Since the last meeting of this committee, we have been experimenting with a different presentation of our forecasts of the MI - Adjusted Monetary Base Multiplier. In the past we have always constructed forecasts directly from the forecasts of the various component ratios, which come out of the ARIMA models that we have estimated. These forecasts are not seasonally adjusted, so comparisons over time horizons shorter than one year are difficult to interpret. Our revised presentation allows us to compute forecasts on a seasonally adjusted basis that are not contaminated by errors in forecasting seasonal factors. The presentations employ the known seasonal factors that are published in the Federal Reserve Bulletin for the various components of ...