LAUSR

Abstract This paper investigates the time-varying impacts of macroeconomic factors on portfolio contagion. Mining and using overlapping portfolio data covering more than 600 open-end mutual funds and 1900 stocks in China's stock market from 2007 to 2015, we construct a directed weighted network and calculate its degree of portfolio contagion. The time-varying parameter VAR model with stochastic volatility (TVP-VAR-SV model) using the stochastic model specification search (SMSS) method is applied to explore the impacts. We find that the stock market cycle and the investor sentiment show a more significantly positive time-varying impact on portfolio contagion during periods of stability. The volatility of portfolio contagion is greater during financial environment turmoil.