T4: Tutorial on Advanced Econometrics
Partners involved
Univ. “La Sapienza” Roma (Local organizer), Univ. of Copenhagen, Univ. of Klagenfurt, Univ. of Lodz.
Date and location:
12/06/2008 - 14/06/2008 in Rome, Italy.
Motivation and Objectives
Heuristic optimization is increasingly employed for building advanced models in econometrics. The tutorial will provide the necessary background on the most recent developments, such as cointegration, conditional heteroscedasticity, structural breaks and co-breaking. The tutorial draws a path from standard time series and econometric models to non-linear and non-stationary models, addressing both probabilistic topics concerning stochastic processes and their properties, and the statistical inference related to identification and estimation of the model.
Organization
Two days with seven lectures of 90 minutes and an afternoon "hands on" session in the computer laboratory.
- Overview of standard time series analysis and linear models: Non-stationary series; trend analysis; motivation for unit roots and their meaning.
- Multivariate time series: Measurement of the relationships between dynamic phenomena; non-stationarity in a multivariate environment (cointegration).
- Conditional heteroscedasticity: ARCH and GARCH models and their structural properties; identification, estimation and testing procedures.
- Multi-regime models, their motivation and use: Threshold autoregressive models; Markov switching models; double threshold models.
- Further non-linearities: Outliers, level changes, structural breaks (detection of, relevance and meaning in macroeconomic data).
- Heuristic optimization for advanced model building: Motivation for heuristic methods; hybridization; examples of implementation.
- Policy applications: Policy simulations; fixed-targets policy application; dynamic optimization; (stochastic) dynamic programming and optimal control.
- "Hands-on" session: building and analyzing non-linear models.