To Steal or Not to Steal

1

• To Steal or Not to Steal
• Firm Attributes, Legal Environment and Valuation
• Art Durnev
  E. Han Kim
• (University of Miami Business
  School) (Ross School of Business at the
  University of Michigan)

2
Corporate Governance Issues at Country Level

• Effects of regulation concerning corporate
  governance issues at the country level
• La Porta et al. (1997, 1998, 2002), Rajan and
  Zingales (1998), Demirgüç-Kunt and Maksimovic
  (1998), Kumar, Rajan, ZIngales (1999), Wurgler
  (2000)
• show legal rights of investors matter
• external financing, valuation, capital
  allocation, financial markets development,
  economic development

3
Corporate Governance Issues at Firm Level

• Do all firms in weak legal regimes suffer from
  poor corporate governance and do firms in strong
  legal regimes practice uniformly high-quality
  governance?
• Does the variation in governance practices
  (especially in weaker legal regimes) reflect
  firms adaptation to poor legal environment, as
  in Coase (1960), resulting in some firms having
  higher-quality governance than is required by
  law?
• If so, is there a systematic pattern in which
  firms choose their quality of governance? What
  are the relevant firm attributes and how are they
  related to the observed governance practices?
• Is the quality of governance priced in stock
  markets, and if so, is it economically
  significant for corporate decision makers to take
  notice?

4
Corporate Governance Issues at Firm Level

• Is the quality of governance priced in stock
  markets, and if so, is it economically
  significant for corporate decision makers to take
  notice?
• Corporate governance and performance
• Recent studies based on US data
• Gompers, Ishi, Metrick (2001), anti-takeover
  provisions
• Gillan, Hartzell, Starks (2003), various
  governance mechanisms
• Growing number of international studies
• Black (2000), Russian companies
• Black, Jang, Kim (2002), Korean companies
• Doidge, Karolyi, Stulz (2003), firms with ADRs
• Doidge, Karolyi, Stulz (2004), governance scores
• Klapper and Love (2003), governance scores

5
A Model with Internal Financing

• Risk neutral controlling shareholder with
  ownership a
• Interest rate 0
• measures investment profitability for firm i
• Endowment with e 0

1
j
6
A Model with Internal Financing

• Quality of Governance Practices 1 d, where
• d the proportion of cash flows that the
  controlling shareholder diverts for private
  benefits
• Includes outright stealing, shirking, perks,
  tunneling
• High d bad corporate governance
• Low d good corporate governance
• Diversion takes place before investment. The
  results do not change if diversion happens after
  investment decision is made (under slightly
  different assumptions).
• c cost of diversion PV of legal
  costs/penalties and costs of diverting and
  converting corporate resources into cash
  equivalents
• Fines, jail term, loss of reputation, bribes
• High c in the U.S.
• Low c in Russia
• The controlling shareholder will invest if
• Personal benefits from investment gt net benefits
  from diversion

7
The Optimal Investment and Diversion

• Investing up to gives
• The optimal diversion,

8
Investment Opportunities

• Hypothesis In strong legal regimes, firms
  divert less and practice higher quality
  governance.
• Good rules and effective enforcement reduce
  thievery.
• Hypothesis Controlling shareholders of firms
  with more profitable investment opportunities
  divert less for private gains and practice
  higher-quality governance.
• One is less likely to commit crime if one has
  something valuable to lose
• One is less likely to commit crime if one has
  something valuable to lose

9
Investment Opportunities

• Hypothesis The impact of profitable investment
  opportunities on corporate governance is greater
  in weaker legal regimes.
• Good governance driven by private incentives
  becomes a more important complement to
  regulation when regulation is less effective.

10
CControlling shareholders payoff as a function
of diversion d   Payoff Sensitivity
of d w.r.t. is smaller in the U.S. than in
Russia  
11
Ownership

• Hypothesis Controlling shareholders of firms
  with higher cash flow ownership practice higher
  quality of governance.
• Restatement of well-known Jensen Mecklings
  agency argument
• One does not steal from oneself
• Hypothesis The impact of ownership on governance
  is stronger in weaker legal regimes.
• Concentrated ownership becomes a more important
  tool to resolve agency conflict between
  controlling and minority shareholders in weaker
  legal regimes

12
External Financing

• Hypothesis For a given level of profitable
  investment opportunities, controlling
  shareholders of firms with greater dependence on
  external financing practice better governance.
• One does not spit into the well one drinks
  from
• Conjecture The relation between reliance on
  external financing and governance is stronger in
  weaker legal environment.
• Firms in weak legal regimes have greater
  incentive to improve their governance practice to
  overcome the deleterious effects of poor legal
  protection on their ability to raise external
  financing.

13
Valuation

• Hypothesis Firms with better governance are
  valued higher.
• Conjecture The effect of governance on valuation
  is stronger in weaker legal regimes.
• Good corporate governance is valued higher
  where it is scarce

14
Governance Data
1.
15
Governance Data
2. StandardPoors
1.
16
Governance Data

• CLSA Corporate Governance Scores in 2000
• 494 companies across 24 emerging markets
• Based on 57 questions in 7 categories
• Managerial Discipline (0-100)
• E.g., incentives towards a higher share price
• Transparency (0-100)
• E.g., timely release of annual reports
• Board Independence (0-100)
• E.g., Chairman who is independent from management
• Board Accountability (0-100)
• E.g., foreign nationals are present on the board
• Responsibility (0-100)
• E.g., share trading by managers is transparent
• Fairness (minority shareholders protection),
  PROTECT, (0-100)
• Measures to protect minority shareholders
• Composite score, COMP, (0-90)
• Social awareness (0-100)

17
(No Transcript)
18
Governance Data

• SP Transparency Ranking in 2000
• 573 companies across 16 emerging markets and 3
  developed countries
• Transparency and disclosure
• Ownership structure and investor relations (0-22)
• Financial transparency and information disclosure
  (0-34)
• Board and management structure and process (0-35)
• Aggregate, TRAN, (0-91)
• Comparing across the two data sets
• CLSA 0.16?SP country and industry dummies
• 0.05
• Both datasets are available at my homepage.

