Academic journal article
By Arize, A. C.; Malindretos, J.; Obi, Z. Ike
Atlantic Economic Journal , Vol. 30, No. 2
This paper formally tests whether short- and long-term interest rates of 19 countries are statistically cointegrated. The single-country studies of Stock and Watson , MacDonald and Speight , McFadyen, Pickerill and Devaney , Hall et al. , Wallace and Warner , and Mandeno and Giles , among others, provide evidence in support of a long-run relationship between the short- and long-term interest rates. Such evidence appears to collaborate the expectations theory of the term structure of interest rates, which argues that these two variables will not deviate from each other in the long-run, but will move closely together. This is so because the theory views long-term rates as an average of the expected short-term rates.
Studies such as Mustafa and Rahman  and Taylor  provide evidence in support of the segmented-market theory, which is at the opposite end of the spectrum from the expectations theory. Unlike the expectations theory, the segmented-market theory of the term structure argues that short- and long-term rates are completely unrelated, because they are determined independently of one another by market forces. Empirical evidence in Mustafa and Rahman  focused on the Engle and Granger (E-G)  method and data for the United States, whereas Taylor's  study employed a unit-root test of the short-long spreadseries and the VAR method, including the data for the United Kingdom.
In summary, the above discussions suggest that there is conflicting evidence in the literature on the relationship between short- and long-term rates. (1) No consistent conclusions emerge from these studies. Nevertheless, which hypothesis prevails can have an effect on the model builder and on the conduct of monetary policy. Many macroeconomic models typically employ a single interest rate in representations of the economy and decision making without relying on a spectrum of differing maturities. Of course, if the expectations theory prevails, then central banks can influence long rates by operating at the short end of the market.
The contribution of this study to the literature is the extension of the analysis to 19 countries, including the United States, using recent advances in time-series econometrics. Despite the expansion in research on the linkage between short- and long-term interest rates over the past few years, it has been limited mostly to the data for the United States and the United Kingdom. The question of the nature of the linkage in other countries has not been adequately examined in the literature. This study reports results for cointegration between long-term and short-term interest rates in 19 countries. As Hall, Anderson, and Granger  point out, there are only a few empirical studies employing the cointegration method in the term structure literature.
Another issue, however, that has not received any attention in this literature relates to the stability of the cointegrating relationship. Previous studies investigating this linkage have presumed (either explicitly or implicitly) that the relationship is stable. It is possible that this may not be the case. There is no reason to believe a priori that, over the past 24 years, the relative importance of factors influencing the linkage between short- and long-term interest rates has remained unchanged. As Hansen  points out:
"One potential problem with time series regression models is that the estimated parameters may change over time. A form of model mis-specification, parameter non-constancy, may have severe consequences on inference if left undetected."
It is believed that credible evidence of such a linkage should be ascertained, not only by testing for statistical cointegration, but also by investigating whether the cointegrating relationship has been structurally stable over the sample period. Therefore, the second objective of this note is to provide new evidence on the stability of the cointegrating relationship between short - and long-term interest rates. …