi, Philip Davies, David De

Angelis, Leo De Bever, Roberto De Santis, Louis Gagnon, Isil Erel, Rudi Fahlenbrach, Kristin Forbes, Ruoran Gao, Di Huang,

Han Kim, April Knill, Henry Kwok, Pedro Matos, Bill Megginson, Derrick Man Yew Meng, Michel Nadeau, Jim Ee Puah,

Stefano Rossi, Samir Saadi, Rex Santerre, Mark Seasholes, Tow Heng Tan, Kelsey Wei, Mike Weisbach, Ying Wu, Bernard

Yeung, Yildiray Yildirim, and participants at the American Finance Association 2011 meetings (Denver), Alberta’s AIMCo

Distinguished Lecture, 2010 Asian Finance Association Meetings, 2010 Dimensional Fund Advisors Investment Conference,

2010 Darden International Finance Conference, NUS Business School’s 2010 CAMRI Lecture, 2010 NBER International

Finance and Macroeconomics Conference, National Taiwan University International Finance Conference and at Arizona State,

Calgary, Concordia, Connecticut, Cornell, Georgetown, HEC Paris, Michigan, Northeastern, Queen’s, and Rutgers. Detailed

suggestions of NBER discussant, Kathryn Dominguez, were especially helpful. Paul Karolyi provided useful research

assistance. This paper has previously circulated under the title of “What is Different about Government Controlled Acquirers in

Cross-Border Acquisitions?” Contact: Professor G. Andrew Karolyi, Johnson Graduate School of Management, Cornell

University, 348 Sage Hall, Ithaca, NY, 14853, Phone: (607) 255-2153, Email: gak56@cornell.edu.

1

Abstract

We examine the motives for and consequences of 4,026 cross-border acquisitions constituting $434

billion of total activity that were led by government-controlled acquirers over the period from 1990 to

2008. We compare this activity with cross-border acquisitions involving corporate acquirers and

government funds, including sovereign wealth funds (SWFs), and find that government-controlled

acquirers are much more like corporate acquirers than SWFs in the characteristics of target countries and

firms they pursue. Positive share-price reactions to the announcements of acquisition deals led by

government acquirers are, like those of corporate acquirers, much larger than deals led by government

funds. Policy implications are discussed, especially in light of recent regulatory changes in the U.S. and

other countries that seek to restrict foreign acquisitions by government-controlled entities.

1

The growing participation of government-controlled firms in the market for cross-border

acquisitions has drawn much attention in the media and among policymakers. Prominent deals include the

failed $19.5 billion investment by Chinalco, China’s state-owned metals group, in Rio Tinto, the U.K. and

Australian dual-listed mining company, in 2009 and the acquisition attempt by Dubai World Ports, a port

management company owned by the government of the United Arab Emirates (UAE) to acquire

Peninsular & Oriental Steam Navigation for $6.8 billion in 2006. Though some of the largest deals

involving sovereign acquirers gaining the most attention did indeed fail, many have been successfully

completed. Some sovereign acquirers involve large sovereign stabilization, savings or wealth funds

(SWFs), like the Abu Dhabi and Kuwaiti Investment Authority, Singapore’s Temasek Holdings and the

China Investment Corporation, but the vast majority of deals involve state-controlled corporations and

agencies, like Malaysia’s Petronas, China’s National Overseas Oil Corporation, and Japan Tobacco Inc.

There are serious and growing concerns about the expanded role of governments in global capital

markets in general, and about foreign acquisitions led by government agencies, in particular. Their

objectives and behavior are not well understood and there is opaqueness surrounding their governance

and activities. Yet, to date, researchers have devoted relatively little attention to their study. What little

has been done has focused just on SWFs and not the many large government-controlled corporations that

have become increasingly active in transactions around the globe.

The main goal of this paper is to remedy this deficiency with a comprehensive global study of all

government-led cross-border acquisitions over the past two decades. Our approach is novel as we are able

to benchmark our analysis of cross-border deals involving government-controlled acquirers against those

of corporate acquirers. Few, if any, of the existing studies of SWFs benchmark their investment decisions

to evaluate their motives and to accurately gauge the economic magnitude of their consequences. We

pursue a number of open questions. Do government-controlled acquirers pursue targets domiciled in

countries that differ from cross-border acquisitions led by corporations from the same home country? Are

the attributes of the target firms different for government-controlled acquirers? Are the target firm’s

share-price reactions around the deal announcement involving government-controlled acquirers higher or

2

lower? To calibrate our findings, we also test whether these motives or consequences differ for SWFs

compared to other government-controlled acquirers though they all belong to a continuum of sovereign

investment vehicles active in cross-border deals (Butt, Shivdasani, Standavad, and Wyman, 2008). Our

surprising finding is that government-controlled acquirers are actually much more similar to corporate

acquirers than to SWFs in the target firms and countries they pursue and in the market’s reactions to those

cross-border announcements. The finding is important for how predictions from existing theory have been

interpreted and for on-going public policy debates.

Theory predicts important differences between government and corporate acquirers should arise.

Target firms become, at least, partially state-owned in such transactions and, as such, a major concern is

they become less efficient or less profitable than if they remained privately-owned firms following the

acquisition. Studies rationalize how public enterprises are inefficient with excess employment/wages and

with goods production that is closer to the needs of self-interested politicians or bureaucrats than any

consumers. This view arises naturally in models of bargaining (through subsidies/bribes) between

politicians and managers in Shleifer and Vishny (1994) and through agency problems in the internal

organization of governments between bureaucrats and politicians. Tirole (1994) refers to these problems

as “dissonant objectives” in the division of labor within government entities (due to information problems

or bad incentive contracts). 2

Another view regards public enterprise objectives as one of maximizing social welfare, curing

market failures, and improving on the decisions of private enterprises when externalities introduce

divergences between private and social objectives (Atkinson and Stiglitz, 1980). Tirole (2001) devises a

model of stakeholder capitalism and investigates how managerial incentives and control structures need to

be modified to deal with deadlocks in decision-making and the lack of clear missions for management.

Most empirical studies of SWFs emphasize this view and hypothesize that social, economic, and political

2 Indeed, there is supportive empirical evidence of this inefficiency view in the poor performance of state-owned banks and

banking systems (by, among others, Berger, Clarke, Cull, Klapper and Udell, 2005; Mian, 2006; Micco, Panizza, and Yanez,

2007; and, Taboada, 2011) and of existing state-owned and newly-privatized firms (such as, Boyko, Shleifer and Vishny, 1993;

Megginson, Nash and Randenborgh, 1994; and, Dewenter and Malatesta, 1997, 2001).

3

objectives could influence how they pursue overseas targets.3 For example, sovereign vehicles may be

used to diversify the industrial base of their home country by seeking overseas targets in global industries

in which their home country is underrepresented (such as in the case of China and access to the mining,

energy, and resource sectors; see Bremmer, 2010).

One of the challenges in implementing tests of the theoretical models of public enterprises is that

they imply that distortions may exist in decisions. Predictions are unreliable without well-defined

benchmarks which limits the power of the tests. To specify alternative hypotheses relative to our null that

the acquisition decisions of public and private corporations are similar, we draw on a large body of

theoretical and empirical work of cross-border mergers by corporate acquirers. For example, Froot and

Stein (1991) rationalize how acquiring companies are more likely from countries with appreciating

currencies and target companies, from countries with depreciating currencies, as favorable currency

moves help acquirers bid more aggressively overcoming potential information problems in financial

contracting. They affirm this prediction at the deal level for U.S. acquisitions abroad, and Erel, Liao, and

Weisbach (2012) do so for world-wide cross-border acquisition flows. If government-controlled

acquisitions exacerbate information problems in contracting because of dissonant objectives or

undefinable stakeholder goals, we would predict even stronger, positive government-controlled acquirer

responses to home-country exchange-rate appreciation, all else being equal.

Conditional on a deal announcement, we also predict a weaker, less-positive market reaction from

target shareholders associated with the alternative hypothesis of dissonant objectives or undefinable

stakeholder goals for government-controlled acquirers. Of course, it is also possible that government-led

acquisitions can circumvent frictions in cross-border deals associated with information problems, so that

weaker, less-positive (and possibly negative) sensitivity of deal flow to exchange-rate movements may

3 Studies by Bernstein, Lerner, and Schoar (2009), Chhaochharia and Laeven (2009), Dewenter, Han, and Malatesta (2010),

Fotak, Bortolotti, Megginson, and Miracky (2010), Knill, Lee, and Mauck (2009), Kotter and Lel (2011), Fernandes (2011),

and Dyck and Morse (2011) compile data on equity investments for SWFs using a variety of sources. SWFs are broadly

defined as public investment agencies which manage part of the foreign-currency assets of national states and are typically

funded by commodity export (e.g. oil) revenues or the transfer of assets directly from official foreign exchange reserves. See

also Jory, Perry, and Hemphill (2008), Balding (2011) and books by Saw and Low (2009) and Bremmer (2010).

4

arise than for corporate acquirers. By the same argument, target shareholders may respond more, rather

than less, positively, to their announcement than for those initiated by corporate acquirers.

We pursue similar tests for specific alternative hypotheses related to differences in market

valuations (Erel, Liao, and Weisbach, 2012), in macroeconomic conditions and levels of development

(Chari, Chen, and Dominguez, 2012), in corporate governance (Rossi and Volpin, 2004; Bris and Cabolis,

2008; and, Ellis, Moeller, Schlingemann, and Stulz, 2011), and in the role of transactions costs, tariffs,

and information barriers linked to geographic distance (see gravity models of Portes and Rey, 2005;

Coeurdacier, De Santis, and Aviat, 2009). In each case, government-led acquisitions may be associated

with worse contracting problems that higher market valuations, faster economic growth, more advanced

economic development, and stronger governance rules are predicted to overcome in corporate-led deals.

Under these circumstances, we predict heightened sensitivity of deal flow to these factors. If government

acquirers alleviate contracting problems, fewer acquisitions will be positively associated with them.

We also investigate alternative hypotheses related to social, economic, and political motives

related to differences in the industrial mix of the acquirer and target country (a diversification motive), to

differences in political constraints that may bind sovereign agencies, to the size of accumulated foreign

currency reserves of the acquirer that may help fund overseas acquisitions, and to the overall presence of

government-controlled firms in the acquirer’s public equity marketplace. These latter tests are particularly

useful in benchmarking the acquisition flows of sovereign-controlled acquirers to those of SWFs, a

difference for which theory gives no guidance.

Our study is important because of recent regulatory developments around the world. Indeed,

another key motivation for us to examine the determinants and consequences of cross-border acquisitions

of government-led entities comes from the heightened regulatory concerns that are now globally

widespread. Consider, for example, the Dubai World Ports deal that was originally blocked by the U.S.

Congress in March 2006 as it involved the potential transfer of 11 terminals in six U.S. ports to a foreign

government agency. One year later, Congress passed the Foreign Investment and National Security Act

(FINSA) of 2007 that gave legal status to the little-known Committee on Foreign Investment in the U.S.

5

(CFIUS), a multi-agency group formed in 1975 to monitor U.S. policy on foreign investments that may

have any effect on national security.4 Following the passage of FINSA, similar types of legislation

promoting foreign review of government acquisitions of domestic targets were passed in China, Germany

and Australia.5 If government-led acquirers do pursue different kinds of targets, or if target firms’

shareholders react less positively in the short-run to the announcements of such acquisitions, then

knowing how these deals differ in type and in terms and conditions can provide useful guidance on the

economic consequences such regulatory restrictions might have. Of course, if no differences are

detectable, then one might wonder about the real value of such regulatory actions at all. The broader

implications of our findings for policy-making are discussed in our concluding remarks.

Our sample includes 127,786 announced cross-border acquisitions between 1990 and 2008 with

a total (constant dollar) transaction value equal to $9.04 trillion. Government-controlled acquirers are

identified as those in which the acquirer’s ultimate parent is flagged as a government entity. Government-

controlled acquirers constitute 4,026 withdrawn and completed cross-border deals constituting over $434

billion over the period from 1990 to 2008. SWFs and other government investment funds add another 733

deals representing over $158 billion over this period.

We do find differences in the cross-border activity that is led by government-controlled acquirers

from that led by corporate acquirers or that by government-controlled funds in terms of the country of

domicile of the acquirers and of the host targets. At the country level, government-led acquirers are more

likely to invest in geographically-close countries with depreciating currencies, with stronger legal

environments, and with more potential to diversify their risk than are corporate acquirers. But these

4 The H.R. 556 Foreign Investment and National Security Act of 2007 was signed into law by President Bush on July 26, 2007.

The Act intends “to ensure national security while promoting foreign investment and the creation and maintenance of jobs, to

reform the process by which such investments are examined for any effect they may have on national security, to establish the

Committee on Foreign Investment in the United States, and for other purposes.” CFIUS was created in 1975 in the Exon-Florio

Amendment to the Defense Production Act of 1950 in which, as the designee of the President, authority was granted to conduct

an investigation into the possible impact on national security of acquisitions involving “foreign persons which could result in

foreign control of persons engaged in interstate commerce in the United States” (Title 50, U.S. Code § Appendix 2170(a)). 5 In August 2008, China began reviewing foreign acquisitions of local companies for national security concerns as an outgrowth

of its 2006 Regulations on Mergers and Acquisitions of Domestic Enterprises by Foreign Investors. See “China forms

committee to review foreign acquisitions, citing security,” (Wall Street Journal, August 26, 2008). A report of the U.S.

Government Accountability Office, entitled “Laws and Policies Regulating Foreign Investment in 10 Countries,” provides a

useful comparison of foreign investment review procedures in different countries (GAO-08-320, Table 3, February 2008).

6

differences are economically small compared to how government-led acquirers differ from SWFs and

other government funds in how we proxy for their social, economic, and political objectives. SWFs and

other government funds are more likely from autocratic countries with higher levels of foreign currency

reserves, whereas government-led corporations are not affected by these factors.

At the deal level, logistic regression analysis shows that both government-led and corporate

acquirers are more likely than government funds to pursue smaller targets in related industries with fewer

growth opportunities. But there is little evidence that any of these acquirers target differently financially-

distressed, capital-constrained, or less profitable firms. Overall, the key take-away from this analysis of

acquisition motives is that the economic magnitudes of the differences between government-controlled

and corporate acquirers are much smaller than those that separate either of them from SWFs.

Even if the patterns of cross-border acquisition activity involving sovereign acquirers is

indistinguishable at the country or deal level, an even more important test is to examine short-term share-

price reactions to deal announcements, which captures the impact on target firms of the various types of

acquirers. We find significant differences in short-term share price reactions to deal announcement. The

median cumulative abnormal market-adjusted returns (CMARs) around announcements (with a three-day

investment horizon) of cross-border deals by corporate acquirers are 2.5%, those of government acquirers

are only 1.5%, and, for SWFs, they are significantly lower at only 0.8%. In cross-sectional tests, the

differences between share-price reactions for corporate and government-controlled deals disappear once

we control for various country-level, firm- and deal-specific factors, but the differences between them and

those for SWF-led deals persist. Buy-and-hold abnormal returns (BHARs) following SWF deals are

significantly more negative than for corporate and other government-controlled acquirers for longer

investment windows (including BHARs as much as -50.3% versus -7.6% at 36 months).

1. Data and Descriptive Statistics

We use Thomson Reuters Security Data Corporation’s (SDC) Platinum Mergers and Corporate

Transactions database to collect data on 127,786 announced cross-border acquisitions between 1990 and

7

2008 with a total (constant dollar) transaction value equal to $9.04 trillion.6 We collect a number of data

items, including the announcement date, whether it succeeded or was withdrawn, the target’s name, its

status (subsidiary, joint venture partner, private, government-owned or publicly-listed company), its 4-

digit Standard Industrial Classification (SIC) code and country of domicile, the name of the acquirer, its

SIC code and country of domicile, its intermediate and ultimate parent firm’s name, status (if either

relevant), the fractional stake in the target that the deal represents, and the deal value, if disclosed. We

only consider deals in which the fractional stake in the target exceeds 5% and classify the deal as a

minority block acquisition if the fractional stake in the target is less than 50%. We also collected other

deal characteristics, including the medium of exchange (cash/stock payment), and the premium paid for

the shares acquired calculated as the offer price relative to the 1-day, 1-week, and 4-week trailing price of

the target’s shares. We convert all deal values reported into U.S. dollars using national exchange rates

from the WM/Reuters prevailing at the time of the deal (WMR quotes are based on 4:00pm London

(Greenwich Mean Time) in U.K. Pound Sterling, which are, in turn, converted into U.S. dollars at the

U.S. dollar/Pound Sterling national exchange rate) and we further report them in Constant 2000 U.S.

dollar terms using the U.S. Consumer Price Index.

Government-controlled acquirers are identified as those in which the acquirer or acquirer’s

ultimate parent is flagged as a government entity including corporations. The variable of interest is

“AUPPUB” and whether it identifies the ultimate parent as government-owned, which SDC defines as

one in which 50% or more of the shares outstanding are government owned. We proceeded to double-

check the ultimate parent’s ownership status at the time of the deal’s announcement by hand using a

variety of company annual reports, regulatory filings, on-line news reports and other resources. We sorted

from highest to lowest all of the government-led acquirers by total U.S. dollar Constant 2000 value (again

6 We exclude leveraged buyouts, spin-offs, recapitalizations, self-tender offers, exchange offers, repurchases and privatizations

and we exclude acquirers from overseas territories of the U.K. and Netherlands that are tax havens, including the Bahamas,

British Virgin Islands, Cayman Islands, Guernsey, Isle of Man, Jersey and Netherland Antilles. This filter on cross-border

acquirers from overseas territories excluded 10,962 corporate deals worth cumulatively $353 billion (in Constant 2000 U.S.

dollars) or 6% of the original sample count and 3% of the total value. We also exclude countries in which there are fewer than

50 cross-border acquisitions, whether led by government-controlled or corporate acquirers. This filter has only a modest impact

on the overall sample and is used to include countries with sufficient acquisition activity to justify its consideration. About 11%

of the deals (4% of total deal value) are excluded as a result.

