本文将研究视角聚焦于富时中国A50股指期货,深度挖掘其价格预测的有效方法,并对其是否具备A股定价因子的属性展开全面探究。研究选取2014年10月31日至2024年12月13日的相关数据,在经过严谨的数据预处理流程后,运用随机森林、长短期记忆...本文将研究视角聚焦于富时中国A50股指期货,深度挖掘其价格预测的有效方法,并对其是否具备A股定价因子的属性展开全面探究。研究选取2014年10月31日至2024年12月13日的相关数据,在经过严谨的数据预处理流程后,运用随机森林、长短期记忆网络(LSTM)以及梯度提升树三种模型对其价格进行预测,结果显示梯度提升树模型在预测精度上表现最为突出。在定价因子检验环节,以沪深300成分股收盘价作为A股的代理变量,通过Fama-Macbeth检验以及滚动回归稳健性检验进行深入分析,最终证实富时中国A50期指对A股市场定价存在显著的正向影响,这一研究成果为投资者的决策制定以及金融领域的学术研究提供了极具价值的参考依据。This article will focus on the FTSE China A50 stock index futures, deeply explore its effective methods for price prediction, and comprehensively explore whether it has the attributes of A-share pricing factors. The study selected relevant data from October 31, 2014 to December 13, 2024. After a rigorous data preprocessing process, three models including random forest, long short-term memory network (LSTM), and gradient boosting tree were used to predict their prices. The results showed that the gradient boosting tree model performed the most outstandingly in terms of prediction accuracy. In the pricing factor testing stage, the closing prices of the Shanghai and Shenzhen 300 constituent stocks were used as proxy variables for A-shares. Through Fama Macbeth test and rolling regression robustness test, in-depth analysis was conducted, and it was finally confirmed that the FTSE China A50 futures index has a significant positive impact on the pricing of the A-share market. This research result provides valuable reference for investors’ decision-making and academic research in the financial field.展开更多
文摘本文将研究视角聚焦于富时中国A50股指期货,深度挖掘其价格预测的有效方法,并对其是否具备A股定价因子的属性展开全面探究。研究选取2014年10月31日至2024年12月13日的相关数据,在经过严谨的数据预处理流程后,运用随机森林、长短期记忆网络(LSTM)以及梯度提升树三种模型对其价格进行预测,结果显示梯度提升树模型在预测精度上表现最为突出。在定价因子检验环节,以沪深300成分股收盘价作为A股的代理变量,通过Fama-Macbeth检验以及滚动回归稳健性检验进行深入分析,最终证实富时中国A50期指对A股市场定价存在显著的正向影响,这一研究成果为投资者的决策制定以及金融领域的学术研究提供了极具价值的参考依据。This article will focus on the FTSE China A50 stock index futures, deeply explore its effective methods for price prediction, and comprehensively explore whether it has the attributes of A-share pricing factors. The study selected relevant data from October 31, 2014 to December 13, 2024. After a rigorous data preprocessing process, three models including random forest, long short-term memory network (LSTM), and gradient boosting tree were used to predict their prices. The results showed that the gradient boosting tree model performed the most outstandingly in terms of prediction accuracy. In the pricing factor testing stage, the closing prices of the Shanghai and Shenzhen 300 constituent stocks were used as proxy variables for A-shares. Through Fama Macbeth test and rolling regression robustness test, in-depth analysis was conducted, and it was finally confirmed that the FTSE China A50 futures index has a significant positive impact on the pricing of the A-share market. This research result provides valuable reference for investors’ decision-making and academic research in the financial field.
