Instructors: Gianluca Cubadda, Tommaso Proietti
Multivariate Time Series (Cubadda): Stationary and Ergodic Multivariate Time Series. Multivariate Wold Representation. Vector Autoregression (VAR) Models. Identification and Estimation of VAR models. Forecasting. Structural VAR Models. Impulse Response Functions. Forecast Error Variance Decompositions. Shocks Identification Using the Choleski Factorization. The Cointegrated VAR. Maximum Likelihood Inference on the Cointegrated VAR. The Common Trends Representation.
State Space Models (Proietti): Review of main concepts in time series (stationarity, autocorrelation, frequency domain analysis). State space models. Unobserved components models for the analysis of economic time series (trends and cycles in macroeconomic time series). Inference for state space models: the Kalman filter, smoothing filter, maximum likelihood estimation. Forecasting with state space models. Topics in business cycle analysis. Filtering economic time series.