Determination of the parameters of automated valuation models for the hedonic property valuation of residential properties: A literature-based approach. Issue 1 (5th December 2017)

• Record Type: Journal Article
• Authors: Metzner, Steffen; Kindt, Andreas
• Is Part Of: International journal of housing markets and analysis. Volume 11:Issue 1(2018)
• Page Start: 73
• Page End: 100
• Publication Date: 2017-12-05
• Subjects: Risk management -- Planning -- Controlling -- Portfolio management -- Hedonic models -- Mass appraisal -- Real estate investments -- Real estate valuation; L85 -- D46 -- R15 -- R30 -- O21 -- G32 -- C50; Residential real estate -- Periodicals; 333.3382205
• DOI: 10.1108/IJHMA-02-2017-0018
• Languages: English
• ISSNs: 1753-8270

Abstract

Abstract : Purpose: The development and testing of the hedonic methods for property valuation require statistical analysis and professional preparation of relevant databases. As a first step, the presumable relevant influencing variables (parameters) have to be determined. Previous studies have shown a large variety of parameters which overlap or deviate from each other. This study aims to collect, systematise and structure different parameters for the further development and testing of hedonic models. Design/methodology/approach: The study comprises a detailed research and deeper analysis of previous studies regarding the hedonic valuation (mainly for residential properties). Flanking areas of examination serve, if they are appropriately suitable, as supplements (e.g. performance analysis and regression). Parameters are extracted from a wide range of literature, compared and integrated into an overall presentation of the results. Findings: In total, 407 parameters were extracted from previous studies on hedonic valuation and performance analysis. Because of various definitions of some parameters in the literature, the current paper combined them in one meaning to avoid misunderstandings in further analysis. Higher-level (global) and lower-level (specific) parameters are contained/described in the final list. The result of this study identifies up to five levels of parameters (within the relevant hierarchy). Research limitations/implications: The parameters have not yet been statistically tested. The relevance of individual parameters has to be tested with relevant corresponding databases and statistical methods (e.g. correlation and regression). Practical implications: To manage larger real estate portfolios, there is a need for regular property valuations. From this perspective, there is a great interest related to the optimisation of the valuation costs, valuation quality and valuation duration. Hedonic methods are considered as an efficient way of performing these valuation tasks. However, further suitable models and parameters are needed. The study describes parameters that can be appropriate for the development of relevant models and creates a structured parameters list, providing the technical basis for the latter. This structured parameter list is substantiated by the evaluation of the existing research. Social implications: Property values represent a significant asset for the national economy and for the individual wealth/welfare. The development of property value in a national economy is also relevant for politics, economy and society. The use of hedonic methods and the knowledge of important individual parameters can contribute towards assessing and substantiating the effect of political decisions on the value of real estate. Originality/value: For the first time, a comprehensive and structured analysis on the value of the relevant parameters used in hedonic methods is performed. Thereby a large number of parameters were identified which question the stability of the results in respective individual studies. In addition, the newly developed hierarchy of parameters can serve as the basis for further research.