New dogs and old tricks: do money and interest rates still provide information content for forecasts of output and prices?

Out-of-sample forecasting experiments are used as an alternative to looking at F-statistics when examining whether money, interest rates or the commercial paper/T-bill spread provide information content for subsequent movements in output, real and nominal personal income, the CPI and the PPI. Here, a variable provides information if it improves the forecast of the explained variable. Employing this procedure we find that the paper-bill spread but not monetary aggregates provide information content for industrial production or real personal income when using data over the 1980-1997 period. In contrast, we find that monetary aggregates provide information content for the CPI and nominal personal income but not the PPI.