| 大数据与时变环境下的原油期货收益率预测研究 |
 点此下载全文 |
| 引用本文：李海奇1,张妮1 ,陈麒同2.大数据与时变环境下的原油期货收益率预测研究[J].财经理论与实践,2026,(3):53-61 |
| 摘要点击次数: 265 |
| 全文下载次数: 71 |
|
|
| 中文摘要:采用时变提升法，系统研究大数据与时变环境下原油期货收益率的预测问题，并识别具有稳健预测能力的宏观经济变量。实证结果表明，在不同预测步长下，时变提升法在预测精度及相应投资组合所获经济收益方面，均显著优于各类传统时间序列模型与主流机器学习方法。进一步分析发现，相较单纯拓展预测变量的维度，合理刻画现实经济的时变特征对改善预测更为关键，且这一优势在经济衰退阶段尤为突出。“货币与信贷”“消费、订单与库存”等宏观变量组对原油期货收益率具有持续且稳健的预测能力。因此，应加强对原油价格短期波动的动态监测及关键宏观指标的即时识别，提升能源市场风险的事前防范能力与异常波动的预警水平，从而为能源价格风险管理及金融市场宏观审慎决策提供有价值的前瞻信息。 |
| 中文关键词:原油期货 大数据 时变提升法 收益率预测 |
| |
| Forecasting Crude Oil Futures Returns Using Big Data in Time-Varying Environments |
|
|
| Abstract:This paper employs the time-varying boosting method to forecast crude oil futures returns using big data in time-varying environments, and further evaluates the predictive contributions of various macroeconomic variables. Empirical results show that the time-varying boosting method significantly outperforms traditional time series models and mainstream machine learning methods across different forecasting horizons, both in terms of forecast accuracy and the economic gains from the resulting portfolios. Further analysis reveals that, compared to simply expanding the set of predictors, properly capturing the time-varying nature of the real economy is more critical for improving forecasting performance, with this advantage being particularly pronounced during economic recessions. Macroeconomic variable groups such as “money and credit” and “consumption, orders, and inventories” exhibit persistent and robust predictive power for crude oil futures returns. Therefore, it is essential to strengthen the dynamic monitoring of short-term oil price fluctuations and the timely identification of key macroeconomic indicators, enhance the ex-ante risk prevention capacity and early warning capability for abnormal volatility in energy markets, thereby providing valuable forward-looking information for energy price risk management and macroprudential decision-making in financial markets. |
| keywords:crude oil futures big data time-varying boosting return forecasting |
| 查看全文 查看/发表评论 下载pdf阅读器 |
|
|
|
