A STUDY OF MARKET EFFICIENCY IN CHINA: COMPARING THE PERFORMANCE OF MUTUAL FUND PORTFOLIOS AGAINST THE SSE COMPOSITE INDEX Pedro M. Kono, D.B.A., Pepperdine University and Temple University Pan Yatrakis, Ph.D., Nova Southeastern University Hao Wang, M.B.A., SAP Japan Abstract This study tests the efficiency of securities markets in China. The analysis compares the performance of mutual fund portfolios with that of the overall SSE Composite Index as of June 30, 2010, December 31, 2010, and June 30, 2011. The portfolios are constructed according to the Modern Portfolio Theory (MPT) developed by Harry Markowitz in 1952 and are based on the funds performance over the prior three years. The study concludes that an optimal mutual fund portfolio developed according to the MPT can outperform the overall market index of Chinese securities when performance is measured as the return per unit of risk (Sharpe ratio) and the risk-adjusted return (M2). 1. Introduction In the second quarter of 2010, the People s Republic of China (China) became the world s second largest economy in terms of gross domestic product. According to the Organization for Economic Cooperation and Development (OECD), China s economy has grown an average of 11.0% per year over the last seven years, and the composite index of the Shanghai Stock Exchange (SSE) has grown in tandem, at an average annual rate of 19.3%. The International Monetary Fund (IMF) now estimates that China will become the world s largest economy some time during the next ten years. Shanghai was the first city in China where securities trading was introduced, back in the 1860s. By the 1930s, Shanghai had emerged as the financial center of the Far East, where both Chinese and foreign investors could trade stocks, debentures, government bonds, and futures. In 1946, the Shanghai Exchange Co. Ltd. was created but, due to civil war and changes in the political environment, it ceased operations in 1949. In 1980, the securities market was reestablished, in line with China s introduction of reforms that created its present socialist market economy. On November 26, 1990 the SSE came into being, and on December 19, 1990 it formally began operations. By the end of 2009, China had 870 listed companies and 1351 listed securities on the SSE. The total market capitalization of tradable shares was US$2,841 billion. Securities listed on the SSE fall into four categories: stocks, bonds, funds, and warrants. The stocks are divided into Class A Shares and Class B Shares, of which the A Shares are available only to domestic investors, while the B Shares can be purchased by both domestic and foreign investors. By the end of 2009, there were 860 A Shares and 54 B Shares listed on the SSE. The SSE Composite Index includes all the A Shares and B Shares listed on the SSE. A mutual fund is a fund managed by an investment management company with the financial objective of generating a high rate of return relative to the risk undertaken. The management company collects money from many investors and invests this money, according to an established strategy, in different stocks, bonds and other financial instruments and in a diversified manner designed to minimize overall risk. The investment company performs thorough research and detailed analysis of market conditions and trends of stock and bond 91
prices. This assists the fund managers in selecting specific securities to be purchased by the fund. Investors who place their money in a mutual fund receive an analogous number of shares of the fund, with the number based upon its daily net asset value (NAV) per share. Investors who sell all or part of their shares receive the value of the NAV per share at the end of the day. The investment management company charges an annual administrative fee and may also charge load (sales) and exit fees. In China, all mutual funds are registered and their NAVs are quoted at the end of every trading day. The administrative fee is typically around 1.5% per year, and purchases are subject to load fees that vary between 0.8% and 1.5%, depending on the amount invested. Sales of mutual fund shares are also subject to a redemption fee of 0.5% if the holding period is less than one year; this fee declines to zero over a twoyear period. Mutual funds were introduced in China on March 16, 1998. As of December 2009, there were 557 funds in existence, representing various asset classes and having a total NAV of US$411 billion. 