서브메뉴
검색
Essays in Econometrics
Essays in Econometrics
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202103210
- ISBN
- 9798315797500
- DDC
- 310
- 서명/저자
- Essays in Econometrics
- 발행사항
- [Sl] : Northwestern University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 314 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-12, Section: B.
- 주기사항
- Advisor: Canay, Ivan A.;Bugni, Federico A.
- 학위논문주기
- Thesis (Ph.D.)--Northwestern University, 2025.
- 초록/해제
- 요약This dissertation consists of three chapters, each of which contributes to the literature on debiased machine learning, local projection, and partial identification.Chapter 1 studies the properties of debiased machine learning (DML) estimators under a novel asymptotic framework, offering insights for improving estimator performance in applications. DML is an estimation method suited to economic models in which the parameter of interest depends on unknown nuisance functions that must be estimated. It requires weaker conditions than previous methods while still ensuring standard asymptotic properties. Existing theoretical results do not distinguish between the two alternative versions of DML estimators, namely, DML1 and DML2. Under a new asymptotic framework, I demonstrate that DML2 asymptotically dominates DML1 in terms of bias and mean squared error, formalizing a previous conjecture based on the simulation results of their relative performance. Additionally, I provide guidance for improving DML2 performance in applications.Chapter 2 proposes a local projection residual bootstrap method to construct confidence intervals for impulse response coefficients of AR(1) models. My bootstrap method is based on the local projection (LP) approach and involves a residual bootstrap procedure applied to AR(1) models. I present theoretical results for my bootstrap method and proposed confidence intervals. First, I prove the uniform consistency of the LP-residual bootstrap over a large class of AR(1) models that allow for a unit root, conditional heteroskedasticity of unknown form, and serially dependent shocks. Then, I prove the asymptotic validity of the confidence intervals over the same class of AR(1) models. Finally, I show that the LP-residual bootstrap provides asymptotic refinements for confidence intervals on a restricted class of AR(1) models relative to those required for the uniform consistency of my bootstrap.Chapter 3 (with Federico Bugni, Mengsi Gao, and Filip Obradovic) studies the statistical power properties of confidence intervals (CIs) for a partially-identified parameter of interest with an interval identified set. We assume the researcher has bounds estimators to construct the CIs proposed by Stoye (2009), referred to as CI1α, CI2α, and CI3α. We also assume these estimators are "ordered'': the lower bound estimator is less than or equal to the upper bound estimator. Under these conditions, we establish two results. First, we show that CI1α and CI2α are equally powerful, and both dominate CI3α. Second, we consider a favorable situation in which there are two possible bound estimators to construct these CIs, and one is more efficient than the other. One would expect that the more efficient bounds estimator yields more powerful inference. We prove that this desirable result holds for CI1α and CI2α, but not necessarily for CI3α.
- 일반주제명
- Statistics
- 일반주제명
- Mathematics
- 키워드
- Local projection
- 키워드
- Econometrics
- 기타저자
- Northwestern University Economics
- 기본자료저록
- Dissertations Abstracts International. 86-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017357342
■00520260202103210
■006m o d
■007cr#unu||||||||
■020 ▼a9798315797500
■035 ▼a(MiAaPQ)AAI32001019
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a310
■1001 ▼aVelez Salamanca, Amilcar Enrique.
■24510▼aEssays in Econometrics
■260 ▼a[Sl]▼bNorthwestern University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a314 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-12, Section: B.
■500 ▼aAdvisor: Canay, Ivan A.;Bugni, Federico A.
■5021 ▼aThesis (Ph.D.)--Northwestern University, 2025.
■520 ▼aThis dissertation consists of three chapters, each of which contributes to the literature on debiased machine learning, local projection, and partial identification.Chapter 1 studies the properties of debiased machine learning (DML) estimators under a novel asymptotic framework, offering insights for improving estimator performance in applications. DML is an estimation method suited to economic models in which the parameter of interest depends on unknown nuisance functions that must be estimated. It requires weaker conditions than previous methods while still ensuring standard asymptotic properties. Existing theoretical results do not distinguish between the two alternative versions of DML estimators, namely, DML1 and DML2. Under a new asymptotic framework, I demonstrate that DML2 asymptotically dominates DML1 in terms of bias and mean squared error, formalizing a previous conjecture based on the simulation results of their relative performance. Additionally, I provide guidance for improving DML2 performance in applications.Chapter 2 proposes a local projection residual bootstrap method to construct confidence intervals for impulse response coefficients of AR(1) models. My bootstrap method is based on the local projection (LP) approach and involves a residual bootstrap procedure applied to AR(1) models. I present theoretical results for my bootstrap method and proposed confidence intervals. First, I prove the uniform consistency of the LP-residual bootstrap over a large class of AR(1) models that allow for a unit root, conditional heteroskedasticity of unknown form, and serially dependent shocks. Then, I prove the asymptotic validity of the confidence intervals over the same class of AR(1) models. Finally, I show that the LP-residual bootstrap provides asymptotic refinements for confidence intervals on a restricted class of AR(1) models relative to those required for the uniform consistency of my bootstrap.Chapter 3 (with Federico Bugni, Mengsi Gao, and Filip Obradovic) studies the statistical power properties of confidence intervals (CIs) for a partially-identified parameter of interest with an interval identified set. We assume the researcher has bounds estimators to construct the CIs proposed by Stoye (2009), referred to as CI1α, CI2α, and CI3α. We also assume these estimators are "ordered'': the lower bound estimator is less than or equal to the upper bound estimator. Under these conditions, we establish two results. First, we show that CI1α and CI2α are equally powerful, and both dominate CI3α. Second, we consider a favorable situation in which there are two possible bound estimators to construct these CIs, and one is more efficient than the other. One would expect that the more efficient bounds estimator yields more powerful inference. We prove that this desirable result holds for CI1α and CI2α, but not necessarily for CI3α.
■590 ▼aSchool code: 0163.
■650 4▼aStatistics
■650 4▼aMathematics
■653 ▼aDebiased machine learning
■653 ▼aLocal projection
■653 ▼aPartial identification
■653 ▼aConfidence intervals
■653 ▼aEconometrics
■690 ▼a0501
■690 ▼a0463
■690 ▼a0405
■690 ▼a0511
■71020▼aNorthwestern University▼bEconomics.
■7730 ▼tDissertations Abstracts International▼g86-12B.
■790 ▼a0163
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17357342▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


