多尺度系统性风险与宏观经济的溢出效应研究

Abstract

It is of great significance to comprehensively explore the spillover effects between systemic risk and macroeconomy factors in the context of coordinating high-quality development and high-level security. This paper first measures a systemic risk index and then employs ensemble empirical mode decomposition (EEMD) to decompose systemic risk into three frequency-domain components, that is the high frequency, the medium frequency, and the low frequency. It measures the LSTVAR-DY spillover index, constructs a two-regime nonlinear spillover network that incorporates systemic risk components of different frequencies and macroeconomic variables, and explores the nonlinear spillover effects between systemic risk and macroeconomy from a frequency-domain perspective to identify the sources and directions of spillovers. Additionally, a nonlinear Granger causality test is conducted to further confirm the nonlinear causal relationships among variables. The results show that systemic risk in China exhibits multi-scale frequency-domain characteristics. There exists a significant bidirectional spillover effect between systemic risk and macroeconomy, with asymmetric total spillover effects across different regimes. In the economic expansion and downturn regime, systemic risk serves as the primary net spillover source, whereas in different systemic risk regimes, both systemic risk and macroeconomic factors act as significant net spillover sources. The low-frequency component of systemic risk is a crucial risk source within the economic-financial network, while the role of the medium-frequency component shifts across different regimes. The high-frequency component consistently acts as a net spillover receiver across all regimes. The research conclusions provide new empirical evidence for improving macro-control policies aimed at balancing growth stabilization and risk prevention.