 Modeling housing price uncertainty: probability distributions and market riskViktorija Cohen and Darius Zabulionis ERES from European Real Estate Society (ERES) Abstract: Housing prices are influenced by multiple factors, from economic trends and demographic changes to construction costs and the unique characteristics of different locations. While many of these factors can be analyzed and forecasted, their influence on price levels introduces an additional layer of uncertainty, making predictions of prices challenging. Recognizing this complexity and understanding the probabilistic nature of house prices is crucial for developers, buyers, or policymakers, as price fluctuations directly impact financial risk, investment decisions, and economic stability. For real estate developers, knowing the probability of housing prices staying below a certain level is crucial—it represents a potential financial risk, as lower prices may lead to reduced profits or project viability concerns. On the other hand, homebuyers face the opposite challenge: the probability that prices will rise beyond their affordability level. Policymakers, meanwhile, must assess the fluctuations to ensure market stability, address housing affordability, and design effective regulations. All these perspectives emphasize the need to understand average price trends and the full range of possible price fluctuations. The nature of the factors influencing housing prices can be considered inherently uncertain. One of the critical aspects of understanding housing price uncertainty is the recognition that the housing market is heterogeneous, comprising various interlinked segments. Borgerse (2014) highlights that this heterogeneous nature leads to different pricing dynamics and risk levels across market segments. Housing price volatility is also essential in shaping buyer behavior and investment decisions. Prior studies, such as Wang et al. (2020), suggest that households adjust their purchasing strategies based on expectations regarding future price trends and volatility, which influence their decisions, and the overall risk premium associated with real estate investments. Similarly, Huang (2020) emphasizes that housing risk factors are time-varying by nature. It means that price fluctuations evolve in response to broader economic and policy changes. Furthermore, housing price uncertainty has broader economic stability and consumer welfare implications, influencing financial risks, mortgage markets, and regulatory policies. Prior studies suggest that price volatility can have asymmetric effects on economic growth, with negative price movements often triggering more severe economic disruptions (Asadov et al., 2023). Fluctuations in housing prices impact consumer confidence and spending behavior, reinforcing their role in broader economic cycles (Nikitina, 2023). Additionally, macroprudential policy adjustments, such as loan-to-value regulations or capital requirements, can sometimes lead to unintended price increases in specific markets (Sun et al., 2016; Wu & Li, 2017). Likewise, monetary policy decisions—particularly interest rate adjustments—directly influence housing price fluctuations (Yang, 2023; Bjørnland & Jacobsen, 2010; Zhang & Zoli, 2014). However, interventions must be carefully designed to avoid increasing market volatility or restricting housing accessibility (Changdong & Jiang, 2021; Iacoviello & Neri, 2010). To capture the uncertainty, housing prices can be viewed as a dynamic process, changing over time in ways that are not entirely deterministic. We must examine how housing prices are distributed over time to understand these fluctuations better. A key tool for this is the cumulative distribution function (CDF), which helps estimate the likelihood of prices falling within specific ranges (Huang et al., 2024). However, capturing this complexity requires more than just general trends. Key indicators like the average price, price volatility (variance), and the overall shape of the price distribution are used to make informed decisions. Despite the importance of these insights, a gap in data availability makes it difficult to conduct analysis: publicly available data lacks data on the properties’ housing prices. While some studies, such as those by Young and Graff (1995) and Bond et al. (2007), have explored price distributions, comprehensive real-world data remains limited. This research emphasizes the need for a deeper understanding of housing price fluctuations. It offers insights into housing price risk assessment by applying probabilistic modeling to capture market uncertainty. These findings enhance our understanding of housing price risk assessment, helping developers anticipate financial exposure, assisting policymakers in designing effective regulations and providing buyers with a clearer understanding of affordability risks. Methodology and sampling To model housing price uncertainty, this study treats housing prices as a stochastic process, where prices evolve over time in a non-deterministic manner. Instead of relying on single-value estimates, we use probability distributions to assess the likelihood of prices falling within specific ranges. The study defines housing prices as a random variable and evaluates their behavior using a cumulative distribution function (CDF). For the empirical analysis, three probability distributions were considered: 1. Normal Distribution – Assumes price fluctuations are symmetrically distributed around the mean, often used in financial modeling but may not fully capture real estate price variations. 2. Lognormal Distribution – Accounts for right-skewed price distributions, where extreme price increases are more frequent than Generalized Lambda Distribution (GLD) – A flexible distribution capable of modeling asymmetry and heavy tails, making it well-suited for real estate price dynamics. The study applies statistical fitting techniques using software tools from the R programming environment to determine which distribution best represents housing price behavior. Specifically, the moment method estimates parameters for the GLD, which allows for a more accurate representation of price fluctuations. The dataset consists of 3,186 transaction records of fully completed flat purchases in Vilnius, Lithuania, spanning from January 2018 to October 2022. Each transaction includes 60 property characteristics. Transactions involving commercial properties were excluded to maintain homogeneity within the sample. Preliminary results The findings demonstrate that the Normal distribution does not adequately describe housing price variability, whereas the Lognormal and GLD distributions provide a better fit. GLD offers the most precise representation of market fluctuations, making it a valuable tool for real estate risk assessment and pricing strategies. The analysis showed that the quarantine did not significantly affect the number of RE transactions. However, the total amount of the RE transaction money is much more considerable immediately after the quarantine than before. The 0.1, 0.25, 0.5, 0.75, and 0.9 quantiles of the price of the sq.m. of flats showed a stable increase with respect to the time during the period under the investigation. However, the increase of 0.75 and 0.9 quantiles of RE price per is steeper than that of lower 0.1 and 0.25 quantiles. The findings highlight the importance of selecting appropriate probability distributions for housing price modeling. The results show that the GLD better captured housing price dynamics compared to Normal and Lognormal distributions. The study contributes a more robust understanding of housing price uncertainty, providing additional insights for real estate stakeholders. extreme declines. 3. Keywords: CDF; Housing Price; probability distribution; Uncertainty (search for similar items in EconPapers) Downloads: (external link) Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:arz:wpaper:eres2025_264 Access Statistics for this paper More papers in ERES from European Real Estate Society (ERES) Contact information at EDIRC. |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.