Academic journal article Review - Federal Reserve Bank of St. Louis

Assessing Monetary Policy Effects Using Daily Federal Funds Futures Contracts/Commentary/Commentary

Academic journal article Review - Federal Reserve Bank of St. Louis

Assessing Monetary Policy Effects Using Daily Federal Funds Futures Contracts/Commentary/Commentary

Article excerpt

(ProQuest: ... denotes formulae omitted.)

James D. Hamilton

This paper develops a generalization of the formulas proposed by Kuttner (2001) and others for purposes of measuring the effects of a change in the federal funds target on Treasury yields of different maturities. The generalization avoids the need to condition on the date of the target change and allows for deviations of the effective fed funds rate from the target as well as gradual learning by market participants about the target. The paper shows that parameters estimated solely on the basis of the behavior of the fed funds and fed funds futures can account for the broad calendar regularities in the relation between fed funds futures and Treasury yields of different maturities. Although the methods are new, the conclusion is quite similar to that reported by earlier researcherschanges in the fed funds target seem to be associated with quite large changes in Treasury yields, even for maturities of up to 10 years. (JEL: E52, E43)

Federal Reserve Bank of St. Louis Review, July/August 2008, 90(4), pp. 377-93.

Economists continue to debate how much of an effect monetary policy has on the economy. But one of the more robust empirical results is the observation that changes in the target that the Federal Reserve sets for the overnight federal funds rate have been associated historically with large changes in other interest rates, even for the longest maturities. This paper contributes to the extensive literature that tries to measure the magnitude of this effect.

One of the first efforts along these lines was by Cook and Hahn (1989), who looked at how yields on Treasury securities of different maturities changed on the days when the Federal Reserve changed its target for the fed funds rate. Let i ^sub s,d^ denote the interest rate (in basis points) on a Treasury bill or Treasury bond of constant maturity s months as quoted on some business day, d, and let ξ^sub d^ denote the target for the fed funds rate as determined by the Federal Reserve for that day. Using just those days between September 1974 and September 1979 on which there was a change in the target, Cook and Hahn estimated the following regression by ordinary least squares (OLS):

(1) i^sub s,d^ - i^sub s,d^ 1 = α^sub s^, + λ^sub s^ (ξ^sub d^ - ξ^sub d-1^ ) + u^sub sd^.

Their estimates of λ^sub s^ for securities of several different maturities are reported in the first column of Table 1. These estimates suggest that, when the Fed raises the overnight rate by 100 basis points, short-term Treasury yields go up by over 50 basis points and there is a statistically significant effect even on 10-year yields.

Subsequent researchers found that the magnitudes of the estimated coefficients for λ^sub s^ were significantly smaller when later data sets were used. For example, column 2 of Table 1 reports Kuttner's (2001) results when the Cook-Hahn regression (1) was reestimated using data from June 1989 to February 2000; see also Nilsen (1998).

However, Kuttner (2001) also identified some conceptual problems with regression (1). For one thing, the market may have anticipated much of the change in the target ξ^sub d^ that occurred on day d many days earlier, in which case those expectations would have already been incorporated into i ^sub s,d-1^. In the limiting case when the change was perfectly anticipated, one would not expect any change in i^sub s,d^ to be observed on the day of the target change. To isolate the unanticipated component of the target change, Kuttner used f^sub d^, the interest rate implied by the spot-month fed funds contract on day d. These contracts are settled on the basis of what the average effective fed funds rate turns out to be for the entire month containing day d. Because much of the month may already be over by day d, a target change on day d will have only a fractional effect on the monthly average. Kuttner proposed the following formula to identify the unanticipated component of the target change on day d:

(2) . …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.