地缘政治风险与中国原油市场波动预测研究  

Forecasting the Volatility of Chinese Crude Oil Market Based on Geopolitical Risk

作　　者：杨坤 魏宇 李守伟[1] 刘亮 Yang Kun;Wei Yu;Li Shouwei;Liu Liang(School of Economics and Management,Southeast University,Nanjing 211189;School of Finance,Yunnan University of Finance and Economics,Kunming 650221)

机构地区：[1]东南大学经济管理学院,南京211189 [2]云南财经大学金融学院,昆明650221

出　　处：《管理评论》2023年第1期16-31,共16页Management Review

基　　金：国家自然科学基金项目(71971055;71971191,71671145);教育部人文社会科学基金规划资助项目(17YJA790015,17XJA790002,18YJC790132,18XJA790002);云南省高校科技创新团队(2019014);云南省科技计划基础研究重点项目(202001AS070018);东南大学优秀博士学位论文培育基金(YBPY1971)。

摘　　要：近年来频发的地缘政治事件常被认为是原油市场剧烈波动的重要原因。因此,引入Caldara和Iacoviello提出的地缘政治风险(geopolitical risk, GPR)指数,将混频的GARCH-MIDAS模型扩展为GARCH-MIDAS-GPR类模型,分析不同国家、类别和严重程度的地缘政治风险对我国原油市场波动及其预测精度的影响,并从不同长度的波动率预测、中国原油期货推出前后的波动率预测、替代基础模型的使用、波动率预测方向、原油风险预测和投资组合管理六个维度探讨结论的稳健性。更进一步,引入3类宏观经济不确定性和6类经济政策不确定性指标,对比分析多种不确定性信息对中国原油市场波动的预测能力。研究发现,首先,各国、总体以及严重的GPR指数对我国原油市场长期波动均存在显著的正向影响。其次,地缘政治风险指标的纳入在不同程度上提高了我国原油市场波动预测精度,其中3类反映世界整体地缘政治风险水平的GPR指数表现相对较好。最后,相比于常见的宏观经济和经济政策不确定性因素,地缘政治风险指标可以为原油波动率预测提供更多有用信息。本文的研究结论在预测统计精度与波动率的应用层面均具备稳健性。Frequent geopolitical events in recent years are often regarded as a main cause of the intense fluctuations in crude oil market. Therefore, this paper first uses the GARCH-MIDAS-GPR-type models which incorporate geopolitical risk(GPR) indexes to analyze the impacts of the geopolitical risks of different countries, categories and severity on Chinese oil market volatility and the forecasting accuracy of the models. Then, the robustness of conclusions is further discussed from six perspectives: volatility forecasting with different lengths, volatility forecasting before and after the launch of Chinese crude oil futures, alternative basic model, direction-of-change of crude oil volatility forecasts, crude oil risk forecasting and portfolio management. Furthermore, three macroeconomic uncertainties and six economic policy uncertainties are introduced to compare how helpful different uncertainties are for prediction. The empirical results show that, first, the country-specific, overall and serious GPR indexes have significantly positive effects on the long-run volatility of Chinese crude oil market. Second, geopolitical risk indicators contribute to improving the accuracy of Chinese oil volatility forecasts to varying degrees, and the three GPR indexes which reflect the overall geopolitical risk of the world perform better than other GPR indexes. Finally, compared with the commonly used macroeconomic uncertainties and economic policy uncertainties, geopolitical risk can provide most useful information for forecasting crude oil volatility. All the above-mentioned conclusions are robust in statistical accuracy and applications.

关 键 词：原油市场 地缘政治风险 波动率预测 GARCH-MIDAS-GPR 

分 类 号：K901.4[历史地理—人文地理学] F426.22[经济管理—产业经济] F764.1F724.5