The expectations hypothesis is a theory about how markets determine long-term interest rates on debt-based assets. The theory is simply that the rates are decided by the short-term expectations plus a fixed additional amount to reflect the inherent increased risk in the long-term. Most tests of the expectations hypothesis do not bear it out, but the reasons for this are disputed.
This hypothesis covers interest rates, which can be viewed from two perspectives. They are the interest rates which investors will receive by buying an asset. In turn, they are also the interest rates which the original issuer of the asset, such as a corporation or government, must pay to borrow in this way.
In most cases the expectations hypothesis is not so much a prediction tool as a way of saying what the relationship between known rates should be. Generally both short-term and long-term rates for a particular asset, or different forms of the same asset such as a 1-year bond and a 3-year bond, are already known. This means we can tell immediately if the expectations hypothesis is correct.
The precise formula used in an expectations hypothesis varies from case to case. There is a consistent principle, which is that the short term and long term rates will vary by a fixed level. The logic is that all the factors that affect the short term rate apply to the long term rate, but that the long term rate also includes a "premium" to cover uncertainty, for example the lengthier period of time during which the issuer may default.
As many studies show the expectations hypothesis is not borne out by reality, its main function is as the starting point for an economic puzzle. Economists believe that figuring out why the hypothesis isn't borne out may help explain more about how markets really work. One theory is that while the basic reasoning of the expectations hypothesis is valid, the "premium" isn't consistent and instead changes over time, possibly at varying rates. Another theory is that the hypothesis falsely assumes it is possible to accurately forecast the short term rates, when in reality there are too many variable factors in play to do so.
There are some studies that suggest that the hypothesis proves more accurate as the time period of the long-term rates increases. At first glance this may seem counterintuitive as there is more opportunity for variation. In practice it may be that a longer time period gives more time for market imperfections to be corrected and for investors to gain additional information, meaning that demand and supply even out to produce a more predictable interest rate.
