Investor heterogeneity and stock market volatility: cross-country evidence from an extended heterogeneous agent model

Abstract

We study how investor heterogeneity relates to stock-market volatility in a cross-country setting using an extended fundamentalist–chartist heterogeneous agent model (HAM). We propose a time-varying heterogeneity intensity measure that combines belief differentials with endogenous investor composition to capture composition-weighted heterogeneity. Using equity index data for 12 developed and emerging markets over 2010–2023 at multiple sampling frequencies, we estimate the model for each market under a harmonized protocol. Across markets and specifications, the belief-gap measure alone is not systematically related to realized volatility, whereas the composition-weighted heterogeneity intensity exhibits a strong and robust positive association with volatility. The results highlight the importance of incorporating both heterogeneous beliefs and participation dynamics in behavioral asset-pricing frameworks and provide a parsimonious metric for characterizing time-varying heterogeneity in international equity markets.