What is the effect of VIX and (un)expected illiquidity on sectoral herding in US REITs during (non)crises? Evidence from a Markov switching model (2014 – 2022)

The study investigates the impact of sector and market-wide illiquidity shocks on herding within US Real Estate Investment Trusts (REITs), on a sub-sector level, including health, hotel, mortgage, residential, retail and Warehouse REITs. Using daily data from January 2014 to February 2022, and consistent with noise trader risk theory, the research confirms the existence of herding behavior within US REITs on a sub-sector level, along with identifying that herding effects are intense on days with negative market returns as compared to days with positive market returns. Assessing the impact of investor sentiments via the VIX index, we find that herding behavior rises with a hike in investor uncertainty and fear. Motivated by the presence of a structural break within our data set corresponding to the Covid-19 outbreak, we use a Markov Switching approach and find significant evidence of sub-sector herding being more intense during the crash regime/covid-19 phase, relative to the expansionary phase. When assessing illiquidity, our results confirm that i) during the expansionary phase only expected illiquidity (market and sector-wide) enhances sub-sector herding within US REITs while ii) during the crash phase only unexpected illiquidity (market and sector-wide) enhances sub-sector herding within US REITs.