Purpose – Many studies have shown that real estate returns are not normally distributed. Instead, the returns were found to be leptokurtic and slightly skewed with fat tails on both ends of the distribution. This specific shape of real estate distributions complicates tasks such as measuring property risk, forecasting real estate markets, and optimizing real estate portfolios. Some of the reasons for the non-normality have been documented, but they cannot fully explain the distribution curve. In this paper we focus on time-related aspects such as the maturities of lease contracts and vacancy periods and their effect on real estate returns. The aim of our research is twofold: (1) providing some new reasons for non-normal distributions of real estate returns and (2) giving some guidance for modelling real estate risk.

Design / methodology / approach – A unique data set of lease contracts from a German residential real estate portfolio is employed. Distributions of contract maturities and vacancy periods are analysed with distribution fitting routines.

Findings – Contract maturities and vacancy periods are not normally distributed. These results shed some light on the return distributions in favour of compound distributions. Additionally a subset of lévy processes is provided for modelling real estate risk.

Research limitations / implications – Only residential real estate in Germany is examined. Other sectors and countries are interesting topics for future research.

Originality – This paper may provide evidence in favour of compound distributions for direct real estate investments. Furthermore, some guidance is provided for modelling real estate risk. This is valuable for both research and management practice.