International Portfolio Diversification - Essay Example

International Portfolio Diversification Introduction Investing on the stock market can be a very risky venture. According to Yavas (2007), both the Capital Asset Pricing Model (CAPM) and the Modern Portfolio Theory (MPT) indicates that investors should hold a well diversified portfolio in order to reduce risk. Beta is used to measure risk. A stocks beta indicates the sensitivity of the stock’s returns to the market returns (Madura 2006, p. 304). Madura (2006, p. 304) states that investors who have a diversified portfolio use beta to determine how well their portfolio reflects movements in the market. Investors believe that favourable characteristics that are related specifically to a particular firm will offset unfavourable characteristics of other firms. This is also true for industries and so it implies that a wide range of stocks spanning various industries should be held. It is expected that certain factors affecting securities on the stock market are either firm or industry specific and so in order to reduce unsystematic risk holding securities from a wide range of industries is recommended. This is also true in relation to national securities. Certain risks are country specific and so in order to reduce risk international portfolio diversification is recommended. This paper examines how International portfolio diversification can result in a reduction in portfolio risk. It looks at various elements of risk that are associated with portfolio. It looks at market returns in five countries and shows how correlation between these markets can impact negatively on portfolio risk reduction. An analysis is also carried out to determine whether of the returns on stock markets in Japan and Canada are integrated based on their trade relationships over the years. The findings from other studies are also explored to determine how they concur with the results from this research. How can international diversification reduce risk? There are differences in the level of economic growth among countries. Some are developed while others are developing. Yavak (2007) states that these differences can lead to portfolio risk reduction as the timing of business cycles are usually different. According to Eiteman et al (2007) the case for international portfolio diversification can be broken down into two components. They are the potential risk reduction benefits of holding international securities and the potential foreign exchange risks that comes with it. Portfolio risk reduction The risk associated with a portfolio of securities is measured by the ratio of the variance of the return on the portfolio in relation to the variance of the market return (Eiteman et al 2007). As noted earlier, this is represented by beta. As the number of securities in the portfolio increases the portfolios beta approaches the market beta. A portfolio that is fully diversified would have a beta that is equal to 1. Therefore the risk that is associated with holding a particular stock can be reduced through diversification. However, risk cannot be eliminated totally (Eiteman et al 2007). This can be explained by the fact that the total risk of a portfolio comprises a systematic and a non-systematic element. The systematic element is associated with the market and unsystematic element is related to the individual elements in the portfolio. Increasing the number of securities in the portfolio reduces the unsystematic element (Eiteman et al 2007). This same approach can be taken in the form of investing in various stock markets across the globe. When investors hold securities in several countries they are able to cushion their portfolio from shocks in any one market. Therefore, if economic conditions in one country are affected by unfavorable factors, any resultant reduction in stock market returns may be offset somewhat by gains in other stock markets or at least be cushioned by the relative weight of other securities held in other parts of the world. This is however, assuming that the markets are not highly integrated. Lessard (1976, p. 33) indicates that because of low levels of correlation among nations only a fraction of national systematic risk elements is systematic in a world context and so international diversification may result in a reduction in portfolio risk. However, results from studies done by Christofi and Pericli (1999) and Valadkhani et al (2008) indicates that investors may not be able to reduce risks and increase returns substantially through international portfolio diversification because the returns on various stock markets across the world are highly correlated. All of these studies were done at different times, between two and three decades apart and so Yavak (2007) points out that national economies have become more integrated recently due to growing international trade and investment flows and