19
Other Data

• Legal Regime
• investor protection ? rule of law
• de jure ? de facto
• virtually no correlation between the two
• corr 0.05 p-val 0.82
• Investment opportunities
• 1999-2000 growth in sales
• Reliance on external financing
• projected need for outside capital as in
  Demirgüç-Kunt and Maksomovic (1998)
• actual growth rate sustainable growth rate
  (maintaining constant debt to assets ratio),
  1999-2000 average
• Ownership concentration
• Cash flow rights
• Control rights
• Control rights gt Cash flow rights
• Pyramidal structure, cross-holdings, dual-class
  shares
• From Claessens et al. (2002), 1996
• WEDGE CONTROL - CASH
• Valuation

20
Control Variables

• Industry dummies
• different accounting practices, competitiveness,
  government regulation
• 13 groups as in Campbell (1996)
• Size
• greater scrutiny
• RD/TA
• related to cost of diversion
• Export/sales
• exposure to globalization pressure
• ADR dummy
• firms have to comply with U.S. stock exchange
  regulations
• Consolidation dummy 1 if a firm consolidates
  its financial statements
• partially owned subsidiaries are treated as fully
  owned ones

21
Summary Statistics by Country
22
CLSA Scores and Governance Scandals

• How reliable are CLSA scores?
• Is it just legal environment noise?
• Look for governance scandals for 84 firms from 14
  countries with at least 11 firms
• Similar to Khanna, Kogan, Palepu (2002)
• Lexis-Nexis, 01/01/1999 12/31/2001
• assets expropriation, accounting misreporting,
  share dilution, insider trading, undertaking
  illegal projects
• manually checked 29,320 articles
• 49 scandals
• Correlation between CLSAs COMP and SCAND is
  -0.36 (p-val 0.00)
• Control for legal environment, of newspapers
  per capita (NEWS), and media attention (TOTAL)
• SCAND -0.017 ? COMP 0.015 ? LEGAL 0.017
  NEWS 0.088 ? TOTAL

23
Firms with more profitable investment
opportunities and greater reliance on external
financing practice higher quality governance and
the relations are stronger in weaker legal
regimes
significant at 1
5 10  
24
Within-Country Variation in Governance

• Is within-country variation in governance larger
  in poor legal environment countries?
• Two-stage estimation
• Run CG on firm-specific attributes
• Run e on firm-specific attributes and LEGAL
• Coefficient on LEGAL is negative and significant

25
Within-country variation in corporate governance
is larger in poor legal regimes Stage
1 Stage 2
significant at 1
5 10
26
Firms with more concentrated cash flow ownership
practice higher quality of governance and the
relation is stronger in weaker legal regimes
significant at 1
5 10
27
Firms with better of governance are valued higher
and the relation is stronger in weaker legal
regimes
significant at 1
5 10  
CLSA COMPOSITE
CLSA SOCIAL
CLSA PROTECT
SP TRAN
28
Economic Significance

• On average, a one standard deviation increase in
  CLSA composite score increases Q by 9
• The relation between corporate governance and
  performance is stronger in weaker legal regime
  countries
• Mexico, LEGAL 1 (de jure) 3.33 (enforcement)
  3.33
• increase in Q is 13.2
• Hong Kong, LEGAL 5 (de jure) 8.33
  (enforcement) 41.65
• increase in Q is 4.6
• May explains the previous mixed findings on U.S.
  data

29
Robustness

• Sample selection
• Use Heckman two-step selection model
• selection variables size, analysts coverage
• Endogeneity
• Simultaneous equations estimation (3-stage least
  squares)
• Panel data estimation (new result, 2 years for
  CLSA, 3 years for SP data)
• Alternative Variables and Specifications
• Use INVESTOR and ENFORCE separately
• Principal component analysis to define LEGAL
• investor protection, creditor protection,
  efficiency of judicial system, rule of law,
  absence of corruption, risk of expropriation,
  risk of contract repudiation
• Control for assets tangibility
• Tobit regression

30
Implications

• Debate over Coases arguments on corporate
  governance
• law matters, but firms adapt to allow efficient
  private contracts
• For policymakers
• to give the controlling shareholders greater
  incentives to improve governance practices
• Pro-growth policies vs. re-distributive policies
• Convergence in governance
• With increasing globalization of trade, national
  boundaries and legal structures become less
  effective in defining corporate policy
• Any debate over convergence must also consider
  firm-specific factors
• Social awareness is neither related to
  firm-specific factors, nor is it valued by
  investors.

31
Research in Progress

• To what extent can corporate charters substitute
  for corporate laws?
• What is the optimal mix of regulation by law
  and self-regulation?
• What is the role of financial markets development
  (Doidge, Karolyi and Stulz, 2004) and political
  connections (Faccio, 2004) in firms choice of
  governance structure?
• Are political connections and corporate
  governance substitutes?
• Is the Sarbanes-Oxley too costly to comply with?
• the Sarbanes-Oxley act and foreign private
  issuers
• How does government grabbing-hand policies
  affect firm governance/ disclosure policy? (Dyck
  and Zingales (2004))
• Do firms place ADRs to protect themselves from
  evil governments?
• Can we differentiate between good governance
  and bad governance trade flows in terms of the
  quality of institutions?
• Is a countrys own institutional development more
  responsive to good governance flows?