Hedonic Price Modelling of Condominium Apartments in Östermalm, Stockholm : A Multiple Linear Regression and Quantile Regression Analysis

Sammanfattning

This study estimates the inherent prices of observable housing attributes in the Östermalm condominium market using a hedonic pricing framework applied to 4 281 transactions sourced from Svensk Mäklarstatistik over a 24-month pre-reform observation window. Attributes examined include living area, monthly housing association fee, elevator access, floor position (in the manner of ground level or vertically higher than ground level), and construction period. Stratified ordinary least squares models are estimated across five room-size segments, with variable selection performed through stepwise AIC optimisation. To assess whether implicit prices vary across price segments within each stratum, quantile regression is applied at τ = 0,25, 0,50, and 0,75. Living area, floor position, and elevator access are statistically significant determinants of contract price across all strata. Monthly fee exerts a consistently negative effect in four of five strata. Construction period is significant throughout, with the pre-1914 building stock commanding the highest implicit valuation relative to all later construction eras. Model reliability is highest in the smaller room-size segments, with adjusted R² values of 0,705 and 0,790 in the one- and two-room strata respectively, and declines progressively for larger apartments. Quantile regression reveals that prices increase systematically with price quantile for living area and prestige attributes (such as above ground, elevator access, etc), while the monthly fee discount is larger for lower-quantile buyers. The primary contribution of this study is the establishment of a methodologically transparent and replicable pre-reform baseline for the condominium pricing development of the April 2026 mortgage deregulation, so that it can be assessed in future research. The quantile regression baseline enables attribution of post-reform price changes across the price distribution, providing analytical insights to segment sensitivity and policy goals.