357 of these funds invest in stocks and focus their strategies on investing styles such as value-added, value investing, income, growth, prudent growth, active growth, dividends, balanced, indexed, value optimized, and index optimized. An additional 113 target investments in bonds and preferred shares, or focus their strategies toward value-added investing, prudent growth, and income investing. The remaining funds invest in currencies and foreign markets, or pursue investment styles such as guaranteed capital and portfolios balanced between stocks and bonds. The SSE indices are the authoritative statistical measures followed widely and used domestically and abroad to measure the performance of China s securities market. The SSE Index Series consists of 58 indices, including 52 equity indices, five bond indices and one fund index. The SSE Index Series measures the market trends in the Shanghai securities market and reflects the performance and price movement of companies across a diverse range of industries. Its individual component indices offer investors a variety of benchmarks for portfolio analysis. The present study uses the SSE Composite Index as its benchmark of the performance of the Shanghai Exchange because it reflects the total market capitalization of all listed stocks, e.g., all A Shares and B Shares listed on the SSE. This index has been published since July 15, 1991. In our analysis, we apply the concepts of Modern Portfolio Theory (MPT) developed by Harry Markowitz in 1952 (Markowitz, 1952, 1959). Even after 60 years, this theory remains the dominant method used in portfolio analysis (Chernoff, 2002). Markowitz s breakthrough was to recognize that risk must not be measured in terms of individual securities, but rather by how the risk of each individual security relates to the risks of the other securities in a portfolio (Chernoff, 2002). Markowitz used a meanvariance optimizer that required the estimation of expected return, variances, and co-variances; this statistical technique remains to this day the basic method for creating efficient portfolios. MPT also formed the foundation for the work of William Sharpe on the Capital Asset Pricing Model (CAPM), which states that the most efficient portfolio is the stock market itself (Chernoff, 2002). Finally, MPT provides a means of calculating the price of risk and applying trade-off theory, i.e., the amount of additional risk that must be assumed in exchange for an increase in a portfolio s expected return. According to the theory, 92
optimal portfolio diversification allows investors to reduce volatility and trade off risk and return by: a) maximizing return and minimizing risk; b) maximizing return for the same level of risk; or c) minimizing risk for the same level of return. The process produces portfolios that provide maximum expected return for minimum variance (Markowitz, 1952). Such diversified portfolios are more efficient because they maximize the combination of input (risk) per unit of output (return), and their optimal combinations form the so-called efficient frontier (Markowitz, 1991). This efficiency is attainable not so much by investing in many securities as by avoiding investments in securities with high co-variances (Markowitz, 1952). This study also applies the M2 statistic developed by Franco Modigliani and his granddaughter Leah Modigliani in 1997. The Modigliani-Modigliani measure or M2 expresses the under- or over-performance, in percentage terms, of a security or portfolio of securities in relation to a benchmark adjusted for relative risk (Bodie, Kane, and Marcus, 2010). Bernstein (2006) stated that portfolio managers constant search for alpha, or excess returns, makes markets more efficient. This quest creates a paradox, since investors would prefer to track an index or to adopt other kinds of passive strategies. But if all investors were to follow such strategies, the market would become less efficient, creating new opportunities for portfolio managers eager to seek excess return (Bernstein, 2006). Many investors are engaged in separating returns attributable to alpha from those relating to beta and attempting to increase exposure to alpha and active risk (Hill, 2006). This is apparent from the growth of hedge funds, where total return on a portfolio can be allocated between the return from selection of securities and the return from bearing risk (Fama, 1972). However, despite such efforts, investing according to the MPT in diversified index funds that carry low management and transaction