March 26-March 29, 2018
Practical DSGE Models
Fabio Canova (Norwegian Business School)
- Bayesian estimation of DSGE models: a refresher. Identification issues.
- Choosing the variables to estimate DSGE models. Dealing with trends/nonbalanced growth paths. Data Rich DSGEs,
- The role of measurement error, singularity and estimation.
- Eliciting databased priors and prior predictive analysis.
- Higher order solution, model pruning and estimation
- Time varying DSGE models: structural breaks, occasionally binding constraints, continuously varying parameters, Markov switching.
- Dealing with model misspecification, evaluation and respecification.
- Composite likelihood and quasi-Bayesian approaches.
April 3-April 6, 2018
New Keynesiam models with search and matching frictions, financial frictions and banks. Transmission and policies.
Paolo Gelain (Cleveland Fed)
Francesco Furlanetto (Norges Bank)
- A basic medium scale New Keynesian model: the transmission mechanism of shocks, monetary policy rules and inflation targeting, discussion of the model-based output gap and monetary policy trade-offs, optimal monetary policy, two-agent New Keynesian (TANK) models, comparison with VAR empirical evidence.
- Extension I, Introducing unemployment: sticky wages, nominal rigidities and to search and matching frictions, the natural rate of unemployment and shifts in the Beveridge curve, optimal monetary policy.
- Extension II, Introducing financial frictions: the financial accelerator mechanism, monetary policy trade offs in the presence of financial frictions, collateral constraints, the role of financial factors in macroeconomic fluctuations, monetary policy and leaning against the wind.
- Extension III, Introducing Banks: Collateral constraints, moral harard models, amplification effects, monetary and prudential policies, the role of capital requirements, global banks.
July 23-27, 2018
Macroeconomics at the Zero Lower Bound
Mathias Trabandt (Freie Universitat Berlin)
- The Canonical New-Keynesian model with search and matching and financial frictions.
- Effects of fiscal policy in normal times vs. deep recessions (zero or effective lower bound).
- The fiscal multiplier in the linearized vs. fully nonlinear New-Keynesian model.
- Importance of real rigidities for the multiplier: Kimball vs. Dixit-Stiglitz aggregators.
- Assessing self-financing fiscal stimulus and self-defeating fiscal consolidations.
- Solution techniques for models with frictions and constraints.
July 30 –August 3, 2018
Econometric techniques for large data sets
Dimitris Korobilis (University of Essex)
- Univariate classical inferencial procedures with many predictors (extreme bounds, dynamic factors, forecast combinations)
- Univariate Bayesian approaches with many predictors (basic methods, Srinkage priors, Bayesian model averaging)
- Multivariate inference (Bayesian VARs, FAVAR, Large Bayesian VARs).
- Machine Learning (ML) algorithms (Supervised learning, Bayesian deep learning, General strategies for inference, Distributed computation)
- Machine Learning (ML) algorithms for high-dimensional inference in regression and VAR models (Random projections, Random forests
- Scalable approximations, Belief propagation and message passing algorithms)
Application deadline: March 10 for the Spring courses, June 30 for the Summer courses.