specifically in relation to the terms of international financial transactions. Yavak (2007) also presents the following factors as contributors to an increased level of integration and correlation among markets. They include: i. the development of global of multinational corporations and organizations facilitating globalization such as the World Trade Organization (WTO); ii. advances in information technology iii. deregulation of the financial systems in major industrialized countries; iv. A high level of growth in international capital flows; and v. the abolishment of foreign exchange controls Yavak (2007) and Valadkhani et al (2008) indicates that recommendations for international portfolio diversification are generally based on the existence of a low level of correlation among national stock markets. If the markets are perfectly correlated, no major benefits in terms of risk reduction can be obtained. Foreign exchange risk According to Eiteman et al (2007) foreign exchange risk can also be reduced through international portfolio diversification. Eiteman et al (2007) also states that an internationally diversified portfolio is essentially the same as a domestic portfolio because the investor’s aim is to combine assets that are not perfectly correlated in order to reduce the level of risk while tapping into a larger pool of investments outside the home market. The only difference between the domestic and the international portfolio is that by investing outside the home country an investment that is denominated in a foreign currency is acquired which implies the acquisition of two additional assets – the currency of denomination and the asset purchased with the currency (Eiteman et al 2007). Data and Results The data used in this study consists of the weekly value of the stock market indices for the Nikkei Index in Japan, the NASDAQ Composite Index in the USA, the FTSE Index in the UK, the S&P/TSX Composite Index in Canada and the DAX Index in Germany. The weekly returns were calculated based on the following formula. Return = (Pt – Pt-1)/Pt-1 The data covered 156 weeks for the period January 2008 to December 2010. The descriptive statistics for the data is shown in the table below. Country Mean Median Maximum Minimum SD Skewness Kurtosis p-Value Japan -0.001 0.002 0.121 -0.243 0.04 -1.341 8.322 0.006 USA 0.001 0.004 0.109 -0.153 0.038 -0.394 2.241 0.006 UK 0.0002 0.001 0.134 -0.21 0.037 -0.837 7.53 0.006 Canada 0.0005 0.004 0.137 -0.161 0.037 -0.625 4.378 0.006 Germany 0.0001 0.0004 0.161 -0.216 0.042 -0.598 5.73 0.007 Table – 1 Description of data Employed January 2008 to December 2010 The table above which is labeled Table -1 shows the descriptive statistics of the data which contains the sample means, medians, maximums, minimums, standard deviations (SD), skewness, kurtosis and p-values relating to the five countries. The highest mean return is 0.001 in the USA and the lowest Japan with a -.0.001. The standard deviation range from 0.037 fro UK and Canada – making them the least volatile to 0.042 for Germany – the most volatile. All the countries are developed and so the closeness of the standard deviations is not considered abnormal as they are closely integrated through trade. The distributions of all the various indices are negatively skewed, indicating that the mean is the lowest of the three measures of central tendency - the mean, median and mode (Mason and Lind 1997). Most of the weekly stock returns have excess kurtosis which indicates that they have very high peaks and are therefore described as leptokurtic (Spiegel and Stephens 1999). The assumptions are rejected at a 5% level of significance for all stock market index returns because the computed p-values are all below ? = 0.05. Five (5) per cent is the lowest level of significance at which the null hypothesis is rejected and indicates that the normality assumptions of mean = 0 and variance = 1 have not been met. The market returns of five countries namely: Japan, USA, Canada, Germany and the UK were examined to determine whether significant levels of correlation exist among them. The table below indicates the levels of correlation among the market indices of these countries.   