Valuation accuracy of external and internal property valuations in Germany

ABSTRACT

While all valuers are obliged to act impartially and transparently to reduce bias, the closer relationship between valuers and clients among internal valuations may raise additional concerns regarding the independence of the valuer and hence the objectivity of the result. This paper analyses how internal and external valuations differ in their ability to mirror market prices. The dataset for the analyses contained 4,805 commercial properties in Germany between 1995 and 2013. The first part of the analysis was a Market-Adjusted Valuation and Actual Sale Price Comparison, based on sold properties. It showed that a majority of both valuation types had a valuation error within the acceptable threshold of 15% but that external valuations were on average significantly closer to sale prices than internal valuations. Due to sample selection issues, a second analysis, called Actual Valuation and Fitted Sale Price Comparison, was carried out. Real transactions were used to derive hedonic prices that could be compared against valuations of held properties. The Heckman Correction was used to mitigate sample selection bias. The results showed that both valuation types produced a majority of observations within the set threshold but that external valuations were on average closer to sale prices than internal valuations.

KEYWORDS:

• Internal valuation
• external valuation
• valuation accuracy
• Germany
• Heckman

Acknowledgments

Many thanks to MSCI/IPD for the access to their databank. Many thanks also to Steffen Sebastian from the University of Regensburg and the participants of the ERES conference 2016 for their support and feedback. No research funding from external sources has been used in this research.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. In subsequent parts, actual is used to refer to the original valuation or sale price from the dataset and hedonic or fitted is used to refer to variables derived by regression.

2. If at least 75% of the market rent derives from one of the following sectors, the property should be allocated to that sector. If at least 25% of the market rent derives from a secondary use, there is a ”mixed use” … (IPD, 2011).

3. Translated as: ‘Market value is determined by the price which may be achieved on the day of valuation under regular trading conditions, depending on legal circumstances and actual characteristics, depending on the nature and the location of the property or other asset without regard to any exceptional or personal circumstances’ (§ 194 BauGB).

4. For simplicity only referred to as fitted or hedonic prices in subsequent parts.

Additional information

Notes on contributors

Jan Reinert

Jan Reinert is currently employed at Heitman in Frankfurt, Germany and an external PhD candidate at Regensburg University, Germany.