This lecture covers the Generalized Method of Moments (GMM). How do you estimate alphas, betas, and lambdas; how do you evaluate if models are any good? GMM is a very flexible econometric framework for lots of problems, and we’ll also explore that a bit.
- Asset Pricing: chapters 11 and 12
- Hansen, Lars Peter, 1982, Large Sample Properties of Generalized Method of Moments Estimators, Econometrica 50, 1029-1054.
- Hansen, Lars Peter and Kenneth J. Singleton, 1982, Generalized Instrumental Variables Estimation of Nonlinear Rational Expectations Models, Econometrica 50, 1269-1286.
- Cochrane, John H. GMM Notes
- Cochrane, John H., A Brief Parable of Overdifferencing
Guide to above readings:
- Hansen 1982 is The Article that defines GMM. Read it (at least) three times. The first time through, just understand the notation and the statement of the theorems. Find the GMM estimate defined, standard errors, test statistics. Get ready to use GMM. The second time through, read the “if” part of the proofs. Understand the stationarity, ergodicity, etc. assumptions. They matter! Finally, try to read the proofs.
- Hansen and Singleton (1982) is the crucial application to the consumption based model.
- GMM Notes is a written version of my notes for the lectures. It’s not exactly one to one, I condensed the lectures. The lectures and pdfs of the whiteboards should be enough. This is one place to turn if those are confusing, and hence just an optional resource.
- The Brief Parable of Overdifferencing is a good example for the “Choosing a W matrix” lecture, showing you how statistical efficiency can lead to bad estimators.