This paper presents the principles of studying global spatial autocorrelation in the land property market, as well as the possibilities of using these regularities for the construction of spatial regression models. Research work consisted primarily of testing the structure of the spatial weights matrix using different criteria and conducting diagnostic tests of two types of models: the spatial error model and the spatial lag model.
The paper formulates the hypothesis that the application of spatial regression models greatly increases the accuracy of transaction price prediction while forming the basis for the creation of cartographic documents including, among others, maps of land value.
In the traditional approach, geostatistical modeling involves analyses of the spatial structure of regionalized data, as well as estimations and simulations that rely on kriging methods. Geostatistical methods can complement traditional statistical models of property transaction prices, and when combined with those models, they offer a comprehensive tool for spatial analysis that is used in the process of developing land value maps. Transaction prices are characterized by mutual spatial correlations and can be considered as regionalized variables. They can also be regarded as random variables that have a local character and a specific probability distribution.
This study explores the possibilities of applying geostatistical methods in spatial modeling of the prices of undeveloped land, as well as the limitations associated with those methods and the imperfect nature of the real estate market. The results are discussed based on examples, and they cover both the modeling process and the generated land value maps.
The article presents a method for developing geographically weighted regression models for analyzing real estate market transaction prices and evaluating the effect of selected property attributes on the prices and value of real estate. The property attributes were evaluated on a grading scale to determine the relative (percentage) indicators characterizing the relationships on the real estate market. The market data were analyzed to evaluate the influence of infrastructure availability on the prices of land in Olsztyn. The results were used to assess the effect of every utility service on the property transaction prices.
The real estate market is a specific and imperfect field of research, and its complex structure and the presence of information gaps necessitate the use of advanced analytical tools. One such tool is simulation modeling, which has a variety of practical applications and can be used to model real-life systems characterized by a high degree of complexity and a high share of random components. In this paper, virtual data was used to simulate transactions on the local real estate market. Simulation tools were applied to generate additional information about transactions, their spatial distribution and transaction prices. The applicability of the iterative Monte Carlo approach with a standard regression model and a spatial regression model was evaluated.
The patterns and relations between real estate prices and the factors which shape them can be presented, among others, in the form of traditional statistical models, as well as by means of geostatistical methods. In the case of research involving the diagnosis and prediction of transaction prices, the key role is played by the spatial aspect, hence the particular significance of geostatistical methods using spatial information.
The main goal of the conducted research is to determine the probability of the occurrence of a price in a given location in space by means of geostatistical simulation - indicator kriging. Indicator kriging does not use the entirety of information included in a dataset, and can, therefore, be useful in situations when the assumptions involving the spatial stationarity of the examined phenomenon are not fulfilled by an entire dataset, but are fulfilled by a certain part of the set. The maps of the probability with which a regionalized variable (price) takes on particular values, limited by arbitrarily selected cutoff values, were prepared by means of indicator kriging. An alternative approach to the preparation of price probability maps is the determination of the spatial distribution of areas in which, with the assumed probability, the value of the price falls within the predetermined ranges.
The paper presents both the essence as well as a theoretical description of the geostatistical simulation of a transaction on the real estate market, as well as the results of an experiment involving the transaction prices of real properties located in the north-western part of the city of Olsztyn.
The result of the research is a set of virtual information about the places in which the transactions have occurred and about the prices of real estate, constituting a reflection of the market processes which may take place in the near future.
Hedonic models, commonly applied for analyzing prices in the property market, do not always fulfil their role, mainly due to the application of simplified assumptions concerning the distribution of variables, the nature of relations or spatial heterogeneity. Classical regression models assumed that the variation of the explained variable (price) is explained by the effect of market features (fixed effects) and the residual component. The hierarchical structure of market data, both as regards market segments and the spatial division, suggests that statistical models of prices should also include random effects for selected subgroups of properties and interactions between variables. The mixed model provides an alternative for constructing various regression models for individual groups or for using binary variables within one model. With its appropriate structure, it makes it possible to take into account both the spatial heterogeneity and to examine the effects of individual features on prices within various property groups. It can also identify synergy effects. The article presents the issue of mixed modelling in the property market and an example of its application in a market of dwellings in Olsztyn. The research used transaction data from the price and value register, supplemented with spatial data. The obtained model was compared with classical regression models and geographically weighted regression. The study also covered the usefulness of mixed models in the mass evaluation of properties, and the possibility of using them in spatial analyses and for the development of property value maps.
