【作者】 周金城；

【导师】 陈乐一；

【作者基本信息】 湖南大学 ， 理论经济学， 2014， 博士

【摘要】 农产品价格波动剧烈,对农业生产、人们生活有着不利影响。准确识别我国农产品价格波动特征,对农产品价格发展趋势进行科学研判,可指导利益相关者采取预控措施规避风险,可为相关部门的宏观调控提供微观层面的理论依据与实证支持。如果农产品价格向均衡调整的速度不是固定不变,而呈现非线性调整走势,则传统的线性模型难以捕捉其动态调整行为。按照发现(提出)问题——分析问题(理论与实证分析)——解决问题的思路,研究了以下问题。首先,探寻了我国农产品价格波动过程中的非线性动态特征。采用ARCH LM检验方法、Quandt-Andrews断点检验方法、H-P滤波法,发现我国各类农产品价格调整过程中确实普遍存在群集性、时变性、非线性转换、非对称性等非线性动态特征。其次,探究了农产品价格波动呈现非线性动态调整的原因。农产品价格波动群集的原因在于自然灾害袭击、农业生产者分散决策、间歇性发生的疫情,农业经济主体的异质预期,农产品的资本化发展。非线性转换的原因在于农产品价格形成机制的变化、不同价格条件下农产品价格反应机制的差异。价格非对称传导的原因在于市场势力、价格保护、不对称调整成本、不对称信息、产品特性、运输成本等。再次,主要运用非线性时间序列建模方法刻画了我国9类农产品价格波动的非线性动态特征。通过TGARCH、EGARCH和CGARCH模型,发现我国大米、猪肉、西红柿等9类农产品价格都存在波动群集特征。除了大米价格,小麦、豆油、猪肉、牛肉、鸡肉、西红柿、辣椒、土豆价格的波动还都存在非对称效应。通过SETAR和STAR模型,发现我国大米、猪肉、西红柿等9类农产品价格都存在非线性转换特征,而且它们都是在两个不同的机制之间转换,转换速度也各有不同。同时,发现非对称ARCH类模型、SETAR模型、STAR模型这三类非线性模型的拟合效果都优于线性AR模型。通过MS-VAR模型,发现9类农产品价格波动可显著分为下跌、基本平稳(或微幅上涨)、上涨三个运行机制,各类农产品价格在不同的运行机制的转换概率、平均持续时间一般明显不同。通过TAR、C-TAR、MTAR、MC-TAR模型,发现我国大豆价格与豆油价格、豆粕价格之间的传导都是非对称的；大豆国际市场价格对国内市场价格的传的影响是对称的,豆油国际市场价格对国内市场价格的传导是非对称的。在规模化前阶段,我国生猪价格与玉米价格传导的是非对称的,而在规模化后阶段,生猪价格与玉米价格传导的是对称的。我国猪肉产业链的四个阶段都存在非对称的价格传导关系。最后,提出相关部门在进行农产品价格调控时,应注意各类农产品价格转换或传导的门限值,注意各类农产品价格在不同机制之间相互转换的概率差异,注意各类农产品价格在不同机制(或阶段)的调整幅度与速度,把握价格调节的时机、节奏与力度,有的放矢进行宏观调控。要进一步完善农产品价格、供需信息平台建设,完善农产品市场机制建设,完善农产品价格期货市场,制定有针对性的农业产业政策,完善农产品价格应急机制。更多还原

【Abstract】 Highly fluctuating price of agricultural products is infaust to either agricultural producing or human’s life. Thus, both accurate identification to characteristics of fluctuating price and scientifically estimating trend of agricultural products in China could help related authorities to take actions to prevent it from risks, and provide department concerned’s with microcosmic theoretical basis and real evidence support for their macro-economic Control. While the adjustment to equalization of agricultural products’price is not in a fixed speed, in a non-linear trend otherwise, it will be hard to obtain its dynamic adjustment process by a traditional linear model. The following questions were researched by rising probolms, analyzing probolms (theory and real evidence analysis),and then solving probolms.Firstly, non-linear dynamic characteristics in the process of price fluctuating of agricultural products were researched. Via ARCH LM test,Quandt-Andrews breakpoint test,Hodrick-Prescott filter, it has been found that there are several non-linear dynamic characteristics including clusting, time varying, Nonlinear Regime-Switching and Asymmetric in the process of various agricultural products price adjustments.Secondly,the reasons of fluctuating price of agricultural products appearing in non-linear dynamic characteristics were researched. The reasons of price clustering of agricultural products lies in natural hazard attacking,agricultural producers’ decentralized decision-making, intermittent epidemic, economic subject of farmer’s heterogeneous Beliefs and agricultural products’capitalization.Nonlinear Regime-Switching are owing to the change of agricultural products’price formation mechanism and the difference between every agricultural prices reaction mechanisms under different price terms.Asymmetric price transmission is as a result of market power, price protection, asymmetric adjustment costs, asymmetric information, the product characteristics and transportation costs.Thirdly, nine kinds of agricultural products’non-linear dynamic characteristics were described Through non-linear time series modeling method. Through TGARCH,EGARCH and CGARCH model, we found that9kinds of agricultural products in China including rice,pork,tomatoes have volatility clustering characteristics. Besides price of rice,there also exists asymmetric effect in price fluctuating of wheat soybean, oil, pork,beef,chicken, tomato, pepper and potato, there is also non-linear transition among these9kinds of agricultural products mentioned above, furthermore,they switch from two different systems with different speed. In the meantime,research results show that fitting results of non-linear model (ARCH models、SETAR model、STAR model) are superior to that of linear model (AR model). Through MS-VAR model, it is also found that price fluctuating operation mechanism of these nine kinds of agricultural products’ can be classified to there kinds-declining, generally stable (slightly increasing) and rising, among which these nine kinds appears different transition rate and average duration. Through TAR,C-TAR,MTAR and MC-TAR model, result shows that price transmission in China between soybean,soybean meal and soybean oil is asymmetric; and impact made by soybean’ price in international market on price transmission in domestic market is symmetrical,while the impact on price transmission of soybean oil is Asymmetric. Before the stage of scaling, price transmission between pork and corn is Asymmetric, but after the stage of scaling,it is symmetrical. Within four stages of pork industry chain, there all exist Asymmetric price transmission.Lastly, while relevant departments proceed to control agricultural products price, it’d be better to pay attention to the threshold of price transmission and transition, the different commutative transition rate among different systems and the range and speed of price adjusting in different systems or stages of all kinds of agricultural products, and control the timing, rhythm and intensity of price adjustment to get better targeted micro-control. It is also in need to further perfect agricultural products’ price, construct supply-demand information platform,improve market mechanism of agricultural products and agricultural products’ futures market, establish targeting agricultural policy, and improve emergency response mechanism of agricultural products.更多还原