Estimation and Updating Methods for Hedonic Valuation

Michael Mayer, Steven C. Bourassa (Corresponding Author), Martin Hoesli (Corresponding Author), Donato Scognamiglio

Research output: Contribution to journalArticle

5 Downloads (Pure)

Abstract

Purpose

The purpose of this paper is to investigate the accuracy and volatility of different methods for estimating and updating hedonic valuation models.

Design/methodology/approach

The authors apply six estimation methods (linear least squares, robust regression, mixed-effects regression, random forests, gradient boosting and neural networks) and two updating methods (moving and extending windows). They use a large and rich data set consisting of over 123,000 single-family houses sold in Switzerland between 2005 and 2017.

Findings

The gradient boosting method yields the greatest accuracy, while the robust method provides the least volatile predictions. There is a clear trade-off across methods depending on whether the goal is to improve accuracy or avoid volatility. The choice between moving and extending windows has only a modest effect on the results.

Originality/value

This paper compares a range of linear and machine learning techniques in the context of moving or extending window scenarios that are used in practice but which have not been considered in prior research. The techniques include robust regression, which has not previously been used in this context. The data updating allows for analysis of the volatility in addition to the accuracy of predictions. The results should prove useful in improving hedonic models used by property tax assessors, mortgage underwriters, valuation firms and regulatory authorities.

Original languageEnglish
Pages (from-to)134-150
Number of pages17
JournalJournal of European Real Estate Research
Volume12
Issue number1
Early online date3 Apr 2019
DOIs
Publication statusPublished - 2019

Fingerprint

Boosting
Prediction
Gradient
Robust regression
Mortgages
Hedonic model
Switzerland
Trade-offs
Valuation model
Machine learning
Scenarios
Least squares
Design methodology
Property tax
Authority
Firm valuation
Underwriters
Neural networks

Keywords

  • Hedonic models
  • Machine learning
  • Robust regression
  • Appraisal accuracy
  • Appraisal volatility
  • Mixed effects models
  • R31
  • C45
  • C53

Cite this

Estimation and Updating Methods for Hedonic Valuation. / Mayer, Michael; Bourassa, Steven C. (Corresponding Author); Hoesli, Martin (Corresponding Author); Scognamiglio, Donato.

In: Journal of European Real Estate Research, Vol. 12, No. 1, 2019, p. 134-150.

Research output: Contribution to journalArticle

Mayer, Michael ; Bourassa, Steven C. ; Hoesli, Martin ; Scognamiglio, Donato. / Estimation and Updating Methods for Hedonic Valuation. In: Journal of European Real Estate Research. 2019 ; Vol. 12, No. 1. pp. 134-150.
@article{d71aafe03e7848328211b4bf4d193373,
title = "Estimation and Updating Methods for Hedonic Valuation",
abstract = "PurposeThe purpose of this paper is to investigate the accuracy and volatility of different methods for estimating and updating hedonic valuation models.Design/methodology/approachThe authors apply six estimation methods (linear least squares, robust regression, mixed-effects regression, random forests, gradient boosting and neural networks) and two updating methods (moving and extending windows). They use a large and rich data set consisting of over 123,000 single-family houses sold in Switzerland between 2005 and 2017.FindingsThe gradient boosting method yields the greatest accuracy, while the robust method provides the least volatile predictions. There is a clear trade-off across methods depending on whether the goal is to improve accuracy or avoid volatility. The choice between moving and extending windows has only a modest effect on the results.Originality/valueThis paper compares a range of linear and machine learning techniques in the context of moving or extending window scenarios that are used in practice but which have not been considered in prior research. The techniques include robust regression, which has not previously been used in this context. The data updating allows for analysis of the volatility in addition to the accuracy of predictions. The results should prove useful in improving hedonic models used by property tax assessors, mortgage underwriters, valuation firms and regulatory authorities.",
keywords = "Hedonic models, Machine learning, Robust regression, Appraisal accuracy, Appraisal volatility, Mixed effects models, R31, C45, C53",
author = "Michael Mayer and Bourassa, {Steven C.} and Martin Hoesli and Donato Scognamiglio",
year = "2019",
doi = "10.1108/JERER-08-2018-0035",
language = "English",
volume = "12",
pages = "134--150",
journal = "Journal of European Real Estate Research",
issn = "1753-9269",
publisher = "Emerald Group Publishing Ltd.",
number = "1",

}

TY - JOUR

T1 - Estimation and Updating Methods for Hedonic Valuation

AU - Mayer, Michael

AU - Bourassa, Steven C.

AU - Hoesli, Martin

AU - Scognamiglio, Donato

PY - 2019

Y1 - 2019

N2 - PurposeThe purpose of this paper is to investigate the accuracy and volatility of different methods for estimating and updating hedonic valuation models.Design/methodology/approachThe authors apply six estimation methods (linear least squares, robust regression, mixed-effects regression, random forests, gradient boosting and neural networks) and two updating methods (moving and extending windows). They use a large and rich data set consisting of over 123,000 single-family houses sold in Switzerland between 2005 and 2017.FindingsThe gradient boosting method yields the greatest accuracy, while the robust method provides the least volatile predictions. There is a clear trade-off across methods depending on whether the goal is to improve accuracy or avoid volatility. The choice between moving and extending windows has only a modest effect on the results.Originality/valueThis paper compares a range of linear and machine learning techniques in the context of moving or extending window scenarios that are used in practice but which have not been considered in prior research. The techniques include robust regression, which has not previously been used in this context. The data updating allows for analysis of the volatility in addition to the accuracy of predictions. The results should prove useful in improving hedonic models used by property tax assessors, mortgage underwriters, valuation firms and regulatory authorities.

AB - PurposeThe purpose of this paper is to investigate the accuracy and volatility of different methods for estimating and updating hedonic valuation models.Design/methodology/approachThe authors apply six estimation methods (linear least squares, robust regression, mixed-effects regression, random forests, gradient boosting and neural networks) and two updating methods (moving and extending windows). They use a large and rich data set consisting of over 123,000 single-family houses sold in Switzerland between 2005 and 2017.FindingsThe gradient boosting method yields the greatest accuracy, while the robust method provides the least volatile predictions. There is a clear trade-off across methods depending on whether the goal is to improve accuracy or avoid volatility. The choice between moving and extending windows has only a modest effect on the results.Originality/valueThis paper compares a range of linear and machine learning techniques in the context of moving or extending window scenarios that are used in practice but which have not been considered in prior research. The techniques include robust regression, which has not previously been used in this context. The data updating allows for analysis of the volatility in addition to the accuracy of predictions. The results should prove useful in improving hedonic models used by property tax assessors, mortgage underwriters, valuation firms and regulatory authorities.

KW - Hedonic models

KW - Machine learning

KW - Robust regression

KW - Appraisal accuracy

KW - Appraisal volatility

KW - Mixed effects models

KW - R31

KW - C45

KW - C53

UR - http://www.mendeley.com/research/estimation-updating-methods-hedonic-valuation

U2 - 10.1108/JERER-08-2018-0035

DO - 10.1108/JERER-08-2018-0035

M3 - Article

VL - 12

SP - 134

EP - 150

JO - Journal of European Real Estate Research

JF - Journal of European Real Estate Research

SN - 1753-9269

IS - 1

ER -