Forecasting models for the German office market

Más información sobre el libro

This work addresses the research gap in forecasting models for the German office market, which is primarily explored by commercially oriented organizations. It offers a scientific examination of various rent and total yield forecasting models across nine major German cities, analyzing their applicability and performance while identifying city-specific and forecasting horizon patterns. Following a literature review focused on Anglo-Saxon research, the theoretical foundations relevant to the empirical investigation are established. This includes a discussion of general real estate market characteristics, time series and panel data specifics, common forecasting models, techniques, and performance measures. Key findings from the empirical analysis reveal that ARIMA, GARCH, and multivariate regression models can effectively forecast rent series in the German office market. Notably, GARCH models outperform single ARIMA models for forecasting horizons of three to five years during periods of increased volatility in city rent series. Additionally, univariate models tend to excel in the short term, while multivariate regression models demonstrate superior performance in the long term. Interestingly, certain cities show a consistent dominance of one forecasting model over others.