8

using the U.S. Consumer Price Index) across all deals in which they were involved. We confirmed by

hand the government-controlled status of the top 72 acquirers which ultimately represented 78% of the

cross-border deal value ($461 billion) and over 526 of the deals.7

Government-controlled investment funds are identified as those acquirers with SIC codes in the

999A – 999G range or government agencies related to investment offices, pension, health and welfare

funds, trusts or holding companies with SIC codes 6019, 6371, 6722, 6726, 6798, and 6799. We also

double-checked the name of the acquirer against the list of names supplied on the SWF Institute website

(www.swfinstitute.org) to confirm its status as an SWF. The largest SWFs in our data sample of cross-

border acquisitions included Singapore’s Temasek Corporation, which was involved in 167 deals totaling

$39.8 billion, Singapore’s Government Investment Corporation (GIC; 81 deals, $19.8 billion), Saudi

Arabia’s SABIC (4 deals, $12.3 billion), China Investment Corporation (7 deals, $7.4 billion), and the

Abu Dhabi Investment Authority (22 deals, $8.7 billion). It is noteworthy that collectively SWFs and

government-controlled investment funds are prominent among the largest cross-border acquirers, but they

comprise 733 deals and less than $159 billion of the total deal value, or one-quarter of our total

government-led acquisition activity. But this relatively low total value may arise from the difficulty of

defining exactly what a SWF is, an important point which Fotak, Bertolotti, Megginson, and Miracky

(2010) aptly point out (see their Panels A and B, Table I comparing SWF classifications by Truman

(2009) and by the SWF Institute).8 Lastly, to benchmark correctly corporate acquirers with government-

controlled corporations, we also screen out any private investment funds in a similar manner.9

Table 1 presents summary statistics on the number and total value of cross-border acquisition

deals involving at least a 5% stake in a target corporation. We present the total number of deals, only

7 Consider, for example, that the EDF (Electricité de France) Group of France, which was a 100% government-controlled until

2004 and is still 84%-owned by the government, initiated 19 acquisition deals between 1992 and 2008 totaling $42.3 billion

and its targets included Constellation Energy Group (U.S., $4.5 billion), Délmagyarórszági Aramszólgálta (Hungary, $3.6

billion), and PowerGen plc (U.K., $3.4 billion). 8 By comparison, Bernstein, Lerner, and Schoar (2009) identify 1,752 deals by SWFs averaging $351 million per deal implying

about $615 billion in cumulative SWF activity, but these are not inflation adjusted and the sample runs from 1983 to 2007.

Fotak, Bertolotti, Megginson, and Miracky (2010) in their SDC sample evaluate 141 deals at $572 million per deal or

cumulatively $80.6 billion. Their Bureau van Dijk “Zephyr” sample is much larger at 314 deals at €1.253 billion per deal or

$1.57 trillion over 1997-2008. Beck and Fidora (2008) report $91.5 billion of deal activity in 2007 and 2008 alone, which is far

larger than any of the other samples. 9 This filter on private investment funds excludes 17,524 corporate-led cross-border deals, less than 15% of the original sample.

9

those in which transaction values are reported, the total and average per deal value across all those for

which data is reported, the number of withdrawn deals, those involving minority stakes (less than 50% of

target shares), and the number of targets that are publicly-listed corporations. As noted above, our sample

of government-controlled acquirers constitutes 4,026 deals totaling $434 billion in value, which

represents about 3% of all corporate-led acquisitions (123,027) and 5% of their total value ($8.4 trillion).

It also represents about five times the count of government-controlled fund deals (733) and three times

their total value ($158 billion). Figure 1 plots the number and total value of cross-border acquisitions by

year and demonstrates that a significant increase in government-led acquisition activity occurred in 2007

and 2008, particularly in terms of total deal value. Almost 20% of all government-led acquisition deals

and 35% of the value were concentrated in those two years alone. Corporate-led acquisition activity (not

shown) was also heightened during 2007-2008, but not as intensely in terms of the value of all deals

($1.61 trillion, or only 20%).

Only 1,613, or about one-third, of the government-led acquisition deals report deal values.10 The

proportion of corporate deals for which values are disclosed is much higher at 43% (51,277 deals) and

that of government-controlled fund acquisitions are even higher at 58%. The average deal value involving

government-controlled acquirers ($269 million) is much larger than those involving corporate acquirers

($165 million) yet smaller than those involving government-controlled funds ($374 million).11 As a

fraction of the total number of deals initiated by government-controlled acquirers, more than one-third is

withdrawn (1,355 out of 4,026) and more than 30 percent initiated by government-controlled funds is

withdrawn; by contrast, only 25% of corporate-led acquisition deals is withdrawn.

Another major difference between the government-led and corporate-led acquisitions is the

proportion of them that involve minority stakes in the target firm; 73% of government-controlled funds

10 This could be for one of three reasons. First, many deals involve subsidiaries, plants, or joint venture transactions in which the

deal value is too small to report. Second, there might be differences across countries in disclosure requirements. Lastly, the

parties to the transaction, both companies themselves and their advisors, might strategically choose not to do so. 11 Large SWF-led acquisition deals took place in the last two years of our sample in major financial institutions. The most

prominent examples include GIC of Singapore’s $9.8 billion stake in UBS, GIC and Abu Dhabi Investment Council’s $6.9

billion and $7.6 billion stakes in Citigroup, China Investment Corporation’s $5 billion in Morgan Stanley and Singapore’s

Temasek Holdings and the Kuwaiti Investment Authority’s $5 billion and $3.7 billion investment in Merrill Lynch.

10

involve minority stakes, compared to 60% of government-controlled corporations whereas less than 40%

of corporate deals do. The finance literature proposes that the motives behind these two types of

transactions differ, so we will separate out the majority control and minority stake deals for both

government and corporate acquirers in most of our analysis. Finally, about 40% of the sample of targets

among government-controlled funds involves a publicly-traded firm, a much higher fraction than 20% for

government-controlled corporations and even higher than that for corporate acquirers (only 12,669 or

11%). This is an important constraint for our analysis at the deal level for which we will need to obtain

financial statement information to evaluate by which attributes the targets of government-controlled

acquirers, corporate acquirers, and government-controlled funds differ.

2. Determinants of Cross-Border Acquisition Activity Led by Government-Controlled Acquirers

Our first goal is to measure whether the level of cross-border acquisition activity led by

government-controlled acquirers differs from that of corporate acquirers and that of government-

controlled funds. Does deal activity that is led by government-controlled acquirers emanate from (pursue

targets in) some countries more intensely than others? If so, in either case, what are the country-level

attributes or market conditions of those countries that dominate (attract) government-led cross-border

acquisition activity? In order to answer these questions, we compute two cross-border ratios of deal

activity: the first measures the fraction of all cross-border acquisition activity involving government-

controlled acquirers that emanates from a given country i and the second measures the fraction of all

acquisition activity involving government-controlled acquirers that targets a particular country j. We

compute similar cross-border ratios of deal activity involving corporate acquirers and government-

controlled funds as benchmarks.

In Table 2, we report the countries in rank order by those which have the highest fractions of

government-led activity in the world measured by total deal value by acquirer country (Panel A) and by

target country (Panel B). The acquirer countries in which government-led acquirers dominate all cross-

border activity include China (23% by deal value, 17% by deal counts), France (18%, 17%), Singapore,

11

Saudi Arabia, UAE, Sweden, Italy, and Norway. The acquirer countries that dominate cross-border

activity by corporate acquirers are different. For example, less than 1% of cross-border corporate

acquirers are based in China. The U.K. and the U.S. alone dominate 40% of all cross-border activities

involving corporate acquirers (both fall into the “Rest of the World” category in Panel A). Though

government-controlled funds have similar patterns in their countries of origin as government-controlled

acquirers, there are still big differences. For example, Sweden is near the top of the list in its activity in

cross-border deals involving government-controlled acquirers but only 0.02% of acquisitions by

government-controlled funds originate from Sweden. In both absolute and relative terms, China is a much

more active in cross-border acquisitions by its government-controlled corporations ($100 billion) than by

its government-controlled funds ($14 billion). By contrast, the government funds based in UAE are much

more active acquirers ($44 billion) than their government-controlled corporations ($22 billion). Many of

the countries at the top of the list involving government-controlled acquirers are those that are typically

identified with large accumulated foreign currency reserves due to oil exports and export-driven trade, but

it is not exclusively so (e.g., France, $80 billion, Singapore, $28 billion, Italy, $19 billion, and Sweden,

$21 billion).

The leading target countries for government-led cross-border acquisition activity are more surprising.

Panel B shows that the U.S. has the highest fraction by deal value (18%), but its fraction by the count of

deals is only 10% indicating that several large deals dominate the U.S. market. Hong Kong is second by

the fraction of deal value (14%), first by fraction of deal count (11%). Much of this activity stems from

the government-led deals emanating from China, as seen in Panel A. The U.K. is third by the fraction of

deal value (11%) and Germany is the third by the fraction of deal count (9%). The other countries that are

primary targets for government-led cross-border acquisitions include Australia, Italy, and the Netherlands,

each of which exceed 100 deals and $20 billion in value. France, Spain, and Sweden are also among the

target countries that attracts over 100 deals in our sample period.

Though it is relatively easy to connect the dominant presence of China’s government-led

acquirers in Hong Kong as the primary target market, it is more complex to discern it for the broader level

12

of activity around the world. We will develop another more refined measure of the proportion of

government-led activity by pairs of acquirer and target countries next, but Figure 2 offers a preliminary

look by region. In Panel A, we report the top countries in declining rank by raw total deal value led by

government-controlled acquirers and indicate which regions they target for their activity. For China’s

$100 billion of deal activity, the largest target component is Developed Asia (about $59 billion, much of

which targets Hong Kong), followed by the U.S./Canada (about $19 billion) and Developed Europe

(about $12 billion). France’s $81 billion of government-led acquisitions mostly target Developed Europe,

then the U.S./Canada, whereas Singapore’s acquisitions target Developed Asia and some in Europe.

UAE’s and Saudi Arabia’ s government-led acquirers total about $25 billion each, but UAE’s prefer

Developed Asia and Europe whereas Saudi Arabia’s tilt their acquisitions toward Developed Europe and

the U.S./Canada. The U.S. is the largest target country for government-led acquisitions ($78 billion) and,

in Panel B, we note that the dominant acquirer countries are from the EMEA region (Emerging Europe,

Middle East, and Africa), followed by Developed Europe and then Emerging Asia. The $46 billion in

government-led acquisition activity targeting the U.K. comes from Developed Europe, Developed Asia

and then the EMEA region. For Hong Kong, the dominance of Emerging Asia (mostly all China) in its

$58 billion of deal activity is quite apparent.

We next estimate panel regressions of the level of government-controlled acquisitions across

acquirer-target country pairs and years to uncover their primary determinants. Our dependent variable is

an acquisition intensity measure by the number of deals led by government-controlled corporations

between an ordered country pair, normalized by the total number of government-led cross-border

acquisitions in the host country.12 We denote this ratio, AGijt, where superscript “G” represents

government-led deals. In Table 3, Models (1), (4), (5), and (6) examine only government-controlled

acquirers. One important advantage of our experimental design is that we can perform the exact same

12 This measure is similar to that employed by Rossi and Volpin (2004), Erel, Liao, and Weisbach (2012), Ferreira, Massa, and

Matos (2010), and others to benchmark cross-border acquisition activity between country pairs relative to domestic acquisition

activity in the host country. For robustness check, we also compute a similar ratio but divide the activity in the numerator by

the total activity by government-led acquirers in the home country. The results are qualitatively similar and available upon

request.

13

computations for corporate-led and government funds-led cross-border acquisition activity between

country pairs in Models (2) and (3), respectively. We denote this ratio, ACijt (A

Fijt), where superscript “C”

(“F”) represents corporate-led (government funds-led) deals. There may not be any government-led

acquirer deal flow between a given pair of countries in a given year though there may be for corporate

acquirers. To facilitate comparisons, we compute differences between the acquisition ratios of

government-led acquirers and corporate acquirers in years in which there is at least some activity for one

of them, which we define as an excess ratio (AGijt – AC

ijt). In this way, we are able to determine whether

government-led acquirers from country i disproportionately seek out targets in country j relative to

corporate acquirers that come from country (t-statistics are denoted under “Diff (1)-(2)”). We also

construct excess ratios for the differences between the acquisition ratios of government-controlled

acquirers and government funds (AGijt – AF

ijt in column “Diff (1)-(3)”). Our objective is to regress different

country-level factors not only on the acquisition ratios themselves to understand the key drivers of the

activity flows but also on the excess ratios in order to test whether factors differentially affect the activity

flows for different types of acquirers.13 (All variable definitions for dependent and explanatory factors,

data sources and summary statistics are furnished in an internet appendix.) The models are all estimated

with year fixed effects using ordinary least squares (OLS) with robust standard errors correcting for

heteroscedasticity and clustered by target country. We report the number of observations, the adjusted R2,

overall and for specifications without year fixed effects and with year/country fixed effects for both target

and acquirer countries.14

Valuation can affect acquisition activity through two channels: “cheap” assets (Froot and Stein,

1991) and cheap capital (Shleifer and Vishny, 2003). Regardless of which channel is at work, theory

predicts that acquirers are more likely from countries with appreciating currencies and for which stock

markets have increased in value. We use annual exchange rate returns using WM/Reuters quotes from

13 With 64 countries represented in our overall sample of cross-border acquisitions and 19 years of activity, the potential number

of country-pair observations is the square of the number of countries by year (64 × 63 or 4,032 × 19 years or 76,608). The

effect of these screens is to limit the number of observations to include about 40 countries. 14 We estimated all of our models using a Tobit specification to recognize the potential effects of censoring with acquisition ratios

ranging from 0% to +100%. All of our basic inferences in Table 3 are intact and results are available upon request.

14

Datastream. In Model (1), we see that currency valuation does matter for government-controlled

acquirers, though the economic effect is small. The statistically-significant positive coefficient of 0.044

implies that a one-standard deviation increase in annual real exchange rate returns for a given country pair

(16% per year) is associated with a 0.70% increase (0.044 × 0.16) in government-led activity, which

represents only 11% of its yearly deal ratio (6%). Similar to Erel, Liao, and Weisbach (2012), currency

valuation also matters to corporate acquirers, but it is only marginally significant. Government-controlled

acquirers are significantly more sensitive to currency valuation than their corporate counterparts (t-

statistic of 2.95), but the differences in economic magnitudes are negligible. We infer from this result that

potential distortions or externalities associated with government-controlled corporations do not matter.

Though differences in annual real stock market returns (national CPI-adjusted index returns from

Datastream) explain little of cross-border activities by government-controlled acquirers or corporate

acquirers, we do see in Model (3) that government-controlled funds are more likely to invest in targets

whose stock market has increased in value.

Governance differences between two countries can also affect the cross-border mergers and

acquisition activity.15 We use the anti-self-dealing index of Djankov, La Porta, Lopez-de-Silanes, and

Shleifer (2008) to proxy for quality of investor protection (“ASDI differences”) and the score of disclosure

quality from the Center for International Financial Analysis and Research as in La Porta, Lopez-de-

Silanes, Shleifer, and Vishny (1998) to proxy for accounting standards (“Accounting Standard

differences”). We find a reliably positive association between corporate-led acquisition activity with

differences in corporate governance and transparency in Model (2), similar to Rossi and Volpin (2004),

Bris and Cabolis (2008) and others. For deals led by government-controlled acquirers, they are also likely

to pursue targets in countries with weaker accounting standards, but the coefficient on ASDI actually

turns negative in Model (1) and it is statistically different from the positive coefficient for corporate

15 Rossi and Volpin (2004), Starks and Wei (2012), Antrás, Desai, and Foley (2009), Bris and Cabolis (2008), Bris, Brisley, and

Cabolis (2008), and Ellis, Moeller, Schlingemann, and Stulz (2011) find that cross-border mergers and acquisition activity

between two countries increases the greater the difference in the quality of investor protections and accounting standards

between the acquirer’s and target’s countries.

15

acquirers, which is consistent with view that government-controlled acquirers can circumvent contracting

problems. Again, however, the economic effects of differences in anti-self-dealing index scores are

small: a one-standard-deviation higher ASDI score of 0.33 between acquirer and target translates into a

0.97% lower rate of acquisition activity for government-controlled acquirers, or 16% of the annual

country-pair deal ratio. Government-controlled fund acquisition flows are not sensitive at all to

differences in accounting standards, but are positively so for ASDI differences like for corporate

acquirers. The differential effect from government-controlled acquirers is significant (t-statistic of -4.31),

but again the economic effects are small. Distortions linking government-led deal activity to

governance/accounting standards are small.

Differences in a country’s wealth and changes in its macroeconomic conditions between targets

and acquirers can matter for cross-border acquisition activity.16 We include differences in the logarithm of

GDP per capita, as a measure of the country’s wealth, and in real GDP growth (“GDP Growth

Differences”) as a proxy for the change in macroeconomic conditions. Differences in the level of

economic development (“Log GDP per capita”) do matter for corporate acquirers and government-

controlled funds, but not for government-controlled acquirers. Acquirers from better developed countries

are more likely to acquire firms in less developed countries, but in each case the economic size of the

relationship is small. The most striking finding here is that government-controlled funds are statistically

and economically significantly different from government-controlled acquirers in their much higher

sensitivity to differences in GDP growth (0.303 coefficient implies a one-standard deviation increase is

associated with a 32% increase in its average deal ratio, eight times higher than the equivalent for other

government-controlled acquirers).

Geographic proximity influences cross-border acquisition flows.17 Closer geographic proximity

significantly increases cross-border deal ratios for all three types of acquirers, though it is only weakly

16 Rossi and Volpin (2009), Chari, Chen, and Dominguez (2012), and Erel, Liao, and Weisbach (2012) show that developed

countries’ firms are, in fact, more likely to acquire less developed countries’ firms. 17 Research on trade, FDI flows, and cross-border acquisitions emphasizes the important role that geography plays (among others,

see the gravity models of Anderson, 1979; Portes and Rey, 2005; Anderson and van Wincoop, 2003; Coeurdacier, De Santis,

and Aviat, 2009; also, Rossi and Volpin, 2004; and, Erel, Liao, and Weisbach, 2012). The role of transactions costs, tariffs, and

16

different across the different types of acquirers. A one-standard-deviation increase in the geographic

distance (about 2,900 miles) between acquirer and target translates into a 1.7% lower rate of acquisition

activity for government-controlled acquirers using Model (1), or 28% of its mean deal ratio. The market

correlation measure is also reliably different from zero, and with a positive sign in most models.

