Study And Application On The Risk Metrics Of Securities Investment

Risk measurement and risk management are key topics worldwide. The nature and behavior of security prices is the basis of the effective risk management.Academicians have spent large amounts of both time and resources attempting to understand the behavior of stock prices over time. Regulations are surely interested in the efficiency of the capital market whereas, practitioners and investors hope to exploit valuable information in asset prices.However, the linear methodology of EMH has limitations inherently as it is invalid to capture complicated "patterns" in stock prices. So, a new research trend, from the point of linear and equilibrium to the nonlinear and evolution view, emerges, and the Fractal Market Hypothesis transcend the Efficient Market Hypothesis. After the nature and behavior of the security prices is discovered gradually, the classic risk metrics show its drawbacks and limitations correspondingly. So how to build new risk metrics which reflect the fractal characteristics of the stock market becomes a new issue for the capital researchers and the risk manager. This paper engages a comprehensive research from this view.In order to conquer the drawbacks of the traditional risk metrics, the paper researches on the nonlinear characteristic and it's mechanism of the fluctuation of the securities price from the piont of interaction of inner factors of the capital market, and explains the foundation reason of volatility and risk from the positive feedback mechanism and hoterogeneous investor of the market. And then establish a new risk metrics, which based on the deterministic characteristics of the market as a whole and its randomness as a part. The new risk metrics include two parts , one is the overall risk represented by its Hurst's exponent, the other is partial risk represented by its fluctuation frequency. The risk metrics is not only in the person of the fluctuation behavior of the security price, but also in the person of the diffferent risk preference of investors. The positive research shows that the new metrics is better than traditional metrics both in theory and application.