Residential Valuation Research Articles

Machine learning approach to residential valuation: a convolutional neural network model for geographic variation

Geographic location and neighbourhood attributes are major contributors to residential property values. Automated valuation models (AVM) often use hedonic pricing with location and neighbourhood attributes in the form of numeric and categorical variables. This paper proposed a novel approach to automated property valuation using a machine learning model with a convolutional neural network (CNN), fully connected neural network layers with numeric and categorical variables. In this study we compare the results of a fused model, which treat geographical data as an input with the performance of the baseline neural network model with only numerically or categorically represented data. Furthermore, the residential valuation by the proposed fused model was tested with actual sold price data in Greater Sydney, Australia. The study found that the fused model produced valuations with a significantly lower mean absolute percentage error (MAPE) (8.71%) than the MAPE of the baseline model (11.59%). The results show that the fused model with CNN significantly improves the accuracy for residential valuation, reducing spatial information loss by data manipulation and distance calibration.

Machine Learning Applications to Land and Structure Valuation

In some applications of supervised machine learning, it is desirable to trade model complexity with greater interpretability for some covariates while letting other covariates remain a “black box”. An important example is hedonic property valuation modeling, where machine learning techniques typically improve predictive accuracy, but are too opaque for some practical applications that require greater interpretability. This problem can be resolved by certain structured additive regression (STAR) models, which are a rich class of regression models that include the generalized linear model (GLM) and the generalized additive model (GAM). Typically, STAR models are fitted by penalized least-squares approaches. We explain how one can benefit from the excellent predictive capabilities of two advanced machine learning techniques: deep learning and gradient boosting. Furthermore, we show how STAR models can be used for supervised dimension reduction and explain under what circumstances their covariate effects can be described in a transparent way. We apply the methodology to residential land and structure valuation, with very encouraging results regarding both interpretability and predictive performance.

Pricing office rents in Sydney CBD: testing the water on automated rent reviews

PurposeAutomated valuation models (AVMs) are statistical asset pricing models omnipresent in residential real estate markets, where they inform property tax assessment, mortgage underwriting and marketing. Use of these asset pricing models outside of residential real estate is rare. The purpose of the paper is to explore key characteristics of commercial office lease contracts and test an application in estimating office market rental prices using an AVM.Design/methodology/approachThe authors apply a semi-log ordinary least squares hedonic regression approach to estimate either contract rent or the total costs of occupancy (TOC) (“grossed up” rent). Furthermore, the authors adopt a training/test split in the observed leasing data to evaluate the accuracy of using these pricing models for prediction. In the study, 80% of the samples are randomly selected to train the AVM and 20% was held back to test accuracy out of sample. A naive prediction model is used to establish accuracy prediction benchmarks for the AVM using the out-of-sample test data. To evaluate the performance of the AVM, the authors use a Monte Carlo simulation to run the selection process 100 times and calculate the test dataset's mean error (ME), mean absolute error (MAE), mean absolute percentage error (MAPE), median absolute percentage error (MdAPE), coefficient of dispersion (COD) and the training model's r-squared statistic (R2) for each run.FindingsUsing a sample of office lease transactions in Sydney CBD (Central Business District), Australia, the authors demonstrate accuracy statistics that are comparable to those used in residential valuation and outperform a naive model.Originality/valueAVMs in an office leasing context have significant implications for practice. First, an AVM can act as an impartial arbiter in market rent review disputes. Second, the technology may enable frequent market rent reviews as a lease negotiation strategy that allows tenants and property owners to share market risk by limiting concerns over high costs and adversarial litigation that can emerge in a market rent review dispute.

An Investigation Into the Use of “Hybrid” Adjustment Techniques in the Application of the Sales Comparison Method in Residential Valuation

Abstract The sales comparison is the most common and universally accepted method in valuation. Although the theoretical entry point of the method is the same across most continents, its application in practice is varied and often determined by local circumstances. This often necessitates the modification of the method. For instance, while Zambian valuation practice uses this method in residential valuation, its application goes beyond the basic valuation model, incorporating a less known technique called the “reduced floor area (RFA)” technique. The RFA technique is a form of relative importance (weight) concept which assesses ancillary buildings on site relative to the main use; for residential properties this is the main house on site. Despite its obscurity in valuation literature, practitioners find its use acceptable within the dictates of local circumstances. Nonetheless, the lack of documentation means knowledge on the technique is transmitted verbally from senior valuers to graduates, and its application is not consistent across the profession, contributing to variances in the assessed values. This necessitates detailed scrutiny of the technique. Data for the study was collected from the Valuation Surveyors Registration Board (VSRB), a statutory body responsible for licensing valuers and regulating valuation practice. This is the first time the RFA technique is being discussed in a scholarly article.

