Abstract
The estimation of parameters of real estate return distributions is affected by the tools used to do the work. Consistent with previous studies, investment risk models with infinite variance describe distributions of individual property returns in the National Council of Real Estate Investment Fiduciaries (NCREIF) database over the period 1980 to 2010. Applying Maximum Likelihood Estimation (MLE) to the historic data shows real estate investment risk to be heteroscedastic, but the characteristic exponent of the investment risk function varies more among property types than previously reported. Apartment properties introduced in this study evidence the same basic performance characteristics as office, retail, and industrial properties.