This paper focuses on the analysis of stock index futures prices. In the paper, Imodel the future price with two-factor stochastic modeling and estimate state andparameters jointly with particle filter.In modeling the futures price, I take into account the characteristics of the stockindex futures to reflect its short-term fluctuations around the mean line and long-termuncertainty because of the industry policy or the change of economic environmentwith two factors.For the established state-space model, we use the particle filter, rather than thecommonly used Kalman filtering. Kalman filtering is mainly applied to linear Gaussiandynamic system, for the case of non-linear, extended Kalman filter and unscentedKalman filter approximation corresponding adjustments were made, and the case fornon-Gaussian noise, the Kalman filter estimated effect will be reduced. However,particle filter on the state space modeling without any restrictions on the face ofnonlinear non-Gaussian models, particle filter will appear at ease.Finally, this paper through the CSI300stock index as an empirical object to thefutures price to implement joint state and parameter estimation. |