Government-controlled acquirers are, however, somewhat more likely to pursue targets in countries that

are likely to diversify their risks (smaller positive coefficient) than both corporate acquirers and

government-controlled funds. There is a good chance that the market correlation measure proxies for the

same kind of proximity measures that other cross-border merger studies have uncovered associated with

regional blocs, religion, culture or language. Government acquirers are less likely to invest in countries

with low potential to diversify their risks, but these differences by type are small.

The explanatory power of each of these models is generally quite weak. So, the statistical

precision with which we observe any differences in individual coefficients by type of acquirer is

overshadowed by the low adjusted R2 in most models. For example, the adjusted R2 in Model (2) of 13%

is low for the corporate-led cross-border deals, but that for the government-controlled acquirers (5%) is

even lower. Similarly, the adjusted R2 for the excess ratio specifications are modest averaging around 2%.

What is notable is that what sizeable incremental explanatory power there is arises just from year and

unobservable target country fixed effects. For example, in Model (1), the adjusted R2 reaches as high as

15% and declines to only 4% without any fixed effects. For the other models, the explanatory power of

these variables without fixed effects ranges between 2% and 12%. This suggests that there is substantial

variation in the differential rates of cross-border acquisition activity among government acquirers,

corporate acquirers, and government-controlled funds that is unexplained and more than half of that

explainable across countries is unobservable even with the variety of variables that existing studies have

collectively put forward to rationalize cross-border deals.

barriers are linked to geographical distance, but they are also linked to commonness of culture, language, ethnicity, and religion

(Stulz and Williamson, 2003; Ahern, Daminelli, and Fracassi, 2011).

17

Several robustness tests are furnished in the tables. We replicate our experiment for government-

controlled acquirers defined in terms of the value of deals (Model 4) instead of their count. The sample

size is adversely impacted by this test, but the previously uncovered factors with respect to currency

valuation, geographic proximity, market correlations, and accounting standard still hold up.18 In Model

(5), we exclude deal activity (again, in terms of the count of deals) that involve majority control

transactions for government -led acquisitions, because government-controlled acquirers are much less

likely to engage in a control transaction, as we saw in Table 1. The overall explanatory power is weaker

than for all deals, but many of the variables that were reliable in Model (1) are similarly so.

Finally, recent studies focusing on SWFs and their investment strategies around the world have

directed their analysis toward other broader political, economic and social objectives formulated from a

theoretical and empirical literature of government enterprises. As an important contribution of our study

is to widen the lens on government acquirers to include not only SWFs, but also the many government-

controlled corporations, it is sensible to investigate whether our large sample of government-controlled

corporations are driven by a political, economic and social factors as those found by recent SWF studies.

Models (6) and (7) present the results of two additional panel regressions of the level of cross-

border acquisitions involving government-controlled acquirers and funds across country pairs and years,

with special focus on proxy variables for such broader political, economic and social objectives. To the

extent that government-controlled cross-border acquisition activity represents an arm of a government’s

industrial diversification program, we may expect that government-led activity would be higher between

countries with more dissimilar industrial structures (Butt, Shivdasani, Standavad, and Wyman, 2008;

Bremmer, 2010).19 For each country pair in each year, we measure as an annual proxy for industry

dissimilarity using the sum-of-squared differences in their respective market capitalization-based weights

of each industry using 48 different global industry categories based on Fama and French (1997). The

18 More than half of our SDC sample does not have a deal value, a finding similar to other studies. Netter, Stegemoller, and

Wintoki (2011) and Di Giovanni (2005) both report that 55% of acquisition deals in SDC do not have a value attached. 19 Dyck and Morse (2011) find that SWFs target specific industries in foreign countries to attempt to influence their country’s

long-term industrial mix.

18

greater is the sum-of-squared differences in industry weights, the more dissimilar is the industrial

structure of the acquirer and target countries. There is no evidence that either government-controlled

acquirers or government funds acquire targets to diversify their home industrial structure. If anything,

they are actually more likely to acquire targets in countries with similar industrial structures.

The political system in a country can influence many aspects of foreign policies on trade, capital

flows, and foreign direct investment (Stulz, 2005) and indeed may influence the acquisitions of

politically-connected, government-controlled acquirers.20 We consider a measure of the autocratic control

or democratic openness of the government as a proxy for the risk of agency conflicts that stem from

politicians or bureaucracies pursuing their self-interests. The coefficient associated with the difference

between target and acquirer country in the PolityIV democracy index is reliably different from zero for

both government acquirers and funds, and with a significant negative sign. Both government acquirers

and funds are more likely from countries with less democratic governments and are similarly more likely

to pursue targets in more democratic systems.21

Government-led activity may be better enabled and financed in acquirer countries with higher

accumulated total foreign currency reserves.22 In fact, SWFs are typically funded by foreign exchange

assets in the form of balance of payments surpluses, official foreign currency operations, the proceeds of

privatizations, governmental transfer payments, fiscal surpluses, and/or receipts resulting from

commodity exports. We obtain annual measures of reserves as a fraction of GDP by country from the

World Bank’s World Development Indicators database. We find that higher accumulated foreign currency

reserves in the acquirer country only weakly positively influence government-controlled funds in

pursuing overseas targets; government-controlled corporations are actually much less likely influenced by

20 Bernstein, Lerner, and Schoar (2009) show SWFs with greater involvement by politicians in fund management are more likely

to invest in their home country and in high P/E industries. Sojli and Tham (2011) find only SWF investments in U.S. targets

that witness an increase in government-related contracts following their acquisition experience valuation increases. 21 We conducted a robustness check using differences between acquirer and target country in the political risk ratings from the

annual scores of the International Country Risk Guide (ICRG), as constructed by the PRS Group. The results were much

weaker, including for the component bureaucratic quality, investor profile, and institutional quality and corruption scores. The

authors thank René Stulz for sharing the ICRG data for this experiment. 22 Butt, Shivdasani, Standavad, and Wyman (2008, p. 78) point out that high foreign currency reserves may be well in excess of

the immediate needs of governments to offer protection against sudden capital outflows and represent a signal of their interest

in promotion of domestic economic development, long-term macroeconomic objectives or even foreign policy objectives.

19

the size of accumulated reserves compared to government funds. This effect is economically large: a one-

standard deviation in reserves as a fraction of GDP is associated with a 34% lower annual deal ratio.

Finally, the more acquisitive government enterprises are in a country, the greater their

involvement in the economy and capital markets, and the more likely they would influence government-

controlled cross-border acquisition activity.23 We measure the scope of a government’s presence in the

economy each year in terms of the fraction of domestic acquisition activity involving government

acquirers over the preceding five year period using the same SDC Mergers database. The coefficients on

prior domestic acquisition activity by government acquirers are not only positive, but economically

larger: a one-standard-deviation increase in domestic activity (0.09) is associated with an 8.6% higher rate

of cross-border acquisition activity by government-controlled acquirers from that country, an increase of

144% of its average annual deal ratio. For government funds in Model (7), the same comparative static

implies an 88% higher average annual deal ratio, which, though also large, is distinctly weaker.

3. Characteristics of Targets of Government-Controlled Acquirers in Cross-Border Deals

One of the problems with our country-level analysis is that it is aggregated activity at the level of

annual country-pair deal flow. Consequently, it fails to account for firm-level and deal-specific factors

that potentially affect the decision to pursue a target overseas. Research shows that the likelihood of being

a merger target or even that involving a minority stake is affected importantly by the target firm’s own

financial and operating conditions, ownership structure, governance, those of the acquirer, as well as the

terms and conditions of the deal. Unfortunately, to control for firm-level factors, we must rely only the

subsample of target firms for which we have public data. This subsample is necessarily unrepresentative

of the overall sample of mergers and acquisitions. Recall from Table 1 how the number of public targets

of government-controlled acquirers falls dramatically (4,026 to 715 deals), corporate deals falls by almost

23 Active domestic acquisition activity by government-controlled agencies could reflect political economy motives. Bernstein,

Lerner, and Schoar (2009) show that the SWFs where politicians are involved in the management of the fund are

disproportionately more likely to invest in domestic firms than externally-managed SWFs, a finding which they interpret as a

distortion in capital allocation from short-term political considerations.

20

90% (123,027 to 12,669 deals), and government funds falls by 60% (733 to 304 deals). The sample

erodes because we exclude target firms in the financial services and utilities sectors and even further

depending on what firm-specific variables are readily available in Thomson Reuters Worldscope

database, our primary source.

With these cautions in mind, we perform the following experiment. We estimate a logistic

regression (logit) model to predict whether an observed cross-border acquisition is initiated by a

government-controlled acquirer (dependent variable equals one) or a corporate acquirer (equals zero). We

also estimate another logit model to evaluate whether an observed cross-border acquisition is initiated by

a government-controlled acquirer rather than a government-controlled fund. Intuitively, this approach

presumes that these other acquisitions represent a reasonable benchmark to help understand the nature of

government-controlled acquirers. Our null hypothesis is that government-controlled acquirers are not any

more likely to be announced than corporate acquirers or government-controlled funds in cross-border

acquisition deals and that the firm-level and deal-specific determinants are not different. To lend power to

our tests, we identify firm-level and deal-specific variables associated with specific alternative hypotheses

that we might be able to reject in favor of the null.

Some of these alternative hypotheses carry over from our analysis at the country level in the

previous section. For example, we can proxy for market valuation both at the country level and at the deal

level. In our tests we use the preceding year’s market-to-book ratio of the target as our main valuation

proxy.24 The governance motive can also be proxied by a variable related to the ownership structure of

the target. The fraction of closely-held shares is often used as a proxy for agency problems (Faccio and

Lang, 2002; Doidge, Karolyi, Lins, Miller, and Stulz, 2009).25 We also investigate other motives for

cross-border acquisitions from the literature on minority block acquisitions and include several firm-level

control variables. Product market relationships between customers and suppliers are often strengthened by

24 Other studies of the determinants of SWF target investments use the lagged stock returns of the target (Kotter and Lel, 2011) or

the Tobin’s q ratio of the firm or the industry of which it is a member (Fernandes, 2011) 25 Leuz, Lins, and Warnock (2009) show a higher fraction of closely-held shares deters foreign portfolio investments in a firm.

21

a partial integration of the two firms.26 Another reason for at least partial equity stakes is that target firms

are financially constrained.27 We use a zero-dividend dummy to proxy for financial constraints, but have

explored a number of alternatives, including those of Whited and Wu (2008) and Hadlock and Pierce

(2010). Kotter and Lel (2011) and Fernandes (2011) measure financial constraints of their sample of SWF

investment targets using a cash-to-assets ratio. Important attributes of a deal can matter. We examine

whether the deal is withdrawn, whether the offer was all cash or mixed stock/cash, and the fraction of

shares in the target the acquirer was after. We also follow other studies of SWF investments and include

target firm-specific control variables from the year prior to the deal including the (logarithm of) total

assets, return on assets, leverage (long-term debt to assets), and sales growth.

Table 4 presents the results of the logit regressions. Panel A evaluates the attributes of the targets

of government-controlled acquirers against those of corporate acquirers, and Panel B, the same against

those of government funds. The main specification (Models 1 and 4) includes all deals. We focus only on

minority deals in Models (2) and (6) and exclude withdrawn deals in Models (4) and (8), as we saw a

larger proportion of these among the sample of government-controlled acquirers in Table 1. Financial

firms and utilities are excluded in all models except Models (1) and (4). Coefficients are reported as

marginal effects, both country and year fixed effects are included, pseudo-R2 (with and without fixed

effects), the unconditional likelihood of the event of a government-controlled deal (denoted “Predicted

Y”), and the number of observations are presented at the bottom of the table. Standard errors are robust to

heteroscedasticity.28 With the data constraints that we impose above, our sample of government-

controlled acquirers versus corporate acquirers includes 6,100 observations, of which 3.37% are

26 Williamson (1979), Grossman and Hart (1986) and Aghion and Tirole (1994) rationalize circumstances in which full (merger)

versus partial integration (minority stakes) can be optimal in information environments in which incomplete contracting arises. 27 Firms facing difficulties in raising capital are more likely to sell partial equity stakes to other firms and empirical studies of the

U.S. markets by Allen and Phillips (2000) and Fee, Hadlock and Thomas (2006) provide support for this idea. Liao (2009)

shows that financial constraints are more important in other countries and especially in cross-border partial equity acquisitions. 28 One concern is that logit models with rare events may produce biased and inefficient estimates (King and Zeng, 2001). The

number of cross-border deals involving government-controlled acquirers is relatively rare at around 2% of the whole sample.

King and Zeng devise a weighting scheme with rare-events corrections. Unfortunately, their procedure is not conformable to

reporting with marginal effects as we prefer. We repeated our logit tests with the King-Zeng corrections and compared our

baseline results (without marginal effects). Most of our inferences obtain. Another concern of logit models with many fixed

effects is that estimates of coefficients could be biased (Greene, 1999). We use conditional logit models to estimate deal

determinants and find similar results to logit models presented in the paper.

22

government-controlled acquisitions, and that of government-controlled acquirers versus funds includes

454 observations of which 86% are government-controlled acquisitions.

We find that few variables are statistically reliable predictors of government-controlled relative to

corporate acquisitions. For example, the zero-dividend proxy in Model (1) for financial constraints is

insignificant, but significantly positive in Model (3) without financials/utilities, which implies that

government-controlled acquirers are more likely to pursue financially-constrained industrial targets.

About 43% of target firms do not pay dividends, but we find that they are associated with a 1.1% higher

likelihood of a government-controlled acquisition, which is a sizeable jump compared to the mean

likelihood (around 3%). This finding is similar to that of Kotter and Lel (2011) which shows that SWFs

are 1.5% more likely to invest in targets with low cash-to-asset ratios against a benchmark set of all

Worldscope firms. There is also some evidence that contracting motives matter, as the related-industry

dummy is significant in Model (1) at the 5% level; however, when we exclude financials and utilities as

target firms, the related-deal dummy coefficient turns negative and sometimes significant. We also find

that the closely-held shares variable has no explanatory power. There is weak evidence in Model (1) that

government-controlled acquirers in cross-border deals are deterred by the fact that more shares of the

target are held closely by institutions or insiders than corporate acquirers are. Among the control

variables, firm size and market-to-book seem to matter most. Government-controlled acquirers are more

likely to be associated with larger target firms with more growth opportunities. But the small marginal

effect implies that this relationship is economically weak. In fact, these attributes together offer little

explanatory power. The pseudo-R2 is only 2.1% when reported for those specifications without target

country fixed effects (compared to almost 19% with them).

We also find few target attributes are significant predictors of government-controlled acquirers

relative to government-controlled funds.29 There is no evidence that the financing motive is at work:

government-controlled corporations and funds pursue target firms with similar financial constraints, as

29 We drop the related-deals dummy for all but Model (5) since SWF and state-controlled funds are less likely to pursue targets in

the same industry when a related-industry definition for the SWFs and state-controlled funds is the financial services sector

(with SIC codes between 6000 and 6999). Most SWFs invest in a wide variety of industries.

23

measured by the zero-dividend dummy. Government-controlled acquirers are somewhat less deterred by

the fact that more shares of the target are held closely by institutions or insiders than government-

controlled funds are; however, the dummy for closely-held shares more statistically reliable for minority

deals only and with financials/utilities excluded. Kotter and Lel (2011) find this attribute is unimportant

in their tests using the Worldscope public equity universe as a benchmark for SWFs. Lastly, though

government-controlled acquirers prefer larger target firms with more growth opportunities when

compared with corporate acquirers, they are associated with smaller target firms with less growth

opportunities when benchmarked to government-controlled funds. The marginal effects are much larger

than those in Panel A. In fact, these two target firm attributes explain proportionally more of the

differences between deals comparing government-controlled acquirers and government funds than those

relative to corporate acquisition targets. The pseudo-R2 in Panel B is more than double that of Panel A,

even without target country fixed effects.

4. Market Reactions to Announcements of Deals Led by Government-Controlled Acquirers

We next examine how shareholders react to the announcements of cross-border deals led by

government-controlled acquirers. As we have done throughout this study, we benchmark the magnitude

of these reactions to acquisitions led by corporate acquirers and by government funds. Under the central

null hypothesis that government-controlled and other cross-border acquirers are not different, we predict

that the shareholders’ reactions to the announcements of such deals should also be indistinguishable. We

also investigate alternative hypotheses related to corporate as well as political, economic and social

motives to rationalize how target shareholders may react differently to acquisition announcements

involving government-controlled acquirers. Whether such deals manifest distortions due to inefficiencies

from bargaining games between politicians and bureaucrats, agency problems among bureaucrats, or

whether they arise due to broader stakeholder goals of industrial diversification, the redeployment of

accumulated foreign currency reserves or the general expansion of government’s role in markets, we

expect the share-price reactions to government-led deals to be more attenuated than for corporate-led

24

deals.30 Of course, to the extent that government acquirers are more likely to be associated with targets

that circumvent cross-border frictions, then the target share-price reactions may be even more positive

than for corporate-led acquisitions. We do have strong priors to guide us. Studies of corporate-led, cross-

border deals – both mergers and partial-block minority stake acquisitions - uncover larger, positive share-

price reactions for the targets, whereas those of SWF investments range from 1% to 3% (Fotak, Bertolotti,

Megginson, and Miracky, 2010; Dewenter, Han, and Malatesta, 2010; and, Kotter and Lel, 2011). The

interesting question then is whether the reactions to deals announced by non-SWF government-controlled

acquirers will be closer in magnitude to those of corporate acquirers or of government funds.

The challenge that we face with this additional analysis is that we need to collect stock returns

data from Datastream for the sample of public targets and this will adversely impact the sample sizes. For

all acquisitions, our sample of government-controlled cross-border deals falls to only 298, those of

corporate acquirers fall to 7,482 observations and, for government funds, to only 138 deals. For minority

deals only, our sample includes 233, 4,767 and 122 deals, respectively. Financial services and utilities

sector target firms are excluded. We compute the cumulative market-adjusted returns for varying-length

windows around deal announcement dates obtained from SDC, report the median reactions for

acquisitions by acquirer type, and perform cross-sectional tests of differences in the reactions using

variables that are related to the various motives we explored above. For as many observations as possible,

we also collect bidder premiums defined as the bid price relative to the closing stock price of the target

some time prior to the announcement date. This additional data helps control for whether the reactions

differ because the terms of the offer are systematically different for government-led acquisitions.

