The Study Of The Value-at-risk Measurement Based On The VaR Mode

Under the influence of factors such as economic globalization and financeintegration, competing and relaxing control and finance innovation and technologicalprogress, and so on, global financial environment and financial market have changedgreatly. Meanwhile, the fluctuation of the financial market and system riskareaggravated greatly. Risk management has become one of the key competitivenessof the industrial and commercial enterprises and financial institution. The foundationof risk management is to measure the risk. As a new tool of risk measuring andmanagement, VaR technique is used widely since it was born, and has already becomethe major technique to measure market risk abroad at present.There are two problems in the methods of traditional VaR measurement, the first isthe hypothetical of distribution，The traditional VaR metering is carried out under theassumption of normal distribution，but the normal distribution can not describe thecharacteristics accurately，Another problem is that some people often use allhistorical data to analysis when measure the value at risk. Doing this，there may be aheavy workload and influence the accuracy of the results。For these two problems，thisarticles mainly to solve the following two parts：First, discuss the yield distribution. In this paper, it use descriptive statistics andfitting figure to research whether the rate of return in line with the normal distribution.Then use the smooth curve of VaR to compare the effect of the normal distributionand hyperbolic Distribution. At last, we negate the assumption of the normaldistribution.Second, in the assumption of hyperbolic distribution to measure the risk in the bullmarket case. In this paper, the yield data is divided into a rising stage and the data offalling stage. First, measure the value-at-risk in the bull market case (the bear market)using the yield historical data and get a group of VaR value. Then, measure thevalue-at-risk in the bull market (the bear market) case using the data of rising stage(the data of falling stage) and get another group of VaR value. Finally, use the smoothcurve of the two groups of VaR to compare the effect in order to get a more efficientmeasurement of value-at-risk.