lecture_note_102316

lecture_note_102316 - Econ 5280 Applied Econometrics...

Info icon This preview shows pages 1–5. Sign up to view the full content.

View Full Document Right Arrow Icon
Econ 5280: Applied Econometrics Instructor: Professor Jin Seo Cho ©Jin Seo Cho, 2016, All Rights Reserved.
Image of page 1

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Chapter 4 Linear Model without Normal Residuals 4.1 Introduction and Preliminaries Recall that the assumptions of the classical linear model. 1. Dataset is a sequence of IID observations: {( Y t , X t ) R k + 1 t = 1 , 2 , ,n } ; 2. For some unknown β * , E ( Y t X t ) = X t β * , where X t = X t 1 X t 2 X tk , and β = β 1 β 2 β k ; 3. n t = 1 ( X t X t ) is invertible; 4. U t X t N ( 0 , σ 2 * ) , where U t = Y t - E ( Y t X t ) . From now, we remove the normality assumption in the fourth. This is too strong in the sense that most data do not reveal distributional properties to researchers. The main goal of this elimination is to acquire the applicability of model analysis. Nevertheless, we cannot expect that data analysis without the final assumption will deliver the same or better results than we could obtain from the CLM case. We instead suppose that the researcher has many observations to complement this missing information. When there are many data observations, we can derive similar results to the CLM case from the OLS estimator although the normality assumption may not hold. 46
Image of page 2
Here are some properties we exploit many times from now: 1. Law of Large Numbers (LLN): If a sequence of random variables { X t } is IID and E [ X t ] < , then 1 n n t = 1 X t a.s. E [ X t ] . 2. Central Limit Theorem (CLT): If a sequence of random variables { X t } is IID and E [ X 2 t ] < , then 1 n n t = 1 ( X t - E [ X t ]) A N [ 0 , var ( X t )] . 3. Cauchy-Schwarz Inequality • For any two sequences { a i } and { b i } ( i = 1 , 2 , ,n ) , n i = 1 a i b i n i = 1 a 2 i 1 2 n i = 1 b 2 i 1 2 . • If f ( x ) and g ( x ) are given such that b a f ( x ) 2 dx < and b a g ( x ) 2 dx < , then b a f ( x ) g ( x ) dx b a f ( x ) 2 dx 1 2 b a g ( x ) 2 dx 1 2 . • For any random variable X with PDF φ ( ) , f ( x ) and g ( x ) are such that f ( x ) 2 φ ( x ) dx < and g ( x ) 2 φ ( x ) dx < , then f ( x ) g ( x ) φ ( x ) dx f ( x ) 2 φ ( x ) dx 1 2 g ( x ) 2 φ ( x ) dx 1 2 . That is, E φ [ f ( X ) g ( X ) ] E φ [ f ( X ) 2 ] 1 2 E φ [ g ( X ) 2 ] 1 2 , where subscript φ indicates that the random variable X has PDF φ ( ) . Unless confusions otherwise arises, we will omit the subscript indicating PDF from now. 4.2 Extension 1: Model with Conditional Homoskedasticity 4.2.1 Assumptions 1. Dataset is a sequence of IID observations: {( Y t , X t ) R k + 1 t = 1 , 2 , ,n } ; 47
Image of page 3

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
2. For some unknown β * , E ( Y t X t ) = X t β * ; 3. E [ X t X t ] is positive definite ; 4. E [ X 2 tj ] < ,j = 1 , 2 ,...,k ; 5. U t X t ( 0 2 * ) with σ 2 * < , where U t = Y t - E [ Y t X t ] . 4.2.2 Properties of OLS Estimator As before, we note that there are two unknown parameters β * and σ 2 * . We estimate this by the OLS estimator and the unbiased estimator defined before: ̂ β n = ( X X ) - 1 X Y ; ̂ σ 2 n = ( n - k ) - 1 ̂ U ̂ U , where ̂ U = Y - X ̂ β n .
Image of page 4
Image of page 5
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern