The Study On Fluctuation Of International Dry Bulk Cargo Freight Rate Index

By using generalized autoregressive conditional heteroskedasticity (GARCH) model and stochastic volatility (SV) model of international dry bulk freight index volatility study separately, portraying fluctuation characteristics sequence, the results were analyzed and compared. By selecting the data to give the corresponding statistical characteristics, and according to statistical characteristics, the use of three types of GARCH model described in the text and the corresponding three SV model are volatility characteristics BCI, BPI and BSI yields sequence analysis, and fitting result obtained by comparing the parameter estimates compare models.Freight Index for BCI, BPI and BSI return series were modeling, analysis, we found that:the yield on three sequences are present "fat tail" and volatility clustering. On the basis of this conclusion, the continued use of the generalized autoregressive conditional heteroskedasticity (GARCH) model and stochastic volatility (SV) model separately for each return series were fitted, and the goodness of fit of each return series are analyzed, compared with. The results showed that:GARCH class model, GARCH-t model, BCI, BPI and BSI return series are the best fitting effect. SV class model fitting results for L-SV model is optimal. Comprehensive comparison, the BPI return series, SV class model simulation is better than GARCH class model, while earnings for the BCI and BSI series, GARCH-t model is better than L-SV models.