Course description

Time Series Econometrics

Introduction

The aim of the course is to give the students a formal understanding of time series econometrics at a level expected among Ph.D students in economics, finance and related disciplines.

Course content

I. Univariate stationary time series

  • Stationary AR and MA processes
  • Forecasting
  • Spectral analysis

II. Models of non-stationary time series

  • Deterministic and stochastic trends, unit root tests, structural change
  • Trend/cycle decompositions (linear filters)
  • Analysis of business cycles in the frequency domain, spurious cycles

III. Vector autoregression (VAR) methodology

  • Granger causality, cointegration.
  • Structural VARs – impulse responses, forecast error variance decomposition
  • Identification: Cholesky, long-run restrictions, sign restrictions, external instruments.

IV. Methods of Estimation

  • Instrumental variables (IV) estimation
  • Maximum likelihood estimation
  • Generalized method of moments (GMM) estimation

V. State space models and the Kalman filter

  • Kalman filter
  • Factor models

VI. Bayesian estimation

Learning outcome knowledge

After taking this course students should have a solid knowledge of the basic techniques used in time series econometrics, so that eventually they can master and produce sophisticated applied econometric analysis. The students will learn univariate and multivariate models of stationary and nonstationary time series, including structural VARs, state space models, the Kalman filter and factor models. The students will learn to master the main estimation methods, such as maximal likelihood, instrumental variables and GMM. The students will also learn Bayesian estimation.

Exam organisation

  • Home exam: 100%