Wharton GIS Lab


Automated Valuation Models

Wharton GIS Lab offers the opportunity to utilize contemporaneous real estate data, which is updated every six months at a minimum, and the ability to link to property specify data, allowing analysts to track change on a property by property basis, aggregating houses into neighborhoods which function as interdependent units.

The Wharton GIS lab has recently developed a hybrid hedonic/spatial autocorrelation model of home pricing for the greater Philadelphia area. Using both state-of-art hedonic pricing techniques combined with the explicit estimation of the spatial covariance structure of Philadelphia-area properties, the model represents a significant step forward in the class of traditional models of property valuation.

The recent rapid growth in both the technologies and methodologies of spatial econometrics have facilitated the development of breakthrough research in real estate economics. While current work has indicated that such hybrid models possess very strong predictive power (predicted sales price is within 5% of actual sales price the majority of the time), the Wharton GIS lab will continue to position itself on the cutting edge in order to further develop, improve and refine such important work in the field of property asset valuation.

Learn more by downloading the Property Valuation Presentation