Application Of Dynamic Portfolio Strategy In China’s Stock Market

In financial investments, risks and benefits co-exist. Whether in reality or in financial investment research, people are risk-averse and the pursuit of profit. To solve the problem of rational use of investor funds to achieve maximization of return while keeping the risk capped, Markowitz founded the portfolio theory in 1952. After that, with the development of financial markets in developed countries, institutional investors or individual investors in developed countries gradually began to apply portfolio theory in the stock markets, foreign exchange markets, bond markets and other major capital markets. However, Chinese capital market development has lagged behind the developed countries. Consequently, both the academic research and practical applications of the portfolio theory have lagged behind the developed countries too. Nevertheless, in recent years, with the improvement of China’s stock market system, the efficiency of the Chinese stock market has improved, institutional investors start to consider using the portfolio theory the portfolio theory has gradually become an indispensable tool for Chinese investors. This thesis proposes a dynamic portfolio theoretical model for the sector indices of the Chinese stock market. Technically, the Mean-Variance(M-V) model of Markowitz is used as a kernel for this new dynamic model. The parameters of the M-V model is dynamically estimated using the most recent historical data of length L,and then the parameterized model is applied throughout the future period of length T. However, in order to find out the optimal range of these two exogenous parameters(L, T), back test is conducted over historical data. And this back testing has to be repeated dynamically too.The thesis starts with a brief introduction(Chapter 1) to the investment methodologies in the stock markets, the definition and significance of the Chinese Stock Index(CSI) 300, the industrial sector classification in the CSI index. Then Chapter 2 provides a literature review on the portfolio theory from Markowitz up to the present days, focusing on the M-V model of Markowitz, the multi-period portfolio models and optimal investment and consumption models. Chapter 3 describes how the CSI 300 index and the sector indices are defined and calculated. Chapter 4 presents the dynamic portfolio model of the thesis which uses the M-V model of Markowitz as a kernel and dynamically adapts this kernel along the time line. The dynamic portfolio model consists of the M-V model as the kernel and two exogeneous parameters: L– the length of the look-back window for estimating the parameters of the kernel and T – the length of the loof-forward window for applying the calibrated kernel for the near future. Chapter 5 implements the dynamic portfolio theory for managing a dynamic portfolio of the ten stock sector indices. It is demonstrated that the optimal range of(L, T) exist in the Chinese stock market, so the dynamic portfolio of the ten sector indices is feasible and significant. Chapter 6 concludes the thesis.