We consider a simple random walk process, a special case ofthe Martingale model, which exhibits a deterministic break in its drift term,for instance, from positive to negative. This particular example can be aplausible model for a time series on exchange rates which displays a persistentcurrency...

The Hodrick-Prescott …lter is often applied to economic series as part of thestudy of business cycles. Its properties have most frequently been exploredthrough the development of essentially asymptotic results which are practicallyrelevant only some distance from series endpoints. Our concern...

In this paper we consider the situation where the deterministiccomponents of the processes generating individual series are linear trendsand the individual series are independent I(0) or I(1) processes. We showthat when those time series are used in ordinary least squares regression,the...

We consider the possibility of estimating a Dickey-Fuller regression, constraining the autoregressive parameter to be at most one, and imposing prior knowledge of the sign of the drift parameter. ...

In this paper we introduce a new test of the null hypothesis of nocointegration between a pair of time series. For a very simple generating model, ourtest compares favourably with the Engle-Granger/Dickey-Fuller test and the Johansentrace test. Indeed, shortcomings of the former motivated the...

It is known that Dickey-Fuller tests can lead to spurious rejections of the unit root nullhypothesis when the true generating process is difference-stationary with a break.Suppose now that an unsuccessful attempt is made to allow for a break, either throughmisplaced dummy variables or through...

The hypothesis that futures price is an unbiased predictor of the future spot price has been oneof the most controversial topics in the empirical literature on market efficiency. The conflictingresults provided so far are not robust to the time period considered or to the econometricmethod...

Futures market efficiency has been one of the most researched topics for a number of years.The huge amount of results produced, highly dependent on the econometric techniquesadopted and on the time period analysed, are often conflicting: for a given market, someauthors find evidence of...