The regularities and relations between real estate prices and the factors that shape them may be presented in the form of statistical models, thanks to which the diagnosis and prediction of prices is possible. A formal description of empirical observation presented in the form of regressive models also offers a possibility for creating certain phenomena in a virtual dimension. Market phenomena cannot be fully described with the use of determinist models, which clarify only a part of price variation. The predicted price is, in this situation, a special case of implementing a random function. Assuming that other implementations are also possible, regressive models may constitute a basis for simulation, which results in the procurement of a future image of the market.
Simulation may refer both to real estate prices and transaction prices. The basis for price simulation may be familiarity with the structure of the analyzed market data. Assuming that this structure has a static character, simulation of real estate prices is performed on the basis of familiarity with the probability distribution and a generator of random numbers. The basis for price simulation is familiarity with model parameters and probability distribution of the random factor.
The study presents the core and theoretical description of a transaction simulation on the real estate market, as well as the results of an experiment regarding transaction prices of office real estate located within the area of the city of Olsztyn. The result of the study is a collection of virtual real properties with known features and simulated prices, constituting a reflection of market processes which may take place in the near future. Comparison between the simulated characteristic and actual transactions in turn allows the correctness of the description of reality by the model to be verified.
The growth of both the construction market and the property market depends on various macroeconomic and legal factors, as well as on demographic, institutional, stock and local conditions. The aim of this research was to determine the spatial differentiation, dynamics and determinants of housing development activity in Poland in the context of historical and current legal conditions. This activity was measured, first of all, by the number of construction contracts and the number of completed buildings and dwelling units.
During the research, an attempt was made to establish determinants of construction activity, by analyzing social, demographic and economic factors concerning individual districts. With this aim in view, the study used statistical panel data models, constructed on the basis of data created as a result of combining time series of observations for cross-sectional units.
The results of the research are presented not only in the form of statistical models, but also as a series of cartographic studies, prepared with the application of GIS tools, presenting the current status of housing development activity in Poland.
Ewa Kucharska-Stasiak, Sabina Źróbek and Radosław Cellmer
Although real estate valuation is supposed to make the market transparent, it has been noted to be partial in many countries. Analysis of literature and results of the statistical analysis of survey responses of Polish property appraisers indicated that: Property valuers operate in an environment which exerts influence on the final result of valuation. The attempts of the client to influence the valuation process and their effectiveness do not depend on the gender and age of the property appraiser. The problem of valuation bias should be seen in the weakness of the system enforcing compliance with ethical standards, with this being an area which requires reinforcement in many countries.
Radosław Cellmer, Mirosław Bełej and Aneta Cichulska
The real estate market, as an open, complex and dynamic system, responds to changes in the environment of economic, legal or social conditions, although the pace and direction of these changes depends on the level of inertia of this system. At the same time, this market stimulates the market environment through prices. This study attempts to identify cause-and-effect relationships in the scope of the impact of selected economic and social indicators on prices of residential premises, as well as to identify the effects of price changes on these indicators. The time horizon of the study covered the years from 2008 to 2018. In the studies, to assess the stationarity of time series, an extended Dickey-Fuller test was used for the model with a free expression and linear trend, a vector autoregression model (VAR) was then constructed and Granger tests and impulse response analysis were performed using the Impulse Response Function (IRF). As a result, it was demonstrated that the response of real estate prices to the impulse from explanatory variables appears between the first and the fourth quarters, and expires after about three years.