| Summary: | 碩士 === 臺灣大學 === 財務金融學研究所 === 95 === An analytic method for financial extremal events is presented in this paper. This method is based on Pareto distribution, copula theory, and a point process. This paper is divided into two parts: static analysis and dynamic application. In the static part, we explore some properties of bivariate Pareto copula, including its frailty, mixture, Archimedean structure, tail dependence, convergence property, Monte Carlo simulation method etc., perform static simulations, and explain the applications in finance. In the dynamic part, bivariate Pareto copula is embedded into a discrete time series model. Then the correlation structure, Markov structure, and the joint distribution of n-periods etc. are discussed. The effects of parameters and correlation characteristics are examined in the dynamic simulations.
|
|---|
