SOVEREIGN ACQUIRERS A NEW FORCE IN GLOBAL MARKETS

SOVEREIGN ACQUIRERS: A NEW FORCE IN GLOBAL MARKETS G. Andrew Karolyi Professor of Finance and Global Business Alumni Chair in Asset Management Keynote Address NTU International Conference on Finance December 10, 2010 Taipei, Taiwan

Sovereign acquirers: A new force in global markets • Sovereign acquirers – which include Sovereign Wealth Funds (SWFs) and government-controlled corporations – are attracting attention in global financial markets as they pursue overseas investments because: – Diversity of objectives and investment philosophies – Potential political and economic ambitions – Different degrees of transparency of objectives & ambitions – Size ($3. 8 tr in SWF assets, another $1 tr in acquisition deals) • SWFs as government-created investment vehicles funded by commodity export revenues or transfers from foreign-exchange reserves get major headlines • But many large, government-controlled corporate acquirers increasingly active in cross-border deals 1

Global Current Account Imbalances Drive Growth Sovereign Acquirer Assets Deficits Surpluses Japan $126 b US $465 b Eurozone $110 b China $364 b Russia $41 b Asia ex Japan, China $124 b Middle East $134 b Source: The Economist, January 2, 2010 2

Who are the important players? 3

Rise of Sovereign Acquirers Affect Several External Market Participants… Global Corporations Global Asset Managers + Potential Source of Funding + Potential Strategic Co-investors - Support for National/Political Goals + Large Pool of Investment Capital + Potential Co-investors - Competition for Assets Source: Citigroup, 2008, Journal of Applied Corporate Finance, Volume 19(1), p. 74. Rise of Sovereign Acquirers Overseas Governments + Source of Global Liquidity - Independence of Strategic Assets - Concerns about Transparency - Lack of Market Access Reciprocity 4

5

Some recent controversial deals 6

Growing media focus & concern 7

Foreign Investment & National Security Act of 2007 8

Recent Legislative Actions in Other Countries… 9

A New Study • G. Andrew Karolyi and Rose Liao, 2010, “What is Different about Government-Controlled Acquisitions in Cross-Border Acquisitions? ” Cornell University working paper (http: //ssrn. com/abstract=1571560) • We examine the motives for and consequences of 4, 026 failed and completed cross-border acquisitions constituting $434 billion of activity over 1990 -2008 involving government-controlled acquirers • Benchmark against 123, 027 corporate-led acquisitions worth $8. 4 tr • 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 firm’s market reactions to announcements of cross-border deals by government-controlled acquirers different? – Is longer-run financial & operating performance different? 10

Cross-border activity of government acquirers 11

Our Findings • There is large cross-country variation in cross-border activity led by government-controlled acquirers by home of acquirers (China, France, Singapore) and by countries of targets (US, HK, UK) – Government acquisitions are relatively more intense for geographicallyclose, industrially less-similar countries, acquirers more likely to come from countries with larger current account surpluses, foreign currency reserves… – BUT…overall explanatory power of differences is low • Government-controlled acquirers are more likely to target larger, more financially-constrained firms and market reactions (CMARs) to acquisition announcements are somewhat lower… – BUT…explanatory power of different target firm choices are very low – BUT…market reactions statistically indistinguishable after controls • Longer-run consequences are not significantly different • Conclusion? Concerns about objectives of government-controlled acquirers may be misplaced 12

The Roadmap • • Sovereign Acquirers: Who & what? Motivation Summary Data – Sample of Firms – Summary Statistics • The Four Experiments – Determinants of Cross-Border Acquisition Activity – Logit regressions of Target Firm Choices – Market reactions and cross-sectional tests – Newer findings: Long-run financial & operating performance • Concluding Remarks 13

Data and Sample • Thomson Reuter’s Security Data Corp (SDC) Platinum Mergers and Corporate Transactions data yields 127, 053 deals between 1990 & 2008 with total value of $8. 9 trillion (in Constant 2000 US $) • Deal characteristics: fractional stake in target >5% (majority if >50%), announcement date, failed/withdrawn, target’s 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 – SDC’s AUPPUB status is government, requires 50%+ control – Double-check by hand using annual reports, filings, news stories, etc – 733 deals ($153 billion) of SWFs and state-controlled funds held back (cross-check with www. swfinstitute. org) 14

Summary Statistics 15

Largest Government-Led Deals in the Sample • Top Govt Acquirers: China Unicom Ltd ($22 b, 5 deals), EDF (France, $21 b, 41), CNOOC ($16 b, 5), Japan Tobacco ($21 b, 5), Petrochemical Inds (Kuwait, $9 b, 3), Telia AB (Sweden, $9 b, 23), Swisscom ($7 b, 11), Petronas ($6 b, 15) 16

