Common features in UK commercial real estate returns

Abstract

This paper examines the degree of short run co-movement in UK commercial real estate returns. The hypothesis is that a large fraction of the fluctuations may result from a small number of core disturbances that are transmitted from one region to another and from one property type to another. It adopts an approach from the business cycles literature which uses common features, and their complement, common cycles. The empirical modelling follows the work by Tiao and Tsay (1985), Engle and Kozicki (1993), Vahid and Engle (1993, 1994) and Engle and Issler (1995). Thirty-nine regional rates of return series are used, covering retail, office and industrial real estate for the economic planning regions of the UK. For the series that exhibit serial correlation, bivariate and multivariate common feature/common cycle tests are performed and reduced dimensional VAR models are estimated. The results suggest a single common cycle for the retail and industrial markets and three common cycles in the office market. The existence of common features is important in the study of commercial returns as, when common features exist, uncovering these enhances understanding of the returns generating process. Moreover, this assists in understanding the scope for diversification within real estate portfolios. The results offer important insights into the links between regional real estate markets, not least because they are consistent with previous studies using very different approaches.