WHAT IS DIFFERENT ABOUT GOVERNMENTCONTROLLED

1
WHAT IS DIFFERENT ABOUT GOVERNMENT-CONTROLLED
ACQUIRERS IN CROSS-BORDER ACQUISITIONS? G.
Andrew Karolyi (Cornell) and Rose C. Liao
(Rutgers)
2
What is this paper about?

• We examine the motives for and consequences of
  5,317 failed and completed cross-border
  acquisitions constituting 619 billion of
  activity over 1990-2008 involving
  government-controlled acquirers
• We benchmark the activity at the country level
  and at the deal level by cross-border
  acquisitions led by corporate acquirers
• We seek answers to the following questions
• Are country factors that drive cross-border
  acquisitions by government-controlled acquirers
  different?
• Are attributes of target firms of cross-border
  acquisitions by government-controlled acquirers
  different?
• Are target firms market reactions to
  announcements of cross-border deals by
  government-controlled acquirers different?

3
Cross-border activity of government acquirers
4
Some Recent Noteworthy Deals
5
Foreign Investment National Security Act of 2007
6
Recent Legislative Actions in Other Countries
7
Theoretical Motivation

• Large literature rationalizes how public
  enterprises are inefficient with excess
  employment wages and goods production closer to
  needs of self-interested politicians
  bureaucrats than consumers
• Shleifer Vishny (1994) model of bargaining
  between politicians and managers via subsidies
  and bribes
• Tirole (1994) dissonant objectives in internal
  organization of government due to information
  problems, incentive contracts
• Atkinson Stiglitz (1980) maximizing social
  welfare, curing market failures due to monopoly
  power, negative externalities
• Our contribution
• Target firms become at least partially
  state-owned in cross-border deals are useful
  setting to evaluate with natural benchmark
• This benchmark confers experimental design
  advantage to enhance power of our tests with
  specific testable alternative hypotheses

8
New Studies on Sovereign Wealth Funds (SWFs)

• We contribute to an emerging literature on SWF
  investments (11 new working papers
  and one book, including Kotter Lel, 2008
  Fotak, Bertolotti, Megginson, Miracky
  Fernandes Chhaochharia Laeven Dewenter, Han
  Malatesta Knill, Lee Mauck Dyck Morse
  Bernstein, Lerner Schoar, all 2009)
• SWFs are public investment agencies which manage
  governments foreign-currency reserves
  accumulated via goods/commodity exports
• Recent papers focus on SWF investments or
  holdings, comparing across funds by quality of
  governance or level of transparency, but with no
  clearly-defined overall benchmarks
• Exception Chhaochharia Laeven
• We widen the lens studying not only SWF-led
  acquisitions but also other government-controlled
  acquirers, including agencies majority controlled
  by SWFs and benchmark all by corporate acquirers

9
Our Findings

• There is large cross-country variation in
  cross-border activity led by government-controlled
  acquirers by home of acquirers (China, 115b
  France, 94b Singapore, 69b) and by countries
  of targets (US, 124b UK, 91b HK, 66b
  Australia, 27b)
• Government acquisitions are relatively more
  intense for geographically-close countries than
  corporate deals, less sensitive to differences in
  economic development, legal institutions,
  accounting standards stringency of FDI
  restrictions
• Overall, explanatory power of differences is low
• Government-controlled acquirers more likely to
  target larger firms with more growth options and
  more financial constraints and market reactions
  (CMARs) to acquisition announcements are similar
  (8)
• Pseudo-R2 of logits of target firm choices are
  very low
• CMARs statistically indistinguishable, especially
  if controls included
• SWFs targets are larger, less constrained deals
  less likely to fail CMARs are positive (1.5),
  but smaller than other govt-led deals

10
The Roadmap

• The Questions Asked in this Paper
• Motivation
• Summary
• Data
• Sample of Firms
• Summary Statistics (Table 2, Figure 2)
• Results
• Determinants of Cross-Border Acquisition Activity
  (Table 3)
• Logit regressions of Target Firm Choices (Tables
  4 5)
• CMARs and cross-sectional tests (Tables 6 7)
• Concluding Remarks

