<![CDATA[This study examines the complex relationship between multi-hazard disaster risks and property prices in the Jakarta Metropolitan Area, one of the world's most disaster-prone urban regions. The research investigates how various natural hazards, including floods, earthquakes, and other environmental risks, influence real estate values across 138 districts encompassing 15,758 property data. This study pioneers the integration of hierarchical Bayesian modeling with causal machine learning techniques to quantify multi-hazard impacts on property prices, providing the first comprehensive analysis of disaster risk interactions in Indonesian real estate markets. We employ methodological triangulation across Bayesian inference, causal forests, and spatial econometrics to ensure robust causal identification. We employ a multi-methodological approach combining spatial analysis, hierarchical Bayesian modeling, and causal forest algorithms on a dataset of 15,758 properties. The analysis includes Moran's I for spatial autocorrelation (0.73 for risks, 0.65 for prices), PyMC for Bayesian inference with 12,000 MCMC samples, and EconML for causal effect estimation with heterogeneous treatment effects. Properties with high disaster risk experience an 12.2% price discount (95% CI: -20.5%, -3.7%), with each unit increase in average risk score reducing prices by 4.3% (95% CI: -7.9%, -0.4%). Spatial clustering is highly significant (Moran's I = 0.73, p < 0.001). Heterogeneous effects reveal progressive impacts from 3.2% in bottom quintile to 9.4% in top quintile. Policy simulation demonstrates that comprehensive flood mitigation could increase total property values by 840.6 billion IDR, generating an average price increase of 14.8% with benefit-cost ratio exceeding 3:1.]]>
