Academic journal article The Journal of Real Estate Research

A Comparative Analysis of the Accuracy and Uncertainty in Real Estate and Macroeconomic Forecasts

Academic journal article The Journal of Real Estate Research

A Comparative Analysis of the Accuracy and Uncertainty in Real Estate and Macroeconomic Forecasts

Article excerpt

(ProQuest: ... denotes formulae omitted.)

Amongst the large number of studies that consider the accuracy of professional forecasts, few have considered how inaccuracy and/or uncertainty in one set of forecasts impacts another. Rather, the majority focus on either empirical tests of forecast accuracy or consider those factors and circumstances that may help to explain and contribute to the differences in forecast accuracy. This is despite the fact that forecaster's frequently have to rely on forecasts of explanatory variables during the estimation of their forecasts. Indeed papers such as Oller and Barot (2000), Hendry and Clements (2003), and Stekler (2007) highlight data as one of the potential causes of forecast inaccuracy.

Real estate is an interesting asset class to use to investigate this issue. The nature of it as both a real and investment asset provides real estate with quite distinct characteristics, in comparison to most mainstream asset classes. In particular, real estate has direct linkages and connections with the macroeconomy. The direct link via the demand for the asset from an occupier perspective means that measures of economic performance are key drivers of real estate, in particular rents. The importance of this linkage is clearly illustrated by the extensive use of macroeconomic variables in econometric models of rental markets.1 It naturally follows that macroeconomic variables are important in a forecasting context. These close linkages provide the motivation and the points of interest for this paper. Using professional forecasts for market rents, capital and total returns, and a variety of macroeconomic variables for the United Kingdom, we compare the relative accuracy of the forecasts. We then extend that initial analysis to consider whether heightened inaccuracy and uncertainty in professional forecasts of macroeconomic series have a corresponding impact on the real estate forecasts.

While many researchers have evaluated the accuracy of macroeconomic series (e.g., Batchelor and Dua, 1991; Zarnowitz and Braun, 1993; Laster, Bennett, and Geoum, 1999; Oller and Barrot, 2000; Loungani, 2001; Batchelor, 2007; Boero, Smith, and Wallis, 2008; Campbell and Sharpe, 2009), a relatively small number have considered the accuracy of published real estate forecasts (e.g., Ling, 2005; McAllister, Newell, and Matysiak, 2008; Bond and Mitchell, 2011; Matysiak, Papastamos, and Stevenson, 2012; Papastamos, Matysiak, and Stevenson, 2015). This has been primarily due to a lack of availability of published professional forecasts. The majority of these researchers have considered the same U.K. data as analyzed in the current paper.2 The data consists of forecasts collated by the Investment Property Forum (IPF), a U.K. industry organization. Since the late 1990s, the IPF has surveyed a variety of property consultancies, research firms, fund managers, and financial institutions and asked them to provide forecasts for the U.K. All property indices for rents, capital values, and total returns are from MSCI/IPD (Investment Property Databank). The forecasts are therefore of aggregated portfolio indices across all property segments. A number of common elements do reveal themselves in the research that has analyzed the IPF data. Firstly, that the forecasters tend to herd and to be conservative (McAllister, Newell, and Matysiak, 2008; Matysiak, Papastamos, and Stevenson, 2012; Papastamos et al., 2015). Due to the extended time frame examined, Papastamos, Matysiak, and Stevenson (2015) are able to analyze accuracy across the extreme market movements observed during the last real estate cycle. Their results confirm the conservative nature of forecasters and the smoothing that results from a tendency to underestimate growth rates during strong market conditions and overestimate when the market is performing poorly. This is consistent with considerable empirical evidence from studies that examine macroeconomic forecasts (e. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.