Multivariate stochastic volatility for herding detection: Evidence from the energy sector

The paper proposes a multivariate asymmetric stochastic volatility approach, allowing for common factors that detect and measure herding behavior conditional on the stylized facts of asset returns and another factor that captures non-herding behavior. Applying our approach to the constituents of the S & P 500 energy sector in periods of high uncertainty, the findings reveal a wealth of information on herding detection related to asset returns' co-movements and volatility encountered by the energy sector. We also examine to what degree macroeconomic indicators' uncertainty influences the common factors on herding detection. We conclude that stylized facts of energy assets experience significant changes, arising from the increased systemic market risk and crude oil prices that are exposed to.

Automated summary

This paper proposes a multivariate asymmetric stochastic volatility approach to detect and measure herding behavior in asset returns. Applying this approach to the constituents of the S&P 500 energy sector, the findings reveal valuable information about herding detection and its relationship with asset returns' co-movements and volatility. The paper also examines the influence of macroeconomic indicators' uncertainty on the common factors of herding detection.