ECON 870, Fall 2011

ECON 870: Advanced Econometrics

(Last updated: Nov 1, 2011)

Instructor: Saraswata Chaudhuri

Email: saraswata_chaudhuri@unc.edu

Office: Gardner 305 B

Office Hours: TBA

TA: Fernando Chague

Email: fchague@email.unc.edu

Lecture Times: MW 10:30 – 11:45

Lecture Location: Gardner 210

Prerequisite:

ECON 770 and ECON 771. Familiarity with Matlab will be helpful.

Course Objective and Description:

ECON 870 is designed to give the students a firm understanding of the theoretical aspects behind the commonly used statistical techniques in Economics. The course is broadly divided into two parts. The first part focuses on the classical large sample methods of estimation and inference, namely, (Quasi) Maximum Likelihood, Least Squares and Generalized Method of Moments. In the second part we will focus estimation and inference based on fully Non-parametric and Simulation based methods. Finally we will discuss asymptotic refinements to all the methods using Bootstrap.

Grading Policy:

Your final grade will be based on five homework assignments (40%), a midterm (30%) and a final (30%) examination.

§ While you are free (in fact encouraged) to discuss the homework questions with your classmates (not senior students or professors), everyone must turn in their own copy of the homework solutions.

§ The midterm and the final will be in-class examinations.

Textbook:

(CT) “Microeconometrics” by Colin Cameron and Pravin Trivedi. I strongly recommend that you solve all the exercises in this book. Online resources for the book are available from the website http://cameron.econ.ucdavis.edu/mmabook/mma.html.

Required Supplemental Reading:

(TA) “Advanced Econometrics” by T. Amemiya.

(NM) Newey, W. K. and McFadden, D. “Large Sample Estimation and Hypothesis Testing,” Handbook of Econometrics, Volume 4, 1994.

(AY) Yatchew, A. “Nonparametric Regression Techniques in Economics,” Journal of Economic Literature, Volume 36, 1998.

(JH) Horowitz, J. “The Bootstrap,” Handbook of Econometrics, Volume 5, 2001.

Course Outline, Readings[i] and Homeworks:

Linear Models: (1.5 lectures)

· (CT): Chapter 4.

· (TA): Chapter 2.

· Homework -1 (due Sep 21)

Large Sample Theory: (1.5 lectures)

· Lecture notes (UWLLN).

· (CT): Appendix A for a list of results.

Extremum Estimation: (8 lectures)

· (NM): p. 2113-2191

· (TA): Chapter 4.

· (CT): Chapters 5-6.

· Homework -2 (due Sep 28)

Hypothesis Testing: (6 lectures)

· (NM): p. 2215-2239.

· (CT): Chapters 7-8.

· Homework -3 (due Oct 21)

MIDTERM EXAMINATION

· Assigned on October 24, 2011

· Due on October 31, 2011 (before the lecture): please type your answers

Nonparametric Methods: (4 lectures)

· (CT): Chapter 9.1-9.6 and 9.8. While section 9.7 is arguably more useful for practical purposes, we will reserve it for the next semester so that we can learn these semiparametric methods in the context of concrete applications. However, the nonparametric methods from section 9.1-9.6 will be very useful then in dealing with the infinite dimension nuisance parameters explicitly.

· (AY): While there are numerous rigorous surveys on nonparametric methods, some of which are classic, in my opinion this paper explains the basic methods in a very intuitive way that suits our level of treatment of this topic.

· Homework - 4 (due Nov 21)

Bootstrap: (5 lectures)

· Lecture notes (Consistency of Bootstrap, Edgeworth Expansion and Asymptotic refinement).

· (JH): p. 3160-3186.

· (CT): Chapter 11.

Simulation-Based Methods: (1 lecture, very much incomplete treatment)

· (CT): Chapter 12.