An Analysis of the Behaviour of Commercial Real Estate Indices

There has been criticism within the real estate professions that the lack of availability of good quality data has led actuarial professionals and academics to ignore real estate for modelling purposes. It has sometimes been asserted that this is one of the reasons for the decline in institutional investment in real estate, as asset allocation experts have ignored the asset class because of the difficulty in modelling it. In general, there is little discussion in the actuarial literature on the subject of real estate indices. The only published stochastic real estate investment model for actuarial use is that by Wilkie (1995) – this was an update of that produced by Daykin and Hey (1990), itself based on advice from Wilkie. In this paper, we discuss the availability of data for real estate stochastic modelling. There have been many developments in the collection, presentation and analysis of real estate data that have not found their way into the actuarial literature. We then review the Wilkie real estate model and use the research work done by real estate finance academics to inform a critique and development of that model. In developing the models different data sets are used, including data from valuation-based and unsmoothed indices in order to find appropriate parameter estimates. The significance (or otherwise) of the parameter estimates is tested for each of the fitted models and the differences between the fitted models examined.