This research demonstrates the substantial benefits obtained by modeling housing price using a- repeat sales factorial model. In particular, the model is able to give accurate forecast of housing returns on a short or medium run. The index is built-up by determining the weight of 9 economic and financial indices (rental index, short or long-term rate, inflation, stocks index, REITs index, population, Disposable income, population and CPI) to explain capital returns and then to represent housing prices dynamics. The index is computed on Paris housing market from transactions. Mainly the results provide empirical evidence of the ability of the model to forecast short and mid-term changes of the housing prices and more importantly of the housing returns dynamics. Also, the proposed model makes possible to analyze deeply the basic elements that govern the housing market. The developed model also offers applications to regulation and credit risk