New Developments in Time Series Econometrics: An Overview

Abstract

Empirical data in economics are typically non-experimental, especially in finance and macroeconomics where researchers usually rely on time series gathered by official agencies or other investigators. This raises two basic problems for econometric modeling: first, to understand the dynamic structure of such series, both individually (e.g., stationarity and persistence properties) and jointly (dynamic relations between series); second, to use these series in order to identify and assess potential explanatory (“structural”) models. Because such data are non-experimental, so that observations cannot be made independent and optimal experimental designs are not available, modeling and inference often require an exceptional degree of sophistication. Fortunately, in recent years, statistical methods for the analysis of time series have developed considerably and several remarkable innovations have been introduced.

This work was supported by the Academic Development Fund of Wilfrid Laurier University, the Institut de Statistique (Université Libre de Bruxelles), the Intercollegiate Center for Management Sciences (Bruxelles), the School of Economics, Kwansei Gakuin University (Japan), the Social Sciences and Humanities Research Council of Canada, the Natural Sciences and Engineering Research Council of Canada, the Government of Québec (Fonds FCAR), and the Ontario-Québec Academic Exchange Program.