11
Data and Sample

• Thomson Reuters Security Data Corp (SDC)
  Platinum Mergers and Corporate Transactions data
  yields 155,696 deals between 1990 2008 with
  total value of 10.6 tr (in Constant 2000 US )
• Deal characteristics fractional stake in target
  gt5 (majority if gt50), announcement date,
  failed/withdrawn, targets name, status (sub, JV,
  private or govt-owned), intermediate ultimate
  parent name/status, deal value (if disclosed),
  SIC code, terms of payment, open market purchase
  or private negotiation
• Excluded buyouts, spinoffs, recaps, self-tender
  offers, exchange offers, repurchases and
  privatizations, acquirers from UK, Dutch
  tax-haven territories (Bahamas, Cayman, Guernsey,
  Dutch Antilles)
• Government-controlled acquirers
• SDCs AUPPUB status is government, requires 50
  control
• Double-check by hand using annual reports,
  filings, news stories, etc
• SWF identified by name from www.swfinstitute.org
  ownership status

12
Summary Statistics (Table 1)
13
Largest Government-Led Deals in the Sample

• Top Govt Acquirers Temasek (Singapore, 40b, 167
  deals), China Unicom Ltd (22b, 2), Japan Tobacco
  (21b, 6), EDF (France, 21b, 20), CNOOC (China,
  18b, 1)
• Top SWF Acquirers Temasek, GIC (Singapore, 23b,
  56 deals), SABIC (Saudi, 12b, 4), ICD (UAE,
  11b, 15), China Development Bank (7b, 4)

14
The Roadmap

• The Questions Asked in this Paper
• Motivation
• Summary
• Data
• Sample of Firms
• Summary Statistics (Table 2, Figure 2)
• Results
• Determinants of Cross-Border Acquisition Activity
  (Table 3)
• Logit regressions of Target Firm Choices (Tables
  4 5)
• CMARs and cross-sectional tests (Tables 6 7)
• Concluding Remarks

15
Determinants of Cross-Border Acquisition Activity

• Does deal activity led by government-controlled
  acquirers emanate from some countries more than
  from others? Are they more likely to pursue
  targets in some countries more than others?
• We compute cross-border ratios of government-led
  deal activity (by counts and by cumulative
  constant dollar value) in two ways
• Between acquirer country i target j relative to
  all activity from country i
• Between acquirer country i target j relative to
  all activity to country j
• We repeat the same computations for corporate-led
  acquisitions and, separately for both types, for
  minority and majority deals
• Finally, we evaluate relative ratios of
  government-led to corporate-led acquisition
  activity between country pairs i and j

16
Cross-Border Acquisition Activity Ratios (Table
2)
17
Cross-Border Acquisitions by Origin of Acquirer
(Fig 2a)
18
Cross-Border Acquisitions By Major Targets (Fig
2b)
19
Determinants of Cross-Border Acquisition Activity

• Our central null hypothesis is that cross-border
  acquisition activity between country pairs
  involving government-controlled acquirers is no
  different than that for corporate acquirers
• Theory/empirical evidence guides us to specific,
  testable alternatives
• Valuation differences between target acquirer
  firms (Mkt returns, FX)
• Wealth shocks (Froot Stein, 1991 Baker et al,
  2008 Erel, Liao Weisbach, 2009)
• Sentiment misvaluation (Shleifer Vishny,
  1993 Moeller et al., 2005 Dong et al, 2006)
• SWF study by Bernstein et al (2009) show relative
  P/E matters
• Corporate Governance (Legal (ASDI), Accounting
  Standards, PolityIV)
• Rossi Volpin (2004), Antras et al (2007),
  Starks Wei (2007) Bris Cabolis (2008), Bris
  et al (2008) for cross-border corporate mergers
• SWF study by Bernstein et al (2009) emphasize
  political affiliations of SWF management
• Geographic proximity (great circle distance)
• Gravity models in trade and FDI flows
  (Anderson, 1979 Portes Rey 2005)
• Culture, language, ethnicity, religion (Kang
  Kim, 2009 for SWFs, Chhaochharia Laeven)
• Control variables
• GDP per capita, real GDP growth differences, Tax
  haven target dummy, market correlation,
  EU-specific deal, FDI restrictiveness (Golub,
  2003 for OECD countries)