To measure the share price reactions, we compute cumulative market-adjusted buy-and-hold

returns (CMARs) over three different windows around the deal announcements: 21 days (from 10 days

before announcement to 10 days after, denoted “(-10, +10)”), 11 days (“(-5, +5)”), and 3 days (“(-1,

+1)”). The market index returns are for Datastream’s capitalization-weighted national market indexes.

30 Ideally, we want to examine how shareholders of acquirers react to the announcements. But many government-controlled

acquirers do not have publicly-traded stocks, which limits the power of such a test. Ellis, Moeller, Schlingemann, and Stulz

(2011) show how important differences in shareholder protection laws matter for acquirer returns in cross-border deals.

25

Table 5 presents summary statistics of the CMARs. In Panel A, we compare the CMARs for

government-controlled, corporate led and government fund-led cross-border acquisitions for all deals and,

in Panel B, we do so for those involving minority acquisition stakes. The median CMARs and the

numbers of observations are presented for each of the three investment windows around deal

announcements. We also report the p-values associated with the Wilcoxon rank-sum z-statistics for

differences in the medians between acquirer types.

Our first step is to calibrate our CMARs for corporate and SWF/government fund acquisitions

with existing findings.31 The target CMARs for our sample of corporate acquirers are around 6.2% for the

21-day window, 5.0% for the 11-day window and 2.5%, for the 3-day window. Each of these medians are

reliably different from zero based on Wilcoxon-rank-sum tests. The sizes of these reactions are

comparable to those in other recent studies of cross-border mergers. For example, in Bris and Cabolis

(2008) for their sample of 420 target firms, they find a positive and statistically significant 14%

cumulative abnormal return (CAR) for a 5-day window and a further 11% CAR for up to 10 days

following the announcement window. Kang and Kim (2008) find 9% CARs in their out-of-state partial

equity acquisitions in the U.S. and Allen and Phillips (2000) uncover a 6.9% reaction for their full sample

of minority block acquisitions. Both studies use long windows (close to 21-day horizon) but focus on U.S.

firms and domestic transactions only. Liao (2009) compares minority block acquisition deals that are

domestic and cross-border around the world and finds that her 4,780 domestic deals (49 countries around

the world) have CMARs of 8.7% for a 21-day horizon and 7.42% for her 1,851 cross-border deals.

The CMARs of the target share-price reactions to acquisitions by SWFs and other government

funds are much smaller, though still reliably positive (for 138 deals). The median 3-day reaction is 0.8%

and it declines to as small as 0.2% for the 21-day event window around the deal announcement. These are

close in magnitude to those of other SWF studies. Fotak, Bortolotti, Megginson and Miracky (2010, their

Table 7, Panel B) find a statistically significant average CMAR of 1.25% (median of 0.17%) for the

equivalent 3-day window for their 688 SWF investment announcements. Kotter and Lel (2011, their

31 Note that most studies report average CMARs, which tend to be higher than median CMARs due to skewness in the data.

26

Table 5) find 3-day CARs of 1.56% for their sample of 309 cross-border SWF investment

announcements. Dewenter, Han, and Malatesta (2010, Table V) find an average 1.72% CAR for a 3-day

horizon around their 202 investment announcements.32

The 298 targets of government-controlled acquirers have median CMARs of 4.7% for the 21-day

window, 2.8% for the 11-day window, and 1.5% for the 3-day window. These are economically smaller

than for the corporate-led acquisitions. The p-values for the tests on the medians reject that they are

different for the 21-day window, but they are significantly different at the 10% level and at the 5% level

for the 11-day and 3-day event windows, respectively. Each of these share-price reactions is economically

larger than those for the government funds and, indeed, the Wilcoxon rank-sums tests for the medians

indicate that these differences are statistically significant.

A large fraction of these deals initiated by government-controlled corporations and funds involve

only minority stakes, so Panel B examines minority deals only. CMARs of targets of corporate cross-

border acquisitions are lower with medians of 4.0% for the 21-day window to as low as 1.5% for the 3-

day window. Those for the government funds are also smaller, and the median CMARs among the 122

minority deals for the 21-day window are no longer significantly different from zero. Among these

minority deals, the CMARs for the government-controlled acquirers appear smaller in magnitude again

than those of the corporate acquirers, but our Wilcoxon rank-sum tests suggest that they are no different

from each other. This result confirms that the significant difference between government-controlled

acquirers and corporate acquirers in the overall sample is largely driven by the fact that corporate

acquirers are more likely to engage in control changing acquisitions which are associated with higher

announcement returns. For 21-day and 11-day windows, the differences between the government-

controlled acquirers and government funds are statistically significant and economically large.

One of the reasons we are unable to reject the null that reactions to government-controlled

acquirers and corporate-led acquisition announcement reactions are no different could be that the former

32 Fotak, Bertolotti, Megginson and Miracky (2010) and Chhaochharia and Laeven (2009) do not indicate how many of the

announcements are associated with cross-border deals by the SWFs, so direct comparison is difficult.

27

bid more aggressively. Cross-sectional regression analysis follows, but first we report summary statistics

in Table 5 on the bidder premiums paid in the government-led deals, at least for the subset of deals we

examine for which the data is readily available in SDC. The bid premium is computed as the bid price as

a percentage of the closing price of the target shares 1-day, 1-week, and 4-week prior to the

announcement date. What we find is that the bid premiums for corporate-led deals are actually higher for

all transactions (Panel A) as well as for minority deals only (Panel B), but the differences with

government-controlled acquirers are not economically large and never different statistically. For the four-

week bid premium, for example, corporate acquires bid 27% and government-controlled acquirers bid

almost 23%, with a p-value of the difference in medians of 0.34. Part of the problem is the reduced

sample size which includes only 105 of 298 government-led deals with bid price information. We do find

that the bid premiums are smaller for the acquisitions of government funds than of government-controlled

acquirers and these differences are statistically significant for the 4-week and 1-week horizons. Of course,

the bid premiums involving minority stakes are smaller for all acquirer types. Here the differences

become too small to make serious inferences with the limited sample sizes.

Table 6 reports results from cross-sectional regressions of the 21-day CMARs of cross-border

acquisition announcements on various country-level, deal- and firm-specific variables. These variables

are the same as what we included in our logit models in the previous section. What we want to know is

whether the lower share-price reactions associated with government fund acquisitions persist when we

account for differences in the types of targets they pursue. Dummy variables for each type of acquirer are

therefore included as well as F-statistics of tests for differences in their respective coefficients (p-values

in bottom two rows of table). For CMARs associated with each event window around announcements, we

estimate four specifications, three of which evaluate subsamples of deals with available explanatory

variables and one for just minority deals. All are estimated with OLS and robust standard errors for the

coefficients are computed with corrections for heteroscedasticity. We include target country and year

fixed effects in all specifications.

28

In Model (1) for the 21-day event window, we confirm the finding in Table 5 that CMARs of

government-controlled acquirers are indistinguishably different from those of corporate acquirers but

significantly higher than those of government funds, though this time with robust standard errors and

country and year fixed effects. In each of the additional specifications with different combinations of

control variables and just minority deals (Models 2 to 4), this finding does not change. There is a negative

coefficient associated with total assets (log), which implies that the share-price reactions are significantly

lower for larger targets. We also see the same for the market-to-book ratio, which implies that the

reactions are smaller for growth-oriented targets. None of the other variables have explanatory power,

except the withdrawn deals dummy, which is negative and economically large at -4.4% as expected from

existing research.33 Lower share price reactions are associated with those deals that are never completed.

There is a positive, but statistically unreliable coefficient on the one-week bid premium in this model, a

finding that we expected based on univariate statistics in Table 5.

For the shorter 11-day and 3-day event windows, we still find statistically and economically-large

differences in share-price reactions between government-controlled acquirers and government funds and

no differences with corporate acquirers in most specifications. The statistical significance of the

differences with government funds disappears when we focus only on minority deals in Models (8) and

(12), but the economic magnitudes of the coefficients still show that the reactions to government fund

deals are lower. Recall that the differences between median reactions to government-controlled and

corporate-led acquisitions were statistically significant in Table 6 for the 3-day window; now, in

multivariate regressions, the difference disappears (p-value of 0.046 in Model (9) without control

variables and at 0.171 in Model 10 with control variables). The statistical significance of the coefficients

on total assets, the market-to-book ratio, and the withdrawn-deals dummy are robust across all

33 Bates and Lemmon (2003, their Table 8), for example, examined corporate acquirers in the U.S. and showed that three-day

cumulative abnormal returns to targets of failed or withdrawn majority-control deals in the U.S. averaged 3.9% lower than

successful deals. When we examine only U.S. domestic corporate acquirers in our sample, we confirm that the failed or

withdrawn majority deals have 7.70% lower three-day CMARs than successful U.S. deals.

29

specifications of these shorter event-windows, but the economic magnitude of the differences in reactions

are smaller.

5. Long-Run Consequences of Government-Led Acquisitions

In this final section, we study the long-run performance of target firms to better understand the

financial consequences of deals by government-controlled acquirers, as some effects may only be

revealed gradually over time. Our central null hypothesis is that the longer-run investment returns or

financial and operating performance of the targets are no different for government-controlled acquirers,

corporate acquirers, and government funds. But concerns about agency problems in bureaucracies or

uncertainties about their pursuit of political, social goals through these acquisitions should be associated

poorer investment returns and weaker financial and operating performance of the target in the longer run

following the closing of the deal. With respect to the acquisitions by SWFs at least, there is evidence that

this is in fact the case (Dewenter, Han and Malatesta, 2010; Fotak, Bortolotti, Megginson and Miracky,

2010; and, Kotter and Lel, 2011). But what we want to know is whether this long-run underperformance

of the target firm also arises for deals involving government-controlled acquirers beyond SWFs and

government funds and whether it is any different from that which arises for equivalent deals led by

corporate acquirers. Of course, our longer-run analysis is restricted to the subset of announced deals that

are successfully completed and those that involve targets that remain public for an extended period after

the deal is completed. Our sample sizes decline to half of those in Table 6.

For each completed acquisition, we compute buy-and-hold abnormal returns (BHARs) cumulated

from one month after the deal completion date for event windows that extend one, two and three years

following it. We benchmark the monthly returns for each target using its capitalization-weighted home

country index obtained from Datastream. We restrict our analysis to those that survive the full event

window of interest and exclude those that drop out for various reasons, such as due to mergers,

acquisitions, or bankruptcies. We employ calendar-time portfolio analysis that will include the returns of

these excluded firms for whatever months for which they are available after the deal is completed.

30

Panel A of Table 7 presents the mean and median BHARs for each horizon and for each type of

acquirer as well as p-values associated with tests for differences in means (two-sided t-statistics) and

medians (Wilcoxon ranked-sum test) across different types. Overall, we find that the median BHARs are

negative for all acquirer types. For targets of corporate acquirers, the median BHARs range from -13%

after one year to -31% after three. This underperformance is statistically different from zero and

economically large. The mean BHARs reflect less negative underperformance ranging from -4.4% after

one year to -7.6% after three years, which implies that there are some influential positive BHAR

outcomes among our large sample of more than 3,000 such deals. The median BHARs of the government-

controlled acquirers are similar to those of the corporate acquirers, but experience lower

underperformance (only -5.4%) to one year following the deal and greater underperformance (as high as -

40.3%) to three years later. The mean and median tests indicate that these differences are insignificant.

The BHARs for the targets of SWFs and other government funds, however, are economically and

statistically larger negative returns. The one-year median BHARs are -10.7% and are insignificantly

different, but the three year median BHARs are -62.8% (compared to the government-controlled

acquirers, p-values equal 0.043 for the differences of medians and 0.026 for the differences of means).

Long-run BHARs of the targets of government-controlled cross-border acquirers are

indistinguishable from those of corporate acquirers, but reveal much lower underperformance than those

of government funds. We can compare our BHARs for government funds with those of other studies of

SWF investments. Fotak, Bertolotti, Megginson, and Miracky (2010, their Table 7, Panel B) compute

median cumulative abnormal returns relative to the local market index ranging from -3.4% for one year to

-9.30% for three years after deal close. Of course, their sample (about 130 deals) includes many domestic

investments; in cross-sectional regressions, they show that the negative returns are much larger for

foreign targets. Kotter and Lel (2011, their Table 9, 172 deals) shows much smaller BHARs (using a

world index) up to -6.9% to three years after deal close and Dewenter, Han, and Malatesta (2010, their

Table 8) show results in between these two (-19.3% for three year horizon, 74 deals).

31

We also carry out a calendar-time portfolio analysis of targets using two windows of one and two

years following the deal. The advantage of this approach is two-fold: we can account for the cross-

correlations in abnormal returns that can bias the estimates of long-run BHARs due to calendar-time

clustering and we can include the deals for which monthly returns arise but do not survive the full one- to

three-year event windows. We report the average monthly portfolio returns (with the number of monthly

observations) as well as the intercept coefficient (“constant”) and slope coefficients from a regression of

the monthly returns on the global factors from Fama and French (2012), as obtained from the website of

Professor Ken French at Dartmouth University. We estimate these time-series regressions using a

seemingly-unrelated regression approach in order to test the equality of the intercept coefficients

(measuring risk-adjusted returns) across portfolios of targets of the different acquirers (note that they are

estimated for a common period of 235 months for which all have available returns).

We find that monthly portfolio returns are positive, on average, and are somewhat larger for those

of government-controlled acquirers than corporate acquirers (1.48% per month versus 1.26% per month

for one-year horizon, 1.01% versus 0.94%, for the two-year horizon). The average monthly returns are

even lower for the targets of government funds (1.14% for one-year and 0.79% for two-year horizons).34

When we account for the Fama-French factors, the alphas (intercept coefficients) are lower, for the targets

of all three types of acquirers. Those for the government-controlled acquirers and corporate acquirers are

significantly different from zero, but the tests (bottom two rows of Panel B) indicate that differences in

alphas across targets by acquirer are insignificantly different (p-values of 0.74 for one-year horizon, 0.11

for two-year horizon). It is noteworthy that the slope coefficients on the size factor (“Small Minus Big”)

of the calendar-time portfolio of targets for the government funds are distinctly higher than those for the

other two types of acquirers, reflecting the fact that their target selections are somewhat tilted toward even

34 Fotak, Bertolotti, Megginson, and Miracky (2010, their Table 7) report average monthly calendar-time portfolio abnormal

returns relative to a local index of 0.72% for a two-year holding period and Kotter and Lel (2011, their Table 9) report

annualized calendar-time portfolio abnormal returns of 1.30% (relative to a world index).

32

smaller capitalization stocks, which runs counter to what we saw in the logit models of Table 4.35 The

slope coefficients on the value factor (“High Minus Low”) are also higher for the targets of government-

controlled acquirers than those of government funds which means that government-controlled acquirers

are more likely to target growth-oriented firms(higher market-to-book ratios), also as evidenced in Table

4.36 Finally, the slope coefficients on the momentum factor (“Winner Minus Loser”) are significantly

negative for government-controlled acquirers and government funds (for two-year horizon), which

implies that government-controlled acquirers and government funds are more likely to pursue distressed

firms, also as evidenced in Table 4.

6. Concluding Remarks

The motives behind the cross-border deals led by government-controlled acquirers differ little

from those involving corporate acquirers as revealed by the country-level factors that influence the deal

flow and in terms of the characteristics of the targets they pursue. More importantly, the magnitudes of

these differences are much smaller than those that separate them from SWFs and other government funds.

There are also negligible differences in the economic consequences of their announcements. Positive

share-price reactions to the announcements of acquisition deals led by government-controlled acquirers

are comparable those of corporate acquirers, and indeed much larger than those for deals led by SWFs

and other government funds. The findings are robust and surprising.

These findings are important on two fronts. First, the large and growing amount of cross-border

acquisition activity that involves government-controlled acquirers has heightened regulatory concern

about such deals in many countries around the world. The Foreign Investment and National Security Act

of 2007 in the U.S. has instituted much tougher scrutiny of potential foreign acquirers that involve a

government entity, and similar legislation is in place or forthcoming in China, Australia and Germany,

35 The difference between the findings in Table 4 and Table 8 could be due to the fact that the Fama and French (2012) factors

are global yet in Table 4, we controlled for country fixed effects. 36 We repeated our seemingly-unrelated regression experiment using the global three-factor model from Hou, Karolyi, and Kho

(2011), with their updated factors to December 2011). Similar inferences about the risk-adjusted returns (the intercept

coefficients) arise.

33

among many other countries. Our findings suggest that these concerns may be unwarranted for most

government-led acquisitions. Greater attention on SWFs and state-controlled funds as a particular type of

government acquirer may indeed be worthy of further scrutiny, but the vast majority of government-led

foreign acquirers in terms of deal count and total value of deal activity appears to be motivated no

differently than corporate-led cross-border deals and the short-run and long-run economic consequences

appear to be indistinguishably different.

Our study also makes an important contribution to the literature on the operational and financial

performance of state-owned enterprises. A number of scholars have argued why and how government

firms are less efficient or less profitable due to the natural conflicts that arise from self-interested

politicians and bureaucrats and there is considerable evidence that government-controlled firms are

indeed associated with weaker operational and financial performance. Our study involves a special

experiment to examine these questions that focuses on transactions in which the target firm becomes, at

least partially, a state-owned enterprise. We exploit a natural benchmark in terms of corporate-led deal

activity and also existing theoretical and empirical research that guides us to different possible motives

for such transactions. These motives furnish testable alternative hypotheses to juxtapose against the null

hypothesis that acquisitions by government-led and corporate acquirers are not different.

We also offer a surprisingly new perspective for the recent growing literature focused on SWF

acquisitions by benchmarking their decisions and outcomes in a novel way, which is relative to cross-

border acquisitions led by other government-controlled acquirers as well as corporate acquirers. The

attributes and characteristics of targets of SWFs are somewhat different and, though the market reactions

are positive to SWF investment announcements (as other studies have shown), they appear to be smaller

than those associated with other government-controlled acquirers. The longer-run investment

performance of the targets of SWF acquisitions also appear to be poorer.

