Reading 45: The Case for Inter....Diversification－LOS g

Lewis correlation statement is incorrect. The issue of the stability of correlations among international markets is still unresolved. Although correlations have risen over time, they also appear to increase when volatility increases. This phenomenon occurs due to the econometrics of the correlation measure. Volatility can mask the true correlation. Academic research has found that previously reported increases in correlations during volatile stock markets were manifestations of the volatility and not increases in the true correlations. Lewis country-specific comment is correct; just because a country has outperformed in the past does not mean it will continue to do so. No one country will always deliver the best performance.

Low correlations between markets reduce volatility, but do not directly affect the level of expected return. The inclusion of emerging markets should result in higher expected returns due to their higher expected economic growth, not because they have low correlations with developed markets.

Although relative value multiples are easy to calculate, the fact that historical data is used is a disadvantage. Historical data may not be relevant to a companys future growth prospects.

The proportion of stock publicly traded (free float) is frequently low because the government is often the primary owner of most of a companys stock.

In developed countries, the correlation between local market returns and currency movements is usually negative, whereas in emerging markets, the local market returns and currency returns are usually positively correlated.

International investing is principally concerned with diversification across countries, while global investing diversifies across industries as well as countries. However, there is no assurance that international diversification will be achieved under a global investing strategyit is possible that the fund manager may decide to commit a disproportionate share of assets to a small set of countries, or even entirely to the investors home country.

The use of historical data will understate the current correlation because correlations have been increasing over time. Higher correlations are common in developed markets. The actions of international investors result in greater capital mobility and higher correlations. Currency risk is about half that of foreign stock risk.

Countries with few multinational corporations and high tariffs will have segmented capital markets and will be less correlated with other country markets. The higher the proportion of multinational corporations and the more open the capital markets, the more integrated the capital market and the higher the correlation with other country markets.

During periods of increased volatility, correlations will be estimated higher due to estimation error, even if correlations have not actually increased.

During volatile periods, the correlation will appear to increase, even if it hasnt, due to the econometrics of the correlation measure. The analyst should examine the standard deviation through time to see if it has increased.

Due to the econometrics of the correlation measure, the correlation will be higher when volatility is higher, even if the correlation structure of the underlying data has not changed.

The benefits of international diversification are overstated if the correlations between markets are higher than expected. If markets were segmented, correlations would be lower. The other responses represent causes of increased correlations between markets.

If the sample of returns between two markets is split in half, the high return and higher volatility half will have a higher correlation than the low return half when in fact the true correlation is best measured by the whole sample correlation. This is due to the econometrics of the correlation measure and illustrates the pitfalls of estimating the correlation statistic during periods of increased market volatility. To determine whether markets really have increased their comovement during periods of increased volatility, the analyst should measure the dispersion (standard deviation) of country portfolios.