基于DDMS模型的BDTI马尔科夫区制转换和持续期依赖特征分析.pdfVIP

第 33 卷第 1 期 上海海事大学学报 Vol. 33 No. 1 20 口年 3 月 Journal of Shanghai Maritime University Mar. 2012 文章编号 :1672 - 9498(2012)01-∞56-05 基于 DDMS 模型的 BDTI 马尔科夫区制转换和 持续期依赖特征分析 蒋边娜 (上海海事大学经济管理学院，上海 201306) 摘 要:为探索国际原油运价在周期性波动过程中的区制转换和持续期依赖特征，采用具有持续期 依赖特征的马尔科夫区制转换 (Duration Dependent Markov Switch , DDMS )模型分析波罗的海原油 运价指数 (Baltic Dirty Tanker Index, BDTI) 的波动规律.结果表明 :BDTI 具有明显的两阶段区制特 征;国际原油运价指数由收缩期(扩张期)状态向扩张期(收缩期)状态的转换概率随着目前状态持 续时间的增长而不断增大;在相同的持续期水平下，收缩期状态向扩张期状态的转换概率超过扩张 期状态向收缩期状态的转换概率 ;BD白的非对称持续期依赖特征与国际原油运输市场的寡头垄断 市场结构密切相关. 关键词:波罗的海原油运价指数;马尔科夫区制转换;持续期依赖;非对称 中图分类号: U69S.28; FSS2.S; 0211.62 文献标志码 :A Analysis of Markov switch and duration dependence of BDTI based on DD岛I S model JIANG Dina (Sch∞，1 of Economics Management , Shanghai Maritime Univ. , Shanghai 201306 , China) Abstract: In order to explore the state switching and the duration dependence in the cycles of internation? al crude oil freight rate , a Duration Dependent Markov Switch (DDMS) model is employed to analyze the fluctuation of Baltic Dirty Tanker Index (BDTI). The results show that BDπhas obvious characteristics of two-phase states; the probability of the switch from contraction to expansion increases with the increase of the duration time of current state; the probability of the switch from contraction to expansion is larger than that from expansion to contraction at the same duration level; the asymm