基于三步回归滤波模型的系统性金融压力指数构建  

Construction of Systemic Financial Stress Index Based on Three Pass Regression Filter Model

作　　者：李妙 Li Miao(School of Statistics,Jiangxi University of Finance and Economics,Nanchang 330013,China)

机构地区：[1]江西财经大学统计学院,南昌330013

出　　处：《统计与决策》2022年第21期131-135,共5页Statistics & Decision

基　　金：江西省研究生创新专项资金项目(YC2019-B087)。

摘　　要：系统性金融风险由金融体系和宏观经济螺旋式发展内生决定,是金融体系和宏观经济动态交互作用的结果。文章基于资产价格波动性、金融市场流动性、风险价差和估值水平等压力信息,从6个金融子市场选取65个基础指标,运用三步回归滤波模型及偏分位数回归,合成一个能够全面反映经济运行情况的系统性金融压力指数,并采用MS-VAR模型对该指数不同风险期进行识别。结果表明:构建的系统性金融压力指数能准确预测宏观经济冲击分布;该指数大多数时间处于低风险状态,而少数高风险状态与我国几次金融大事件相吻合,是金融风险监测预警的有效定量指标。Systemic financial risks are determined by the spiral development of financial system and macro economy,and are the results of the dynamic interaction between financial system and macro economy.Based on the pressure information such as asset price volatility,financial market liquidity,risk spread and valuation level,this paper selects 65 basic indicators from 6 financial sub-markets,and uses three pass regression filter(3PRF) model and partial quantile regression to synthesize a Systemic Financial Stress Index(SFSI) that can comprehensively reflect the economic operation,and finally establishes MS-VAR model to identify different risk periods of the index.The results show that the constructed SFSI can accurately predict the distribution of macroeconomic shocks,and is mostly in a low-risk periods,while a few high-risk periods are consistent with several major financial events in China,and that the SFSI is an effective quantitative index for financial risk monitoring and early warning.

关 键 词：系统性金融风险 系统性金融压力指数 三步回归滤波模型 

分 类 号：O212[理学—概率论与数理统计] F832.5[理学—数学]