Slowly varying (SV) regressors arise commonly in empirical econometric work, particularly in the form of semilogarithmic regression and log periodogram regression. These regressors are asymptotically collinear. Usual regression formulas for asymptotic standard errors are shown to remain valid, but rates of convergence are affected and the limit distribution of the regression coefficients is shown to be one dimensional. Some asymptotic representations of partial sums of SV functions and central limit theorems with SV weights are given that assist in the development of a regression theory. Multivariate regression and polynomial regression with SV functions are considered and shown to be equivalent, up to standardization, to regression on a polynomial in a logarithmic trend. The theory involves second-, third-, and higher-order forms of slow variation. Some applications to the asymptotic theory of nonlinear trend regression are explored.The author thanks two referees and Pentti Saikkonen for comments and suggestions, Sidney Resnick for references on second-order regular variation, and a Kelly Fellowship and the NSF for partial research support under grants SBR 97-30295 and SES 04-142254. An original draft of the paper was written in June 2000 and circulated under the title “Regression with Slowly Varying Regressors.”

This paper presents the likelihood ratio (LR) test for the number of cointegrating relations in the I(2) vector autoregressive model. It is shown that the asymptotic distribution of the LR test for the cointegration ranks is identical to the asymptotic distribution of the much applied test statistic based on the two-step estimation procedure in Johansen (1995, Econometric Theory 11, 25–59), Paruolo (1996, Journal of Econometrics 72, 313–356), and Rahbek, Kongsted, and Jørgensen (1999, Journal of Econometrics 90, 265–289). By construction the LR test statistic is smaller than the non-LR test statistic from the two-step procedure, and application of the LR test may change rank selection in empirical work. Based on a study of existing empirical applications and related Monte Carlo simulations we conclude that the LR test has much better size properties when compared to the two-step-based test. Overall, we propose use of the LR test for rank determination in I(2) analysis.Discussions with Søren Johansen, Hans Christian Kongsted, and Bent Nielsen are gratefully acknowledged. We also thank the editor and three anonymous referees for very constructive comments that have led to a much improved version of the paper. This research was supported by Danish Social Sciences Research Council grant 2114-04-0001.

We consider the cointegration tests of Johansen (1988, Journal of Economic Dynamics and Control 12, 231–254; 1991, Econometrica 59, 1551–1580) when a vector autoregressive (VAR) process of order k is used to approximate a more general linear process with a possibly infinite VAR representation. Traditional methods to select the lag order, such as Akaike's information criterion (AIC) or the Bayesian information criterion, often lead to too parsimonious a model with the implication that the cointegration tests suffer from substantial size distortions in finite samples. We extend the analysis of Ng and Perron (2001, Econometrica 69, 1519–1554) to derive a modified Akaike's information criterion (MAIC) in this multivariate setting. The idea is to use the information specified by the null hypothesis as it relates to restrictions on the parameters of the model to keep an extra term in the penalty function of the AIC. This MAIC takes a very simple form for which this extra term is simply the likelihood ratio test for testing the null hypothesis of r against more than r cointegrating vectors. We provide theoretical analyses of its validity and of the fact that cointegration tests constructed from a VAR whose lag order is selected using the MAIC have the same limit distribution as when the order is finite and known. We also provide theoretical and simulation analyses to show how the MAIC leads to VAR approximations that yield tests with drastically improved size properties with little loss of power.We are grateful to two referees for especially useful and constructive comments.

This paper considers the information available to invariant unit root tests at and near the unit root. Because all invariant tests will be functions of the maximal invariant, the Fisher information in this statistic will be the available information. The main finding of the paper is that the available information for all tests invariant to a linear trend is zero at the unit root. This result applies for any sample size, over a variety of distributions and correlation structures, and is robust to the inclusion of any other deterministic component. In addition, an explicit upper bound upon the power of all invariant unit root tests is shown to depend solely upon the information. This bound is illustrated via a brief simulation study that also examines the impact that different invariance requirements have on power.Thanks are due to Francesco Bravo, Giovanni Forchini, Les Godfrey, Robert Taylor, participants at seminars at the Universities of Birmingham and York and at the ESRC Econometric study group conference, Bristol, 2004, and also to Bruce Hansen, Joel Horowitz, and five anonymous referees. Revisions of this paper have greatly benefited from comments and suggestions from Grant Hillier and Peter Phillips.