fees has proven to be the most efficient investment strategy (Malkiel, 2003). Modern Portfolio Theory revolutionized the investment world by enabling portfolio managers to focus on quantifying the investment risk and expected return of a managed portfolio. Fabozzi and Markowitz (2002) stated that MPT complements the subjective art of investing by providing a scientific and objective analysis of the risks and returns of an entire portfolio. Insights from MPT facilitated other developments in modern financial theory, including the Efficient Market Hypothesis or EMH (Stewart, 2006). Fama (1970) stated that the primary role of capital markets is the efficient allocation of ownership of the economy s capital stock. He also described an efficient market as one where security prices reflect all available information, thereby providing accurate criteria for capital allocation. Under equilibrium of risk and return in capital markets, portfolio performance may then be best measured against passive benchmarks (Fama, 1991). The present study applies Markowitz s (1952) Modern Portfolio Theory (MPT) to Chinese mutual funds selected on the basis of age, market capitalization and market representativeness, and compares the performance of optimal mutual fund portfolios against the SSE Composite Index, which represents the entire securities market in China. The results of this study provide insights to practitioners and academics on the outcomes of investing in portfolios made up of Chinese mutual funds as compared to investing in a broad Chinese market index. 93
2. Methodology Our analysis consisted of constructing an optimal mutual fund portfolio and finetuning it over two additional semi-annual periods according to the principles of the MPT. The analysis, therefore, covers the performance of three optimal portfolios in relation to the market index. Only mutual funds meeting strict criteria relating to maturity, liquidity, and market representativeness as of December 31, 2009 were considered in constructing the optimal portfolio. The selection criteria required: a) more than three years of existence, b) an NAV of more than US$1 billion, and c) a position among the top 18% of mutual funds, which represent 50% of the market s NAV. From among the existing 557 mutual funds traded on the SSE, only 44 mutual funds complied with the required criteria. The study applied the following methodology: 1) Asset classes: Considered the sample of 44 stock and bond mutual funds that satisfied the selection criteria on December 31, 2009. 2) Statistical data: The mutual funds returns, variances, standard deviations, correlations, and co-variances were based on data from the prior three years. The first optimal portfolio was based on data from January 1, 2007 to December 31, 2009. The second optimal portfolio considered data from July 1, 2007 to June 30, 2010. The third optimal portfolio used data from January 1, 2008 to December 31, 2010. 3) Optimal Portfolios: According to the MPT, these were determined from the selected 44 mutual funds, based on their statistical data. The estimated risk-free rate, based on the average of the oneyear deposit rates over the previous three years, was 3.23% per year for the first portfolio, 3.36% per year for the second portfolio, and 3.09% per year for the third portfolio. The estimated market return was 13.0 % per year for the three portfolios, and was based on the average growth rate of the SSE Composite Index over the previous five years. 4) Performance: The performance of the optimal mutual fund portfolios used the Sharpe ratio, which measures the return per unit of risk, and the M2, which determines the risk-adjusted return. The performance of each portfolio was assessed and compared with the market index on June 30, 2010, December 31, 2010, and June 30, 2011. 5) Statistical Tests: The returns per unit of risk of the mutual fund portfolios and of the market index were statistically tested using correlation analysis. The following null hypothesis was tested: A portfolio of Chinese mutual funds constructed according to the MPT provides a higher return per unit of risk and riskadjusted return than the SSE Composite Index. The first optimal portfolio was constructed using data as of December 31, 2009. It was revised as of June 30, 2010 creating the second optimal portfolio. It was revised again as of December 31, 2010 forming the third optimal portfolio. The mutual funds codes, strategies, prices on the construction date, weights, and number of shares for each portfolio were as follows: 94
