Detailed Programme

Day 1 (Monday 04/07/2016):

09:15-10:45:
Lecture 1: Solving stochastic dynamic programming models
Lecturer: Rust, Schjerning
Part I: Introduction to Structural Estimation: The Limits of Inference with and without Theory

Part II: Introduction to dynamic decision problems: Basic theory and numerical tools

  •     Finite/infinite horizon DP problems (VFI and Policy Iterations)
  •     Discrete/continuous choice (Numerical optimization methods)
  •     Discrete/continuous states (Interpolation methods)
  •     Stochastic taste shocks and states (Numerical Integration)
  •     Simple example: Cake eating.

10:45-11:15: Coffee break

11:15-12:45:
Lecture 2: Structural Estimation
Lecturer: Rust, Schjerning

  •     Simulation of the model
  •     Estimation methods: MLE, Indirect Inference, Simulated Method of Moments
  •     Identification
  •     Simple examples: Coin flipping and Cake eating.

12:45-14:15: Lunch break:

14:15-16:15:
MATLAB exercise 1: The cake-eating problem

Day 2 (Tuesday 05/07/2016):

09:15-10:45:
Lecture 3: Structural estimation and discrete decision problems (DDPs)
Lecturer: Rust, Schjerning

  •     The nested fixed point algorithm (NFXP)
  •     Mathematical Programming with Equilibrium Constraints (MPEC)
  •     Counter factual simulations and implied demand (bottom up approach)
  •     Empirical example: Engine replacement

10:45-11:15: Coffee break

11:15-12:45:
Lecture 4: Alternative Approaches to Structural Estimation for DDPs
Lecturer: Schjerning

  •     Sequential Estimation: Nested Pseudo Likelihood and CCP estimator
  •     Structural Estimation of Discrete Markov Decision Models by Sieve Approximations

12:45-14:15: Lunch break:

14:15-16:15:
MATLAB exercise 2: Bus-engine replacement model

Day 3 (Wednesday 06/07/2016):

09:15-10:45:
Lecture 5: Empirical application of discrete decision problems: demand for cars
Lecturer: Iskhakov, Rust, Schjerning

  •     Background DSGE models and stationary/non-stationary recursive competitive equilibria
  •     A Dynamic Model of Vehicle Ownership, Type Choice, and Usage

10:45-11:15: Coffee break

11:15-12:45:
Lecture 6: Solving and Estimating models with Discrete-Continuous choice
Lecturer: Iskhakov
Part I: Solving and Estimating models with only continuous choice using EGM

  •     EGM method and its generalizations in one and multiple dimensions
  •     Comparison of solution and estimation methods: VFI vs EGM, and EGM/NFXP vs MPEC
  •     Example: Consumption and savings with income uncertainty and liquidity constraints

Part II: Solving and Estimating Discrete-Continuous Choice Models using DC-EGM

  •     Generalization of EGM for discrete-continuous problems
  •     Simple example: Consumption and savings model with retirement

12:45-14:15: Lunch break:

14:15-16:15:
MATLAB exercise 3: Consumption, savings and labor supply

Day 4 (Thursday 07/07/2016):

09:15-10:45:
Lecture 7: Empirical Application of DC-EGM
Lecturer: Iskhakov

  •     A dynamic programming model of consumption and savings, human capital, labor supply and retirement in Australia

10:45-11:15: Coffee break

11:15-12:45:
Lecture 8: Solving and estimating static games of incomplete information
Lecturer: Schjerning

  •     Methods: NFXP, MPEC, CCP estimator and Nested Pseudo Likelihood
  •     Example: Simple entry game

12:45-14:15: Lunch break:

14:15-16:15:
MATLAB exercise 4: Estimating static games with multiple equilibria

Day 5 (Friday 08/07/2016):

09:15-10:45:
Lecture 9: Solving and estimating dynamic games of incomplete information
Lecturer: Iskhakov, Rust
Part I: Empirical applications of Dynamic Games

Part II: Finding all Markov Perfect Equilibria of Finite State Directional Dynamic Games

  •     Recursive Lexicographical Search (RLS)
  •     Main example: Bertrand price competition with leapfrogging investments

10:45-11:15: Coffee break

11:15-12:45:
Lecture 10: Solving and estimating dynamic games of incomplete information
Lecturer: Iskhakov, Rust, Schjerning

  •     Methods: NFXP-RLS, MPEC, CCP-estimator and Nested Pseudo Likelihood
  •     Example 1: Dynamic exit/entry model.
  •     Example 2: Bertrand price competition with leapfrogging investments

12:45-14:15: Lunch break:

14:15-16:15:
MATLAB exercise 5: Solving and estimating dynamic games with multiple equilibria

Readings

A detailed reading list will be provided to the participants upon acceptance to the course.