检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:杨政[1] 吴浩成 张靖 马永开[1] YANG Zheng;WU Haocheng;ZHANG Jing;MA Yongkai(School of Economics and Management,University of Electronic Science and Technology of China,Chengdu 611731,China)
机构地区:[1]电子科技大学经济与管理学院,成都611731
出 处:《计量经济学报》2023年第3期828-847,共20页China Journal of Econometrics
基 金:国家自然科学基金(71531003)。
摘 要:本文提出了降维数据预测方法和组合预测相结合的一种新方法,旨在通过模型组合避免选择降维数据预测方法造成的不确定性和降维过程中信息损失.以122个宏观经济因子和14个技术指标因子预测SP500股指超额收益为实证对象,利用单变量预测、单变量预测组合、降维数据预测和降维数据预测组合等进行预测评估.研究发现:第一,样本外拟合优度(R_(OoS)^(2))表明降维数据预测组合优于三种降维预测方法,落后于单变量VOL(1,9)和单变量预测组合.确定性等价收益(CER)显示降维数据预测组合优于单变量组合预测,落后于单变量VOL(1,9)预测和主成分分析预测.两方面综合结果说明降维数据预测组合均衡兼顾了统计上的预测精度和经济上的收益.第二,不同经济周期和不同风险厌恶系数的结果表明降维数据预测组合方法是均衡稳健的.最后,针对组合预测包含的信息提出一个简单的计量检验方法,检验结果表明组合预测能够利用降维数据预测最优的模型的信息,避免单一模型造成的信息损失.We propose a new method combining dimensionality reduction data forecasts and combination forecasts in this study.The purpose is to avoid the uncertainty caused by the method selection of dimensionality reduction data forecasts and the information loss caused by the dimensionality reduction process.This paper forecasts the excess return of S&P 500 index based on a data set consisting of 122 macroeconomic factors and 14 technical index factors.We compare the forecasting results of univariate forecast,univariate combination forecasts,dimensionality reduction data forecasts and dimensionality reduction data combination forecasts.The conclusion of this paper includes three aspects.First,the out-of-sample goodness of fit(R_(OoS)^(2))shows that the dimensionality reduction data combination forecasts is superior to three dimensionality reduction data methods,and is inferior to the univariate VOL(1,9)forecasts and univariate combination forecasts.Certainty equivalent return(CER)indicates that the dimensionality reduction data combination forecasts is better than the univariate combination forecast,and is worse than the univariate VOL(1,9)forecasts and the principal component analysis forecasts.The comprehensive results show that the dimensionality reduction data combination forecasts take into account both forecasting accuracy and economic benefits.Second,the performance of the dimensionality reduction data combination forecasts is robust in different economic cycles and different risk aversion levels.Finally,we propose an econometric test method for the information contained in the combined results.The test results show that the dimensionality reduction data combination forecasts make full use of the forecasting information with the best dimensionality reduction data forecasts,and avoids the information loss caused by a single model.
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.222