VaR Algorithm Of Delta-Normal And Historical Simulation

With the outbreak of "subprime mortgage crisis "and" European debt crisis.", Financial risks, especially the importance of market risk measurement and control,is particularly prominent. VaR, which is approved by the Basel Committee, is the most basic measurement tool of market risk. This research is based on High-frequency data of stock index futures, with methods of Delta-Normal and Historical Simulation, to build the forecast model and back test it, and finally find the reasonable VaR model on stock index futures.There is an introduction about the background of this article and the importance of VaR in finance risk management in preface.The literature review section briefly describes the progress of VaR research. The theoretical analysis section first introduces the basic concept of VaR, then introduces the algorithm of Delta--Normal method and historical simulation method, finally brings in the VaR back test.The empirical analysis elaborated on Delta--Normal VaR model and Historical VaR model, then details the back test of Failure rate model and Independence model. If passed the test, the VaR model is proved to be a reliable model else the VaR model is needed to be improved. The conclusion section describes the effective VaR model, as well as some new ideas we found in the process of modeling and model checking.Reasons for using data Frequency and time scope. In order to improve the versatility of the VaR model, we using the data from April16,2010to December30,2012,because there are VaRieties of market structure. High-frequency data for1minute,3minutes,5minutes,10minutes,15minutes,30minutes,45minutes of data are commonly used in daily transactions. We select the most commonly used High-frequency data of1minute,3minutes,5minutes and15minutes.The modeling ideas. First modeling VaR based on classical assumptions, then back test it to analyze the effectiveness of the model. The model needs to be improved if it cannot pass through the back--test; If the model can not pass through a special case of event independence test, you need to carry out further improvements to the model.Only when the model is acceptable after back test, it is proved to be a reliable and effective model.Innovation of this paperPreviously VaR articles were mainly concentrated on the marginal VaR and the composition of VaR. There are few of papers on high frequency futures data analysis of the effectiveness of the VaR model especially on stock index futures.Data innovation. The data sources of existing research are usually derived from the Mandarin finance or trading blazer, and the accuracy of the software itself is the problem of these data. The data source used in this paper is the API provided by the Shanghai Futures IT comprehensive transaction platform. Thus the accuracy of the data used in this article is better than others.Model innovation. Previous studies were mostly applied to stock market, the stock market is a one-way transaction, thus the risk analysis faced by equity investors is prices went down. Futures is a two-way trading mechanism, so this is a long VaR model to measure the downside risk faced by investors holding long positions in futures price, and a short VaR model to measure the risk of investors holding short positions in futures prices.