Does Better Corporate Governance Lead to Stock Market Development and Capital Accumulation A Case St

1
Does Better Corporate Governance Lead to Stock
Market Development and Capital Accumulation? A
Case Study of India

• Comments by Walter Novaes
• PUC-Rio

2
Corporate Governance and Capital Accumulation

• Development of capital markets
• Financing Lowers cost of capital (Rajan and
  Zingales, AER)
• Diversification Fosters entrepreneurship
• Allocation of capital
• Monitoring by stock markets (Holmstrom and
  Tirole, JPE)

3
Time Series Data

• Evolution of 60 legal variables in India
  1970-2005
• Centre of Business Research, University of
  Cambridge
• Annual data on stock prices, volume of trading of
  trading, market cap 1976 2005
• IMF and World Bank
• Gross Private and Public Capital Accumulation,
  GDP 1950-2004
• Government of India (Economic Survey, 2005-06)

4
Empirical Strategy Cointegration tests

• Example PPP
• logP(Turkey)logelogP(Italy)
• logP(Turkey) I(1) stochastic trend
• logP(Italy) I(1) stochastic tend
• Log(e) I(1) stochastic trend
• a(1)logP(Turkey) a(2)logP(Italy) a(3)loge
  I(0) stationary
• Interpretation In the long-run,
  zlogP(Turkey)-logP(Italy)-loge cannot deviate
  from zero too much
• Likewise Market cap, Capital Accumulation and
  legal variables should not deviate too much from
  each other in the long-run

5
Cointegration test
6
Findings and Interpretations

• Main Results No long term relationship between
  legal variables and stock market development or
  capital accumulation. (H0 not rejected)
• Interpretation 1 Stock markets are like casinos
  not important for capital accumulation.
• Interpretation 2 variation of law on the books
  does not capture the variation of the rule of law

7
Is cointegration a good way to answer the
question?

• Sample size Important for cointegration tests
  (35 observations?)
• Legal variables Flat in several years (enough
  variation?)
• Should we expect legal variables and size of
  stock market to be I(1)?
• If not Cannot rely on cointegration theory to
  ignore potential endogeneity problems (e.g.,
  unobservable variables)
• Cointegration vs Multicountry (LLSV) vs Natural
  Experiments (exogenous shocks with control
  samples)
• LLSV and natural experiment papers find that
  the law is relevant.
• What is the problem that the cointegration
  approaching addresses?