MeanVariance Convergence around the World

1
Mean-Variance Convergence around the World

• Discussion by,
• Bhaskaran Swaminathan, PhD
• Director of Research
• LSV Asset Management
• Professor of Finance
• Cornell University

2
Motivation of the paper

• Correlations among international equity markets
  have increased over time suggesting increasing
  financial market integration.
• Is there also convergence in the risk-return
  tradeoffs in the various markets?
• Are these two phenomena the same or distinct?

3
Key conclusions in the paper

• Proposes a new measure of risk-return tradeoff
  derived from cluster analysis.
• Concludes risk-return characteristics of 17
  developed stock markets have converged over time.
• 14 emerging markets are converging toward
  developed markets.
• This phenomenon is distinct from the convergence
  in correlations.
• Evidence is consistent with increasing financial
  market integration.

4
Discussion questions

• How can we test for convergence in risk-return
  tradeoffs across markets?
• Measures based on asset pricing theory Sharpe
  ratio.
• Alternate non-parametric measures.
• Discussion of the measure used in this paper
• Economic intuition and interpretation
• Robustness
• Sharpe ratios of various MSCI benchmark indices
  using data from 1970 to 2005.
• Is there a convergence in risk-return tradeoff?

5
Testing for convergence

• International CAPM predicts, in equilibrium,
  marginal price of risk should be equal across
  assets
• One could test whether the Treynor ratios have
  converged over time across different markets.
• One could use a conditional CAPM and test whether
  there is an improvement in the pricing of the
  various equity market portfolios over time.
• Beta measurement can be problematic and a
  single-factor model may be mis-specified.

6
Testing for convergence

• The Capital Market Line (CML) suggests, in
  equilibrium, Sharpe (reward-to-risk) ratios
  should be the same across all diversified
  portfolios
• One could test whether the Sharpe ratios have
  converged over time across different markets.
• Sharpe ratios are widely used and intuitive to
  understand.

7
A simple version of the measure used in this paper

• The return distance is the deviation of each
  markets returns from the return of an
  equal-weighted market portfolio.
• How do we interpret the risk distance? It is not
  the deviation from the standard deviation of an
  equal-weighted market portfolio.
• Average of standard deviations is not equal to
  the standard deviation of the average return!
• This definition ignores the correlation structure
  among the various market returns and the
  resulting diversification benefits!
• Compare a given markets risk to the risk of the
  world market portfolio! Consider both
  equal-weighted and value-weighted portfolios.

8
Other comments on this measure

• Provide economic intuition for this measure and
  try to relate it to how international investors
  would view changing risk-return profiles.
• Compare and contrast this measure with more
  traditional measures like Sharpe ratio and check
  whether you get consistent results.
• The cross-market distance measure equally weights
  each country. US and Japan constitute more than
  70 of the global market cap and they show little
  convergence in their return distance (see Table
  3).
• Examine the robustness of your results under
  value-weighting!
• Bottom line What is the advantage of using this
  measure? Is it robust? Does it show international
  convergence?

9
Increasing correlations among MSCI benchmark
indices
Increasing international stock market
correlations!
10
Sharpe ratios of MSCI indices
Sharpe ratios diverged in the last 5 years across
US, EAFE, and Emerging Markets compared to the
prior 30 years!
11
(No Transcript)
12
Final thoughts..

• In a world in which market inefficiencies can
  persist over time and are not corrected
  instantaneously as assumed in rational asset
  pricing, there can be systematic differences in
  risk-return tradeoffs across securities and
  markets.
• The actions of active money managers trying to
  exploit these differences is likely to force
  convergence but only slowly over time.
• Overall, the paper addresses an interesting and
  relevant issue. Well written and econometrics
  well executed.
• Examine the robustness of the risk-return
  distance measure and compare to more traditional
  measures!