【作者】 张立杰；

【导师】 寇纪淞；

【作者基本信息】 天津大学 ， 信息管理与信息系统， 2012， 博士

【摘要】 中国棉花流通体制改革始于1998年，政府希望通过建立棉花需求与供给的宏观调控机制，采用优化棉花生产空间布局、建立国家储备制度、利用进出口贸易调节机制等手段调节国内棉花供给和需求，稳定棉花价格。棉花流通体制改革以来，棉花价格出现了几次大的波动，特别是2008开始的价格波动，在各种因素的共同作用下，棉花价格波动幅度与波动周期与前几次有明显不同。本文从系统科学方法角度，采用计量经济研究方法，以中国棉花价格及其影响因素为研究对象，分析影响中国棉花价格的一些重要因素并模拟预测棉花价格，重点进行了以下几个方面的工作：1、分析了中国棉花流通体制改革以来，棉花价格波动的特征。通过对波动系数等指标的分析，确认了棉花价格具有周期性、季节性波动特征。2、分析了中国棉花价格及价格波动与棉花种植成本及国内国际棉花产量的关系。通过相关性分析确认了中国棉花价格与国内棉花种植成本具有较强相关性，中国棉花价格波动与国内棉花产量变动相关性不高但与国际棉花产量变动具有较高相关性。3、纺织行业纱线产量是国内棉花主要需求方，通过对纱线产量与棉花价格的协整分析来看，纺织行业生产需求不是棉花价格变动的主要因素，棉花价格变动对纺织行业产量影响不大。4、分析了其他纺织材料价格变动对棉花价格的影响。通过对棉花、涤纶、锦纶、锦纶、粘胶及羊毛价格的时间序列进行协整分析，定量的分析了纺织原材料价格间的相关性。5、分析了进出口贸易对棉花价格的影响。通过对1999年至2011年间不同阶段棉花进出口量与棉花价格的关系进行分析，认为通过进出口贸易调节国内棉花供需，稳定棉花价格的政策在2009年前是有效地，2009年后，这一政策的效果不够明显。对比了国内棉花价格CCindex与世界主要棉花价格指数CotlookA（FE）,CCindex对CotlookA的影响大于CotlookA对CCindex的影响。6、研究了短期棉花价格的模拟与预测，通过分析，确认ARIMA(1,1,1)模型对短期棉花价格的模拟与预测效果较好，误差率低于4%，在ARIMA模型的基础上，引入ARIMA与SVM组合模型对棉花价格进行预测，该模型预测结果在均方根误差上优于ARIMA模型。更多还原

【Abstract】 China’s cotton circulation system has been changed since1998, Chinese government wants to balance cotton supply&demand and to stable cotton price in macro level, some policies has been applied in cotton circulation system, which including cotton industry transmission optimization, national reserve system, import and export policies. There have been several price volatility cases after the reform, especially from2009to2010.This thesis focuses on China’s cotton price and its key influenced factors by econometrics method in systematic views, especially on following issues:1. The thesis study the characteristics of cotton price after cotton circulation reform, use some calculated index such as coefficient and fluctuation index to analyze features of China’s cotton price, result shows that China’s cotton price has volatility and seasonal features.2. The thesis study the correlations among cotton price, cotton price volatility, cotton production cost, domestic cotton supply and world cotton supply, result shows that China’s cotton price has correlations with cotton production cost in some extend, China’s cotton price volatility has high correlations with world cotton yield variety.3. Textile industry is the following chain of cotton industry, the thesis study the correlations between cotton price and textile industry output. Co-integration analysis on cotton price and yarn production shows that textile industry demand is not the key factor of cotton price volatility.4. The thesis studies the correlations between cotton price and other alternatives products’ price, such as Polyester, Nylon, Viscose Acrylic and Wool, by using co-integration analysis on time series. Result shows that chemical fiber is not an effective cotton alternative.5. The thesis evaluate China’s cotton import and export policy and its influence to cotton price, result shows that export quantity has nothing to do with cotton price volatility, effects of import quantity to cotton price are difference in various times, there has no influence between cotton import quantity and cotton price from1999to2003, this also happened after2009, but from2003to2009, there did exist co-integration between import quantity and price, amplitude of cotton price from2003to2009is less than other times, which means that form2003to2009, the policy of using import and export to adjust domestic cotton supply and to stable cotton price is effective, but the policy has less effect after2009.Comparison of time series of CCindex and CotlookA shows that influence of CCindex to CotlookA is greater than CotlookA to CCindex.6. The thesis use ARIMA (1,1,1) model to simulates cotton price, results show the proposed model is effective, error rate of forecasting is less than4%,beside ARIMA model, the thesis also use hybrid model of ARIMA and SVM to simulate and forecast cotton price, result shows that the hybrid model is better in root mean square error.更多还原