Jap Return USA Return Lon Return Can Return Ger Return Japan 1 USA 0.737286866 1 UK 0.750120801 0.851637328 1 Canada 0.70065627 0.816661774 0.852861569 1 Germany 0.770784202 0.882655695 0.928925387 0.813291075 1 Table – 2 Correlations between Stock Market Indices The table above which is labeled Table – 2 represents the correlation of the weekly stock market returns for three years from January 2008 to December 2010 for the Nikkei 225 index in Japan, the NASDAQ composite index in the USA, the FTSE 100 in the UK, the S&P/TSX COMPOSITE index in Canada, and the DAX index in Germany. They represent two European countries – UK and Germany; two North American countries – Canada and the USA; and one Asia country – Japan. The results indicate that all the markets are highly positively correlated with Germany and the UK, both of whom play a vital role in the European Union, except on currency matters, showing the highest correlation of 93%. USA and Germany also showed a significantly high level of correlation at 88%. Canada and Japan though also very significant, showed the lowest level of correlation at 70%. At both the 95% and 99% confidence level where ? = 0.05 and 0.01 the computed p-values were above the critical value and so the null hypothesis that there is no correlation between the stock market indices for the five countries would be rejected at both the 5 per cent and 1 per cent level of significance. The alternative hypothesis which indicates the existence of statistically significant levels of correlation would therefore be accepted. These correlations are not perfect and so possibilities exist for portfolio risk reductions. A cursory look at the data indicates that there were many weeks when the indices did not move in the same direction and so there is sufficient room to maneuver. Depending on how well the market is understood, investors could reap major gains if the use all the information that is available to their advantage. Scenario relating to Japan and Canada The data in the table below provides information on the daily stock market indices relating to the period October 1, 2010 to March 10, 2011 and the period March 12, 2011 to August 3, 2011 for the Nikkei 225 index for Japan and the S&P/TSX Composite index for Canada. The period October 2010 to March 10, 2011 represents the period immediately preceding the tsunami while the period March 12 to August 3 represents the period immediately following the tsunami.   Mean Median SD Maximum Minimum Kurtosis Skewness p-value Before March 11   Canada 0.001 0.001 0.007 0.017 -0.018 -0.012 0.061 0.001 Japan 0.001 0.001 0.01 0.029 -0.021 0.0612 0.105 0.002 After March 11   Canada -0.001 0.0001 0.008 0.016 -0.02 -0.085 -0.31 0.002 Japan 0.0004 0.0004 0.017 0.057 -0.106 18.076 -2.119 0.003 Table - 4 – Descriptive data for the peri0ds before and after tsunami The table labeled Table – 4 indicates that the highest mean return was 0.1% for both countries before the tsunami and a negligible 0.04% after the tsunami. The standard deviation was 1% for Japan and 0.7% for Canada before the tsunami and 0.017 for Japan and 0.008 for Canada after the tsunami. This indicates that there was a significant increase in volatility in Japan’s index after the tsunami than there was for Canada. . The maximum returns for Japan was consistently higher after the tsunami than before where it increased from 2.5% to 5.7%. Canada’s index experienced the opposite when the index maximum return before the tsunami was more (1.7%) before the tsunami and 1.6% after the tsunami. The lowest return for Japan before the tsunami was a minus 2.1% while that for Canada was a minus 1.8%. However, the lowest return for Japan after the tsunami reached double digits of minus 10.6% with Canada experiencing a minus 2%. The skewness for both Canada and Japan was positive in the period before with percentages of 6.1% and 10.5% respectively. There were negative rates of skewness for Canada and Japan after the tsunami a negative 31% and 211% respectively. The kurtosis for Canada was negative both before and after the tsunami with higher levels in the period after. Those levels are considered moderate. On the other hand, kurtosis of Japan’s index was significantly higher during the period after the tsunami than before. The p-values indicate that the assumptions regarding normality have not been met. The correlation results for the two between Japan and USA both before and after the tsunami were very different for both periods and in comparison to the correlation obtained in Table – 2. The table below provides information on the results. Table 4 – Correlation before the tsunami Table 4 shows a very weak correlation of 0.7% before the tsunami between both markets. The p-value suggests that there was no statistically significant correlation between the Canadian and the Japanese stock market index before the tsunami. This suggests that risks can be significantly reduced by holding securities in both markets. The table below shows the correlation between both indices during the period after the tsunami.   