There are still many open questions. We readily admit that there are several possible alternative

explanations for government-led acquisitions that we have not yet considered. For example, we have not

yet tried to identify characteristics of the different types of government agencies that represented these

34

acquirers. We have also not tried to separate out the types of SWFs or state-controlled funds by

governance and transparency attributes, as the other studies of SWFs have done. Our study does not

examine longer-term operational and financial consequences from their newly-acquired stakes that more

appropriately reflect the political, economic or social goals of government acquirers. Other analysis that

we have largely ignored is at the policy level. A number of countries have instituted rules and legislation

for foreign investment reviews and we have not evaluated what, if any, are the consequences of those rule

changes for cross-country acquisition activity or for terms and conditions at the deal level. Finally, we

have not evaluated any positive or negative externalities of cross-border government-led deal activity for

other social, political, and economic objectives. After all, decision-makers that influence the government-

controlled acquirers that we study likely have a broader set of stakeholder concerns than just which

targets are chosen and how shareholders react to their announcement.

35

References

Aghion, Philippe and Jean Tirole. "The Management Of Innovation," Quarterly Journal of Economics,

1994, v109(4), 1185-1209.

Ahern, Kenneth, Daniele Daminelli, and Cesare Fracassi, “Lost in Translation? The Effect of Cultural

Values on Mergers around the World,” University of Michigan Working Paper, 2011, Available

at: http://ssrn.com/abstract=1545514.

Allen, Jeffrey W. and Gordon M. Phillips. "Corporate Equity Ownership, Strategic Alliances, And

Product Market Relationships," Journal of Finance, 2000, v55(6,Dec), 2791-2815.

Anderson, James E. "A Theoretical Foundation For The Gravity Equation," American Economic Review,

1979, v69(1), 106-116.

Anderson, James E. and Eric Van Wincoop. "Gravity And Gravitas: A Solution To The Border Puzzle,"

American Economic Review, 2003, v93(1,Mar), 170-192.

Antrás, Pol, Mihir A. Desai, and C. Fritz Foley, Multinational Firms, FDI Flows and Imperfect Capital

Markets, Quarterly Journal of Economics, 2009, v124(3), 1171-1219.

Atkinson, A.B. and Stiglitz. “J. E. Lectures in Public Economics,” London: McGraw-Hill, 1980.

Baker, Malcolm, C. F. Foley and Jeffrey Wurgler. "Multinationals As Arbitrageurs: The Effect Of Stock

Market Valuations On Foreign Direct Investment,” Review of Financial Studies, 2008, v22(1),

337-369.

Balding, Christopher. “A Portfolio Analysis of Sovereign Wealth Funds,” Regulating Foreign Capital:

The Rise and Implications of Sovereign Wealth Fund and State-Owned Enterprise Investment,

May 2011, Imperial College Press.

Bates, Thomas W. and Michael L. Lemmon. "Breaking Up Is Hard To Do? An Analysis Of Termination

Fee Provisions And Merger Outcomes," Journal of Financial Economics, 2003, v69(3,Sep), 469-

504.

Beck, Roland, and Michael Fidora. “The Impact of Sovereign Wealth Funds on Global Financial

Markets,” Intereconomics, 2008, v43(6), 349-358.

Berger, Allen N., George R. G. Clarke, Robert Cull, Leora Klapper, and Gregory F. Udell. “Corporate

Governance And Bank Performance: A Joint Analysis Of The Static, Selection, And Dynamic

Effects Of Domestic, Foreign, And State Ownership,” Journal of Banking and Finance, 2005,

v29(8/9,Aug/Sep), 2179-2221.

Bernstein, Shai, Josh Lerner, and Antoinette Schoar. “The Investment Strategies of Sovereign Wealth

Funds,” Harvard Business School working paper, April 2009, Available at:

http://www.nber.org/papers/w14861.

Bremmer, Ian. The End of the Free Market: Who Wins the War between States and Corporations?

Penguin Books, 2010, London, U.K.

36

Bris, Arturo, Neil Brisley, and Christos Cabolis. "Adopting Better Corporate Governance: Evidence From

Cross-Border Mergers," Journal of Corporate Finance, 2008, v14(3,Jun), 224-240.

Bris, Arturo and Christos Cabolis. "The Value Of Investor Protection: Firm Evidence From Cross-Border

Mergers," Review of Financial Studies, 2008, v21(2,Mar), 605-648.

Buckley, Peter J., and Mark Casson, The Future of the Multinational Enterprise, 1976, Holmes and Meier

Publishers, New York, N.Y.

Butt, Shams, Anil Shivdasani, Carsten Standavad, and Ann Wyman. “Sovereign Wealth Funds: A

Growing Global Force in Corporate Finance,” Journal of Applied Corporate Finance, 2008, 20(1),

73-83.

Caves, Richard E. "International Corporations: The Industrial Economics Of Foreign Investment,"

Economica, 1971, v38(149), 1-27.

Chan, Yuk-Shee. "On The Positive Role Of Financial Intermediation In Allocation Of Venture Capital In

A Market With Imperfect Information," Journal of Finance, 1983, v38(5), 1543-1568.

Chhaochharia, Vidhi, and Luc Laeven. “Sovereign Wealth Funds: Their Investment Strategies and

Performance,” University of Miami working paper, 2009, Available at:

http://ssrn.com/abstract=1262383.

Chari, Anusha, Wenjie Chen, and Kathryn M.E. Dominguez. “Foreign Ownership and Firm Performance:

Emerging-Market Acquisitions in the United States,” IMF Economic Review, 2012, v60, 1-42.

Coeurdacier, Nicolas, Roberto De Santis, and Antonin Aviat. “Cross-Border Mergers and Acquisitions

and European Integration,” Economic Policy, 2009, v57, 57-89.

Cushman, David O. "The Effects Of Real Wages And Labor Productivity On Foreign Direct Investment,"

Southern Economic Journal, 1987, v54(1), 174-185.

Desai, Mihir A., C. Fritz Foley and James R. Hines, Jr. "The Costs Of Shared Ownership: Evidence From

International Joint Ventures," Journal of Financial Economics, 2004a, v73(2,Aug), 323-374.

Desai, Mihir A., C. Fritz Foley and James R. Hines, Jr. “A Multinational Perspective On Capital Structure

Choice And Internal Capital Markets," Journal of Finance, 2004b, v59(6,Dec), 2451-2487.

Desai, Mihir A., C. Fritz Foley and James R. Hines, Jr. "Capital Controls, Liberalizations, And Foreign

Direct Investment," Review of Financial Studies, 2006, v19(4,Winter), 1433-1464.

Dewenter, Kathryn L. "Do Exchange Rate Changes Drive Foreign Direct Investment?" Journal of

Business, 1995, v68(3), 405-433.

Dewenter, Kathryn L. and Paul H. Malatesta. "Public Offerings Of State-owned And Privately-Owned

Enterprises: An International Comparison," Journal of Finance, 1997, v52(4,Sep), 1659-1679.

Dewenter, Kathryn L. and Paul H. Malatesta. "State-Owned And Privately Owned Firms: An Empirical

Analysis Of Profitability, Leverage, And Labor Intensity," American Economic Review, 2001,

v91(1,Mar), 320-334.

37

Dewenter, Kathryn L., Xi Han and Paul H. Malatesta. “Firm Values and Sovereign Wealth Investments,”

Journal of Financial Economics, 2010, v98(2), 256-278

Di Giovanni, Julian. “What Drives Capital Flows? The Case of Cross-border M&A Activity and

Financial Deepening,” Journal of International Economics, 2005, v65(4), 127–149.

Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer. "The Law And

Economics Of Self-Dealing," Journal of Financial Economics, 2008, v88(3), 430-465.

Doidge, Craig G., Andrew Karolyi, Karl V. Lins, Darius P. Miller and Rene M. Stulz. "Private Benefits

Of Control, Ownership, And The Cross-listing Decision," Journal of Finance, 2009, v64(1), 425-

466.

Dunning, John H. International Production and the Multinational Enterprise, 1981, Allen and Unwin

Publishers, London, U.K.

Dyck, I. J. Alexander and Morse, Adair, “Sovereign Wealth Fund Portfolios,” Chicago Booth Research

Paper No. 11-15, February 2011, Available at SSRN: http://ssrn.com/abstract=1792850.

Ellis, Jesse, Sara B. Moeller, Frederik P. Schlingemann, and René M. Stulz. “Globalization, Governance,

and the Returns to Cross-Border Acquisitions,” Ohio State University working paper, 2011,

Available at SSRN: http://ssrn.com/abstract=1737224.

Erel, Isil, Rose C. Liao, and Michael S. Weisbach. "Determinants of Cross-border Mergers and

Acquisitions,” Journal of Finance, 2012, v67(3), 1045-1082.

Faccio, Mara and Larry H. P. Lang. "The Ultimate Ownership Of Western European Corporations,"

Journal of Financial Economics, 2002, v65(3,Sep), 365-395.

Fama, Eugene F. "What's Different About Banks?," Journal of Monetary Economics, 1985, v15(1), 29-40.

Fama, Eugene F. and Kenneth R. French. "Size, Value, and Momentum in International Stock Returns,"

Journal of Financial Economics, 2012, v105(3), 457–72.

Fee, C. Edward, Charles J. Hadlock, and Shawn Thomas. "Corporate Equity Ownership And The

Governance Of Product Market Relationship," Journal of Finance, 2006, v61(3,Jun), 1217-1251.

Fernandes, Nuno. “Sovereign Wealth Funds: Investment Choices and Implications Around the World,”

IMD working paper, 2011, Available at: http://ssrn.com/abstract=1341692.

Ferreira, Miguel A., Massimo Massa, and Pedro P. Matos. “Shareholders at the Gate? Institutional

Investors and Cross-Border Mergers and Acquisitions,” Review of Financial Studies, 2010, 23(2),

601-644.

Fotak, Veljko, Bernardo Bortolotti, William L. Megginson, and William Miracky. “The Financial Impact

of Sovereign Wealth Fund Investments in Listed Companies,” University of Oklahoma Working

Paper, 2010, Available at: http://ssrn.com/abstract=1108585.

Froot, Kenneth A. and Jeremy C. Stein. "Exchange Rates And Foreign Direct Investment: An Imperfect

Capital Markets Approach," Quarterly Journal of Economics, 1991, v106(4), 1191-1218.

38

Golub, Stephen S. “Measures of Restrictions on Inward Foreign Direct Investment for OECD Countries,”

OECD Economic Studies, 2003, 36(1).

Graham, Edward M. and Paul R. Krugman. “Foreign Direct Investment in the United States,” Washington

D.C.: Institute for International Economics, 1995.

Grossman, Sanford J. and Oliver D. Hart. "The Costs And Benefits Of Ownership: A Theory Of Vertical

And Lateral Integration," Journal of Political Economy, 1986, v94(4), 691-719.

Hadlock, Charles J., and Josh R. Pierce. “New Evidence on Measuring Financial Constraints: Moving

Beyond the KZ Index,” Review of Financial Studies, 2010, v23(5), 1909-1940.

Hertzel, Michael and Richard L. Smith. "Market Discounts And Shareholder Gains For Placing Equity

Privately," Journal of Finance, 1993, v48(2), 459-485.

Hou, Kewei, G. Andrew Karolyi, and Bong-Chan Kho. “What Factors Drive Global Stock Returns?”

Review of Financial Studies, 2011, v24(8), 2527-2574.

Jory, Surendranath, Mark Perry, and Thomas Hemphill. “An Analysis of Sovereign Wealth Funds’

Recent Venture in Developed Capital Markets,” Management Online Review, October 2008,

Oxford Management Publishing,

Kang, Jun-Koo and Jin-Mo Kim. "The Geography Of Block Acquisitions," Journal of Finance, 2008,

v63(6,Dec), 2817-2858.

Kho, Bong-Chan, Rene M. Stulz and Francis E. Warnock. "Financial Globalization, Governance, And

The Evolution Of The Home Bias," Journal of Accounting Research, 2009, v47(2), 597-635.

King, Gary, and Langzhe Zeng. “Logistic Regression in Rare Events Data,” Political Analysis, 2001,

v9(2), 137-163.

Klein, Michael W., Joe Peek and Eric P. Rosengren. "Troubled Banks, Impaired Foreign Investment: The

Role Of Relative Access To Credit," American Economic Review, 2002, v92(3,Jun), 664-682.

Klein, Michael W. and Eric Rosengren. "The Real Exchange Rate And Foreign Direct Investment In The

United States: Relative Wealth Vs. Relative Wage Effects," Journal of International Economics,

1994, v36(3/4), 373-389.

Knill, April, Bong-Soo Lee, and Nathan Mauck. “'Sleeping with the Enemy' or 'An Ounce of Prevention':

Sovereign Wealth Fund Investments and Market Destabilization,” Journal of Financial

Intermediation, 2012, v21(2), 315-340.

Kotter, Jason, and Ugur Lel. “Friends or Foes? The Stock Price Impact of Sovereign Wealth Fund

Investments and the Price of Keeping Secrets,” Journal of Financial Economics, 2011, v101(2),

360-381.

La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny. "Law And

Finance," Journal of Political Economy, 1998, v106(6,Dec), 1113-1155.

Leuz, Christian, Karl V. Lins, and Francis E. Warnock. "Do Foreigners Invest Less In Poorly Governed

Firms?,” Review of Financial Studies, 2009, v22(8), 3245-3285.

39

Liao, Rose C. “Corporate Block Acquisitions around the World,” Ohio State University working paper,

2009, Available at: http://fisher.osu.edu/~liao_102/.

Megginson, William L., Robert C. Nash, and Matthias Van Randenborgh. "The Financial And Operating

Performance Of Newly Privatized Firms: An International Empirical Analysis," Journal of

Finance, 1994, v49(2), 403-452.

Mian, Atif. "Distance Constraints: The Limits Of Foreign Lending In Poor Economies," Journal of

Finance, 2006, v61(3,Jun), 1465-1505.

Micco, Alejandro, Ugo Panizza, and Monica Yanez. "Bank Ownership And Performance: Does Politics

Matter?,” Journal of Banking and Finance, 2007, v31(1,Jan), 219-241.

Moeller, Sara B., Frederik P. Schlingemann, and Rene M. Stulz. “Wealth Destruction On A Massive

Scale? A Study Of Acquiring-Firm Returns In The Recent Merger Wave," Journal of Finance,

2005, v60(2,Apr), 757-782.

Morck, Randall and Bernard Yeung. "Why Investors Value Multinationality," Journal of Business, 1991,

v64(2), 165-188.

Morck, Randall, Bernard Yeung, and Minyuan Zhao. "Perspectives on China’s Outward Foreign Direct

Investment," Journal of International Business Studies, 2008, v39(3), 337-350.

Netter, Jeffry M., Mike Stegemoller, and Babajide M. Wintoki. “Implications of Data Screens on Merger

and Acquisition Analysis: A Large Sample Study Of Mergers And Acquisitions From 1992–

2009,” Review of Financial Studies, 2011, v24, 2316-2357.

Portes, Richard and Helene Rey. "The Determinants Of Cross-Border Equity Flows," Journal of

International Economics, 2005, v65(2,Mar), 269-296.

Rhodes-Kropf, Matthew and S. Viswanathan. "Market Valuation And Merger Waves," Journal of

Finance, 2004, v59(6,Dec), 2685-2718.

Rossi, Stefano and Paolo F. Volpin. "Cross-Country Determinants Of Mergers And Acquisitions," Journal

of Financial Economics, 2004, v74(2,Nov), 277-304.

Saw, Swee-Hock, and Linda Low. Sovereign Wealth Funds, 2009, Saw Centre for Financial Studies,

NUS Business School, National University of Singapore Publishers, Singapore.

Shleifer, Andrei and Robert W. Vishny. "Politicians And Firms," Quarterly Journal of Economics, 1994,

v109(4), 995-1025.

Shleifer, Andrei and Robert W. Vishny. "Stock Market Driven Acquisitions," Journal of Financial

Economics, 2003, v70(3,Dec), 295-311.

Sojli, Elvira, and Wing Wah Tham. “The Impact of Foreign Government Investment: Sovereign Wealth

Fund Investments in the US,” International Review of Finance, 2011, v12, 207 – 243.

Starks, Laura, and Kelsey Wei. “Cross-Border Mergers and Differences in Corporate Governance,”

University of Texas working paper, 2012, forthcoming International Review of Finance.

40

Stulz, René. “The Limits of Financial Globalization,” Journal of Finance, 2005, v60(4), 1595-1638.

Stulz, Rene M. and Rohan Williamson. "Culture, Openness, And Finance," Journal of Financial

Economics," 2003, v70(3,Dec), 313-349.

Taboada, Alvaro. “The Impact of Changes in Bank Ownership Structure on the Allocation of Capital,”

Journal of Banking & Finance, 2011, v35(10), 2528-2543.

Tirole, Jean. "The Internal Organization Of Government," Oxford Economic Papers, 1994, v46(1), 1-29.

Tirole, Jean. “Corporate Governance,” Econometrica, 2001, v69(1), 1-35.

Truman, Edwin M. “A Blueprint for Sovereign Wealth Fund Best Practices,” Revue d'Économie

Financière, Programme National Persée, 2009, v9(1), 429-451.

Whited, Toni M. and Guojun Wu. "Financial Constraints Risk," Review of Financial Studies, 2006,

v19(2,Summer), 531-559.

Williamson, Oliver E. "Transaction-Cost Economics: The Governance Of Contractual Relations," Journal

of Law and Economics, 1979, v22(2), 233-262.

41

Table 1. Summary Statistics.

This table presents summary statistics on cross-border block acquisitions involving at least a 5% stake in the target

corporation announced during the period of 1990 to 2008. The data are obtained from Thomson Reuter’s Security

Data Corporation’s (SDC) Platinum Mergers and Corporate Transactions database. We exclude leveraged buyouts,

spin-offs, recapitalizations, self-tender offers, exchange offers, repurchases and privatizations and deals in which

acquirers are domiciled in overseas territories of the U.K. (Bahamas, British Virgin Islands, Cayman Islands,

Guernsey, Isle of Man) and the Netherlands (Netherland Antilles) and we exclude countries in which there are fewer

than 50 cross-border acquisitions, whether led by government-controlled acquirers, corporate acquirers or

government-controlled funds, over the period from 1990 to 2008. The acquirer’s ultimate parent public status is used

to identify government controlled acquirers, which is defined as at least 50% cash flow ownership. We define

government-controlled funds as those that are sovereign wealth funds (SWFs) or state-controlled funds. SWFs are

identified as a financial acquirer in Securities Data Corporation under ACQUIROR_TYPE data item and matched

by name (using SDC data item AN) to a list at the SWF Institute website, http://www.swfinstitute.org/funds.php).

