We empirically examines the financial portfolio choice of households as a function of their exposure to real estate risk as a possible background risk. Using Panel Study of Income Dynamics data from 1984 to 2001, our estimation results control for sample selection and unobservable time-invariant...