ECS570 Empirical IO: Dynamic Structural Models

• Topics

  Day 1
  1. Introduction to structural estimation of dynamic discrete choice models
  2. Nested fixed point (NFXP) estimator
  3. Mathematical programming with equilibrium constraints (MPEC)
  
  Day 2
  1. Inversion theorem and CCP estimator
  2. Nested pseudo-likelihood (NPL) estimator
  3. Bajari, Benkard, and Levin (BBL) estimator
  
  Day 3
  1. Dynamic models of equilibrium
  2. Stationary equilibrium model of durable goods market
  3. Non-stationary equilibrium models
  
  Day 4
  1. Berry–Levinsohn–Pakes (BLP) estimator
  2. Modelling static games
  3. MLE, MPEC, CCP and NPL estimators for static games
  
  Day 5
  1. Estimation of dynamic games with MPEC, CCP, NPL estimators
  2. Solving dynamic games with multiple equilibria, recursive lexicographical search (RLS)
  3. Nested MLE estimator for dynamic directional games with multiple equilibria (NRLS)

• Learning outcome

  Knowledge - upon completing the course students will:

  • • Apply structural econometric models used in IO to analyze market dynamics
    • Evaluate different estimation technuqies used to estimate games and single-agent dynamic problems
    • Understand the notion of stationary and nonstationary Markov Perfect Equilibrium

  Skill - upon completing the course students will:

  • Be able to develop computer code for the estimation routines in a selected programming language
  • Evaluate the scope of application of the methods within IO

  General competence - upon completing the course students will:

  • Evaluate and choose appropriate econometric methods and apply them to study a variety of questions related to industry dynamics and market competition

• Teaching

  Plenary lectures and lab sessions

• Restricted access

  • PhD candidates from Department of Economics, NHH
  • PhD candidates from other Departments at NHH
  • PhD candidates from University of Bergen
  • PhD candidates from other higher educational institutions

  Promising master students if approved by course responsible

• Recommended prerequisites

  PhD-level preparation in econometrics, e.g. ECS508 and ECS509 and microeconomics (e.g. ECS504), including static and dynamic game theory.

• Compulsory Activity

  Attendance in lectures and lab sessions

• Assessment

  Group project with two to four students.

  The project is due 6 weeks after the last lecture.

• Grading Scale

  Pass -fail

• Computer tools

  Python/Matlab

• Literature

  Lecture notes and a list of papers will be provided by the instructors.