Marcelo Cajias and Joseph-Alexander Zeitler
The paper employs a unique online user-generated housing search dataset and introduces a novel measure for housing demand, namely “contacts per listing” as explained by hedonic…
Abstract
Purpose
The paper employs a unique online user-generated housing search dataset and introduces a novel measure for housing demand, namely “contacts per listing” as explained by hedonic, geographic and socioeconomic variables.
Design/methodology/approach
The authors explore housing demand by employing an extensive Internet search dataset from a German housing market platform. The authors apply state-of-the-art artificial intelligence, the eXtreme Gradient Boosting, to quantify factors that lead an apartment to be in demand.
Findings
The authors compare the results to alternative parametric models and find evidence of the superiority of the nonparametric model. The authors use eXplainable artificial intelligence (XAI) techniques to show economic meanings and inferences of the results. The results suggest that hedonic, socioeconomic and spatial aspects influence search intensity. The authors further find differences in temporal dynamics and geographical variations.
Originality/value
To the best of the authors’ knowledge, it is the first study of its kind. The statistical model of housing search draws on insights from decision theory, AI and qualitative studies on housing search. The econometric approach employed is new as it considers standard regression models and an eXtreme Gradient Boosting (XGB or XGBoost) approach followed by a model-agnostic interpretation of the underlying effects.
Details
Keywords
Brano Glumac and Thomas P. Wissink
This paper aims to report on homebuyers’ preferences and willingness to pay for installed home photovoltaic systems. Their influence on the market position of a dwelling is…
Abstract
Purpose
This paper aims to report on homebuyers’ preferences and willingness to pay for installed home photovoltaic systems. Their influence on the market position of a dwelling is relatively unknown. Considering that expected lifespan of photovoltaic systems is at least 25 years, it is likely that many dwellings with a photovoltaic system will enter the housing market.
Design/methodology/approach
Few houses with installed photovoltaic systems have been sold in the market to date. Lack of real market data imposes a method based on the stated preference data. Therefore, the general preferences toward photovoltaic systems are determined by a discrete choice model based on responses of 227 homebuyers in the Eindhoven region, The Netherlands. Further, the model estimates were used to assess the indirect willingness to pay for home photovoltaic systems. This initial willingness to pay is further reassessed with the direct willingness to pay collected in an open-ended questionnaire format.
Findings
Results of the model show that the homebuyers’ preferences for home photovoltaic systems are large and significant. In addition to general preferences, this article reports on the taste heterogeneity carried out by separating observations based on the respondents’ characteristics. For example, photovoltaic systems are more appealing to homebuyers in more urban or central neighbourhoods. Further, the results of the direct survey lead to the conclusion that people are probably willing to pay close to the replacement value of the system and only 22 per cent of all respondents did not want to pay anything for the installed photovoltaic system.
Research limitations/implications
These findings are exploratory and they raise a number of questions for further investigations, such as those regarding the real estate value of the installed photovoltaic systems. The reported findings must be regarded as local, thus further research is necessary to understand the impact on European housing markets.
Practical implications
Preferences and willingness to pay for home photovoltaic systems can provide a variety of economic, social and political recommendations to different interested parties such as homeowners, buyers, realtors, retailers, energy companies and governments. For instance, a homeowner would like to know what would be the effect of a photovoltaic system on the housing market.
Originality/value
As per the knowledge of authors, this is the first paper to estimate the impact of an installed photovoltaic system on housing choice, measured by stated choice data in the local housing market. It expands the existing body of knowledge for increasingly important issues of valuing and measuring preferences for photovoltaic systems installed on dwellings.
Details
Keywords
António M. Cunha and Júlio Lobão
This paper studies the dynamics and elasticities of house prices in Spain and Portugal (Iberia) at the Metropolitan Statistical Area (MSA) level, addressing panel regression…
Abstract
Purpose
This paper studies the dynamics and elasticities of house prices in Spain and Portugal (Iberia) at the Metropolitan Statistical Area (MSA) level, addressing panel regression problems such as heterogeneity and cross-sectional dependence between MSA.
