Expectations and the Treasury Bill-Federal Funds Rate Spread over Recent Monetary Policy Regimes

Abstract

This paper shows that the spread between the 3-month Treasury bill and the federal funds rate has significant predictive power for the future change in the federal funds rate during the volatile nonborrowed reserves operating regime, but it has less and no predictive power during the borrowed reserves regime and the federal funds targeting regime, respectively. These findings suggest that Treasury bill rates forecast future federal funds rates most accurately when the Federal Reserve follows a well-defined rule that does not smooth the impact of shocks on the federal funds rate. an anonymous referee, and the editor for helpful comments, Robert Cumby and John Huizinga for making their software available to me, and Peter Nagan for survey data. The views expressed here are those of the author and do not necessarily reflect the views of the Federal Reserve Board or any other members of its staff. Any errors are my own. 1 As pointed out by a referee, a third possibility is that bill rates may be unbiased but not very powerful predictors of the level of future federal funds rates. 2 Melton and Roley (1988) show that the standard deviation of weekly percentage changes in the federal funds rate increased about four times during the nonborrowed reserves period from October 567

Tests of the expectations hypothesis: Resolving the anomalies when the short-term rate is the federal funds rate

“…(5) are presented in Table 1. 6 Consistent with the findings of Hardouvelis (1988), Simon (1990) and Roberds et al (1996), the estimate of b is close to and not significantly different from one during the period of monetary aggregate targeting. This is true whether the equation is estimated using only reserve settlement days or other days.…”

“…While a number of ways of testing the EH have been proposed, perhaps the most popular test is the test that Hardouvelis (1988), Simon (1990) and Roberds et al (1996) used. This test is derived under the null hypothesis that the EH holds.…”

“…Evidence when the short-term rate is the effective federal funds rate Hardouvelis (1988), Simon (1990) and Roberds et al (1996) test the EH using Eq. (5) with the effective federal funds rate as the short-term rate and the 3-month T-bill, or similar rate, as the long-term rate.…”

“…In 1994 the Fed began announcing its target for the funds rate. 5 3 Hardouvelis (1988) and Simon (1990) use the 3-month T-bill rate exclusively, while Roberds et al (1996) use several long-term rates, including the 3-month T-bill rate. 4 Note that the estimation period is shorter than the sample period of the data.…”

“…Indeed, it is critical to many analyses of monetary policy where shorter-term rates are believed to be determined by the marketÕs expectation for the overnight federal funds rate. Hardouvelis (1988), Simon (1990) and Roberds et al (1996) investigated the expectations hypothesis (EH) using the effective federal funds rate as the short-term rate and the 3-month T-bill rate (or similar rate) as the long-term rate. With two exceptions these authors find little evidence to support the EH when the short-term rate is the federal funds rate.…”