infonet-economy

This paper uses a sample of German commercial and residential property returns to estimate parameters for stable distribution functions. A quantile-based estimation methodology is used to examine distributions of income, capital growth, and total returns. There are controls for the effects of property characteristics and for possible differences between appraisal-based and transaction-based return distribution parameters. The findings reveal that the assumption of normality in return distributions can be rejected for virtually all subsamples of all property types, and for all years from 2000 to 2009. However, income return distributions are found to be less leptokurtic than those of capital growth. Building characteristics do not have a strong influence on distribution parameters, while transaction-based returns do differ from appraisal-based returns for retail and residential properties.