Data on monetary aggregates are subject to periodic redefinitions, presumably in part to improve their link to measures of output. Money data are also revised on a regular basis. Taking these data imperfections into account, we reassess the evidence on the marginal predictive content of M1 and...