Improving the Accuracy of Recent Survey Forecasts of the
T-bill Rate
Other Survey Information Can Help
By Hamid Baghestani
Hamid Baghestani teaches economics at the American University
of Sharjah, UAE. He received his Ph.D. from the University of Colorado,
Boulder. His primary research interests are in time-series analysis,
forecasting, and empirical finance. The author may be contacted at
hbaghestani@ausharjah.edu.
This study concentrates on the Survey of Professional Forecasters
(SPF) to demonstrate a way to improve the consensus forecasts of interest
rates. It promotes the notion that, in improving the survey forecast
accuracy of a variable, one should investigate the usefulness of the
predictive information contained in the survey forecasts of other theoretically
relevant variables. This idea has been applied to the SPF forecasts
of the 3-month Treasury-bill rate, which are shown to be one-sided
for 2001.1-2003.4. We improve the accuracy of these forecasts by exploiting
the predictive information contained in the SPF forecasts of inflation
and output growth. We thus recommend that the possible improvement
should be investigated before such interest rate forecasts are utilized
for decision-making.
By Hamid Baghestani