This dissertation examines the influence of property type and country factors on the cross-sectional volatility and correlation structure of national real estate securities index returns. Using constrained, cross-sectional regressions to extract pure country and property type factor returns for real estate indices covering seventeen countries and four property types from February 1990 through November 2002, I find that property type specialization explains very little of the variance of national real estate securities index returns. Because property type effects are so small, country diversification is a more effective tool for achieving risk reduction than property type diversification. However, the relative importance of country effects is declining over time.; This dissertation also evaluates contagion among national real estate securities indices. Applying a logistic regression to daily returns of the 250 most liquid property companies in twenty-one countries during the period January 3, 2000 to November 29, 2002, I measure the coincidence of extreme return shocks within a region and across regions. I characterize the extent of contagion and its determinants. |