Design/methodology/approach
The authors develop a two steps study. First, five distinct estimation methodologies are applied to estimate the long-term house price equilibrium of the Iberian MSA house market: Mean Group (MG), Fully Modified Ordinary Least Square (FMOLS) MG (FMOLS-MG), FMOLS Augmented MG (FMOLS-AMG), Common Correlated Effects MG (CCEMG) and Dynamic CCEMG (DCCEMG). FMOLS-AMG is found to be the best estimator for the long-term model. Second, an additional five distinct estimation methodologies are applied to estimate the short-term house price dynamics using the long-term FMOLS-AMG estimated price in the error-correction term of the short-term dynamic house price model: OLS Fixed Effects (FE), OLS Random Effects (RE), MG, CCEMG and DCCEMG. DCCEMG is found to be the best estimator for the short-term model.
Findings
The results show that in the long run Iberian house prices are inelastic to aggregate income (0.227). This is a much lower elasticity than what was previously found in US MSA house price studies, suggesting that there are other factors explaining Iberian house prices. According to our study, coastal MSA presents an inelastic housing supply and a price to income elasticity close to one, whereas inland MSA are shown to have an elastic supply and a non-significant price to income elasticity. Spatial differences are important and cross-section dependence is prevalent, affecting estimates in conventional methodologies that do not account for these limitations, such as OLS-FE and OLS-RE. Momentum and mean reversion are the main determinants of short-term dynamics.
Practical implications
Recent econometric advances that account for slope heterogeneity and cross-section dependence produce more accurate estimates than conventional panel estimation methodologies. The results suggest that house markets should be analyzed at the metropolitan level, not at the national level and that there are significant differences between short-term and long-term house price determinants.
Originality/value
To the best of the authors' knowledge, this is the first study applying recent econometric advances to the Iberian MSA house market.
Details
Keywords
The residential real estate market represents the entire complex of economic and social relations at the macroeconomic and microeconomic levels. This publication aims to evaluate…
Abstract
The residential real estate market represents the entire complex of economic and social relations at the macroeconomic and microeconomic levels. This publication aims to evaluate developments in the field of financing and the development of the residential real estate market in Slovakia, with a special focus on the determinants affecting the demand, supply, and prices of residential real estate in Slovakia and Bratislava. Owner-occupied housing is the dominant type of housing and has a significant impact on the development of housing issues. Research into the issue of multiple ownership of real estate provides answers to questions about the growth of real estate prices in the conditions of Bratislava. The influence of financial indicators, especially interest rates, the availability of housing loans, and the regulation of the banking sector explain the essential connections in the area of the development of residential real estate prices and factors of their development. The analysis points to the need for the development of rental housing in Slovakia.
Details
Keywords
Masatomo Suzuki and Chihiro Shimizu
Houses are durable, so an imbalance between demand and supply occurs after time has passed since initial construction. The purpose of this study is to quantify the extent of this…
Abstract
Purpose
Houses are durable, so an imbalance between demand and supply occurs after time has passed since initial construction. The purpose of this study is to quantify the extent of this imbalance for existing houses, focusing on the heterogeneity across property segments.
Design/methodology/approach
This study uses a unique data set on the “inquiry volume” that each property received from an online real estate portal to measure the volume of demand in relation to supply. Simple regressions are conducted in the resale condominium market across the Tokyo metropolitan area.
Findings
The inquiry volume successfully tracked a recent expected trend in which demand relative to supply is stronger for condominiums in reasonably priced areas, condominiums in convenient, accessible locations, condominiums built within the last 20 years and compact and spacious units. This study also confirms that these trends cannot be captured through heterogeneity in price levels, which has been widely used in previous studies on measuring housing preferences.
Practical implications
As an indicator of conditions in the housing market, the property-level inquiry volume has strong potential to provide useful information for supply strategies and for the sustainable use of existing housing stocks.
Originality/value
The originality of this paper is the use of information on the buyer side, which is typically unobservable.
