Research Of Portfolio Investment Risk In Foreign Exchange Based On Vine Copula-SV Model

With the pace of our economy and financial internationalization speeding up,foreign trade and investment has become an important part of economic activities in China,the fluctuation of foreign exchange price directly influences the market participants' vital interests.However,with the breakup of "Bretton Woods System",the prevailing of managed floating exchange rate system and the surge of global transnational investment activities,the fluctuation of foreign exchange market strengthens gradually,the risk of foreign exchange investment increases gradually.How to predict the foreign exchange price accurately,and how to measure and control the risk of foreign exchange investment has become the topic of business circles and academics,this paper conducts the risk analysis of foreign exchange portfolio in the framework of risk minimization using CVa R as the risk measure.In view of the inaccuracy of risk measure which is caused by the insufficiency of existing time series statistical model for high dimensional financial time series.this paper chooses heavy tail SV model to describe the fluctuation heteroscedasticity and heavy tail of foreign exchange assets,chooses Vine Copula to describe the complex correlation of four kinds of foreign exchange asset,establishes Vine Copula-SV model for the joint conditional distribution of four foreign exchange assets.Based on the four main foreign exchange assets(USD,HKD,EUR and HKD),The empirical results show that Vine Copula-SV model could describe the fluctuation heteroscedasticity and complex correlation of the foreign exchange well.Considering the Vine Copula-SV model's goodness-of-fit test usually only judge the the overall distribution of fitting situation,however,this article is to measure the risk of portfolio,which is closely related to the fitting capacity of the model's distribution tail mainly.Thus,it is necessary to check the tail fitting ability of Vine Copula-SV model.This paper uses Kupiec failure inspection method to test the Va R which is based on Vine Copula-SV model,The result of the test shows that Va R which is based on Vine Copula-SV model using dual variable method of Monte Carlo simulation passed the test.Finally,we conduct the risk analysis of portfolio in foreign exchange using CVa R as the risk measure and find that the effective frontier of HKD and JPY coincides with USD,HKD,EUR,JPY at the reasonable expected return region.Therefore,while constructing foreign exchange portfolio for four foreign exchange assets,investors could choose the portfolio of HKD and JPY in the framework of CVa R.The risk analysis of foreign exchange investment portfolio which is based on Vine Copula-SV model not only further describes the complex correlation between four kinds of foreign exchange assets but also provides a new idea for the measure of foreign exchange investment portfolio risk,moreover,this has strong theoretical significance for the actual foreign investment.