Cointegration, Causality, and Forecasting

Clive Granger is a seminal figure in modern econometrics. His name is inextricably bound up with advances in our knowledge of forecasting methods, causality, integration and co‐integration, specification testing and nonlinear models, but this is not an exhaustive list of the areas he has influenced; he has often written papers in applied econometrics and statistics that pertain to many of the key areas of today's research e.g. the nature of ARCH models. He has influenced a large number of econometricians, both directly as a supervisor and indirectly through being either a colleague or a co‐author. Consequently, it is not surprising to find that there was no shortage of interesting topics to be treated and eminent econometricians willing to participate in a Festschrift for him. It is clearly impossible for me to comment on all the papers in this volume so I offer snapshots of it instead. One such ‘still' would reveal that there are four articles on prediction, one on causality, six on theoretical aspects of integration and co‐integration, two on specification tests, two on applied methods of integration and co‐integration, two on ARCH models, and three on miscellaneous topics. The prediction papers range from a comparative exercise by Stock and Watson to theoretical work on the evaluation of density forecasts by Deibold et al. Causality analysis is represented by a piece by Hendry and Mizon that runs through the myriad of ways that Granger causality comes up in econometrics. Inter alia, they point out that such an idea is fundamental to valid impulse response analysis. Stock provides a point optimal test for unit roots while Hatanaka and Yamada analyse the impact of structural change on unit root tests. Multi‐cointegration is explored by Engsted and Johansen while Ermini goes back to the DHSY consumption function and reconsiders it in the framework of seasonal co‐integration. Since co‐integration is often analysed within a VAR, determination of its lag order and the number of variables entering it are important issues, so it is not surprising that Lutkepohl and Saikkonen offer an analysis of VAR order selection and Gonzalo and Pitarakis concentrate on the dimensionality of the VAR system

Causality (physics · Cointegration · Econometrics · Economics · Granger causality · Mathematical economics · Watson · Artificial Intelligence · Computer Science · Monetary Policy and Economic Impact