An Empirical Note on the Term Structure and Interest Rate Stabilization Policies

Abstract

The expectations theory of the term structure of interest rates supplemented by the rational expectations and time-invariant risk premium assumption implies that the spread between the long and the short rate has in general predictive power for the short rate. This implication was consistently rejected in recent studies Time variations of the risk premium, which are probably important for the behavior of the yield on long-term bonds, are not an entirely satisfactory explanation for these findings. In a more recent article of Mankiw and Miron [1986], who analyzed quarterly data for three-and six-month interest rates over the period 1890-1979, an interesting new explanation for the failure of the expectations theory emerged. They showed that the spread had substantial predictive power for changes in the short rate in the period before the founding of the Federal Reserve (1890-1914), whereas for all other periods considered (1915-1933, 1934-1951, 1951-1958, and 1959-1979), the spread did not help to predict the short rate. Thus, Mankiw and Miron suggested that the rejection of the expectations theory with recent data is a consequence of the commitment of the Federal Reserve to stabilize interest rates resulting in random walk behavior of the short rate. Under these circumstances, expected future short rates are equal to the actual short rate and variations of the spread are only brought about by changes in the risk premium.