Data description of aggregate investor trading.

This paper suggests an alternative approach to measuring systemic risk in financial markets by examining the interconnectedness among heterogeneous investors. Utilizing variance decomposition and a trading database from the Korea Stock Exchange spanning 2002-2018, we find that systemic risk, as quantified by total connectedness based on microlevel investor activity, intensifies during both domestic and global financial crises. In addition, our analysis indicates that retail investors, often termed noise traders, are pivotal contributors to the propagation of financial shocks. We also find that portfolios constructed by the sensitivity of total connectedness yield additional returns. This study could enhance our understanding of the contagion effect by incorporating the investor perspective, and the findings could offer valuable insights for policy-makers and regulators.