基于多层因子模型的中国金融稳定综合指数构建  

Construction of China Financial Stability Composite Index Based on Multi-level Factor Model

作　　者：司颖华 董文娟 SI Ying-hua;DONG Wen-juan(School of Statistics and Data Science,Lanzhou University of Finance and Economics,Lanzhou 730020,China)

机构地区：[1]兰州财经大学统计与数据科学学院,甘肃兰州730020

出　　处：《数学的实践与认识》2024年第7期100-112,共13页Mathematics in Practice and Theory

基　　金：国家自然科学基金《贝叶斯潜在阈值TVP-FAVAR模型的构建及其应用》(72063022)。

摘　　要：利用多层因子模型重建中国金融稳定综合指数(Financial stability composite index,简称FSCI),测度了近十年国内外典型事件造成的波动.另外,还采用主成分分析法对FSCI进行了测算,以便凸显多层因子模型测算FSCI的科学性和优势.研究表明:多层因子模型构建出的FSCI显示,“金融失衡”和“金融不稳定”时期与样本期内发生的国内外金融事件基本相耦合。此外,两种方法计算出的FSCI走势特征基本一致,但是多层因子模型比主成分分析法的估计结果稳定.This article uses a multi-layer factor model to reconstruct the Financial Stability Composite Index(FSCI)of China,measuring the volatility caused by typical events both domestically and internationally in the past decade.In addition,principal component analysis was also used to calculate FSCI,in order to highlight the scientific nature and advantages of multi-layer factor models in calculating FSCI.Research shows that the FSCI constructed by a multi-layer factor model indicates that the periods of"financial imbalance"and"financial instability"are basically coupled with domestic and foreign financial events that occur during the sample period.In addition,the trend characteristics of FSCI calculated by the two methods are basically consistent,but the estimation results of the multi-layer factor model are more stable than those of the principal component analysis method.

关 键 词：金融稳定综合指数 多层因子模型 序贯最小二乘法 主成分分析法 

分 类 号：F832[经济管理—金融学] F224