科研管理

科研管理 ›› 2017, Vol. 38 ›› Issue (12): 144-154.

• 论文 • 上一篇    下一篇

基于经验模态分解的中国出口集装箱运价指数波动特性

汤霞1,2,匡海波2,孟斌2,冯文文2   

  1. 1.珠海城市职业技术学院 经济管理学院，广东 珠海519090；2.大连海事大学 综合交通运输协同创新中心，辽宁 大连116026
  • 收稿日期:2017-05-11 修回日期:2017-10-25 出版日期:2017-12-20 发布日期:2017-12-19
  • 通讯作者: 匡海波
  • 基金资助:

    国家自然科学基金项目：自贸区港口生态圈演化、平衡及评价机制研究(71672016，2016-2020);长江学者和创新团队发展计划：港口协同发展与绿色增长(IRT13048，2013-2017); 河北省交通运输厅重点项目：河北港群绿色增长发展模式研究 (ZJT2015037，2015-2017);中国物流学会课题：广东自贸区对珠海港的影响与对策研究（2016CSLKT3-154，2016-2017）;广东省高职交通运输教指委课题：协同创新理念下高职港口与航运管理专业课程体系构建研究——基于工作过程系统化方法（jtysglzw201502，2015-2018）。

Dynamic volatility of China’s containerised freight index based on EMD

Tang Xia1,2, Kuang Haibo2, Meng Bin2, Feng Wenwen2   

  1. 1.School of Economics and Management, Zhuhai City Polytechnic Institute, Zhuhai 519090, Guangdong, China; 
    2. Collaborative Innovation Center for Transport Studies, Dalian Maritime University, Dalian 116026, Liaoning, China
  • Received:2017-05-11 Revised:2017-10-25 Online:2017-12-20 Published:2017-12-19
  • Contact: kuang-haibo kuanghaibo

摘要： 中国出口集装箱运价指数（CCFI）能有效反映集装箱海运市场变化，是航运企业经营决策、政府部门航运产业政策制定的重要参考。本文基于经验模态分解（EMD）方法数据驱动的特征及处理非线性、非平稳、多尺度数据序列的优势，将CCFI有效分解为若干本征模态（IMFs）；然后提出综合考虑各IMFs的T检验P值和波动频率对IMFs进行重构，并基于希尔伯特变换和统计分析探讨了各重构项经济内涵；最后从内在影响因素角度探讨了CCFI长期发展趋势、重大事件影响、季节波动、短期市场不均衡的波动特性。研究表明：CCFI长期呈缓慢下降趋势，由市场船舶运力供给大于货运需求决定；CCFI的暴涨暴跌由金融危机等重大事件引起，影响程度大，持续时间长；CCFI的季节波动受季节性生产等因素影响，呈较规则正弦波动，高峰在9、10月份前后，低谷在3、4月份左右；CCFI短期市场不均衡波动受油价、汇率等因素影响，影响程度小，持续时间短；CCFI波动存在约4年的大周期、1年的小周期。

关键词: 中国出口集装箱运价指数, 波动特性, 经验模态分解, 运价预测

Abstract: The volatility of China containerised freight index (CCFI) can reflect the fluctuation of shipping market effectively. It is an important reference for the business decision-making of shipping enterprises and the policy making of shipping industry in government departments. In this paper, CCFI has been decomposed into several independent intrinsic modes (IMFs) effectively by empirical mode decomposition (EMD) method for its characteristics of data-driven and effectiveness on the nonlinear, non-stationary and multi-scale data processing. Then, composition on IMFs is proposed by considering P-values of T-test and fluctuation frequencies, and the economic connotation of each component is discussed based on the Hilbert transform and statistical analysis. Finally, we study the long-term trend, the effect of a significant event such as financial crisis, seasonal fluctuations and short-term fluctuations caused by market disequilibrium of CCFI from the perspective of intrinsic factors. The conclusions are as follows. CCFI has been declining slowly for a long time, because the capacity of transportation is greater than the demand for freight. A sharp rise or fall of CCFI has been caused by significant events such as financial crisis with big impact and long duration. Its seasonal fluctuation is affected by seasonal production, which shows a regular sinusoidal fluctuation. The peak is around September and October, and the trough is around march and April. Its market disequilibrium for short is caused by oil price, exchange rate and other factors, which have no serious impact. The fluctuation of CCFI has a large cycle of about 4 years and a small cycle of 1 year.

Key words: China’s containerised freight index, volatility, empirical mode decomposition, freight forecasting