|
Economics
8346 |
|
For the first class, I will talk loosely about what we do. I will also talk a little about structural/non-structural modeling, I have linked below before class). It would be good if you express your interests as the content of the course is somewhat plastic.
Scheduling: Classes to be rescheduled so far: November 4, no class on November 30th.
Make-up classes: Monday October 26th, 3.30-5.00 (no seminar this week, let me know if you can make this time)
Prensentations:
Nov 9th: Sergiy---Asset Pricing (Ljungqvist-Sargent)
Nov 16th: Lalita---Growth (Ljungqvist-Sargent)
Nov 18th: Shoumen---Risk Sharing in more general models (Obstfeld-Rogoff)
Nov 23rd: Tung---Search (Ljungqvist-Sargent)
Final: December 2nd (in class)
Midterm: Monday October 26th,
in class.
Topics (chapters in Ljungqvist and Sargent, unless otherwise noted):
You might want to consult the WEB-page for the 2013 class to see what we did last year. I plan to not change it, but there is a possibility I will do---I am willing to follow student demand to some extent.
Course structure (tentative):
A midterm and 6-10 homeworks followed by in-depth student presentations (a full class each, which in practice means that I do half the talking to help make things understandable to the rest of the class).
(I insist on homeworks and midterm because my experience is that too many people do not not study properly without these proddings.)
Topics covered (I will start with the first one which is just one lecture to get you to think about the role of empirical work).
1) Introduction (empirical methods in macro,)
2) VAR (focus on interpretions of structural VARs and applications in macro)
3) GMM (how to do, applied perspecitive---I will not prove any econometrics results)
4) Panel data
5) Introduction to dynamic programming (Sargent and Ljungquist)
6) A Midterm
7) In previous years, we have done student presentations after midterm – with students choosing a chapter from Sargent-Ljungquist (large chapter can be 2 students) because this is an influential text written in a different style than most students are used to so it is good if we get exposed as much as possible to this. Or a recent working paper that fits with the material of the class. This year I want to have some computational stuff which I haven’t done before so we will decide as we go along how much time to spend on that.
Some potential topics (meaning that I will do it if you want) I have covered in past graduate classes:
Obstfeld-Rogoff Chapter 6 on Sovereign Debt (contract theory, really)
What drove the great recession? (See list of papers here http://www.uh.edu/~bsorense/ecDynMacro2_14.html)
Gabaix (and others) work on Granularity (a current fad) (see also list of papers here http://www.uh.edu/~bsorense/ec83442013.html)
More of my own work (some of the papers you could use a basis for thesis chapters, I have the data etc.)
Short note on Beneviste-Scheinkman versus (generalized) Euler Equation
Panel Data Program from Ostergaard, Sorensen, Yosha JPE 2002. the data
Interaction Effects in Econometrics (Ozer-Balli and Sorensen, Empirical Economics 2013)
The data are in GAUSS format and you should download to the PC and unzip. You will need to change the paths for loading and the outfile.
GMM Notes part 3 (The first 2 pages are the most important. The theory is not going to be on exam.)
Hansen-Singleton GMM program, the data
Some VAR notes by Christopher Sims and Stock&Watson (skim these, or read them, I will not expect you to know all details)
Blanchard-Quah paper (you don’t need to know the detail, but know the logic of imposing long-run constraints)
Paper by Barsky and Eric Sims (this way of using VARs may become more common)
Homework 1: Read the articles in Journal of Economic Perspectives Symposium Spring 2010 called ``Con out of Econometrics.’’ (You can also read the article The Scientific Illusion in Empirical Macroeconomics by Lawrence H. Summers,
Homework 2: Estimate a VAR (I suggested looking at real GDP growth, unemployment, and interest rates) but you can chose other data. Plot impulse response functions. Explain how you determine the number of lags.