1 The Background Elementary courses in statistics introduce at an early stage the key assumption of “random sampling”. In more technical language, the data set is assumed to be identically and independently distributed (i.i.d.). In this framework a range of simple and elegant results can be derived, for example, that the variance of the mean of n observations is 1/n times the variance of the ob...

Cointegration among interest rates for instruments with different maturities has been widely tested with mixed results. This paper proposes an extension to the Engle-Granger testing strategy by permitting asymmetry in the adjustment toward equilibrium in two different ways. We demonstrate that our test has good power and size properties over the Engle-Granger test when there are asymmetric depa...

This paper proposes Bayesian estimation of cointegrated VAR systems and a simple method of estimating the cointegration rank using the Bayes factors for the adjustment term. Monte Carlo experiments show that the method proposed is more powerful in selecting the rank, especially with a small sample size, than Johansen's LR test. This is due to the fact that Bayesian analysis uses the exact distr...

This paper provides a theoretical fractional cointegration analysis in a nonparametric framework. We solve a generalized eigenvalues problem, extending Bierens’ (1997) approach. To this end, a couple of random matrices is constructed, distinguishing between stationary and nonstationary part of the fractional integrated process. The asymptotic behavior of such matrices is studied and convergence...