State-controlled or public investment funds are defined by a primary Standard Industrial Classification (SIC) code in

the 999A-G, 9000 range and/or government acquirers with any of the following SIC codes related to investment

offices, pension, health and welfare funds, trusts, or holding companies: 6019, 6371, 6722, 6726, 6798, 6799.We

report the total number of deals, the subset with disclosed values, the average deal value (measured in millions of

constant US dollars as of 2000), the total deal value, the number of withdrawn deals, the number involving minority

stakes (less than 50%), and the number of publicly-traded target firms.

Government-

controlled Acquirers

Corporate

Acquirers

Government-

controlled Funds

Total No. of Deals 4,026 123,027 733

No. of Deals with Values Disclosed 1,613 51,277 423

Average Deal Value (Constant 2000 US$) 269 165 374

Cumulative Deal Value (Constant 2000 US$) 434,549 8,447,883 158,176

No. of Withdrawn Deals 1,355 30,474 229

No. of Minority Acquisitions 2,222 46,581 536

No. of Public Targets 715 12,669 304

42

Table 2. Intensity of Cross-Border Acquisition Activity Led by Government-Controlled Acquirers by Country of Acquirers and Targets.

This table presents the number of deals and the total deal value (in 2000 Constant US$ millions) of cross-border block acquisitions led by government-controlled

acquirers (“Government acquirers”), corporate acquirers, government-controlled funds (“Government funds”) involving at least a 5% stake in the target

corporation announced over the period from 1990 to 2008. See Table 1 for data sources, identification of type of acquirer and exclusions by type of deal. Results

are reported by country ranked by the fraction of total deal value that government-led acquirers comprise for leading acquirer countries (Panel A) and for the

leading target countries (Panel B).

Number of Deals Total Deal Value (2000 Constant US$ millions)

Government Acquirers

(1)

Corporate Acquirers

(2)

Government Funds

(3)

Government Acquirers

(1)

Corporate Acquirers

(2)

Government Funds

(3)

Count Percent Count Percent Count Percent Deal Value Percent Deal Value Percent Deal Value Percent

Panel A: By Acquirer Country

China 673 16.7% 557 0.5% 160 21.8% $99,929 23.0% $24,980 0.3% $14,295 9.0%

France 691 17.2% 7,571 6.2% 26 3.5% $81,067 18.7% $736,125 8.7% $12,620 8.0%

Singapore 203 5.0% 2,415 2.0% 187 25.5% $27,834 6.4% $55,405 0.7% $41,030 25.9%

Saudi Arabia 34 0.8% 71 0.1% 3 0.4% $23,149 5.3% $3,103 0.0% N.A. N.A.

Utd Arab Em 85 2.1% 124 0.1% 100 13.6% $22,432 5.2% $8,155 0.1% $44,470 28.1%

Sweden 195 4.8% 4,075 3.3% 7 1.0% $21,366 4.9% $149,214 1.8% $38 0.0%

Italy 115 2.9% 2,574 2.1% 7 1.0% $19,017 4.4% $224,177 2.7% $3,423 2.2%

Norway 254 6.3% 1,521 1.2% 26 3.5% $17,634 4.1% $50,530 0.6% $3,504 2.2%

Finland 233 5.8% 1,647 1.3% 8 1.1% $12,433 2.9% $67,592 0.8% $63 0.0%

Germany 236 5.9% 8,523 6.9% 8 1.1% $12,252 2.8% $863,474 10.2% $196 0.1%

Kuwait 44 1.1% 106 0.1% 14 1.9% $11,218 2.6% $6,995 0.1% $2,960 1.9%

Malaysia 66 1.6% 1,566 1.3% 13 1.8% $9,654 2.2% $42,888 0.5% $1,215 0.8%

Spain 36 0.9% 1,808 1.5% 2 0.3% $9,034 2.1% $309,190 3.7% $1 0.0%

Japan 36 0.9% 4,866 4.0% 10 1.4% $8,711 2.0% $224,285 2.7% $12,682 8.0%

Switzerland 52 1.3% 3,981 3.2% 2 0.3% $7,991 1.8% $434,518 5.1% $1 0.0%

Portugal 34 0.8% 396 0.3% 13 1.8% $6,283 1.4% $24,131 0.3% $663 0.4%

Rest of the World 1,039 25.8% 81,226 66.0% 147 20.1% 44,544 10.3% 5,223,120 61.8% 21,018 13.3%

Panel B: By Target Country

United States 413 10.3% 21,370 17.4% 89 12.1% $77,752 17.9% $2,406,000 28.5% $45,861 29.0%

Hong Kong 446 11.1% 2,296 1.9% 137 18.7% $58,671 13.5% $72,970 0.9% $7,467 4.7%

United Kingdom 304 7.6% 12,415 10.1% 79 10.8% $47,982 11.0% $1,350,127 16.0% $43,223 27.3%

Australia 187 4.6% 5,283 4.3% 31 4.2% $24,124 5.6% $236,021 2.8% $3,242 2.0%

Italy 153 3.8% 3,388 2.8% 11 1.5% $22,956 5.3% $269,566 3.2% $262 0.2%

Netherlands 106 2.6% 3,371 2.7% 13 1.8% $21,744 5.0% $375,262 4.4% $63 0.0%

Germany 350 8.7% 8,993 7.3% 20 2.7% $18,327 4.2% $599,131 7.1% $1,726 1.1%

Finland 72 1.8% 1,629 1.3% 3 0.4% $11,809 2.7% $43,155 0.5% $468 0.3%

Spain 125 3.1% 3,390 2.8% 15 2.0% $11,089 2.6% $214,682 2.5% $12,710 8.0%

Canada 88 2.2% 6,865 5.6% 9 1.2% $10,550 2.4% $509,033 6.0% $1,715 1.1%

Norway 79 2.0% 1,732 1.4% 2 0.3% $10,125 2.3% $86,076 1.0% $14 0.0%

Sweden 174 4.3% 2,838 2.3% 18 2.5% $9,775 2.2% $187,105 2.2% $2,082 1.3%

Argentina 38 0.9% 1,409 1.1% 2 0.3% $7,582 1.7% $64,150 0.8% $3,416 2.2%

Russian Fed 52 1.3% 1,303 1.1% 4 0.5% $6,727 1.5% $65,789 0.8% $221 0.1%

France 134 3.3% 6,362 5.2% 19 2.6% $6,544 1.5% $372,947 4.4% $2,938 1.9%

Singapore 49 1.2% 1,649 1.3% 12 1.6% $6,533 1.5% $50,488 0.6% $1,063 0.7%

Rest of the World 1,256 31.2% 38,734 31.5% 269 36.7% 82,259 18.9% 1,545,381 18.3% 31,706 20.0%

World 4,026 100.0% 123,027 100.0% 733 100.0% 434,549 100.0% 8,447,883 100.0% 158,176 100.0%

43

Table 3. Cross-Country Determinants of Cross-Border Acquisition Activity Led by Government-Controlled Acquirers.

This table presents panel regressions of country-level determinants on the total number (or deal value in constant 2000 US$ millions) of cross-border block

acquisition deals led by government-controlled acquirers from 1990 to 2008. The dependent variable is the total number of deals (or total deal value) between an

ordered country pair, normalized by the total number of cross-border acquisitions in the host country j. See Table 1 for data sources, identification of type of

acquirer and exclusions by type of deal. Models (1), (4), (5), and (6) include only government-controlled acquirers, denoted AGijt,. Model 2 includes corporate

acquirers (ACijt ) and Models (3) and (7) government-controlled funds (AF

ijt). Model 4 measures activity by the total constant-dollar deal activity instead of counts,

Model (5) considers only those deal counts involving minority stakes between 5% and 50% block purchases in targets. See Table IA1 for details on variable

construction and Table IA2 for summary statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels using year fixed effects and robust

standard errors clustered by target country and associated t-statistics are in parentheses. “Diff (1)-(2)” (“Diff (1)-(3)”) denotes a t-statistic from a regression of the

excess fraction of the number of deals between government-controlled acquirers, AGijt, and corporate acquirers, AC

ijt, (government funds, AFijt).

By Total Number of Deals By Value

of Deals

Minority

Deals only

By Number

of Deals

By Number

of Deals

(AGijt) (AC

ijt) (AFijt) Diff Diff (AG

ijt) (AGijt) (AG

ijt) (AFijt)

(1) (2) (3) (1) - (2) (1) - (3) (4) (5) (6) (7)

Annual Exchange Rate Return Differences 0.044*** 0.023* 0.007

0.033* 0.050*** 0.053*** 0.026

(4.35) (1.77) (0.37) (2.95) (1.11) (1.98) (4.18) (4.19) (1.00)

Annual Real Stock Market Return Differences 0.003 -0.007* -0.024***

0.010 0.005 -0.009 -0.019

(0.52) (-1.91) (-2.77) (1.70) (2.74) (0.93) (1.12) (-1.51) (-1.59)

Log GDP Per Capita Differences 0.002 0.012*** 0.010***

0.004 0.001 0.006** 0.018***

(0.82) (5.43) (5.29) (-3.22) (-2.80) (1.67) (0.57) (2.24) (4.86)

GDP Growth Differences 0.050 -0.072 0.303***

0.123* 0.030 -0.090 -0.013

(0.85) (-1.62) (3.06) (1.08) (-2.02) (1.80) (0.58) (-1.49) (-0.13)

Geographic Proximity 0.006*** 0.003** 0.004***

0.006*** 0.005*** 0.004*** 0.003**

(4.91) (2.44) (3.16) (1.80) (1.17) (5.66) (4.34) (3.88) (2.23)

Market Correlation 0.056*** 0.092*** 0.080***

0.060*** 0.044** 0.082*** 0.103***

(2.91) (4.68) (5.11) (-2.81) (-1.98) (2.90) (2.38) (3.67) (4.77)

ASDI differences -0.028** 0.022*** 0.057***

-0.010 -0.018 -0.026** 0.041***

(-2.31) (3.23) (5.40) (-4.50) (-4.31) (-0.85) (-1.60) (-2.05) (2.82)

Accounting Standard Differences 0.112*** 0.083*** -0.016

0.088** 0.078** 0.125*** -0.054

(3.00) (4.37) (-0.32) (1.23) (1.76) (2.20) (2.13) (3.36) (-1.01)

European Union Dummy 0.027** -0.017* 0.005

0.023 0.037*** 0.009 0.006

(2.27) (-2.01) (0.40) (4.57) (1.14) (1.63) (3.37) (0.77) (0.42)

Industry Dissimilarity

-0.421* -0.664*

(-1.99) (-1.95)

Polity IV Democracy Differences

-0.008*** -0.010**

(-3.55) (-2.58)

Total Reserve % GDP

-0.098*** 0.046

(-3.06) (0.64)

Govt Domestic Acquisition Activity

0.961*** 0.490**

(6.70) (2.19)

Observations 8,485 9,613 3,998 8,485 3,748 6,617 7,512 7,100 3,125

Adjusted R2 0.047 0.126 0.046 0.026 0.019 0.028 0.035 0.095 0.123

Adjusted R2 (without year fixed effects) 0.043 0.115 0.042 0.026 0.017 0.027 0.032 0.093 0.119

Adjusted R2 (with year and country fixed effects) 0.150 0.544 0.216 0.243 0.103 0.110 0.122 0.166 0.231

44

Table 4. Logistic Regression Analysis of Probability of Firm Targeted by Government-Controlled Acquirer.

This table presents logistic regressions of the probability that a firm is targeted by a government-controlled acquirer

in a given year. In Models (1) to (4) (Models 5 to 8), the dependent variable equals one if the firm is targeted by a

government-controlled acquirer in any given year between 1990 and 2008 and zero, if it is targeted by a corporation

(Government-controlled funds). Financials and utilities as target firms are excluded in all models except (1) and (4).

Firms with total assets smaller than US$1 million (in 2000 constant dollars) and with negative book values of equity

are also excluded. See Table 1 for data sources, identification of type of acquirer and exclusions by type of deal.

Models (2) and (6) present results for minority stake acquisitions (above 5% but below 50% of target firm’s shares

acquired). Models (4) and (8) present results for successful deals. See Table IA1 for details on variable construction

and Table IA2 for summary statistics. Coefficients are reported as marginal effects. ***, **, and * denote statistical

significance at the 1%, 5% and 10% levels using robust standard errors that allow country and year fixed effects as

indicated and associated t-statistics are in parentheses below the coefficients.

Panel A: Government-controlled acquirers

versus corporate acquirers

Panel B: Government-controlled acquirers

versus government funds

All Deals

Minority

Deals

Only

Exclude

Financials

&

Utilities

Excluding

Withdrawn

Deals

All Deals

Minority

Deals

Only

Exclude

Financials

&

Utilities

Excluding

Withdrawn

Deals

(1) (2) (3) (4)

(5) (6) (7) (8)

Related deals dummy 0.009** -0.007 -0.007* -0.009**

0.458***

(2.18) (-1.34) (-1.80) (-2.34)

(12.06)

Zero-dividend dummy 0.002 0.016** 0.011** 0.009*

-0.038 -0.072 -0.045 -0.201

(0.40) (2.31) (2.43) (1.77)

(-0.85) (-0.37) (-0.30) (-0.71)

High closely-held share dummy 0.010* -0.003 -0.002 -0.001

0.059 0.429*** 0.246** -0.265

(1.78) (-0.40) (-0.37) (-0.11)

(1.62) (3.05) (2.06) (-1.00)

Total assets (log) 0.006*** 0.007*** 0.006*** 0.003**

-0.031*** -0.135*** -0.109*** -0.188**

(5.21) (4.51) (4.80) (2.38)

(-2.72) (-2.65) (-2.72) (-2.18)

Market-to-book 0.002* 0.002 0.002* 0.002

-0.004 0.070 -0.080** -0.104*

(1.70) (1.48) (1.74) (1.55)

(-0.46) (1.12) (-2.48) (-1.90)

Return on assets -0.011 -0.000 -0.008 -0.004

0.064 -0.380 -0.001 -0.065

(-1.42) (-0.00) (-1.35) (-0.65)

(1.15) (-0.82) (-0.01) (-0.16)

Long-term debt/assets 0.009 -0.025 -0.020 -0.021

-0.115 -0.331 0.019 -1.574

(0.76) (-1.34) (-1.47) (-1.26)

(-1.06) (-0.48) (0.04) (-1.36)

Sales growth 0.004** 0.003 0.002 -0.003

0.006 0.023 0.010 0.040

(2.37) (1.17) (0.93) (-1.07)

(0.53) (0.26) (0.19) (0.25)

All Cash Payment Dummy -0.007* 0.004 -0.000 0.001

-0.035 -0.136 -0.130 0.042

(-1.71) (0.62) (-0.10) (0.31)

(-0.81) (-0.80) (-0.96) (0.17)

Observations 6,100 2,281 3,791 2,527

454 104 153 74

Pseudo R2 0.1588 0.1831 0.1882 0.2125

0.4131 0.3160 0.3467 0.4002

Pseudo R2 (without fixed effects) 0.0166 0.0241 0.0213 0.0236

0.2529 0.0473 0.0492 0.0922

Predicted Y 0.0337 0.0235 0.0196 0.0159

0.8671 0.4747 0.6104 0.5210

45

Table 5. Cumulative Market-Adjusted Returns (CMARs) and Bidder Premiums

This table reports the cumulative market-adjusted buy-and-hold returns (CMARs) and bid premiums in percentages around the announcement dates of cross-

border acquisitions led by government-controlled acquirers, corporate acquirers and government-controlled funds. Buy-and-hold returns are cumulated over

three different returns horizons around the announcement date (t=0), including from days t=-10 to t=+10 (“CMARs(-10,+10)”), days t=-5 to t=+5 (“CMARs(-

5,+5)”), and days t=-1 to t=+1 (“CMARs(-1,+1)”). Bid premium is the bid price measured relative to the closing stock price of the target four weeks prior to the

announcement date, expressed as a percentage (defined using SDC codes ((HOSTPR – HOSTC4WK) / HOSTC4WK × 100). We denote it “Bid Premium (-

4w,0).” A similar premium bid defined relative to the closing stock price of target one week prior (-1w, 0) and one day prior (-1d, 0). Panel A (B) present results

for all deals (minority stake acquisitions in which above 5% but below 50% of target firm’s shares acquired). See Table 1 for data sources, identification of type

of acquirer and exclusions by type of deal. Financial services and utilities target firms are excluded. Median values are reported with p-values for the Wilcoxon

rank-sum tests associated with differences in medians between groups are presented in parentheses. ***, **, and * denote statistical significance at the 1%, 5% and

10% levels.

Panel A: All Deals

Panel B: Minority Deals Only

Government-

controlled

Acquirers

(1)

Corporate

Acquirers

(2)

Government

Funds

(3)

Difference

in

medians

(1) - (2)

Difference

in

medians

(1) - (3)

Government-

controlled

Acquirers

(1)

Corporate

Acquirers

(2)

Government

Funds

(3)

Difference

in

medians

(1) - (2)

Difference

in

medians

(1) - (3)

Number of Obs. (298) (7,482) (138)

(233) (4,767) (122)

CMARs (-10, +10) 0.047*** 0.062*** 0.002**

(0.183) (<0.01)

0.033*** 0.040*** 0.001

(0.724) (<0.01)

CMARs (-5, +5) 0.028*** 0.050*** 0.005**

(0.082) (<0.01)

0.020*** 0.034*** 0.002*

(0.322) (0.018)

CMARs (-1, +1) 0.015*** 0.025*** 0.008**

(0.033) (0.058)

0.010*** 0.015*** 0.009*

(0.284) (0.429)

Number of Obs. (105) (3,077) (56)

(67) (1,426) (50)

Bid Premium (-4w,0) 0.232*** 0.268*** 0.040**

(0.338) (<0.01)

0.079*** 0.153*** 0.019*

(0.211) (0.188)

Bid Premium (-1w,0) 0.191*** 0.233*** 0.037**

(0.531) (0.025)

0.078** 0.125*** 0.029*

(0.320) (0.565)

Bid Premium (-1d,0) 0.113*** 0.194*** 0.071*

(0.132) (0.335)

0.035** 0.098*** 0.055*

(0.074) (0.585)

46

Table 6. Regression Analysis of Cross-section of Cumulative Market-Adjusted Returns (CMARs) to Announcements of Cross-Border Acquisitions

This table reports the results for cross-sectional regressions of cumulative market adjusted buy-and-hold returns (CMARs) around the announcement dates of

cross-border acquisitions led by government-controlled, corporate acquirers and government-controlled funds on a variety of firm-specific and country-level

variables. Buy-and-hold returns are cumulated around the announcement date (t=0) for days t=-10 to t=+10 (“CMARs (-10,+10)”), days t=-5 to t=+5 (“CMARs

(-5,+5)”), and days t=-1 to t=+1 (“CMARs (-1,+1)”). Models (4), (8), and (12) present results for minority stake acquisitions (above 5% but below 50% of target

firm’s shares acquired). Financial services and utilities target firms are excluded. See Table IA1 presents details on variable construction and Table IA2 for

summary statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels using robust standard errors that allow year and country fixed effects

and associated t-statistics are in parentheses below the coefficients. The p-values at the bottom of each model denote a test of the hypothesis (“Test H0”) that the

coefficients on government-controlled acquirer and corporate acquirer or government fund dummy variables are equal.

