Revisiting Herding Behavior in REITs: A Regime-Switching Approach

Employing a dynamic model that captures herding under different market regimes we provide novel evidence on the herding behavior of US-listed Real Estate Investment Trusts (REITs). Our sample is extensive and covers the period from 2/1/2004 to 28/6/2013. Estimates of herding behavior are derived using a Markov regime-switching model. The preliminary analysis confirms the existence of three market regimes (low, high and extreme or crash volatility) with transition ordered as `low, high and crash volatility’. Although static herding model rejects the existence of herding in REITs markets estimates of the regime-switching model reveal substantial evidence of herding behavior under the crash regime for almost all sectors. Most interestingly we observe a shift from negative herding behavior during low and high volatility regimes to positive herding behavior under crash regime for almost all REITs sectors.