Details
Keywords
M.K. Francke and F.P.W. Schilder
This paper aims to study the data on losses on mortgage insurance in the Dutch housing market to find the key drivers of the probability of loss. In 2013, 25 per cent of all Dutch…
Abstract
Purpose
This paper aims to study the data on losses on mortgage insurance in the Dutch housing market to find the key drivers of the probability of loss. In 2013, 25 per cent of all Dutch homeowners were “under water”: selling the property will not cover the outstanding mortgage debt. The double-trigger theory predicts that being under water is a necessary but not sufficient condition to predict mortgage default. A loss for the mortgage insurer is the result of a default where the proceedings of sale and the accumulated savings for postponed repayment of the principal associated to the loan are not sufficient to repay the loan.
Design/methodology/approach
For this study, the authors use a data set on losses on mortgage insurance at a national aggregate level covering the period from 1976 to 2012. They apply a discrete time hazard model with calendar time- and duration-varying covariates to analyze the relationship between year of issue of the insurance, duration, equity, unfortunate events like unemployment and divorce and affordability measures to identify the main drivers of the probability of loss.
Findings
Although the number of losses increases over time, the number of losses relative to the active insurance is still low, despite the fact that the Dutch housing market is the world’s most strongly leveraged housing market. On average, the peak in loss probability lies around a duration of four years. The average loss probability is virtually zero for durations larger than 10 years. Mortgages initiated just prior to the beginning of the financial crisis have an increased loss probability. The most important drivers of the loss probability are home equity, unemployment and divorce. Affordability measures are less important.
Research limitations/implications
Mortgage insurance is available for the lower end of the market only and is intended to decrease the impact of risk selection by banks. The analysis is based on aggregate data; no information on individual households, like initial loan-to-value and price-to-income ratios; current home equity; and unfortunate events, like unemployment and divorce, is available. The research uses averages of these variables per calendar year and/or duration. Information on repayments of insured mortgages is missing.
Originality/value
This paper is the first to describe the main drivers of losses on insured mortgages in The Netherlands by using loss data covering two housing market crises, one in the early 1980s and the current crisis that started in 2008. Much has changed between the two crises. For instance, prices have risen steeply as has household indebtedness. Furthermore, alternative mortgage products have increased in popularity. Focusing a study on the drivers of mortgage losses exclusively on the current crisis could therefore be biased, given the time-specific circumstances on the housing market.
Details
Keywords
Peng Zhou, Yue Gai and Chaowei Wang
This study conducts a systematic literature review on the determination of urban land value to offer a comprehensive understanding of the methods, datasets, themes and factors of…
Abstract
Purpose
This study conducts a systematic literature review on the determination of urban land value to offer a comprehensive understanding of the methods, datasets, themes and factors of land value. The study aims to identify research gaps and suggest directions for future research in urban land value.
Design/methodology/approach
The study adopts the systematic literature review (SLR) approach to synthesize the literature on urban land value. The SLR is structured according to a three-stage protocol, involving planning, conducting and reporting to ensure transparency and reproducibility. Quantitative bibliometric analysis and qualitative hierarchical thematic analysis are used to assess the evolution of research and to identify key themes and factors.
Findings
The study reveals an unbalanced research focus on developed economies and residential land in urban land value literature. A hierarchical framework categorizes 644 factors into 25 subthemes and 8 themes across physical dimensions (e.g. land attributes and structure attributes) and market dimensions (e.g. land market and macroeconomic conditions). Two primary estimation methods – regression and residual – are identified, each suitable based on data availability. The literature’s evolution is driven by advances in empirical methods. An extensive catalog of databases is compiled, and a corresponding menu of methods is discussed with a focus on empirical identification strategies.
Research limitations/implications
The study is limited by the focus on urban land value and the exclusion of agricultural, recreational and transportation land. Future research should expand to other land types and integrate new data sources and advanced methodologies such as machine learning to enhance empirical robustness.
