Start mastering your classes. Access 3M+ class-specific study documents.

Generalized Least Squares and Feasible Generalized Least Squares (2 pages)

Previewing page 1 of actual document.
View Full Document

Previewing page 1 of actual document.

View Full Document
View Full Document

Generalized Least Squares and Feasible Generalized Least Squares

603 views

18

Lecture number:
18
Pages:
2
Type:
Lecture Note
School:
Cornell University
Course:
Econ 3120 - Applied Econometrics
Edition:
1
Documents in this Packet
• 1 pages

• 1 pages

• 2 pages

• 1 pages

• 2 pages

• 1 pages

• 1 pages

• 1 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

• 1 pages

• 2 pages

• 3 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

• 6 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

• 2 pages

Unformatted text preview:

Lecture 18 Outline of Current Lecture I. Heteroskdasticity Current Lecture II. Generalized Least Squares and Feasible Generalized Least Squares Generalized Least Squares and Feasible Generalized Least Squares As described above, we can deal with heteroskedasticity using OLS robust standard errors, but this is not the most efficient way to estimate theÎ²â€™s. This section outlines how to perform more efficient estimation. 4.1 Generalized Least Squares Suppose (somewhat unrealistically) that we know the form of the heteroskedasticity. We will consider heteroskedasticity of the form Var(u|x) = Ïƒ 2 h(x) so that the variance can be expressed as some function of x. As an example, suppose our model is savei = Î²0 +Î²1inci +ui (1) where inci is the income of household i in a given year, and savei is savings in that year. (What economic parameter does Î²1 represent)? 4 With this model, one can imagine that Var(ui |inci) is increasing in income. The higher someoneâ€™s income, it makes sense that there is higher variance in the unexplained component of the model. In particular, suppose Var(ui |inci) = Ïƒ 2 inci Armed with this information, we can transform the model to one with homoskedastic standard errors. Suppose we divide every term in (1) by âˆš inci : savei/ âˆš inci = Î²0/ âˆš inci +Î²1inci/ âˆš inci +ui/ âˆš inci In this equation, the final term (which is still an error term) has variance Var(ui/ âˆš inci |inci) = 1 inci Var(ui |inci) = Ïƒ 2 The errors are now homoskedastic! Our transformed model satisfies assumptions MLR.1-MLR.4, so that we now have a model whose estimates will be best linear unbiased. This procedure is typically called generalized least squares. The generic form of this starts with the model: y = Î²0 +Î²1x1 +Î²2x2 +...+Î²kxk +u that satisfies MLR.1-MLR.4, and that Var(u|x) = Ïƒ 2 h(x) Then, the transformed model y/ âˆš h = Î²0/ âˆš h+Î²1x1/ âˆš h+Î²2x2/ âˆš h+...+Î²kxk/ âˆš h+u/ âˆš h will satisfy MLR.1-MLR.5, and the OLS estimates will therefore be best linear unbiased. 4.2 Feasible Generalized Least Squares The vast majority of the time, we donâ€™t know the form of the heteroskedasticity. In that case, we need to estimate it. But once we do that, we can do a procedure similar to the above. Suppose we think there is heteroskedasticity of the form Var(u|x) = Ïƒ 2 h(x) 5 which we model as Var(u|x) = Ïƒ 2 exp(Î´0 +Î´1x1 +Î´2x2 +...+Î´kxk) (2) This is not the only possible model, we could use other functions of the xâ€™s as well. Exponentiating is convenient because it imposes that the variance will always be positive. Since we need to estimate (2), we write the estimating equation as u 2 = Ïƒ 2 exp(Î´0 +Î´1x1 +Î´2x2 +... +Î´kxk)Î½ where E(Î½) = 1 and E(Î½|x) = 0. Taking logs, we get log(u 2 ) = Î±0 +Î´1x1 +Î´2x2 +...+Î´kxk +e Now weâ€™re ready to apply Feasible Generalized Least Squares: 1. Estimate the full model using OLS, collect Ë†ui 2. Create log(uË† 2 ...

View Full Document

Unlocking...