The expectations hypothesis of the term structure

In fixed-income markets, the efficient markets hypothesis is called the expectations
hypothesis of the term structure. As is the case for efficient markets models of other
asset prices, the expectations hypothesis can be readily formulated in terms of the
forward interest rate, the price at which a future interest rate exposure can be locked in. Forward interest rates are often interpreted as a forecast of the future spot interest
rate. Equivalently, the term premium or the slope of the term structure – the spread
of a long-term rate over a short-term rate – can be interpreted as a forecast of changes
in future short-term rates. The interpretation of forward rates as forecasts implies
that an increase in the spread between long- and short-term rates predicts a rise in
both short- and long-term rates.
The forecasting performance of forward rates with respect to short-term rates has
generally been better than that for long-term rates. Central banks in industrialized
countries generally adopt a short-term interest rate as an intermediate target, but
they also attempt to reduce short-term fluctuations in interest rates. Rather than
immediately raising or lowering interest rates quickly by a large amount to adjust
them to changes in economic conditions, they change them in small increments over
a long period of time. This practice, called interest rate smoothing, results in
protracted periods in which the direction and likelihood, but not the precise timing,
of the next change in the target interest rate can be guessed with some accuracy,
reducing the error in market predictions of short-term rates generally.
Central banks’ interest rate smoothing improves the forecasting power of shortterm
interest rate futures and forwards at short forecasting horizons. This is in
contrast to forwards on foreign exchange, which tend to predict better at long
horizons. At longer horizons, the ability of forward interest rates to predict future
short-term deteriorates. Forward rates have less ability to predict turning points in
central banks’ monetary stance than to predict the direction of the next move in an
already established stance.