CMARs (-10,+10)

CMARs (-5,+5)

CMARs (-1,+1)

All Deals Minority

Deals All Deals

Minority

Deals All Deals

Minority

Deals

(1) (2) (3) (4)

(5) (6) (7) (8)

(9) (10) (11) (12)

Government-controlled Acquirer dummy 0.037 0.181*** 0.251*** 0.172***

0.026 0.138*** 0.228*** 0.117***

0.029* 0.103*** 0.153*** 0.088***

(1.48) (4.37) (3.39) (4.49)

(1.28) (4.09) (3.63) (3.63)

(1.80) (4.07) (2.99) (3.34)

Corporate acquirer dummy 0.044** 0.166*** 0.237*** 0.148***

0.041** 0.143*** 0.208*** 0.127***

0.047*** 0.118*** 0.159*** 0.094***

(2.08) (4.81) (3.71) (4.92)

(2.37) (5.17) (4.08) (5.16)

(3.37) (5.15) (3.54) (4.05)

Government-controlled fund dummy -0.065** 0.101** 0.121 0.096**

-0.048** 0.084** 0.110* 0.084***

-0.015 0.087*** 0.095* 0.078***

(-2.41) (2.38) (1.59) (2.46)

(-2.16) (2.43) (1.76) (2.70)

(-0.89) (3.29) (1.86) (2.99)

Zero-dividend dummy

-0.012 -0.008 -0.007

-0.002 -0.007 0.005

0.000 -0.008 0.010

(-1.17) (-0.56) (-0.54)

(-0.33) (-0.57) (0.57)

(0.01) (-0.73) (1.64)

High closely-held share dummy

0.006 -0.011 0.007

0.007 -0.013 0.007

0.004 -0.017 0.002

(0.59) (-0.71) (0.50)

(0.86) (-1.00) (0.79)

(0.60) (-1.52) (0.21)

Total assets (log)

-0.017*** -0.017*** -0.016***

-0.012*** -0.014*** -0.010***

-0.008*** -0.012*** -0.006***

(-7.50) (-4.23) (-5.58)

(-6.54) (-4.11) (-5.10)

(-5.75) (-3.95) (-4.45)

Market-to-book

-0.014*** -0.015*** -0.011***

-0.011*** -0.014*** -0.007***

-0.011*** -0.013*** -0.007***

(-5.29) (-3.73) (-3.96)

(-5.08) (-4.06) (-2.95)

(-6.75) (-3.86) (-5.08)

Long-term debt/assets

0.009 0.065 0.013

-0.013 0.045 -0.023

-0.016 0.023 -0.010

(0.30) (1.36) (0.44)

(-0.49) (1.11) (-0.89)

(-0.88) (0.70) (-0.59)

Sales growth

0.001 -0.003 0.007

-0.002 -0.000 0.003

-0.003 -0.002 -0.002

(0.23) (-0.44) (1.49)

(-0.46) (-0.03) (0.70)

(-1.07) (-0.28) (-0.60)

Bid premium (-1w, 0)

0.004

0.004

0.003

(1.07)

(1.13)

(1.30)

Withdrawn deals dummy

-0.044***

-0.031**

-0.018*

(-3.13)

(-2.56)

(-1.69)

Country fixed effects Yes Yes Yes Yes

Yes Yes Yes Yes

Yes Yes Yes Yes

Observations 7,075 3,834 1,656 2,442

7,058 3,834 1,656 2,442

6,997 3,820 1,651 2,431

R2 0.153 0.173 0.394 0.106

0.224 0.257 0.405 0.178

0.236 0.267 0.401 0.181

Test H0: Govt. Acquirers = Corp. Acquirers 0.646 0.463 0.680 0.249

0.242 0.751 0.543 0.594

0.046 0.171 0.791 0.587

Test H0: Govt. Acquirers = Govt. Funds <0.001 0.011 0.009 0.010

<0.001 0.042 0.009 0.199

<0.001 0.290 0.052 0.489

47

Table 7. Long-run Investment Returns of Target Firms

This table reports long-run investment returns of cross-border deals led by government-controlled acquirers,

corporate acquirers and government-controlled funds. Panel A reports the mean and median of buy-and-hold

abnormal returns (BHARs) cumulated from one month after the deal completion date (t = 0) for one year (t = +1 to t

= 12 months, or “12m”), two years (“24 m”), and three years (“36 m”). Median (mean) values are reported with p-

values for the Wilcoxon rank-sum tests (two-sided t-statistics) associated with differences in medians (means)

between groups are presented in parentheses. Panel B presents calendar-time portfolio monthly returns one year and

two year following the cross-border investment. We use equal weights in forming calendar-time portfolios and run

seemingly-unrelated regressions (SUR) using Fama-French (2012) four-factor model using global factors. See Table

1 for data sources, identification of type of acquirer and exclusions by type of deal. Financial services and utilities

target firms are excluded. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels. The p-values

reported at the bottom of Panel B are based on tests of the null hypothesis (“H0”) that the intercept coefficients

(“Alpha”) from SUR regressions using the Fama-French four-factor model are equal between portfolios that buy

targets of government-controlled acquirers compared to those of corporate acquirers or government funds.

Panel A: Buy-and-Hold Abnormal Returns (BHARs)

Government-

controlled

Acquirers

(1)

Corporate

Acquirers

(2)

Government

Funds

(3)

Median/

Mean Tests

(1) - (3)

Median/

Mean

Tests

(1) - (3)

BHAR(+1, 12m) Median -0.054 -0.130*** -0.107

(0.336) (0.892)

Mean -0.040 -0.044*** -0.043

(0.924) (0.959)

Observations (124) (3,772) (71)

BHAR(+1, 24m) Median -0.220 -0.228* -0.244**

(0.877) (0.364)

Mean -0.057 -0.040* -0.184**

(0.859) (0.330)

Observations (113) (3,473) (63)

BHAR(+1, 36m) Median -0.403 -0.306** -0.628***

(0.644) (0.043)

Mean -0.094 -0.076** -0.503***

(0.907) (0.026)

Observations (106) (3,211) (57)

Panel B: Calendar Time Portfolio Returns

One-year portfolio Two-year portfolio

Government

-controlled

Acquirers

Corporate

Acquirers

Government-

controlled

Funds

Government-

controlled

Acquirers

Corporate

Acquirers

Government-

controlled

Funds

(1) (2) (3)

(1) (2) (3)

Mean returns 0.0148 0.0126 0.0114

0.0101 0.0094 0.0079

Observations 270 275 237

271 275 265

Rm-Rf 0.8714*** 0.7225*** 1.3075***

0.9715*** 0.7730*** 1.1535***

(7.73) (17.44) (7.01)

(12.25) (22.58) (8.82)

Small Minus Big 0.7639*** 0.7987*** 1.1182***

0.8846*** 0.7317*** 1.2505***

(4.02) (11.43) (3.55)

(6.80) (13.04) (5.83)

High Minus Low -0.0057 0.0146 -0.1079

0.4564*** 0.1039* 0.3822*

(-0.03) (0.22) (-0.37)

(3.61) (1.90) (1.83)

Winner Minus Loser -0.0082 -0.1368*** 0.1087

-0.0699*** -0.0004 -0.1160***

(-0.07) (-3.36) (0.59)

(-3.38) (-0.04) (-3.40)

Alpha 0.0089** 0.0102*** 0.0010 0.0002 0.0048*** -0.0054

(2.12) (6.62) (0.14) (0.08) (3.65) (-1.07)

R2 0.306 0.721 0.269

0.460 0.768 0.334

Test H0: (1) = (2)

0.74

0.11

Test H0: (1) = (3)

0.13

0.15

48

Figure 1. Number of and Total Deal Value of All Cross-Border Acquisitions Led by Government-Controlled Acquirers by Year.

This figure exhibits the number of and total deal value (in 2000 Constant US$ millions) of cross-border block acquisitions led by government-controlled

involving at least a 5% stake in the target corporation announced over the period from 1990 to 2008. The data are obtained from Thomson Reuter’s Security Data

Corporation’s (SDC) Platinum Mergers and Corporate Transactions database. We exclude leveraged buyouts, spin-offs, recapitalizations, self-tender offers,

exchange offers, repurchases and privatizations and deals in which acquirers are domiciled in overseas territories of the U.K. (Bahamas, British Virgin Islands,

Cayman Islands, Guernsey, Isle of Man) and the Netherlands (Netherland Antilles) and we exclude direct investments by sovereign wealth funds and other

government-sponsored investment funds. We further exclude countries in which there are fewer than 50 cross-border acquisitions, whether led by government-

controlled or corporate acquirers or government-controlled funds, over the period from 1990 to 2008. The acquirer’s ultimate parent public status is used to

identify government controlled acquirers, which is defined as at least 50% cash flow ownership.

0

50

100

150

200

250

300

350

$0

$10,000

$20,000

$30,000

$40,000

$50,000

$60,000

$70,000

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Tota

l N

um

ber

of

Cro

ss-B

ord

er D

eals

wit

h G

ov

ern

men

t-

Co

ntr

oll

ed A

cqu

irer

s

Tota

l D

eal

Va

lue

of

Cro

ss-B

ord

er D

eals

wit

h G

ov

ern

men

t-

Co

ntr

oll

ed A

cqu

irer

s (i

n C

on

sta

nt

20

00

U.S

. $

mil

lio

ns)

Total Deal Value of Cross-Border Deals (Left Axis) Number of Cross-Border Deals Right Axis)

49

Figure 2. Total Deal Value of All Acquisitions Led by Government-Controlled Acquirers by Country of Acquirer and of Target Firms.

This figure exhibits the total deal value (in 2000 Constant US$ millions) of cross-border block acquisitions led by government-controlled and corporate acquirers

involving at least a 5% stake in the target corporation announced over the period from 1990 to 2008. The data are obtained from Thomson Reuter’s Security Data

Corporation’s (SDC) Platinum Mergers and Corporate Transactions database. We exclude leveraged buyouts, spin-offs, recapitalizations, self-tender offers,

exchange offers, repurchases and privatizations and deals in which acquirers are domiciled in overseas territories of the U.K. (Bahamas, British Virgin Islands,

Cayman Islands, Guernsey, Isle of Man) and the Netherlands (Netherland Antilles) and we further exclude countries in which there are fewer than 50 cross-

border acquisitions, whether led by government-controlled or corporate acquirers or government controlled funds, over the 1990-2008 period. The acquirer’s

ultimate parent public status is used to identify government controlled acquirers, which is defined as at least 50% cash flow ownership. The results are reported

by country in order of total deal value by government-led acquirers comprise for leading acquirer countries and their target country regions (Panel A) and for the

leading target countries and the home country region of their acquirers (Panel B).

Panel A: By Country of Domicile of Government-Led Acquirers and Target Regions

$- $20,000.00 $40,000.00 $60,000.00 $80,000.00 $100,000.00 $120,000.00

Switzerland

Japan

Spain

Malaysia

Kuwait

Germany

Finland

Norway

Italy

Sweden

United Arab Emirates

Saudi Arabia

Singapore

France

China

Tota

l D

eal

Va

lue

of

Cro

ss-B

ord

er D

eals

of

Go

ver

nm

ent-

Co

ntr

oll

ed A

cqu

irer

s (i

n C

on

sta

nt

20

00

U.S

. $

mil

lio

ns)

By

Larg

est

Acq

uir

er C

ou

ntr

y a

nd

Ta

rget

Reg

ion

s

Developed Americas

Latin America

Developed Asia-Pacific

Emerging Asia-Pacific

Europe Developed

EMEA

50

Figure 2. Total Deal Value of All Acquisitions Led by Government-Controlled Acquirers by Country of Acquirer and of Target Firms. (continued)

Panel B: By Country of Domicile of Target Firms of Government-Led Acquirers and Home Country Regions of Acquirers

$- $20,000.00 $40,000.00 $60,000.00 $80,000.00

France

Russian Federation

Argentina

Sweden

Norway

Canada

Spain

Finland

Germany

Netherlands

Italy

Australia

United Kingdom

Hong Kong

United StatesT

ota

l D

eal

Va

lue

of

Cro

ss-B

ord

er D

eals

of

Go

ver

nm

ent-

Co

ntr

oll

ed

Acq

uir

ers

(in

Co

nst

an

t 2

00

0 U

.S. $

mil

lio

ns)

By

La

rges

t T

arg

et

Co

un

try

an

d A

cqu

irer

Reg

ion

s

Developed Americas

Latin America

Developed Asia-Pacific

Emerging Asia-Pacific

Europe Developed

EMEA

51

INTERNET APPENDIX

State Capitalism’s Global Reach:

Evidence from Foreign Acquisitions by State-owned Companies

G. Andrew Karolyi and Rose C. Liao

February 2013

52

Internet Appendix Table of Key Hypotheses for Cross-Border Acquisition Activity

Hypotheses

Explanations

Valuation Valuation can affect acquisition activity through two channels: cheap assets (Froot and Stein, 1991) and cheap capital (Shleifer and

Vishny, 2003). Regardless of which channel is at work, we would observe that acquirers are more likely from countries with

appreciating currencies and whose stock markets have increased in value. Erel, Liao, and Weisbach (2010) find differences in real

stock market returns and in real exchange rate changes explain much of the level of cross-border merger activity between country

pairs and argue that it largely stems from changing underlying economic conditions. We include both valuation factors.

Corporate Governance Rossi and Volpin (2004), Starks and Wei (2012), Antrás, Desai, and Foley (2009), Bris and Cabolis (2008), Bris, Brisley, and Cabolis

(2008), and Ellis, Moeller, Schlingemann, and Stulz (2011) find that cross-border mergers and acquisition activity between two

countries increases the greater the difference in the quality of investor protections and accounting standards between the acquirer’s

and target’s countries. We use the anti-self-dealing index of Djankov, La Porta, Lopez-de-Silanes, and Shleifer (2008) to proxy for

quality of investor protection and the score of disclosure quality from the Center for International Financial Analysis and Research as

in La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998) to proxy for accounting standards.

Geographic Proximity The empirical literature on trade, FDI flows, and cross-border mergers and acquisitions emphasizes the important role that geography

plays (among others, see the gravity models of Anderson, 1979; Portes and Rey, 2005; Anderson and van Wincoop, 2003;

Coeurdacier, De Santis, and Aviat, 2009; Rossi and Volpin, 2004; and, Erel, Liao, and Weisbach, 2010). The arguments are based on

the role of transactions costs, tariffs, and barriers that are linked to bilateral geographical distance, although they can be linked to

commonness of culture, language, ethnicity, and religion (Stulz and Williamson, 2003; and, Ahern, Daminelli, and Fracassi, 2011).

Wealth and

Macroeconomic

Conditions

Rossi and Volpin (2009), Chari, Chen, and Dominguez (2012), and Erel, Liao, and Weisbach (2010) show that developed countries’

firms are, in fact, more likely to acquire less developed countries’ firms. We include differences in the logarithm of GDP per capita,

as a measure of the country’s wealth, and in real GDP growth as a proxy for the change in macroeconomic conditions.

European Integration Couerdacier, De Santis, and Aviat (2009) show that the European integration process – through joining the European Union (EU)

and/or the European Monetary Union (EMU) – led to a doubling of merger and acquisition activity towards their members and away

from the rest of the world, so we include a dummy variable for those country pairs that involve both as members of the EU.

Risk Diversification The lower the returns correlation between countries, the more important the risk diversification motive for the acquirer. We include a

measure of correlation of returns between the national indexes of the two countries to evaluate the motive of risk diversification.

Industrial

Diversification

Government-led activity would be higher between countries with more dissimilar industrial structures (Butt, Shivdasani, Standavad,

and Wyman, 2008; Bremmer, 2010). Dyck and Morse (2011) find that SWFs target specific industries in foreign countries to attempt

to influence their country’s long-term industrial mix.

Political Democracy The political system in a country can influence many aspects of foreign policies on trade, capital flows, and foreign direct investment

(Stulz, 2005) and indeed may influence the acquisitions of politically-connected, government-controlled acquirers. Indeed, Bernstein,

Lerner, and Schoar (2009) shows that SWFs with greater involvement by politicians in fund management are more likely to invest in

their home country and in high P/E industries. Sojli and Tham (2011) find that only SWF investments in U.S. targets that witness an

increase in government-related contracts following their acquisition experience significant valuation increases.

53

Details on construction of our variables are in Appendix Table IA1 and summary statistics are in Table IA2.37

37 In supplementary tests, we also explored the impact of restrictions on FDI. For a subset of OECD countries, Golub (2003) devised a scoring system for the overall restrictiveness

on inward FDI for each country, based on foreign ownership limits on equity, mandatory screening, licensing and approval, nationality restrictions on board members, and input

and operational restrictions. Government-controlled acquirers are, after all, more likely to be impacted by FDI restrictions because of political concerns related to threats to

national security and excessive political influence (Graham and Krugman, 1995).