Practical implications
The systematic review provides a foundation for practical applications and policy discussions on land value estimation and taxation. It offers a useful catalog of land value databases and a menu of land evaluation methods. They are useful for real estate businesses to perform accurate land evaluations and investment appraisals. They can also assist governments in determining precise land value for tax assessments and public policy formulation.
Originality/value
This paper is among the first to apply the SLR approach to urban land value – the price of an essential asset owned by households, businesses and governments. A key contribution is the identification of two distinct evolution patterns of literature: a “pine tree” pattern, showing linear, accumulative growth using homogeneous methodologies (e.g. regression methods) and a “palm tree” pattern, where diverse methods (e.g. residual methods) form independent branches. This analysis provides a new perspective on how methodological homogeneity influences the structure of research themes, offering insights into the dynamics of knowledge development in the field and in general.
Details
Keywords
Swedish house prices have risen rapidly since the mid‐1990s. How can this be explained? Are houses overpriced? In this paper the author tries to answer these questions.
Abstract
Purpose
Swedish house prices have risen rapidly since the mid‐1990s. How can this be explained? Are houses overpriced? In this paper the author tries to answer these questions.
Design/methodology/approach
The author estimates an error correction model (ECM), and sees if the model can explain the house price developments.
Findings
The model suggests that increasing household disposable income and falling mortgage rates are the most important factors behind the upswing in prices. There is no evidence of overpricing.
Originality/value
Compared to earlier Swedish studies, this study is based on new data and new variables. Furthermore, the estimation period is restricted to the more recent period when Swedish credit markets have been unregulated.
Details
Keywords
This paper aims to develop a methodology to accurately and timely measure movements in housing markets by constructing a continuously estimated house price index.
Abstract
Purpose
This paper aims to develop a methodology to accurately and timely measure movements in housing markets by constructing a continuously estimated house price index.
Design/methodology/approach
The continuous index, which is extracted from an additive model that includes the temporal and the locational effects as smooth functions, can be interpreted as an extension of the classical hedonic time-dummy method. The methodology is applied to housing sales from Sydney, Australia, between 2001 and 2011, and compared to three types of discrete indexes.
Findings
Discrete indexes turn out to approach the continuously estimated index with decreasing period lengths but eventually become wiggly and unreliable because of fewer observations per period. The continuous index, in contrast, is stable, has favourable robustness properties and is more objective in several ways.
Originality/value
The resulting index tracks movements in the housing market precisely and in “real-time” and is hence suited for monitoring and assessing housing markets. Because turbulence in housing markets is often a harbinger of financial crises, such monitoring tools support policymakers and investors in tailoring their decisions and reactions. Additionally, the index can be evaluated arbitrarily frequently and therefore is well suited for use in property derivatives.
Details
Keywords
This paper aims to deal with the construction of seasonal price indices for the housing market, based on Rosen's hedonic equations and using spatial econometric autoregression…
Abstract
Purpose
This paper aims to deal with the construction of seasonal price indices for the housing market, based on Rosen's hedonic equations and using spatial econometric autoregression (SAR) techniques.
Design/methodology/approach
More precisely, the hedonic equations are estimated using disaggregated data, and the extracted indices are averaged over zip code areas. Then the seasonality, which is considered deterministic, is extracted after eliminating the spatial effects. The data set used consists of 8,685 valuations of dwellings, detached dwellings and detached houses that took place in Attica on behalf of a commercial bank during the period 2000‐2009.
Findings
The paper concludes that evidence exists to support the hypothesis that property prices are affected by seasonal and spatial effects beyond structural effects and the effects of the general economic situation. Property valuations are strongly connected with deterministic exogenous variables, such as the size, age and location of the property, the general economic situation, and to a lesser effect the spatial system and the season during which the valuation took place. The estimated spatial effect is positive and quite large in value, indicating a landscape consisting of large homogeneous sub‐areas, while the results demonstrate a seasonal upturn during the first semester and downturn towards the end of the year.
Originality/value
This paper provides a framework for incorporating spatial and seasonal effects in property price index construction.