On the equilibrium of the magnetopause