The wandering energy stars: The challenges of valuing energy efficiency in Australian housing

In the context of curbing greenhouse gas emissions to meet the ambitious but much-needed targets to limit global warming to reasonable levels, residential housing can play a significant role. By enhancing the energy efficiency of residential housing, emission reductions, reduced costs and better health and wellbeing of occupants may be achieved. Despite gradual improvements in home energy efficiency, homeowner engagement could be greatly enhanced by the demonstration of how energy efficiency and energy ratings can be reflected in residential property values. Ascertaining the value of energy efficiency may enable greater investment and potential for increased availability of finance for added energy efficiency options in new and existing housing. Extant research suggests a significant relationship between house values, rental prices and energy efficiency certifications, particularly where mandatory disclosure operates. However, limited research has examined residential valuers’ (appraisers) perspectives and practice considerations.This paper examines the challenges for valuers in identifying and considering energy efficiency initiatives and ratings on residential property value and the capacity of residential valuation professionals to incorporate energy efficiency into the valuation (appraisal) process.The research identified limited means available to valuers in Australia in identifying energy efficiency attributes and ratings; challenges in awareness, knowledge, and understanding of energy efficiency ratings; and, subsequently, limitations in consideration of initiatives and ratings in valuation. This research affirms findings of other Australian and international studies, suggesting mandatory disclosure and certification are essential if energy efficiency is to be considered by the market and reflected in residential valuation.

Online information, data workflows and the role of technology in residential valuation

Technology marches on, and throughout the built environment we are starting to see some of the innovations and developments that have first targeted consumers begin to find their place in the various workflows around more business-focused applications. The availability of property and environmental data is driving workflow opportunities for lenders and their valuation partners as well as influencing the wider property transaction as data moves up the chain and becomes available to drive decisions earlier in the mortgage and conveyancing process to the benefit of consumers. This has significant implications for valuation surveyors; technology is shaping workflows now and will continue to do so.

The Effects of Housing Price Volatility on Mortgage Rates

In this paper, we develop a residential mortgage valuation tree model that incorporates housing price volatility, along with interest rates and interest rate volatility as determinants of mortgage yield spreads (above Treasury yields). When initial loan-to-value (LTV) ratios are low (e.g., 80%), the sensitivity of theoretical mortgage yield spreads to housing price volatility occurs for housing price volatility at annualized rates of 9%–11%. Annual price volatilities of 9%–11% are primarily observed in ‘hot’ residential markets. When initial LTV ratios are high (e.g., 95%), we find the sensitivity of theoretical mortgage yield spreads to housing price volatility begins at annualized rates of 3%–8%, levels consistent with normal real estate markets. These sensitivities are generally irrespective of interest rate levels and interest rate volatilities. Ignoring the sensitivity of mortgage values to housing price volatility can cause lenders to misprice mortgages, and can cause insurance companies to misprice mortgage insurance. This can cause perverse incentives and unintended consequences, with substantial public policy implications including unsafe lending practices, excessive real estate speculation, and systemic risk.

Investigating the UAE Residential Valuation System: A Framework for Analysis

The development of the United Arab Emirates (UAE) into a regional trade, tourism, finance and logistics hub has transformed its real estate markets. However, development has been accompanied by significant real estate prices variance. UAE residential market values (MV) are exposed to fluctuations in capital flows and migration which, in turn, are affected by geopolitical uncertainty, oil price volatility and global investment market sentiment. Internally, UAE residential property markets are fragmented by a complex interplay between tenure, building and evolving location characteristics. In short, the UAE Residential Valuation System (UAE-RVS) confronts multiple challenges to collect, filter and analyze relevant information in complex and dynamic spatial and capital markets. A robust (RVS) can mitigate the risk of unhelpful volatility, speculative excess or investment mistakes. The research outlines the institutional, ontological, dynamic and epistemological issues at play.