Internet Appendix Table of Key Hypotheses for Cross-Border Acquisition Activity (continued)

Hypotheses

Explanations

Total Reserve Government-led activity may be better enabled and financed in acquirer countries with higher accumulated total foreign currency

reserves. Butt, Shivdasani, Standavad, and Wyman (2008, p. 78) point out that high foreign currency reserves may be well in excess

of the immediate needs of governments to offer protection against sudden capital outflows and represent a signal of their interest in

promotion of domestic economic development, long-term macroeconomic objectives or even foreign policy objectives.

Government Domestic

Acquisition Activity

The more acquisitive government enterprises are in a country, the greater their involvement in the economy and capital markets, and

the more likely they would influence government-controlled cross-border acquisition activity. Active domestic acquisition activity by

government-controlled agencies could reflect political economy motives. Bernstein, Lerner, and Schoar (2009) show that the SWFs

where politicians are involved in the management of the fund are disproportionately more likely to invest in domestic firms than

externally-managed SWFs, a finding which they interpret as a distortion in capital allocation from short-term political considerations.

54

Table IA1. Variable Definitions.

Variable Definition

Government-controlled acquirer

deal ratio between target country j

and acquirer country i (AGijt)

The ratio of the number of deals in which the target is from country j and the acquirer is a government-controlled

corporation from country i (where i ≠ j) relative to the total number of cross-border deals with government-controlled

acquirers either targeting country i or from acquiring country j. The ratio is computed in some instances using the total

value of the deals (in Constant 2000 U.S. dollars) instead of the number of deals, but this obtains only for the subset of

deals for which deal value is reported. (Source: SDC Mergers and Corporate Transactions database).

Government-controlled acquirer

deal ratio between target country j

and acquirer country i (ACijt)

The ratio of the number of deals in which the target is from country j and the acquirer is a corporation from country i

(where i ≠ j) relative to the total number of cross-border deals with corporate acquirers either targeting country i or from

acquiring country j. (Source: SDC Mergers and Corporate Transactions database).

Government-controlled acquirer

deal ratio between target country j

and acquirer country i (AFijt)

The ratio of the number of deals in which the target is from country j and the acquirer is a government-controlled funds

from country i (where i ≠ j) relative to the total number of cross-border deals with government-controlled funds either

targeting country i or from acquiring country j. (Source: SDC Mergers and Corporate Transactions database).

Annual Exchange Rate Return

Levels and Differences

Levels of and differences between the annual real bilateral U.S. dollar exchange rate return of the acquirer and target

country. We use national exchange rates from Datastream from WM/Reuters (WMR). WMR quotes are based on 4:00pm

London (Greenwich Mean Time).We obtain National Exchange Rates for the U.K. Pound Sterling and manually convert

these currency quotes to get the quotes for the U.S. dollar. These indices are then deflated using the 2000 constant dollar

Consumer Price Index (CPI) in each country to calculate real exchange rate returns (in U.S. dollars).

Annual Real Stock Market Return

Levels and Differences

Levels of and differences between the annual local real stock market return of the acquirer and target country. We obtain

Datastream total return indices in local currency for each country (Datastream code: RI) and deflate these indices using the

2000 Consumer Price Index (CPI) in each country to calculate real stock returns. (Source: Datastream)

Log GDP per capita Levels and

Differences

Levels of and differences between target and acquirer firm’s country of domicile in the average logarithm of Gross

Domestic Product (GDP, in U.S. dollars) divided by the population (Source: World Bank Development Indicators)

GDP Growth Levels and

Differences

Levels of and differences between target and acquirer firm’s country of domicile in the annual real growth rate of the Gross

Domestic Product (Source: World Bank Development Indicators)

Geographic Proximity The negative of the great circle distance between the capitals of countries i and j. We obtain latitude and longitude of

capital cities of each country. We then apply the standard formula: 3963.0 × arccos [sin(lat1) × sin(lat2) + cos (lat1) × cos

(lat2) × cos (lon2 - lon1)], where lon and lat are the longitudes and latitudes of the acquirer (“1” suffix) and the target

country (“2” suffix) locations, respectively. (Source: http://www.mapsofworld.com/utilities/world-latitude-longitude.htm)

Market Correlation The correlation coefficient using monthly country-level stock index returns denominated in US dollars (Datastream code:

RI) computed annually. (Source: Datastream)

55

Table IA1. Variable Definitions. (continued)

Variable Definition

Anti-Self-Dealing Index Levels

and Differences

Levels of and differences between acquirer and target firm’s country of domicile in the Anti-Self-Dealing Index, a survey-

based measure of legal protection of minority shareholders against expropriation by corporate insiders. (Source: Djankov,

La Porta, Lopez-de-Silanes, and Shleifer (2008)).

Accounting Standards Index

Levels and Differences

Levels of and differences between acquirer and target firm’s country of domicile in the index created by the Center for

International Financial Analysis and Research to rate the quality of 1990 annual reports on their disclosure of accounting

information (Source: La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1997, 1998)).

European Union Dummy Equals 1 if both target and acquirer firm’s country of domicile belongs to the European Union, and equals zero, otherwise.

(Source: The World Factbook)

Industry Dissimilarity The difference in the industrial composition between the acquirer and target country of domicile is computed as the square

root of an equally-weighted sum of squared differences in the relative weights of each industry in each country in each year.

The industry weights are measured as the fraction of the total market capitalization comprised by the publicly-listed stocks

in that industry in that country in that year. An industry is defined as one of 48 different categories according to Fama and

French (1997) which are governed mostly by the first two or three digits of a Standard Industrial Classification (SIC) code.

(Source: Datastream and Professor Kenneth French’s website at Dartmouth University,

http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/index.html) PolityIV Democracy Levels and

Differences

Levels of and differences between acquirer and target country of domicile in the measure of regime democracy and/or

autocracy, ranging from -10 (high autocracy) and +10 (high democracy). The PolityIV Project is led by Monty Marshall

(George Mason University) and Keith Jaggers (Colorado State) and was founded originally by Ted Robert Gurr (University

of Maryland). We use the Polity IV Data Series Version 2009 with annual time-series for up to 163 countries from 1800

through 2009. (Source: http://www.systemicpeace.org/polity/polity4.htm).

Total Reserve % GDP Levels and

Differences

Level of and differences between acquirer and target country of domicile in the measure of total reserves as percentage of

GDP (includes gold, defined in current U.S. dollars) (Source: World Bank Development Indicators)

Government Domestic

Acquisition Activity (Acquirer)

For an acquirer, a measure of a government’s presence in the domestic economy calculated as the ratio of the number of

domestic acquisitions led by government-controlled corporations and funds to that of acquisitions by all domestic

corporations in the previous year five years including the year under consideration. We use all domestic acquisitions

(excluding spinoff, LBO, recap, self-tender, exchange offers, repurchases, similar to what we did for cross-border sample)

including both minority block purchase and majority acquisitions between 1986 and 2009. (Source: SDC Mergers and

Corporate Transactions database).

Government Acquirer Dummy Equals 1 if acquirer's ultimate parent is defined as a government and 0 otherwise. We exclude all transactions if the acquirer

is a state fund as defined below (see “State-Controlled Fund Dummy”). (Source: SDC Mergers and Corporate Transactions

database).

56

Table IA1. Variable Definitions. (continued)

Variable Definition

Related Industry Dummy Equals 1 if the target firm’s Standard Industrial Classification (SIC) code equals that of the acquirer at three-digit level (Source: SDC

Mergers and Corporate Transactions database).

Withdrawn Deals Dummy Equals 1 if the deal is announced but not completed, and 0, otherwise. (Source: SDC Mergers and Corporate Transactions database).

Zero-dividend Dummy Equals 1 if the firm pays no dividends, and 0, otherwise. (Source: Worldscope item WC04551)

Percent of Shares Acquired The percentage of shares of the target ultimately owned by the acquirer (Source: SDC Mergers and Corporate Transactions database).

All Cash Payment Dummy Equals 1 if the deal is 100% paid in cash and 0 otherwise; when the payment is unknown, it is set to missing (Source: SDC Mergers and

Corporate Transactions database).

High Closely-Held Shares Dummy Equals 1 if the firm's insider ownership is in the upper quartile of all Worldscope firms (WC08021) in that calendar year.

Total Assets (log) Book value of total assets in millions of constant 2000 US dollars (Source: Worldscope item WC07230)

Return on Assets (Net Income before Preferred Dividends + ((Interest Expense on Debt - Interest Capitalized ) * (1-Tax Rate))) / Average of Last Year's

and Current Year's (Total Capital + Last Year's Short Term Debt & Current Portion of Long Term Debt) * 100 (Worldscope item

WC08376)

Market-to-Book (Book value of total assets (Worldscope item WC02999)-book value of equity (WC05491*WC05301)+ market value of equity

(WC08001))/book value of assets (WC02999)

Long-term Debt/Assets Ratio of long-term debt to book value of assets (Worldscope items WC03251/WC02999)

Sales Growth One-year local country CPI inflation-adjusted sales growth (Worldscope item WC01001)

State-Controlled Funds Dummy Equals 1 if the firm is targeted by a sovereign wealth fund (SWF) or state-controlled funds. A SWF is identified as a financial acquirer in

Securities Data Corporation under ACQUIROR_TYPE data item and matched by name (using SDC data item AN) to a list of SWFs at

the SWF Institute website, http://www.swfinstitute.org/funds.php). Government-controlled or public investment funds are defined by a

primary Standard Industrial Classification (SIC) code in the 999A-G, 9000 range and/or government acquirers (defined for the ultimate

parent) with any of the following SIC codes related to investment offices, pension, health and welfare funds, trusts, or holding

companies: 6019, 6371, 6722, 6726, 6798, 6799. (Source: SDC Mergers and Corporate Transactions database, SWF Institute).

Minority Block Acquisition Dummy Equals 1 if the deal is a minority block purchase (less than 50% of target firm’s shares) and 0 if the deal is majority control acquisition

(Source: SDC Mergers and Corporate Transactions database)

57

Table IA1. Variable Definitions. (continued)

Variable Definition

CMARs (-10, +10),

CMARs (-5, +5),

CMARs (-1, +1)

Market-adjusted cumulative abnormal returns for the (-10, +10) interval are cumulated from 10 days before to up to 10 days

after when centered on the date of announcement of an acquisition. (Source: SDC Mergers and Corporate Transactions

database and Datastream). Similarly defined for (-5,+5) and (-1,+1) intervals centered on the announcement date.

Premium (-4w, 0),

Premium (-1w, 0),

Premium (-1d, 0)

Bid premium of the bid price measured relative to the closing stock price of the target four weeks prior to the

announcement date, expressed as a percentage (defined using SDC codes (HOSTPR – HOSTC4WK) / HOSTC4WK) *

100). Similar for premium relative to closing stock price of target one week prior (-1w, 0) and one day prior (-1d, 0).

(Source: SDC Mergers and Corporate Transactions database)

BHAR (+1, 12m),

BHAR (+1, 24m),

BHAR (+1, 36m)

Market-adjusted cumulative buy-and-hold abnormal returns for the (+1m, +12m) interval are cumulated from one month to

twelve months after the acquisition when centered on the date of announcement of an acquisition. (Source: SDC Mergers

and Corporate Transactions database and Datastream). Similarly defined for (+1m, +24m) and (+1m, +36m) intervals

centered on the announcement date.

Calendar Time Portfolio Returns Equally weighted market-adjusted monthly return of the portfolio containing firms targeted by government-controlled

acquirer (corporate acquirers or government-controlled funds) within the past 12 months (and 24 months). For those

calendar months when no firms were targeted in the past 12 (and 24 months), we assume zero for market adjusted returns

and market returns for raw returns.

WML (Winner Minus Loser) The average return on the two winner portfolios for 23 countries minus the average return on the two loser portfolios.

(Source: Professor Kenneth French’s website, http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html)

SMB (Small Minus Big) The average return on the three small portfolios for 23 countries minus the average return on the three big portfolios.

(Source: Professor Kenneth French’s website, http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html)

HML (High Minus Low) The average return on the two value portfolios for 23 countries minus the average return on the two growth portfolios.

(Source: Professor Kenneth French’s website, http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html)

Rm-Rf The excess return on the market, which is the value-weighted return on a portfolio of all stocks (from 23 countries) minus

the one-month Treasury bill rate (from Ibbotson Associates). (Source: Professor Kenneth French’s website,

http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html)

58

Table IA2. Summary Statistics on Variables.

Variable Obs Mean StdDev Min Max Variable Obs Mean StdDev Min Max

Table 3 (Variables of Interest) Table 3 (Explanatory Variables)

By Acquirer Annual Exchange Rate Return Difference 13,624 0.02 0.16 -1.10 1.10

Government-controlled acquirer deal ratio 11,713 0.05 0.16 0.00 1.00 Annual Real Stock Market Return Differences 13,121 0.00 0.26 -1.45 1.52

Corporate acquirer deal ratio 15,984 0.07 0.13 0.00 1.00

Log GDP per capita Differences 14,564 0.47 1.54 -4.58 4.69

Government-controlled funds deal ratio 4,807 0.04 0.18 0.00 1.00

GDP Growth Differences 14,266 0.00 0.04 -0.26 0.35

Government-controlled acquirer deal ratio (Deal Value) 9,167 0.05 0.19 0.00 1.00

Geographic Proximity (’000) 15,433 -3.54 2.91 -12.35 -0.03

Government-controlled acquirer deal ratio (Minority) 9,988 0.05 0.17 0.00 1.00

Market Correlation 13,812 0.39 0.18 -0.18 0.78

Anti-Self-Dealing Index Differences 11,326 0.00 0.33 -0.86 0.86

By Target

Accounting Standards Differences 10,507 0.03 0.13 -0.54 0.59

Government-controlled acquirer deal ratio 13,403 0.06 0.18 0.00 1.00

European Union Dummy 15,916 0.10 0.30 0.00 1.00

Corporate acquirer deal ratio 16,013 0.07 0.12 0.00 1.00

Industry Dissimilarity 12,706 0.05 0.02 0.01 0.16

Government-controlled funds deal ratio 6,147 0.05 0.19 0.00 1.00

PolityIV Democracy Differences 12,904 0.51 4.89 -20.00 20.00

Government-controlled acquirer deal ratio (Deal Value) 10,007 0.05 0.20 0.00 1.00

Total Reserve % GDP 12,885 -0.01 0.21 -1.02 1.02

Government-controlled acquirer deal ratio (Minority) 11,774 0.05 0.19 0.00 1.00

Govt Domestic Acquisition Activity (acquirer) 14,683 0.05 0.09 -0.05 1.34

59

Table IA2. Summary Statistics on Variables. (continued)

Variable Model Obs Mean StdDev Min Max

Table 4 Explanatory Variables for Model (1) and (5)

Related Industry Dummy (1) 6,327 0.41 0.49 0.00 1.00

Zero-dividend dummy (1) 6,327 0.38 0.49 0.00 1.00

High Closely-Held Shares Dummy (1) 6,327 0.19 0.40 0.00 1.00

Total Assets (log) (1) 6,327 6.17 2.20 -2.01 11.39

Market-to-Book (1) 6,327 1.67 1.57 0.39 25.88

Return on Assets (1) 6,327 0.02 0.30 -3.14 0.67

Long-term Debt/Assets (1) 6,327 0.15 0.17 0.00 0.89

Sales Growth (1) 6,327 0.28 0.92 -0.77 7.46

Withdrawn Deals Dummy (1) 6,327 0.30 0.46 0.00 1.00

All Cash Payment Dummy (1) 6,327 0.34 0.47 0.00 1.00

Related Industry Dummy (5) 508 0.31 0.46 0.00 1.00

Zero-dividend dummy (5) 508 0.37 0.48 0.00 1.00

High Closely-Held Shares Dummy (5) 508 0.23 0.42 0.00 1.00

Total Assets (log) (5) 508 6.81 2.26 0.98 11.39

Market-to-Book (5) 508 1.72 2.03 0.42 25.88

Return on Assets (5) 508 0.03 0.29 -3.14 0.64

Long-term Debt/Assets (5) 508 0.16 0.17 0.00 0.89

Sales Growth (5) 508 0.40 1.20 -0.77 7.46

Withdrawn Deals Dummy (5) 508 0.39 0.49 0.00 1.00

All Cash Payment Dummy (5) 508 0.26 0.44 0.00 1.00

60

Table IA2. Summary Statistics on Variables. (continued)

Variable Obs Mean Std. Dev. Min Max

Table 6 Variable of Interest and Explanatory Variables for All Deals CMAR (-10, +10) 4,241 0.13 0.44 -0.91 10.67 CMAR (-5, +5) 4,241 0.11 0.28 -0.83 5.26 CMAR (-1, +1) 4,227 0.08 0.19 -0.61 2.08 Government-controlled Acquirer dummy 4,241 0.04 0.18 0.00 1.00

Corporate acquirer dummy 4,241 0.95 0.23 0.00 1.00 Government-controlled fund dummy 4,241 0.02 0.13 0.00 1.00 Zero-dividend dummy 4,241 0.43 0.49 0.00 1.00 High Closely-Held Shares Dummy 4,241 0.18 0.39 0.00 1.00 Total Assets (log) 4,241 5.56 1.89 -2.01 11.39 Market-to-Book 4,241 1.75 1.61 0.39 25.88

Long-term Debt/Assets 4,241 0.13 0.15 0.00 0.89 Sales Growth 4,241 0.28 0.92 -0.77 7.46

Table 7 Variable of Interest and Explanatory Variables Corporate Acquirers Calendar Time Portfolio Returns (1-year) 275 0.01 0.04 -0.30 0.12 Government-controlled Acquirers Calendar Time Portfolio Returns (1-year) 270 0.01 0.07 -0.28 0.33

Government-controlled Funds Calendar Time Portfolio Returns (1-year) 237 0.01 0.10 -0.30 0.43 Corporate Acquirers Calendar Time Portfolio Returns (2-year) 275 0.01 0.04 -0.27 0.11 Government-controlled Acquirers Calendar Time Portfolio Returns (2-year) 271 0.01 0.06 -0.29 0.33 Government-controlled Funds Calendar Time Portfolio Returns (2-year) 265 0.01 0.08 -0.25 0.33 Rm-Rf 231 0.00 0.04 -0.19 0.11

Small Minus Big 231 0.00 0.02 -0.10 0.10 High Minus Low 231 0.00 0.02 -0.10 0.11 Winner Minus Loser 231 -0.01 0.14 -1.00 0.18