Abstract/Summary

Spatial linkages in returns have not yet received much attention in an asset pricing context, however, they can capture important information about idiosyncratic externalities associated with firms’ holdings. We explain returns of real estate companies by modelling the spatial comovement across their underlying assets. We connect stocks of real estate firms using the location of their property portfolios and show that spatial linkages across real estate assets explain some of the variation in abnormal returns, controlling for exposure to systematic factors and firm characteristics. We propose a trading strategy that exploits the information contained in the spatial linkages of the underlying assets. We show that a long-short hedge that sells the stocks that experience a drop in the price if their connected stocks have also gone down in price and buys the stocks that experience an increase in the price if their connected stocks have also gone up delivers a non-market return of 9.7% per year.