Table 1 Portfolio 1 on December 31, 2009 Code 160706 510050 050002 590001 360001 160505 519300 000011 Strategy Indexed Indexed Indexed Value Growth Growth Indexed Value Price on 09/12/31 3.476 3.3082 3.0528 3.2053 3.2168 5.4081 2.6223 11.7011 Weight 0.02% 28.27% 11.79% 1.94% 12.76% 22.04% 6.07% 17.13% Shares 446 854479 386209 60447 396588 407524 231302 146368 Table 2 Portfolio 2 on June 30, 2010 Code 110003 510050 320003 050002 070003 160607 000001 519300 162607 000011 Strategy Indexed Indexed Growth Indexed Growth Value Growth Indexed Value Value Price on 10/6/30 2.0736 2.3865 3.3026 2.2228 2.9237 2.3164 4.3753 1.9079 2.5626 10.5755 Weight 6.15% 5.77% 13.07% 19.59% 6.16% 6.29% 6.05% 32.86% 0.06% 4.01% Shares 296436 241712 395615 881412 210602 271596 138199 1722154 2520 37920 Table 3 Portfolio 3 on December 31, 2010 Code 160706 110003 510050 000021 519001 070003 360001 160607 002001 162703 000011 Strategy Indexed Indexed Indexed Growth Value Growth Growth Value Dividends Growth Value Price on 10/12/31 3.0779 2.3016 2.5877 3.1197 5.9713 3.2951 3.1148 2.5788 5.8837 4.7913 14.5372 Weight 29.93% 7.47% 16.43% 3.52% 0.04% 5.17% 9.87% 3.96% 1.97% 7.41% 14.22% Shares 972570 324588 634951 112991 591 156874 316965 153609 33561 154591 97810 The following tables show the six-month return and risk of the mutual fund portfolios and the market index. This information, in combination with the six-month risk-free rate, established the return per unit of risk (Sharpe ratio) and the risk-adjusted return (M2) on June 30, 2010, December 31, 2010, and June 30, 2011. Table 4 Portfolio 1: Six-Month Return per Unit of Risk and Risk-Adjusted Return on June 30, 2010 Portfolio # of Securities Return Risk-free Rate Risk Return/Risk M2 MF Portfolio 1 8-23.46% 1.13% 15.63% -1.57 3.02% Market Index 100% of market -26.82% 1.13% 15.85% -1.76 Table 5 Portfolio 2: Six-month Return per Unit of Risk and Risk-Adjusted Return on December 31, 2010 Portfolio # of Securities Return Risk-free Rate Risk Return/Risk M2 MF Portfolio 2 10 19.36% 1.25% 14.93% 1.21 2.76% Market Index 100% of market 17.08% 1.25% 15.32% 1.03 Table 6 Portfolio 3: Six-Month Return per Unit of Risk and Risk-Adjusted Return on June 30, 2011 Portfolio # of Securities Return Risk-free Rate Risk Return/Risk M2 MF Portfolio 3 11-0.81% 1.50% 12.38% -0.19 0.95% Market Index 100% of market -1.64% 1.50% 11.73% -0.27 95
3. Summary and Conclusions Our analysis indicates that, in 2010, the diversification effect on Portfolios 1 and 2 reduced their overall risks to a level below that of the overall market. It is also notable that the risk of all three portfolios, as well as that of the market index, diminished from January 1, 2010 to June 30, 2011. As expected, the optimal portfolios required the inclusion of a relatively small number of mutual funds. The number varied from eight in the first portfolio to ten in the second and eleven in the third. This, also, was Table 7 Portfolio 1: Regression Results on June 30, 2010 anticipated because mutual funds are already diversified securities, and their use represents an efficient and cost-effective way of building optimal portfolios (Kono, Yatrakis, Simon, and Segal, 2007). Our analysis concluded that portfolios of mutual funds in China, constructed according to the MPT, had a better ratio of performance to risk than the market index. This was observed throughout the period of January 1, 2010 to June 30, 2011. The beta coefficient was significant at the 0.1% confidence level, although the coefficient of the intercept was not significant at the conventional 5% level. Regression Statistics Multiple R 0.9919 R Square 0.98387 Adjusted R Square 0.98373 Standard Error 0.00184 Observations 118 ANOVA Df SS MS F Significance F Regression 1 0.02385 0.02385 7076.6 8E-106 Residual 116 0.00039 3.4E-06 Total 117 0.02424 Coefficients Standard Error t Stat P-value Lower 95% 95% Lower Intercept 0.00032 0.00017 1.87922 0.06272-1.7E-05 0.00066 3.8E-05 0.00061 X Variable 1 0.97831 0.01163 84.1225 8E-106 0.95528 1.00135 0.95903 0.99759 Table 8 Portfolio 2: Regression Results on December 31, 2010 Regression Statistics Multiple R 0.98536 R Square 0.97093 Adjusted R Square 0.97069 Standard Error 0.0023 Observations 124 ANOVA df SS MS F Significance F Regression 1 0.02151 0.02151 4074.4 1.4E-95 Residual 122 0.00064 5.3E-06 Total 123 0.02216 96
