开放条件下我国粮油期货价格波动的极值风险测度与传导效应研究

Accurate risk measurement is critical to improve the level of market risk quantitative management. The indicator of VaR has been widely applied in the measurement of agricultural market risk. However, the basic application research of VaR estimation techniques is not in-depth enough to efficiently estimate the increasingly complicated market risks. Taking grain and oil futures varieties as object，such as corn, soybean, soybean oil, rapeseed oil and palm oil, the project will conduct research following theory study- method exploring - analysis application pattern. This project, from international perspectives, will effectively analyze the features of price fluctuation in grain and oil futures market, explore the extreme risk incentive factors triggering agricultural futures price fluctuation, establish extreme risk measurement model by applying the semi-parametric estimation based on extreme value theory and quantile theory, compare the flexibilities among different model methods, discuss the extreme risk differences between China and other international grain and oil futures markets, such as U.S, and reveal the change law of dynamic extreme risk transmission effect in grain and oil futures markets at home and abroad. Through the transmission effect analysis of domestic and international grain and oil futures markets, the project can provide a scientific theory basis for the government to enhance risk management of agricultural futures market. What's more, an efficient method to measure market risks proposed by the project, will offer valuable analysis tools of quantitative risk management for the investors.

准确测度风险是提高市场风险量化管理水平的关键。在农产品市场风险测度方面，VaR指标是目前业界的主流，然而有关VaR估计技术的基础应用研究尚不够深入，导致难以有效测度日益复杂的市场风险。本项目以玉米、大豆、豆油、菜籽油、棕榈油等粮油期货品种为研究对象，遵循理论依据-方法探索-分析应用的范式开展研究。基于国际化视角，有效分析粮油期货市场价格波动特征，探讨农产品期货价格波动的极值风险诱因，采用基于极值理论和分位数理论的半参数估计方法，建立极值风险测度模型，比较不同模型方法的适应性，分析中国与美国等国际粮油期货市场的极值风险差异，尝试揭示国内外粮油期货市场间动态极值风险传导效应的变化规律。通过国内外粮油期货市场极值风险的传导效应分析，可以为政府加强农产品期货市场风险管理提供科学的理论依据；研究提出的市场风险有效测度方法，可以为投资者风险量化管理提供有价值的分析工具。

准确测度农产品期货市场极值风险是信息化、市场化和全球化深入发展背景下农业市场主体关注的热点和风险管理决策理论研究的难点。目前业界有关农产品市场极值风险测度模型方法的基础应用研究较为薄弱，对日益复杂的市场风险进行有效测度存在技术瓶颈，难以实现高风险管控前移。本项目研究选择小麦、玉米、大豆、油菜籽、豆油、棕榈油等粮油期货品种为对象，基于近10年来的时序数据分析了我国粮油期货市场的波动特征，深入剖析了粮油期货市场波动风险的内因与外因；采用非参数核密度法科学拟合了主要粮油期货品种价格波动风险的分布形式，解决了传统经验分布并非最优分布的市场波动风险分布形式估计难题；采用极值理论、分位数回归方法、波动性分析方法、历史模拟法和蒙特卡洛模拟等构建了多类高适应性的极值风险测度模型与方法，创新了农产品市场风险管理决策量化研究的技术路径，满足了不同风险偏好者对农产品期货价格波动极值风险的准确测度；针对全球化背景下国内外农业融合互动加强，基于动态极值风险的VaR(Value at Risk)模型，开展了中国与代表性国际粮油期货市场波动的极值风险传导效应研究，将有助于提高管理者应对国际市场重大突发风险冲击的掌控能力。