Double-jump diffusion model based on the generalized double exponential distribution of the random jump and its application.

The generalized double exponential distribution is proposed by combining the asymmetric and biased double exponential distributions, which could fully disclose the features of bias, asymmetry, and steep-peak and heavy tails in financial markets. By referencing to the KDJ model, the generalized double exponential jump diffusion model (GDED-KDJ) is proposed and analyzed. Taking into account the heteroskedasticity and volatility jump, and the double jump diffusion model proposed by Eraker, we further extend the GDED-KDJ to the generalized double exponential double-jump diffusion model, i.e., GDED-SVIJ model. Then, we analyze the features of the new model, such as the general, biased, asymmetric and steep-peak and heavy tails, which also present its superiority. Also, we study the conditions likelihood function and MCMC iterative algorithm of the new model. At last, by the empirical study about three metal futures in SHEF, the feasibility and superiority of the new model are showed and proved. [ABSTRACT FROM AUTHOR]