... E L + u , where ρ is a spatial autocorrelation parameter, W is a spatial weight matrix, and u is a vector of independent and identically distributed (i.i.d.) errors. The spatial lag model is useful when house values are affected by their own attribute as well...

... suggests very low risk and unexplained excess returns. However, many academics suggest the residential real estate market is unique and standard asset pricing models may not fully capture the risk associated with the housing market. The purpose of this paper is to extend the asset pricing literature...