Risk Analysis Of Portfolio By Copula-SV Model

Finance risk management developed rapidly in the press of finance turbulence since 70years last century. Globalized financial instruction annex and mixed operation need good financial risk management and asset risk measurement is important in financial theory. The virtue of VaR method is that it generalizes the asset risk to be a simple statistics which can express the core of risk management—the most potential loss. Traditional VaR calculation assumes that single asset income is normal distribution and different asset incomes are linearly dependent. But in reality, asset income is heavy-tailed and maybe each asset income is nonlinear. The difference of Copula is that it modeling to the whole joint distribution, so it provides more useful information, especially to tail of non-normal and asymmetry distribution,thereby, Copula method can improve the precision of VaR.At first, this paper reviews the process of financial risk management development, points out the background and meaning of the subject, then analyses the current situation of the project. In the text, this paper introduces the calculation methods of VaR, Copula theory and its application in finance risk analysis, discusses the parameter estimation of Copula, SV model as well as time-varying Copula.The body of this paper is that it puts forward and establishes Copula-SV and time-varying Copula-SV model to compulate VaR. This paper also compares the VaR between Copula-SV model and Copula-GARCH model. The demonstrations indicate that Copula-SV model and time-varying Copula-SV model are feasible and available in finance risk.Finally, the paper puts forward the meaning of copula theory in finance fields, summarizes the contents and gives the research prospect.