The Measurement Of Stock Market Systemic Risk Based On The Investment Horizon

In the modern capital market theory, the risk of the securities market usually can be decomposed into two parts of the systematic risk and unsystematic risk. Among them, the systemic risk refers to the part of the changes caused by market factors among the total change in securities gains, which is not to be evaded. While the unsystematic risk is caused by the special factors related to securities, and it can be reduced as much as possible through portfolio diversification. In theory, systemic risk is the only source of effective combination yields uncertainty. A combination with unsystemic risk is invalid. In practice, portfolio managers should eliminate all of unsystemic risk, usually only a systemic risk can reflect the real risk of portfolio. Therefore, the study of systemic risk is of great theoretical and practical significance, and is one of the hot topics in the study of capital market risk pricing theory and practice for nearly30years.In the capital asset pricing model CAPM which is derived on the basis of Markowitz mode by Sharp (1964) and Lintener (1965), Single securities or only a single combination of systemic risk can be described with beta. Beta represents the degree of a single securities or portfolio being affected by the overall market fluctuation. However the CAPM model assumes a single period of investment (i.e., for static portfolio theory), but not makes a specific provision and describe the period length of investment. The accuracy of the investment behavior and the effectiveness of the portfolio analysis is largely dependent on buying and selling point selection, and the length of the investment period depends on investors’ expectations of earnings. The differences of the estimated investment horizon and the investment period of real realize the investment income effect inevitably exist, and the differences will inevitably lead to systemic risk measurement error, so some scholars join the investment period difference to the CAPM model to discuss its influence on systemic risk measure. But, as a result of CAPM model constrained by demanding assumptions, the applicability of the CAPM model in practice is great challenge, Thus with the efforts of scholars, the famous Fama-French three factors model was born. Similarly, in the FF three factors model, beta also represent market systemic risk factors. So, it becomes the concern of scholars both at home and abroad whether the beta parameters of the model also exist the investment horizon effect or not. In this foundation, this article is to invest the research of the systemic risk and the investment period factors.The purpose of this paper is to apply FF investment scale model, which will consider the investment horizon factors to the measurement of systemic risk, to improve the measurement accuracy of systemic risk. In this article, we used copula bayes approach to further analysis the relationship between systematic risk and investment horizon, combining with the empirical analysis to explore the impact of investment period on systemic risk measure. In this paper, the structure of the arrangement is as follows:The first part firstly discussed the research background and significance of securities market systemic risk measurement, and then summarized the existing related research. Then, we pointed out the research targets and contents with pointing out the deficiency of the existing related research, and described the framework of the paper.The second part reviewed the systemic risk comprehensively, including the definition of systemic risk, the source of the securities market systemic risk and it’s measurement method. We examined systemic risk from multiple perspectives, reflecting the need for accurate measurement of systemic risk, and laid a solid foundation for the research of this paper.In the third part, in view of the shortage of the research on risk measurement in the existing securities portfolio model, the FF three factors investment horizon model was put forward on the basis of the FF three factors model, combining with the CAPM investment time limit model. What’s more, we stated the principle and effectiveness of copula bayes approach in this paper, laying a theoretical foundation for empirical analysis in the next part.In the fourth part, on the basis of the FF three factors investment horizon model, this paper analyzed the relationship of investment horizon and the systemic risk parameters beta. In this part, we firstly verify the conclusion that the figures of beta parameter are different when we selected different frequency earnings data in the FF three factors model and there is a certain trend, which reflects the impact of the investment horizon factor in measuring systemic risk. On this basis, we joined the investment horizon to FF three factors model, and estimated parameters using SUR method. By analyzing the parameter estimation results, it is concluded that FF three factors investment horizon model is effective in the research, then we used copula connecting bayes methods to analyze the relationship of the investment scale, i.e. investment horizon factor, and systemic risk parameters.In the fifth part, the last part of this paper, we summarized the empirical analysis conclusions, and put forward some related suggestions from the perspective of investors and regulators.The innovation point of this article is that we joined the investment horizon factor (investment horizon) to the FF three factors model to explore the influence of investment horizon factor on the systemic risk measurement. This article examined the FF three factor investment horizon model in the real diagnosis analysis in the China stock market serviceability, then analysis the relevance of the investment horizon and the systematic risk using copula bayes approach. Due to the limited time and ability, personally think that this paper still has the following several aspects: there may be some errors in our conclusion using monthly data to research the affect of investment horizon factor (investment horizon). We also can use longer cycle of data (such as quarterly or annual data) to study the subject, and compare the conclusion; the model used in this article will be limited by conditions of the original model, which may have a certain gap with the actual situation.