Apartment Transaction Volume and Price Volatility: Construction and Application of a Transaction-Volume-Based Price Uncertainty Index

Observed housing price volatility comprises both fundamental value changes and sampling noise arising from transaction scarcity; however, conventional indicators fail to distinguish between these two components. This study aimed to addresss this problem by constructing a novel transaction- volume-based price uncertainty index (PUI), defined as the rolling standard deviation of hedonic model residuals divided by the square root of the number of transactions. Using apartment transaction records and building registry data, a quarterly panel of 93 districts across the Seoul metropolitan area and five major metropolitan cities was constructed. The correlation between the PUI and realized volatility (RV) was only 0.38, confirming that the two indicators captured distinct information. The explanatory power of transaction volume on volatility operated through time-varying pricing noise, and the PUI significantly predicted future transaction volume changes that RV failed to capture. Asymmetric analysis revealed that additional transactions had no effect on volatility during boom periods, whereas volatility amplification was pronounced during market contractions; this asymmetry was more clearly observed in the PUI than in RV.