Manfred Deistler was born in St. Pölten, Austria, on September 23, 1941, as the younger of two brothers. After attending elementary school and high school (Gymnasium) in St. Pölten, he studied electrical engineering at the University of Technology Vienna, then known as the Technische Hochschule Wien. He graduated with the degree of “Diplom-Ingenieur” (which is roughly equivalent to a master's degree) in 1964. After working in industry for a year and a half, he attended the Institute of Advanced Studies Vienna and earned a diploma in economics from this institution in 1968. During this period he started work on a Ph.D. thesis in applied mathematics at the Technischen Hochschule Wien, obtaining the Ph.D. degree in 1970. After academic appointments at the University of Regensburg as an assistant professor and at the University of Bonn as an associate professor, he was appointed full professor of econometrics at the University of Technology Vienna in 1978, where he has been ever since. He is married to Dr. Gerda Deistler-Russ, and they have a daughter, Gerda.

Manfred Deistler has worked extensively in econometrics and systems identification, which has resulted in numerous papers and an important book with Ted Hannan [1]. He is also a Fellow of the Econometric Society, a Fellow of the Journal of Econometrics, and a Fellow of IEEE (The Institute of Electrical and Electronic Engineers). He is a corresponding member of the Austrian Academy of Sciences. He is or has been a member of the editorial boards of numerous scientific journals, such as Econometrica, Econometric Reviews, Econometric Theory, Empirical Economics, Journal of Econometrics, Journal of Forecasting, Journal of Mathematical Systems and Control, Journal of Multivariate Analysis, Journal of Time Series Analysis, SIAM Journal on Control and Optimization, SIAM Journal on Matrix Analysis and Applications, and Surveys on Mathematics for Industry. He also is a member of the advisory board of Econometric Theory. Manfred Deistler has also been a most important force in establishing econometrics in Austria.

The interview was conducted on February 17, 2006, in Manfred Deistler's office at the University of Technology Vienna in our common native language, German, and later was transcribed into English. The help of Ingmar Prucha in preparing the interview and comments by Hannes Ledolter, Hannes Leeb, and Richard Nickl on a first draft are gratefully acknowledged.

In this paper, we provide a Doob-style consistency theorem for stationary models. Many applications involving Bayesian inference deal with non independent and identically distributed data, in particular, with stationary data. However, for such models, there is still a theoretical gap to be filled regarding the asymptotic properties of Bayesian procedures. The primary goal to be achieved is establishing consistency of the sequence of posterior distributions. Here we provide an answer to the problem. Bayesian methods have recently gained growing popularity in economic modeling, thus implying the timeliness of the present paper. Indeed, we secure Bayesian procedures against possible inconsistencies. No results of such a generality are known up to now.The authors are grateful for the comments and suggestions of two referees. Antonio Lijoi and Igor Prünster were supported by the Italian Ministry of University and Research, grants 2006134525 and 2006133449, respectively. The research of Stephen G. Walker was funded by an EPSRC Advanced Research Fellowship.

For use in asymptotic analysis of nonlinear time series models, we show that with (Xt,t ≥ 0) a (geometrically) ergodic Markov chain, the general version of the strong law of large numbers applies. That is, the average converges almost surely to the expectation of φ(Xt,Xt+1,…) irrespective of the choice of initial distribution of, or value of, X0. In the existing literature, the less general form has been studied.We thank Paolo Paruolo (the co-editor) and the referee for valuable comments. Also we thank the Danish Social Sciences Research Council (grant 2114-04-0001) for financial support.

We study the properties of the multi-period-ahead least-squares forecast for the stationary AR(1) model under a general error distribution. We find that the forecast is unbiased up to O(T−1), where T is the in-sample size, regardless of the error distribution and that the mean squared forecast error, up to O(T−3/2), is robust against nonnormality.The author is grateful to the co-editor Paolo Paruolo and two anonymous referees for helpful comments. The author is solely responsible for any remaining errors.