Research On Risk Measurement In Portfolio

During the past few decades, the world’s economic and financial environment has tremendously changed. As a direct result, the overall volatility of the financial markets has increased considerably. Since its accession to the WTO, China’s self-protection of its capital markets continues to suffer from the pressure of further opening up, which will inevitably lead to harsh intensity of the market competition and gradual increase of the overall risk to the financial industry. Therefor, there is a growing emphasis on the management of financial risk and on investment risks.The basic work of financial risk management is how to measure risk. Similarly, the basic work of modern financial pricing theories including investment theories can also be referred to pricing of risk based on measuring risk. With no exception, Modern Portfolio Theory is itself entirely built on risk measurement and pricing.Markowitz’s mean-variance model has for the first time constructed the most basic and complete analysis framework for making investment decisions, thus it has certainly been, over the last50years, one of the main methods to the investment industry,both theoretically and practically, especially in the fund industry. Mean-variance model, however, exhibits a significant defect in the aspect of risk measurement, which naturally gives birth to new risk measurements. Indeed, he quest for a more reasonable risk measurement, since the variance(1952) to the value at risk(1994), has never stopped.This paper examines the major risk measurements, both raised by Western scholars and by financial practice in recent years, from three particular perspectives as follows:the evolution of each measurement, the basic properties of risk, and the coherence of risk measurement. In this paper, a comprehensive and systematic list of the measurements is made by theoretical arrangement from the three perspective enumerated above, and a comparison is also made referred to their respective advantages and disadvantages. Theoretically, the conditional value at risk(CVaR) measurement is relatively a good one. To further identify the application effect of CVaR, a portfolio selection model is constructed and then applied. In fact, obvious proofs strongly support the above conclusion about CVaR.