The small sample properties of estimators and tests are frequently too complex to be useful or are unknown. Much econometric theory is therefore developed for very large or asymptotic samples where it is assumed that the behaviour of estimators and tests will adequately represent their properties in small samples. Refined asymptotic methods adopt an intermediate position by providing improved approximations to small sample behaviour using asymptotic expansions. Dedicated to the memory of Michael Magdalinos, whose work is a major contribution to this area, this book contains chapters directly concerned with refined asymptotic methods. In addition, there are chapters focusing on new asymptotic results; the exploration through simulation of the small sample behaviour of estimators and tests in panel data models; and improvements in methodology. With contributions from leading econometricians, this collection will be essential reading for researchers and graduate students concerned with the use of asymptotic methods in econometric analysis.
• Brings together original papers by leading econometricians • Use of refined asymptotic methods improves understanding of analytical tools in econometrics • Methodological improvements provide stronger underpinning for theory
List of figures; List of tables; Contributors; Preface; Michael Magdalinos 1949–2002; Acknowledgements; Introduction Garry D. A. Phillips and Elias Tzavalis; 1. Conditional heteroscedasticity models with Pearson disturbances Michael A. Magdalinos and George P. Mitsopoulos; 2. The Instrumental Variables method revisited: on the nature and choice of optimal instruments Aris Spanos; 3. Nagar-type moment approximations in simultaneous equation models: some further results Garry D. A. Phillips; 4. Local GEL methods for conditional moment restrictions Richard J. Smith; 5. Limit theory for moderate deviations from a unit root under weak dependence Peter C. B. Phillips and Tassos Magdalinos; 6. The structure of multiparameter tests Christopher L. Cavanagh and Thomas J. Rothenberg; 7. Cornish-Fisher size corrected t and F statistics for the linear regression model with heteroscedastic errors Spyridon D. Symeonides, Hellen Kandiloriou and Elias Tzavalis; 8. Non-parametric specification testing of non-nested econometric models Qi Li and Thanasis Stengos; 9. Testing for autocorrelation in systems of equations Phoebus J. Dhrymes; 10. Alternative approaches to estimation and inference in large multifactor panels: small sample results with an application to modelling of Asset Returns G. Kapetanios and M. Hashem Pesaran; 11. Judging contending estimators by simulation: tournaments in dynamic panel data models Jan F. Kiviet; 12. A statistical proof of the transformation theorem Karim M. Abadir and Jan R. Magnus; 13. On the joint density of the sum and sum of squares of nonnegative random variables Grant Hillier; 14. Conditional Response Analysis Grayham E. Mizon and Anna Staszewska; Index.
Review of the hardback: 'This volume is extremely wide ranging and has papers in the areas of interest of all econometricians. I highly commend the book. Econometricians will find it of interest both for their own areas of specialization and also for the opportunity to read well done papers in other areas.' Jerry Hausman, MacDonald Professor of Economics, MIT
Review of the hardback: 'This remarkable collection of papers has been assembled in tribute to Michael Magdalinos, a much loved and admired econometrician who died tragically young. The highly distinguished group of contributors include among them former teachers, colleagues, collaborators, friends, and his son Tassos. While the book is a touching and fitting memorial to Michael, the papers cover such a range of topics, and are of such depth and quality, as to also constitute a valuable primer in modern econometric thinking. Researchers, practitioners and students alike will find much in these pages to profit and interest them.' James Davidson, Professor of Econometrics, University of Exeter
Review of the hardback: 'This book is a tribute to one of the great contributors to the subject of econometrics. The excellent papers contributed here reflect on the research of the late Michael Magdalinos and acknowledge his solid and beneficial impact on the foundational issue of finite and asymptotic sample econometrics. They will make the book widely read and referenced.' Aman Ullah, Professor of Economics, University of California, Riverside