ng model for its' forecast which produced a Durbin Watsin statistic of 1.85, and standard error statistic of 1.211. This model eventually proved to be the superior model because of its lower than others error statistics. The combination model produced lower MAD, MSE, RMSE statistics than did the automatic method, but smoothing model was more accurate in that it produced a significantly lower MAPE. The summary of method errors, as well as forecasting models, are contained in appendix 6a. Therefore, using these crude methods I have been able to determine that Smart's single exponential smoothing model provides the most accurate forecasting tool for considering this type of numerical data. Based on this model, the forecasted values of inflation for the third and fourth quarters of 1999 are as follows: Q3 = -3.166*.258*3.682 Q4 = -3.216*.258*3.732 Smart software estimates these value ranges with 95% confidence and an average forecast error of 1.689. By considering some current events that are taking place in the domestic and global economy one might be able to more reasonably estimate this range, and thus assert some greater probabilities upon it. As of August 24, 1999 the Federal Reserve Board took a stance to reduce the leverage of some contributive inflationary aggregates. These actions included a .25% increase in the federal funds rate, bringing the total to about 5.25%. As discussed in Money and Banking, this will have a direct impact on the reserve positions and actions for lending institutions. The FOMC helped to accommodate this position stance by selling treasury securities in the secondary market. This is but one of the FOMC directives that can produce this effect. By doing so it detracts funds from the banks, thus further tightening their positions. On November 3, 1999, the Federal Reserve Bank of Minneapolis released a document prepared with information accumulated before October 25, 1999. These findings were summarized and plac...
