Time Series Econometrics: methods, computations and applications

Department of Econometrics

Programme Leader: Professor S.J. Koopman
E-mail s.j.koopman@vu.nl
Website: http://staff.feweb.vu.nl/koopman

Summary
The programme mainly concerns the development of classical and Bayesian methods for the analysis of time series, with the objective to apply them to empirical problems in economics and finance. Further, the programme focusses on the implementation of econometric methods. The research involves the development of new econometric methods that are based on simulation techniques. Such developments support the work on empirical applications such as volatility modelling, detecting relationships between economic variables and crime, measuring business cycles, seasonal adjustment and modelling of electricity spot prices. The main objective is to publish state-of-the-art theoretical results and their applications in high-quality international journals in the fields of econometrics, statistics and time series analysis.

The programme aims to be involved in high-quality theoretical, empirical and computational research developments in econometrics generally but with much emphasis on time series econometrics. The wide orientation allows the group to be active in different research subjects. Many research activities are concerned with the development of classical and Bayesian statistical methods for the analysis of time series. Theoretical developments are carried out with the objective of applying them to relevant empirical problems in economics and finance. The programme also focuses on numerical issues and the implementation of econometric methods in computing environments.

In particular, much of our research concentrates on (i) the development of simulation techniques for handling non-Gaussian and nonlinear features, from both classical and Bayesian perspectives, (ii) the simultaneous analysis of panels of time series, (iii) the study of dynamic features in long-memory and fractional integration, (iv) learning models and the econometrics of bounded rationality, and (v) empirical analysis and forecasting of financial time series with volatility, discrete and nonlinear features. The theoretical and methodological developments are typically integrated in empirical studies for volatility modeling in financial markets, analyses of dynamics in credit ratings and default data, measuring the business cycle and economic growth, indicators of financial systemic risk, relations between economic policy and crime, and forecasting of volatility, inflation and interest rates.

We primarily aim to publish innovative research in international peer-reviewed journals which are leading in the fields of econometrics, statistics and time series analysis. Although a sufficient quantity of papers need to be written, we regard quality and potential scientific impact as equally important for our research output. Furthermore we aim to give short courses elsewhere, to write textbooks and to develop user-friendly software products with hands-on user guides. In this way we make results available to applied researchers and practitioners in order to aim for establishing a worldwide impact on applied time series econometrics.

Research environment
We currently have a young team of active and enthusiastic researchers. All members have their own international network of colleague research workers at universities such as Oxford and Cambridge in the UK, Chicago and Pennsylvania in US, Bonn and Wiesbaden in Germany, Aarhus in Denmark and European University Institute in Florence. We further have contacts with institutions such as European Central Bank in Frankfurt, US Federal Reserve, US Bureau of the Census in Washington DC, and Eurostat in Luxembourg. We visit these and other universities and institutes on a regular basis but we also receive the colleagues as (senior) research visitors and as postdocs in our department. We further have good research connections with econometricians in the Benelux countries and we have been part of the European Science Foundation Network (EMM) for Econometric Methods and the Modeling of Non-stationary data, Policy Analysis and Forecasting. Finally, members of our groups participate actively in different societies including the Econometric Society, Society for Financial Econometrics, European Seminar on Bayesian Econometrics, International Conference on Computational Financial Econometrics, and the Netherlands Econometrics Study Group.

 

Publications