Canada After Japan After Canada After 1 Japan After 0.024137 1 Table 5 - Correlation after the tsunami Table – 5 indicates that the correlation coefficient has improved substantially but there is still a week correlation between but markets. The p-value suggests that there is no statistically significant correlation between the Canadian and the Japanese Index since the tsunami. This is an indication that systematic risks can be significantly reduced by holding securities in both markets. According to World Political Review (2011) Canada sells approximately $9bn worth of primary resources to Japan, including: coal, canola, lumber, copper, pulp and paper, aluminum, wheat, meat and fish. In return Japan sells approximately $13bn worth of goods consisting primarily of motor vehicles and spare parts, nuclear reactors and parts, and consumer electronics (World Political Review 2011). Japan has been trading with Canada since 1973 and from that time until mid 2000’s Japan was Canada’s second largest export market. The value of Canada’s exports has however remained constant and Canada is now selling more of its products to the UK and China (World Political Review 2011). Based on the results there seems to be little or no integration since the later part of 2010 to present. It therefore implies that Canada was not seriously impacted by the effects of the tsunami in Japan. It also showed that even though Japan has been through some tough straits during the period after the tsunami they were not overly dependent on Canada for support because of their relationships with other countries in the developed and developing world. Conclusion and Recommendation The results for the analysis of stock market results indicate that high levels of correlation exist between the stock indices of the developed countries. The results appear to concur with studies done by Login and Solnik 1995; Fratzscher 2002; and Fernandez-Izquierdo and Lafuente 2004 who used the generalized auto-regression conditional heteroscedasticity (GARCH) model to perform their work which points to high degree of integration between stock markets. High levels of correlation suggests that the returns that are to be gained from international diversification may be limited, However, there are still opportunities for reducing risk as none of the markets are perfectly correlated with a product moment correlation coefficient equal to 1. It must be emphasized however that there are periods that show negative correlation that may suggest that adequately responding to information relating to ones portfolio of securities could reduce unnecessary risks. The risk of sudden foreign exchange fluctuations that may negatively impact an internationally diversified portfolio cannot be over emphasized. It is important that the portfolio be properly diversified in order to avoid huge losses. Additionally, hedging on the foreign exchange futures market may be an option if the size of the portfolio is not too sensitive in relation to transactions costs and other costs of investing outside of the home country. International investing may only be practical or feasible for institutional investors whose task it is to invest funds on behalf of different types of clientele. The statistical tests carried out were very simple and therefore not as rigorous as those carried out by the other researchers mentioned thus far. It therefore implies that more work can be done to determine if these results hold. References Christofi, A. and Pericli, A. (1999). Correlation in price changes and volatility of major Latin American stock markets. Journal of Multinational Financial Management: 9(1). p. 79-93. Eiteman, D.K., Stonehill, A.I. and Moffett, M.H. (2007). 11th ed. USA: Pearson Education Inc. Lessard, D.R. (1976) World, Country, and Industry Relationships in Equity Returns Implications for Risk Reduction Through International Diversification. Financial Analyst Journal: 32(1). p. 32-38 Fernandez-Izquierdo, A. and Lafuente, J.A. (2004). International transmission of stock exchange volatility: empirical evidence from the Asian crisis. Global Finance Journal: 15(2). p. 125-37 Longin, F. and Solnik, B. (1995). Is the correlation in international equity returns constant: 1960-1990? Journal of International Money and Finance: 14(1). p. 3-26. Madura, J (2006). Financial Markets and Institutions. 7th ed. USA: Thomson South-Western Mason, R.D. and Lind, D.A. (1996). Statistical Techniques in Business & Economics. 9th ed. USA: Irwin. Solnik, B.H. (1995). Why Not Diversify Internationally Rather Than Domestically? Financial Analyst Journal. January-February 1995 Spiegel, M.R. and Stephens, L.J. (1999). Schaum’s Outline of Theory and Problems of Statistics. 3rd ed. USA: McGraw-Hill. Valadkhani, A., Chancharat, S and Harvie, C. (2008). A factor analysis of international portfolio diversification. Studies in Economics and Finance. 25(3). p. 165-174 World Political Review. (2011). Global Insider: Canada-Japan Trade Relations: Retrieved from: http://www.worldpoliticsreview.com/trend-lines/8415/global-insider-canada-japan-trade-relations. Last accessed 2nd Aug 2011 Yavas, B.F. (2007). Benefits of International Portfolio Diversification. Graziadio Business Review: 10(2). Retrieved: http://gbr.pepperdine.edu/2010/08/benefits-of-international-portfolio-diversification. Last Accessed 2nd Aug 2011 NB: Market information was obtained from: http://finance.yahoo.com/ Read More