Coefficients Standard Error t Stat P-value Lower 95% 95% Lower Intercept 0.0002 0.00021 0.98913 0.32456-0.00021 0.00062-0.00014 0.00055 X Variable 1 0.96101 0.01506 63.8311 1.4E-95 0.9312 0.99081 0.93605 0.98596 Table 9 Portfolio 3: Regression Results on June 30, 2011 Regression Statistics Multiple R 0.98647 R Square 0.97312 Adjusted R Square 0.9729 Standard Error 0.00187 Observations 119 ANOVA df SS MS F Significance F Regression 1 0.01485 0.01485 4236.45 9.8E-94 Residual 117 0.00041 3.5E-06 Total 118 0.01526 Coefficients Standard Error t Stat P-value Lower 95% 95% Lower Intercept 7.6E-05 0.00017 0.445 0.65714-0.00026 0.00042-0.00021 0.00036 X Variable 1 1.04194 0.01601 65.088 9.8E-94 1.01023 1.07364 1.0154 1.06848 These results shed light on the question of whether investors in Chinese mutual funds could increase their returns over and above the market index by constructing portfolios according to the MPT. We conclude that such an increase of return per unit of risk and return on a risk-adjusted basis is indeed possible using optimal portfolios of mutual funds that track securities in China. The first optimal portfolio and the two revised portfolios showed lower negative returns per unit of risk and higher positive returns per unit of risk than the market index for an eighteen-month period. Additionally, for the same period, these portfolios also provided a higher risk-adjusted return than the market index. The six-month revision of the optimal portfolio may have contributed to this better performance. Israelsen (2009) stated that portfolio revision is likely to be an efficient method of improving a portfolio s performance and protecting the portfolio during periods of market distress. However, the optimal revision frequency has to be determined in light of the transaction and administrative costs incurred (Kono et al., 2007). In an academic sense, this study tested the efficiency of the Chinese stock market. It validated the use of MPT in this context for the construction of efficient portfolios. Our analysis applied MPT to the construction of optimal portfolios of Chinese mutual funds and concluded that an optimal mutual fund portfolio can outperform the most comprehensive market index of stocks trading on the SSE Exchange. This observation runs counter to the findings of other researchers who have concluded that passive management produces better results than active management (Prondzinski, 2010). Our conclusion also challenges the validity of the semi-strong form of the EMH as applied to the Chinese market. This may be due to several factors: securities prices 97
may not reflect all the information available to investors; transaction costs still appear to be very high; and information is not likely to be available without cost (Fama, 1970). It is probable that these conditions reflect the still relatively immature stage of the mutual fund industry in China, which has been in existence for only thirteen years. Our conclusion also differs from that of Cumby & Glen (1990), who evaluated the performance of international mutual funds and found no evidence that those funds, individually or as a group, perform better than a broad international equity index. One possible explanation may lie in the fact that most indexes are weighted by market capitalization, thus overweighting overvalued stocks and underweighting undervalued stocks, and thereby causing capitalization-weighted indexes to underperform (Hsu, Li, Meyers and Zhu, 2007). While this effect may influence the performance of market indexes in more developed markets, it may be even more severe in presumably less efficient markets such as China s. The results of this study may be of benefit to practitioners investing in mutual funds in China and seeking to achieve higher returns per unit of risk and higher risk-adjusted returns than the market index. We found that this is indeed possible with portfolios of Chinese mutual funds constructed according to MPT and revised periodically. As mutual funds already contain the benefits of diversification, practitioners may simplify considerably the task of constructing optimal portfolios by using mutual funds and thereby reduce the time, research efforts, and administrative costs required to achieve the necessary broad diversification of securities portfolios (Kono et al., 2007). China s mutual fund industry is growing rapidly in terms of numbers and NAV as a consequence of the growth of the economy, securities markets, the benefits of cost effectiveness, tax efficiency, liquidity, and regulatory transparency. As of July 19, 2011, the