The Roadmap • • Sovereign Acquirers: Who & what? Background Summary Data – Sample of Firms – Summary Statistics • The Four Experiments – Determinants of Cross-Border Acquisition Activity – Logit regressions of Target Firm Choices – Market reactions and cross-sectional tests – Newer findings: Long-run financial & operating performance • Concluding Remarks 17

Experiment #1: 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 yearly cross-border ratios (AGit) of government-led deal activity (by counts and by cumulative 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 (ACit) and, separately for both types, for minority and majority deals • Finally, we evaluate excess ratios of government-led to corporate-led acquisition activity between country pairs i and j, or Excess. AGit = AGit - ACit 18

Cross-Border Target Acquisitions by Origin of Acquirer 19

Cross-Border Acquisitions By Major Target Regions 20

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 • Specific, testable alternatives guide us where to look for differences: – Valuation Motives (Market index returns, foreign currency returns) – Corporate Governance Motives (Legal, Accounting Standards, Polity. IV) – Political Economy Motives (Industry dissimilarity, WB Govt effectiveness, Govt expenditures per GDP, Govt domestic acquisition activity, Current account surplus, Foreign currency reserves per GDP) – Control Variables (Market correlation, GDP per capita, real GDP growth differences, Geographic distance, EU dummy, Tax haven dummy) • What do we find? – Govt-led and corporate cross-border deal activity influenced by exchange & market returns, but strongly positively related to proximity & correlations – BUT… indistinguishably between two types of deals (low R 2 for excess ratios) 21

Experiment #2: What Target Attributes Attract Government Acquirers? • Country-level analysis ignores many target firm characteristics & dealspecific factors that can affect the decision to acquire a target • Deal-level analysis focuses on public target firms for which sample drops by 90% among corporate deals (123, 027 to 12, 669), 70% for govtled deals (4, 026 to 715), even more depending on Worldscope • Logit regressions; dep variable = 1 for govt-led deals and 0, for corp • Same valuation, governance, political economy motives, additional target-specific ones to guide alternative hypotheses: – Product market relationship & contracting costs – Financial constraints – Firm-specific control variables (B/M, ROA, Sales growth, Log assets, Debt/assets, Worldscope’s Closely-held shares) 22

Logit Regressions of Government Led Deals (Table 4) Gov’t deals more likely to fail. . Gov’t acquirers prefer targets with more assets, higher M/B, more financial constraints Big economic effect? A “highly” constrained target firm (21% of sample) has 0. 6% greater likelihood of a govt-led acquisition Political economy variables weak (not shown here) The headline finding? 23

Are SWFs and Other State-Controlled Funds Different? SWFs less likely to fail SWFs, state funds are less likely to pursue targets in developed, more democratic economies that are industrially dissimilar Less theory to guide, but more explanatory power 24

Experiment #3 Target Firm’s Market Reactions to Government Deals? • How do shareholders of target firms react to announcement of crossborder 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 • We also compare market reactions to SWF acquisitions to those of other government-controlled acquirers 25

Market Reactions to Government Deals (Tables 6 & 7) • Table 7 cross-sectional regressions show differences are negligible when financial constraints, ownership and controls added 26

Market Reactions to SWFs & State-Controlled Funds? • How do shareholders of target firms react to announcement of crossborder deal led by SWFs and state-controlled funds? 27

Experiment #4: Is Long-Run Performance of Target Firms Different? • We compare percentage changes from Year -1 to +1 in: • Capital Expenditures-to-Assets (CAPEX) • Sales growth • Profitability (EBIT) • Likelihood of equity capital raising within two years • Long-run 12 - to 24 -month buy-and-hold-returns (BHARs) • Government-controlled acquirers have smaller increases in CAPEX (3. 5% vs 5. 9%), slower sales growth (12. 2% vs 22. 7%), more negative 12 -month BHARs (-3. 5% vs -2. 6%), bigger increase in EBIT (13. 4% vs 3. 7%), higher likelihood of equity capital raising (22. 8% vs 15. 6%) • But differences are statistically insignificant 28

The Roadmap • • Sovereign Acquirers: Who & what? Background Summary Data – Sample of Firms – Summary Statistics • The Four Experiments – Determinants of Cross-Border Acquisition Activity – Logit regressions of Target Firm Choices – Market reactions and cross-sectional tests – Newer findings: Long-run financial & operating performance • Concluding Remarks 29

Concluding Remarks • Cross-border acquisition activity involving government-controlled acquirers over the past 20 years is substantial and growing • Our study is the first assess whether the motives 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 • Conclusion? Concerns about overreaching goals of governmentcontrolled acquirers may be greatly overblown • 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, many avenues for further study 30