It is commonly believed that while central banks closely control
short-term interest rates, long-term interest rates, which are heavily
influenced by expectations, play a more important role in affecting
aggregate demand. Central banks thus face the problem of setting
short-term interest rates in such a way as to move longer-term interest
rates in the desired direction. A considerable amount of empirical
evidence has been presented suggesting that the explanatory power of the
expectations hypothesis may be greater than previously thought. Two
findings, most likely interrelated, are of particular interest. First,
it appears that the poor empirical performance of the expectations
hypothesis may be due to the fact that short rates are typically
difficult to predict, thus the theory may not be fundamentally flawed.
Second, the expectations hypothesis appears to be better able to account
for the behaviour of interest rates outside the United States.
In case that the failure of the expectations hypothesis stems from
the limited predictability of short rates, one would expect that further
light can be shed on the expectations hypothesis by using data for a
number of currencies with potential substantial differences in the
time-series behaviour (and thus in the predictability) of short rates.
In discussion paper No 1258, Research Fellow Stefan Gerlach and
Research Affiliate Frank Smets explore this possibility by using
one-, three-, six- and twelve-month Euro interest rates for a sample of
17 currencies and find that for all currencies the term spread does
predict future movements in the short rate. Moreover, in a majority of
cases they cannot reject the hypothesis that the estimated slope
coefficient is different from unity as implied by the pure expectations
theory. Finally, they show that the failure of the expectations theory
to hold in a number of countries, most notably the United States, is due
to a lack of predictability of the short-term rate which introduces a
downward bias in the estimates in the presence of a time-varying term
premium. They conclude that, despite the presence of a time-varying term
premium, for many countries the expectations hypothesis is broadly
compatible with the data.
The Term Structure of Euro-Rates: Some Evidence in
Support of the Expectations Hypothesis
Stefan Gerlach and Frank Smets
Discussion Paper No. 1258, October 1995 (IM)