混频数据有序多分类模型及应用  

An MIDAS-OLogit Model with Applications

作　　者：蒋翠侠[1] 聂玉冰 许启发[1,2] JIANG Cuixia;NIE Yubing;XU Qifa(School of Management,Hefei University of Technology,Hefei 230009;Key Laboratory of Process Optimization and Intelligent Decision-Making,Ministry of Education,Hefei University of Technology,Hefei 230009)

机构地区：[1]合肥工业大学管理学院,合肥230009 [2]合肥工业大学过程优化与智能决策教育部重点实验室,合肥230009

出　　处：《系统科学与数学》2024年第12期3606-3625,共20页Journal of Systems Science and Mathematical Sciences

基　　金：国家社会科学基金一般项目(21BJY255)资助课题。

摘　　要：在有序多分类分析中,混频数据观测越来越普遍.为了解决大频率倍差下的有序多分类问题,文章将混频数据采样(MIDAS)技术与有序Logit (OLogit)模型相结合,构建MIDAS-OLogit模型.MIDAS-OLogit模型运用高频解释变量来预测低频的有序多分类结果,扩大了OLogit模型的应用范围,能够适应大频率倍差下混频数据有序多分类分析.为了验证其有效性,文章进行Monte Carlo数值模拟,结果表明MIDAS-OLogit模型的预测性能优于竞争模型.此外,文章运用MIDAS-OLogit模型,对2008-2021年中国上市公司发行的公司债进行信用评级,结果进一步验证了其分类预测与实时预报的优越表现.In ordered multi-classification analysis,mixed data observation is becoming more and more common.In order to solve the problem of ordered multi-classification under a large frequency ratio,we combine the mixed data sampling(MIDAS)with ordered Logit(OLogit)model to construct an MIDAS-OLogit model.The MIDAS-OLogit model uses high-frequency explanatory variables to predict the ordered multi-classification results of the low-frequency variable,which expands the application range of OLogit model and can adapt to the ordered multi-classification analysis of mixed data under large frequency ratio.In order to verify its effectiveness,Monte Carlo numerical simulation is carried out,and the results show that the prediction performance of MIDAS-OLogit model is better than that of competitive models.In addition,we use the MIDAS-OLogit model to credit ratings the corporate bonds issued by Chinese listed companies from 2008 to 2021,and the results further verify its superior performance of classified forecasting and real-time forecasting.

关 键 词：有序多分类 有序LOGIT模型 混频数据 MIDAS-OLogit模型 信用评价 

分 类 号：F832.51[经济管理—金融学] TP181[自动化与计算机技术—控制理论与控制工程]