A Heuristic Robust Approach for Real Estate Valuation in Areas with Few Transactions

FIG Working Week 2017 Surveying the world of tomorrow From digitalisation to augmented reality Helsinki, Finland, May 29–June 2, 2017 The demand for reliable market values in real estate valuation has been increasing over the last decade. One reason is the last subprime crisis, which was caused by a false assessment of the real estate market. The German market transparency is mainly realised by results of analysing purchase prices. Often, the purchases are analysed with a regression approach, which allows for an adequate examination of the real estate market. The results are only reliable in areas with large numbers of purchases. However, in areas with only few transactions the solution of regression is not satisfactory – even if applicable. A small number of purchase prices cannot represent the heterogeneity of the real estate market. Furthermore, the purchase prices may contain outliers. Especially in areas with few transactions, the detection of outliers is a challenging task, because the number of purchases is small. In addition, each detected outlier reduces the sample size. An alternative for dealing with outliers is the application of robust estimation approaches, which handle outliers by down-weighting their influence on the estimated values.