Indices from frontier, emerging, and developing markets are modeled with a multi-factor capital asset pricing model with a seemingly unrelated regression. Statistically significant coefficients are assembled into an adjacency matrix of common risk exposures, and positional network analysis techniques are employed to understand the structure of these risk exposures. We present Eigenvalue, in-degree, and inverse closeness centralities with interpretations, and show that frontier markets are effectively separated from developed markets through these positional analyses.
展开▼
