Panel Estimates Of Office Risk Premia In Europe

Significantly falling and rising property yields, which characterized the European real estate investment scene in the last 10 years, raise the key question of whether market pricing has been in line with fundamentals and rationality. Was the magnitude of the yield shifts, in particular the fall in yields by 2007, the result of investor over-reaction which now continues to affect pricing but in the opposite direction? A growing number of researchers argue that time-series patterns in returns are due to investor attitudes. This paper is concerned with measuring this impact, that is what premia office investors were prepared to accept for their investments. A sizeable research literature, and not only in the real estate field, tells us that these premia are time varying reflecting the changing factors and weights investors take into account in estimating them. This empirical paper seeks to determine the relative impact of investor attitudes on yields in the past ten years. Within a model of yields which contains underlying market and capital market variables to control for the cyclical influences on yields we aim to provide empirical estimates of the additional impact on yields and thus risk premia. There is a plethora of explanations how access to debt and behavioural factors were driving pricing. More specifically the paper has a number of objectives: - to estimate time varying risk premia in the office market as spelled out by the panel model; - to examine whether the pattern of risk premia has been similar or distinctive across European markets; - to assess whether this analysis can be used for generating warning signals for over- and under-pricing; - discuss the implications of this analysis for expected risk premia in European office markets. // The paper makes use of the data collated by Property & Portfolio research and it focuses on twenty large office centres in Europe. A panel model is deployed which contains key capital market and real estate market series as determinants of yields. This specification of this model draws on the literature in yield determination. The predictive ability of this fundamentals methodology is assessed. Subsequently the panel model is augmented with common time specific components to capture the cross cities dependency and investor attitudes common to all cities in the sample. Statistically significant time components would support the argument that investors had a further impact on the yield variation.