Measuring The GEM Risk Based On Price Range-GARCH Model

Recently,some China's companies of small and medium size are growing rapidly.To solve problems shown in the financing process,the central government launched the GEM(Growth Enterprise Market)in 2009.Whereas,GEM is filled with relatively high risk,and is vulnerable in the real operation.For the reasons that the GEM consists of small company scales and these public companies are not capable enough to resist risk of the market.it is enormously necessary to effectively predict the risk in GEM.To achieve this,this paper concentrates on the risk in GEM by a method of determination of quantity and put forward some practical policy suggestion accordingly.This article separate fitting progress using Normal Distribution,T Distribution and GED,measuring the distribution of the index yield of GEM,are tested and compared to more exactly.What's more,a new range-GARCH model is constructed.With the rate of fluctuation gotten from the three distribution-fitting models mentioned above separately,VaR valued is calculated correspondingly.Then we come to a conclusion that range-GARCH has the lowest failure rate and is more effective than the traditional GARCH model.Though a lot of study,we find that the use of GARCH model to calculate the conditional variances,mostly in the closing price,while ignoring the highest and the lowest price for this same fluctuations containing important information,or directly useing the highest and the lowest price to build range——garch model,such as AV,CARR model,without considering the closing prices.This article is trying to put the closing price and the highest and the lowest price together to estimate conditional variances,looking to whether can improve the prediction accuracy.This article is mainly separateted in two parts:first,the range-GARCH model was constructed.According to previous research conclusion:price is an effective proxy variables of volatility,to introduce the price is very poor and GARCH model,improved the basic variables contained in the information,to improve the conditions of variance estimation precision.Second,according to the principle of the calculation of VaR,using range-GARCH model to calculate the VaR,and compared with the traditional GARCH model to calculate the value of contrast.Introduction found very range-GARCH model can better reflect the growth enterprise market yields in risks.By studying the two conclusions:First,the GEM yields distribution mainly show the obvious peak fat-tailed features and cluster effect,but there is no leverage.Second,introduce the price extremely to GARCH model,found that poor price factor coefficient is significant,that price is very poor,as the price fluctuations to upper and lower limits of issue of yield under the influence of volatility.And by comparing the model of the AIC value and the likelihood function,found that poor-GARCH model to the model has better fitting,and explain the effect is better.Third,through the GARCH model and range-GARCH model to calculate the volatility,and.through the VaR calculation principle,calculated the growth enterprise market of the value at risk(VaR),founding that the use of range--GARCH model to calculate the value at risk(VaR)can better reflect the growth enterprise market risks.By Kupiec likelihood ratio test,and also illustrates the ranger-GARCH model is more effective.Compared with the similar articles,this article has the following features:First of all,quantitative measure the risk of financial markets,has become a trend in today's society.But the GEM as a new type of financial market,price fluctuation is bigger,also,able to quantify the risk of measure the growth enterprise market is very important,and domestic studies on the GEM risk is mostly qualitative,but,quantitative measure is less.Secondly,in terms of methods,this paper will be introduced to the range-GARCH model,and on this basis to measure the growth enterprise market risk in our country.