20
Determinants of Acquisition Activity (Table 3)
21
What Factors Drive Government Acquirer Deals?

• Country-level analysis ignores many target firm
  characteristics deal-specific factors that can
  affect the decision to acquire a target
• But deal-level analysis focuses on public target
  firms for which sample drops by 88 among
  corporate deals (150,379 to 17,845) 50,
  govt-led deals (5,317 to 2,261), even more
  depending on Worldscope
• Logit regressions dep variable 1 for govt-led
  deals and 0, for corp deals (country, year fixed
  effects), all reported in marginal effects
• We evaluate market timing, governance motives
  additional ones
• Product market relationship contracting (same
  industry, high RD)
• Partial integration helps with information
  environments with incomplete contracting
• Allen Phillips (2000), Fee, Hadlock Thomas
  (2006), Liao (2009)
• Financial constraints (Zero-dividend, Whited-Wu,
  Hadlock-Pierce dummy)
• Allen Phillips (2000), Fee, Hadlock Thomas
  (2006), Liao (2009)
• Firm-specific control variables
• B/M, ROA, Sales growth, Log assets, Debt/assets,
  Worldscopes Closely-held shares, All-cash deal
  (majority deals only), Fraction of target shares
  (majority deals only)

22
Logit Regressions of Government Led Deals (Table
4)
Economics? A non-dividend- paying target firm
(21 of sample) has 1.5 greater likelihood of
a govt-led deal (4.44 average) A s increase in
12-mo. Mkt. Ret. (31) is associated with 0.81
lower likelihood of a govt-led deal (4.44
average)
23
Are SWF-led Acquisitions Different? (Table 5)
24
Target Firms Market Reactions to Government
Deals?

• How do shareholders of target firms react to
  announcement of cross-border deal led by
  government-controlled acquirer? Differently than
  corporate-led acquisition?
• Cumulative market-adjusted returns (CMARs) using
  local VW market indexes over three windows
  (21-day, 11-day 3-day) for both
  government-controlled corporate acquirers,
  minority majority acquisition deals separately
• Benchmark results in other empirical studies
• Majority-control cross-border corporate
  acquisitions
• 5-day CARs around 14 (Bris Cabolis, 2008),
  another 11 for days 2 to 10
• 11-day CARs of 28.24 (Starks Wei, 2004)
• Minority stake (lt50) corporate acquisitions
• 6.9 US-only sample (Allen Phillips, 2000)
• 9 for out-of-state deals in US (Kang Kim,
  2008)
• 8.7 23-day CMARs for domestic deals globally,
  7.4 in cross-border (Liao, 2009)

25
Market Reactions to Government Deals (Table 6)
26
Cross-sectional Regressions of CMARs (Table 7)
27
Target Firms Market Reactions to SWF Acquirers?

• How do shareholders of target firms react to
  announcement of cross-border deal led by
  sovereign wealth fund? Differently than if by
  other government-controlled acquirers?
• We find CMARs of SWFs are much lower (1.41
  versus 8.05) and the differences are
  statistically significant (Table 6, Panel B) and
  also robust to controlling for various firm-,
  deal-specific and country-level variables (Table
  7, Panel B)
• Benchmark results show positive reactions in
  other SWF studies
• 3-day CARs of 0.46 (Fotak et al, 2009, 212
  deals)
• 3-day CARs of 2.15 (Kotter Lel, 2008, 163
  deals)
• 5-day CARs of 0.89 (Chhaochharia Laeven, 2009,
  89 deals)

28
The Roadmap

• The Questions Asked in this Paper
• Motivation
• Summary
• Data
• Sample of Firms
• Summary Statistics (Table 2, Figure 2)
• Results
• Determinants of Cross-Border Acquisition Activity
  (Table 3)
• Logit regressions of Target Firm Choices (Tables
  4 5)
• CMARs and cross-sectional tests (Tables 6 7)
• Concluding Remarks

29
Concluding Remarks

• Cross-border acquisition activity involving
  government-controlled acquirers over the past 20
  years is substantial and growing and our study is
  the first assess whether the motives for and
  consequences of such deals are different from
  those of corporate acquirers
• While there is significant cross-country
  variation in government-led acquisition activity,
  it is difficult to explain much of it and it is
  also hard to distinguish differences in target
  firm characteristics and market reactions to
  acquisition announcements at the deal level
• SWF-led government deals do pursue different
  kinds of targets and the market reactions to
  their announcements are significantly lower
• There are important implications for recent
  policies adopted in the US and other countries to
  restrain foreign acquisitions, in general, and by
  government entities, in particular
• There are many avenues for further study
  (longer-run consequences, exploring other
  alternative hypotheses, deals by types of SWFs,
  possible positive/negative externalities)