The BSS-FM Estimation of International Assets Allocation for China Mainland Investors

ABSTRACT

Combining the Bayes–Stein shrinkage estimation and the factor model (BSS-FM), we study the international assets allocation issue for investors in China. Our empirical results indicate that the traditional estimation of coefficients for the Markowitz mean-variance model is not stable. Moreover, the BSS-FM method could improve the robustness of estimation, thus leading to a more robust efficiency frontier. Compared to most foreign markets, the Shanghai and Shenzhen stock markets are more profitable but also much more volatile and risky, with a significant correlation. However, their relation to other international exchange markets, except the Hong Kong market, is much lower than the global average. Because low correlation could significantly improve the effects of international diversification, more openness by the Chinese stock market benefits both Chinese and foreign investors. Our simulation indicates that it could greatly reduce investment risk for Chinese investors, if they make their portfolios more internationally diversified.

KEY WORDS:

• Bayes–Stein shrinkage estimates
• factor model
• international assets allocation
• Markowitz mean-variance model

Acknowledgemnt

We gratefully acknowledge partial research support from Yangchen Scholarship (201831837), Social Science Foundation of Guangdong (GD18JRZ01), National Social Science Foundation of China (18BJL065), National Natural Science Foundation of China (71503056) and China Scholarship Council (201608440100).

Notes

1. The data ranking the size of stock exchanges are from a report by the World Federation of Exchanges.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [71503056];National Social Science Foundation of China [18BJL065];Yangchen Scholarship [201831837];China Scholarship Council [201608440100]; and Social Science Foundation of Guangdong (GD18JRZ01).