Author(s) :

Yufei Lu.

Volume/Issue :

Abstract :

This study investigates several forecasting approaches for the United States federal funds rate and consumer price index (CPI) inflation rates in the years 1980-2025. This study compares univariate and joint models under trend and autoregressive specifications, implementing each model in connection with a historical data set and a three-year holdout test sample (2023-2025). The primary emphasis is discovering which approach to forecasting will provide greater predictive accuracy and relevance for macroeconomic policy purposes. The results of empirical work show that while these trend models serve well in capturing long-term trends in monetary and price behavior, autoregressive models perform better in forecasting short-term values because of their greater flexibility in taking into account variations and external factors. Joint modeling of interest rates and inflation through a vector-autoregressive (VAR) system provides small yet significant increases in forecasting accuracy but demonstrates that intermingling between monetary policy and price stability improves forecast accuracy. These findings suggest that accuracy in forecasting is not purely a statistical problem but a crucial policy problem which illustrates that even slight improvements can avoid premature policy tightening or delayed economic easing. Thus, this study seeks to indicate the importance of integrating dynamic, feedback-giving modeling systems into central bank databases in order to bring greater data usefulness and greater stability in financial markets.