Predicting FOMC Decisions Using an Ordered-Probit Model

The Federal Funds Rate is one of the most closely watched interest rates by equity traders and consumers alike. Because of the depth and breadth of the impact of changes in this key interest rate on the economy, it would be beneficial to know in advance of scheduled meetings of the FOMC what the Committee might decide regarding the target Federal Funds Rate. This information would provide opportunities for interest-rate risk management as well as for speculative trading. Many individuals in financial markets do, in fact, attempt to predict this rate to profit from or hedge based on their predictions. This paper uses an ordered probit model along with information on the price of Federal Funds Futures traded on the Chicago Board of Trade (CBOT) to predict the FOMC action at different forecast horizons. The aim of the paper is to predict this rate in real-time. To this end, we use the change in the Fed Funds target rate predicted by Federal Funds Futures (FFF). We also control for “risk premium” in the Federal Funds Futures by using the year-over-year change in Non-Farm Payrolls to account for the business cycle. We control for change of monetary policy regime using a dummy variable that reflects whether the Fed chairman at the time of the meeting was Alan Greenspan or Ben Bernanke. Finally, we include in our model the level of the Fed Funds Rate prior to each FOMC meeting. The sample period runs from January 1994 through September 2008. The prediction accuracy for 1-day, 7-days, 14-days, 21-days, and 28-days forecast horizons are 84%, 78%, 72%, 60% , and 56%, respectively.