Comparable evidence in residential property valuations

When residential property valuation files are critically examined, usually after some time has elapsed since the valuation was completed, the supporting documentation relating to the comparable evidence used to justify the valuation is frequently found to be incomplete or inadequate. This paper reviews the process of selecting and applying comparable evidence with particular emphasis on the need for structured analysis and recording. It contrasts historic practice with current requirements and draws on the RICS Information Paper Comparable Evidence in Property Valuation (2012) and some relatively recent court cases. It suggests that for many residential valuation file documents the term ‘site notes’ is a misnomer and the entire file should more accurately be regarded and described as the valuation record. It suggests adopting a structured approach for dealing with comparables to reduce the risk of potential difficulties when files are later reviewed for any purpose.

Investigating the UAE Residential Valuation System: A Framework for Analysis

The development of the United Arab Emirates (UAE) into a regional trade, tourism, finance and logistics hub has transformed its real estate markets. However, development has been accompanied by significant real estate prices variance. UAE residential market values (MV) are exposed to fluctuations in capital flows and migration which, in turn, are affected by geopolitical uncertainty, oil price volatility and global investment market sentiment. Internally, UAE residential property markets are fragmented by a complex interplay between tenure, building and evolving location characteristics. In short, the UAE Residential Valuation System (UAE-RVS) confronts multiple challenges to collect, filter and analyze relevant information in complex and dynamic spatial and capital markets. A robust (RVS) can mitigate the risk of unhelpful volatility, speculative excess or investment mistakes. The research outlines the institutional, ontological, dynamic and epistemological issues at play.

Residential Estate Valuation Index (REVI): A Consumer Perspective

Motivation/Problem: Real estate has emerged as one of the most profitable investing options in Indian economy, and its valuation plays an important role in the process. Valuations of real estate are a debatable issue due to lack of research and support knowledge available to the stakeholders and market makers, namely, owners, such as sellers, buyers, brokers and valuers, and lenders, such as bankers and financial institutions. In this article, the author seeks to develop a residential estate valuation index, a consumer perspective. Methodology: In the first stage the author has identified consumer-centric factors of residential valuation from literature review and grounded theory. In the second stage SPSS & MS Excel techniques are used to find out consumer-driven weights of each factor and to subsequently develop an index matrix. Output: A residential estate valuation index has been developed. This study attempts to develop a residential estate valuation index for measuring the consumer behaviour with residential estate, residential estate seen as habitat that is an output/product architecture. This index will be helpful for stakeholders, market makers, sellers, buyers, brokers, valuers, bankers and financial institutions in high-involvement decision-making.

How to comprehend Residential Value and Valuation

How to comprehend Residential Value and Valuation

The impact of proximity to cell phone towers on residential property values

PurposeThis paper aims to determine if proximity to cell phone towers influences house prices with a focus on isolating the impact of tower aesthetics on nearby property values.Design/methodology/approachSales transaction data from the Auckland Region during 2005‐2007 were analysed using a series of hedonic models testing various proximity specifications across two populations of cell towers: residential‐only and global (all towers).FindingsThe study could not establish a relationship between cell towers and house prices with the exception of armed monopole towers located in residential areas due to such towers' acute visual disamenity.Practical implicationsThe study's findings can be directly applied to residential valuation practice and can assist government regulators and telecommunication companies in siting new cell towers.Originality/valueThis research provides three distinct methodological improvements: unconventional geocoding that improves spatial accuracy, separate analysis of towers in residential areas that enhances internal validity and inclusion of tower mast data to isolate the impact of tower aesthetics.

The influence of submarkets on water view house price premiums in New Zealand

PurposeThis paper aims to evaluate the impact of submarkets on water view premiums of residential properties and investigate the correlation between submarket view premium and socio‐economic status.Design/methodology/approachOver 53,000 residential sales transactions from 2004 to 2006 are analysed using the hedonic method. The Auckland region is divided into 17 submarkets with similar water view scarcities. The region is analysed along with each individual submarket in order to determine if significant differences in view premiums exist.FindingsThe empirical results indicate that the regionwide model chronically over‐ or under‐estimates view premiums, for example, the regionwide model estimates that a wide water view adds 18 per cent to a home's value while the same view amenity adds only 5 per cent in modest West Harbour but 54 per cent in posh Mission Bay.Practical implicationsThe study's findings can be directly applied to residential valuation practice and in particular mass appraisal systems.Originality/valueThis research fills a gap in the body of knowledge relating to water view externalities by investigating the differing price impacts across submarkets.

