Research On Volatility Risk Of International Tanker Freight Index Based On SV Models

Recent years, the international tanker freight index which reflected the freight trend of the international tanker shipping market, fluctuate more and more severely, however, the research on volatility risk of freight index in shipping academia mainly focused on the dry bulk market, the traditional risk measurement method was not meet the requirements under the new competent situation, therefore, finding out a optimization model to evaluate the volatility risk of the international tanker freight index in theory is quite significant, so that we can realize its volatility fully and provider better theories for the tanker enterprises'risk control.Stochastic volatility (SV) model is better than the Generalized Autoregressive Conditional Heteroscedasticity (GARCH) model in terms of portraying the financial returns serials'volatility characteristics, therefore, this article is trying to adopt the SV model combined with GPD of Extreme Value Theory to analyze the volatility of the tanker freight index, constructed the dynamic VaR model to measure the volatility risk of the international tanker freight index, the main contents are as follows:Firstly, this article mainly introduced the development, constitution, computation method and main characteristics of the international tanker freight index, pointed out that it has the feature of sharp-peak and fat-tail, volatility clustering, long memory and persistence, and the leverage effect which was reacted differently to the positive and negative impact, established the index basis for the following model comparison.Secondly, it introduced three SV and GARCH models respectively in theory, which were standard model, fat-tail model and leverage model, mainly elaborated the MCMC method for the parameters estimate of SV models and the DIC evaluation criterion, on this basis, it combined the GPD of extreme value theory with VaR method to constructed the risk measurement model, then provided the failure rate way for testing the efficiency of the model.Finally, this article selected enough BDTI and BCTI data to carry out the empirical study, constructed the SV model and GARCH model respectively, and then compared them with each other, pointed out that the SV model had the advantages of describing the volatility of the international tanker freight index, and the conclusion is that the best model is L-SV, meanwhile, if under T distribution, GARCH is the best. After that, it adopted the L-SV model to compute the VaR,and the failure rate test indicated that L-SV model can measure the volatility risk of the tanker freight market well, the results were especial suitable for the crude oil tanker market, and the risk of crude oil tanker market is higher than the product oil tanker market.