total number of China s mutual funds stood at 839, and the total NAV at US$355 billion. Since the end of 2009, the number of mutual funds has increased by 50%, although their total NAV has declined by 13.7%. As more mutual funds in China meet the maturity and liquidity criteria used in this study and competition forces their costs down, it may be possible, in the future, to construct even more efficient mutual fund portfolios by following the principles of MPT. According to Carhart (1997) the expense ratios, transaction costs, and load fees [of mutual funds] have a direct negative impact on performance. Other studies have also found that a portfolio of index funds, selected on the basis of low expense ratios and high past returns, outperformed portfolios of index funds selected by investors, thus questioning the rationality of investors selection criteria (Elton, Gruber, and Busse, 2004). With continued growth of China s securities market and increases in competition among market participants, efficiency is likely to approach that envisioned by the semi-strong form of the EMH. In view of this, future research on mutual funds in China could examine the relationship between risk and return, cash flows and performance, and diversified and market portfolios. References Bernstein, P. L. (2006). The Paradox of the Efficient Market.Journal of Portfolio Management, 32(21). Bodie, Z., Kane, A., and Marcus, A. (2010).Essentials of Investments.McGraw-Hill. 98
Carhart, M. M. (1997). On Persistence in Mutual Fund Performance.The Journal of Finance, 3(1). Cai, J., Chan, K. C., Yamada, T. (1997). The Performance of Japanese Mutual Funds.The Review of Financial Studies, 10(2). Chernoff, J. (2002). Physics, Math and Managing Money.Pensions & Investments, 30(9), 3-8. Cumby, R. E. & Glen, J. D. (1990). Evaluating the Performance of International Mutual Funds.The Journal of Finance, 119(2). Elton, E. J., Gruber, M. J., and Busse, J. (2004). Are Investors Rational? Choices Among Index Funds.The Journal of Finance, 59(1). Fabozzi, F. J. & Markowitz, H. M. (2002).The Theory and Practice of Investment Management. Hoboken, NJ: John Wiley & Sons, Inc. Fama, E. F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of Finance. Fama, E. F. (1972). Components of Investment Performance.The Journal of Finance, 27(3). Fama, E. F. (1991). Efficient Capital Markets: II. TheJournal of Finance, 46(5). Harper, J., Madura, J., and Schnusenberg, O. (2006). Performance Comparison Between Exchange-Traded Funds and Closed-End Funds. Journal of International Financial Markets, Institutions and Money. 16(2). Hill, J. M. (2006). Alpha as a Net Zero- Sum Game.Journal of Portfolio Management,32(4). Hsu, J., Li, F., Myers, B., and Zhu, J. (2007). Accounting-Based Index ETFs and Inefficient Markets.Institutional Investor, 2007(1). http://stats.oecd.org/index.aspx?queryid= 556 http://www.sse.com.cn/sseportal/en/c05/p11 10/c15010502_pl1110.shtml. http://www.oecdilibrary.org/economics/data/oecdstat_data-00285-en Israelsen, C. L. (2009). Crash Helmet; Proper rebalancing can help protect a portfolio when markets fall apart. Financial Planning. New York: Feb 2009. 39(2). Kono, P. M., Yatrakis, P. G., Simon, H. K., Segal, S. (2007). Comparing Risk- Adjusted Performance of ETF Portfolios vs. S&P 500 index.financial Decisions (formerly Journal of Financial Decisions), Winter 2007 (2). Malkiel, B. G. (2003). A Random Walk Down Wall Street. New York: W. W. Norton & Company, Inc. Markowitz, H. M. (1952, March). Portfolio Selection.The Journal of Finance,12, 77-91. Markowitz, Harry M. (1959). Portfolio Selection, Efficient Diversification of Investments. New York: John Wiley & Sons, Inc. Markowitz, Harry M. (1991). Portfolio Selection: Efficient Diversification of Investments (2 nd ed.). Cambridge, UK; Basil Blackwell. Miffre, J. (2007). Country-Specific ETFs: An Efficient Approach to Global Asset Allocation. Journal of Asset Management (2007), 8. Poterba, J. M. & Shoven, J. B. (1992). Exchange-Traded Funds: A New Investment Option for Taxable Investors. The American Economic Review, 92(2). Prondizinski, D. A. (2010). Passive versus Active Management of Mutual Funds: Evidence from 2005-2008 99
Period. Nova Southeastern University. Rompotis, G. (2005). An Empirical Comparing Investigation on Exchange Traded Funds and Index Funds Performance, Social Science Research Network (SSRN), Working Paper No. 903110, Dec. 15. Stewart, P. (2006). Behavioral Finance Not to Be Ignored.Trusts & Estates, 145(6). 100