Judging a house by its cover

PurposeThe purpose of this paper is to quantify leaky building stigma associated with monolithic claddings, explore how this stigma has likely been amplified by media coverage, estimate the number of affected properties and quantify the collective house price impact on homeowners of monolithic‐clad dwellings in the Auckland region.Design/methodology/approachResidential sales transaction data organised in two subgroups (single‐family houses and multi‐unit dwellings) from 1997 through 2006 are analysed using a series of annual hedonic pricing models to empirically test for the presence of stigma. This is coupled with a descriptive analysis of leaky building media coverage to understand how this coverage may be influencing the stigma.FindingsThe empirical results show that a leaky building stigma exists and is discounting prices of the Auckland Region's monolithic‐clad single family houses by 5 per cent and multi‐unit dwellings by 10 per cent. Approximately 37,500 monolithic‐clad dwellings have been built in the region since 1992 and their homeowners have suffered an estimated $1 billion reduction in property values due to leaky building stigma.Research limitations/implicationsAlthough leaky building stigma primarily relates to monolithic claddings, this stigma reflects elevated weathertightness risks associated with several Mediterranean‐style architectural features.Practical implicationsThe study's findings can be directly applied to residential valuation practice and can assist the New Zealand government more accurately assess the full economic cost of the nation's leaky building problem.Originality/valueThis research provides an initial empirical study on stigma associated with leaky building syndrome. The findings offer direction to further research on other domestic and international housing markets that are experiencing similar stigma phenomenon.

Vertical Tax Equity: An Analysis of Residential Valuation in a Major U.K. City

This study uses a sample of houses sold in 2005 in the Belfast metropolitan area (UK) to test the accuracy of single-family residence valuations and whether the valuation for tax purposes is significantly different than market value as evidenced by sale price. This study examines how accurately the assessed values reflect market value, whether there are differences across statistical distributions and whether differences between assessed value and market value can be attributed to particular property characteristics. The results indicate that there is a tendency to over-value lower priced properties and under-value higher priced property demonstrating regressive vertical tax inequity. In addition, the net economic effect of valuation inaccuracy is shown to be a modest gain in revenue but one that is unequally distributed across house owners.

Towards the 4th generation - an essay on innovations in residential property value modelling expertise

This essay discusses a variety of possibilities for improving current residential property valuation methodology. At present the inertia is unfavourable for innovative research within this field. It could, however, be argued that sustainability and other goals justify the effort to improve this state of affairs. With the focus on a relatively undeveloped European residential valuation academia in terms of innovations in modelling expertize, the proposition is to accommodate some of the following approaches: strict quantitative methods based on parametric hedonic regression; flexible (i.e. non/semiparametric) quantitative methods, that are still formally accommodated within mainstream economic modelling; flexible quantitative methods based on machine learning; rule-based expert systems of valuation; and methods, that use judgemental data (4th generation). Each of the techniques is evaluated in order to identify problems and subsequently appropriate directions for research are proposed. While having its obvious limitations, a method based on interviewing may serve a purpose in certain well-specified problem settings.

Residential valuation behaviour in the United States, the United Kingdom and New Zealand

This paper investigates the valuation processes employed by United States (US), United Kingdom (UK), and New Zealand (NZ) residential valuation experts. Professional appraisers from all three countries participated in a series of valuation experiments that revealed (i) switching from familiar to unfamiliar property settings did not alter valuation behaviour, (ii) valuation behaviour was consistently non-normative, and (iii) the processes employed by US appraisers differed from those of UK experts but not NZ experts. These behavioural differences are attributed to differences in the normative training and business cultures of the countries. Descriptive models of valuation behaviour are developed and compared.

Residential Valuation Theory and Practice

Residential Valuation Theory and Practice

Property valuation and the structure of urban housing markets

Since the 1980s UK academics have promoted the use of multiple regression analysis in property valuation. Recently, however, there has been growing recognition that regression models will be subject to aggregation bias if they fail to accommodate the existence of housing market segmentation (submarkets). In this study, we compare the empirical performance of a standard hedonic house price regression model for the city of Glasgow with a segmented model which recognises the importance of understanding the underlying market structure and, in particular, the existence of submarkets for different dwelling types. The results show that the (weighted) standard error of the segmented model is significantly lower than that of the market wide model. Consequently, we propose a two‐stage approach to the application of MRA techniques to residential valuation. First, following traditional institutional analysis of housing markets, the market should be subdivided into distinct structurally differentiated market segments. These segments can usefully be identified by principal components factor analysis which allows the identification of the most important common components in the housing bundle. Second, separate house price equations should be estimated for each market segment. Although the best‐fit equation may vary from sector to sector this is likely to reflect the behavioural realities of the property market, and will provide the basis for more accurate valuations.