|
As the financial market continuously develops towards international, financial risk becomes more and more global. From the Asian financial crisis, to the U.S. sub-prime mortgage crisis, financial risk has been widely concerned. Measuring risk is the most important step of risk management, only if you have an accurate measure of risk, the risk management would be effective. Thus, it has a great significance to study the method of measuring risk, which would enhance the effectiveness of risk management, develop the risk management mechanism, and even improve the financial system of China.In this paper, we primary discuss Value at Risk and Conditional Value at Risk, i.e. VaR and CVaR. VaR model was developed by JP. Morgan bank, because of its simple form and economic explanation, it was widely used soon. But, when statistical data show that return distribution of financial assets are not normal, VaR becomes an incoherent measure. Therefore, R. Tyrrell Rockafellar and Stanislav Uryasev put forward VaR to CVaR, which is the expectation of loss when the loss exceeds VaR. CVaR is coherent measure, which has excellent statistical characteristic, and takes tail risk into account. Thus, CVaR is a better method of measuring risk.This paper firstly has a study on both VaR and CVaR, and then compares VaR and CVaR on their characters, proving that the CVaR model is better than VaR in statistics. Secondly, we choose 10 shares in stock market to calculate VaR and CVaR, trying to demonstrate that the CVaR model is steadier than VaR. After that, we use minimizing CVaR to optimize the investment and finger out the average-CVaR effective frontier. At last, we give some suggestion of using CVaR model in China and have a prospect for the development of CVaR model.
|