An online real estate valuation model for control risk taking: A spatial approach

Abstract

The private sector and financial regulators have improved their ability to monitor and control risk taking. The Basel II Accord binds credit institutions to revise the collateral values of their credit loans frequently and states that statistical methods may be used. Although valuation of residential properties has traditionally been based solely on consultation with experts, this article presents an expert system for estimating collateral value that reduces costs and avoids expert consultations by using spatial statistical methods (kriging). This expert system allows for immediate, accurate and affordable valuations of large portfolio properties in the residential market. Using data from the Spanish market, we obtain sufficient goodness-of-fit in comparing actual and estimated prices. Particularly, results indicate that eight out of ten residential properties in Madrid are valued with less than 20% error and eight out of ten in Barcelona. Finally, the price of a housing portfolio is evaluated at two different dates.