Essays on Panel Cointegration
(2005) In Lund Economic Studies no. 129.- Abstract
- This thesis develops new techniques for analyzing cointegrated relationships in panel data. The first chapter is introductory while the remaining six contain the main contributions.
The second chapter is concerned with the estimation of a cointegrated panel relation with endogenous regressors, in which case the least squares estimator is biased unless it is conditioned on the lags and leads of the differences of the regressors. The problem is how to choose the appropriate number of lags and leads. This issue is illuminated by examining the performance of several information criteria that facilitate a data dependent choice. The Monte Carlo evidence suggests that the criterion with the best performance also leads to the... (More) - This thesis develops new techniques for analyzing cointegrated relationships in panel data. The first chapter is introductory while the remaining six contain the main contributions.
The second chapter is concerned with the estimation of a cointegrated panel relation with endogenous regressors, in which case the least squares estimator is biased unless it is conditioned on the lags and leads of the differences of the regressors. The problem is how to choose the appropriate number of lags and leads. This issue is illuminated by examining the performance of several information criteria that facilitate a data dependent choice. The Monte Carlo evidence suggests that the criterion with the best performance also leads to the best performing estimator.
Although cointegration is usually considered to be the most natural choice of null hypothesis, most existing panel tests are based on the null hypothesis of no cointegration. The third chapter therefore develops a new test with cointegration as the null. Asymptotic properties of the test are derived and verified in small samples via Monte Carlo simulations, and implementation is illustrated through an application of the test to international R&D spillovers.
Relationships that span extensive periods of time are prone to structural breaks. Yet, there is presently no test that is general enough to allow for such breaks. The fourth chapter takes a step in this general direction by proposing a test that allows for multiple structural breaks in the cointegration relation. Asymptotic distribution of the test is derived and critical values are provided to permit accurate testing even in small samples, which is verified using Monte Carlo simulations. An application of the test to the Feldstein-Horioka puzzle is also provided.
Empirical evidence suggests that the Fisher hypothesis does not hold, which seems at odds with many theoretical models. The fifth chapter argues that these results can be attributed to the low power of conventional time series tests and that the use of panel data can generate more powerful tests. For this purpose, two panel cointegration tests are developed that allow for cross-sectional dependence, and are shown to be more powerful than existing tests. The empirical results suggest that, based on the new tests, the Fisher hypothesis cannot be rejected.
In the sixth chapter, four new error correction based tests for the null hypothesis of no cointegration are proposed. These tests are less restrictive than most existing tests and are therefore more widely applicable, which implies that they are also expected to be more powerful. This is illustrated via simulations. The empirical application shows evidence of cointegration between health care expenditures and GDP.
The seventh chapter develops two panel cointegration tests that allow for very general forms of serial correlation structures without the need for any kind of adjustment. This makes them very simple in comparison to existing tests, which do not share this invariance property. Asymptotic distributions are derived and Monte Carlo evidence suggests that the new tests compare favorably with several other popular tests. (Less) - Abstract (Swedish)
- Popular Abstract in Swedish
Denna avhandling utvecklar nya tekniker för analys av kointegrationssamband i paneldata. Det första kapitlet är en introduktion medan de resterande sex innehåller avhandlingens huvudsakliga bidrag.
Det andra kapitlet behandlar estimering av kointegrationssamband när regressorerna är endogena, vilket medför att minsta kvadratmetoden inte längre är väntevärdesriktig om den inte betingas på tidsförskjutna värden av differensen av regressorerna. Problemet är hur antalet tidsförskjutna värden ska väljas. Det andra kapitlet belyser denna frågeställning genom att utvärdera egenskaperna hos flera informationskriterier, vilka möjliggör ett databeroende val. Simuleringsresultaten visar... (More) - Popular Abstract in Swedish
Denna avhandling utvecklar nya tekniker för analys av kointegrationssamband i paneldata. Det första kapitlet är en introduktion medan de resterande sex innehåller avhandlingens huvudsakliga bidrag.
Det andra kapitlet behandlar estimering av kointegrationssamband när regressorerna är endogena, vilket medför att minsta kvadratmetoden inte längre är väntevärdesriktig om den inte betingas på tidsförskjutna värden av differensen av regressorerna. Problemet är hur antalet tidsförskjutna värden ska väljas. Det andra kapitlet belyser denna frågeställning genom att utvärdera egenskaperna hos flera informationskriterier, vilka möjliggör ett databeroende val. Simuleringsresultaten visar att kriteriet med bäst egenskaper även leder till den bästa estimatorn.
Trots att kointegration vanligtvis betraktas som det mest naturliga valet av nollhypotes är de flesta existerande test i paneldata baserade på nollhypotesen om ingen kointegration. Det tredje kapitlet utvecklar därför ett nytt test med kointegration som nollhypotes. Testets asymptotiska egenskaper härleds och verifieras i små stickprov via simuleringsmetoder. Den praktiska implementeringen av testet illustreras genom en applikation till internationella forsknings- och utvecklingseffekter.
Benägenheten att påverkas av strukturella skift är stark hos samband som sträcker sig över långa tidsperioder. Trots detta finns det för närvarande inget test som är generellt nog för att tillåta sådana skift. Det fjärde kapitlet tar ett steg i denna generella riktning genom att utveckla ett test som tillåter flera skift i kointegrationssambandet. Testets asymptotiska fördelning härleds och kritiska värden tillhandahålls, vilka möjliggör goda testegenskaper även i mycket små stickprov. Simuleringsresultat och en applikation till Feldstein-Horioka-problemet tillhandahålls också.
Empiriska resultat tyder på att Fisher-hypotesen måste förkastas, vilket inte är förenligt med många teoretiska modeller. Det femte kapitlet argumenterar för att dessa resultat kan förklaras av den låga styrkan hos konventionella tidsserietest och att paneldata kan generera test med bättre styrka. I detta syfte utvecklas två kointegrationstest för paneldata, vilka tillåter tvärsnittskorrelation och har högre styrka än existerande test. De empiriska resultaten visar att Fisher-hypotesen inte kan förkastas baserat på dessa test.
I det sjätte kapitlet utvecklas fyra test baserade på felkorrigering, vilka testar nollhypotesen om ingen kointegration. Dessa test är mindre restriktiva än existerande test och är därför mer tillämpbara, vilket även innebär att de även förväntas ha högre styrka. Detta illustreras via simuleringar. I den empiriska applikationen frambringas bevis för att hälsovårdsutgifter och BNP är kointegrerade.
Det sjunde kapitlet utvecklar två test för kointegration i paneldata, vilka tillåter väldigt generella former av autokorrelation och som inte kräver någon form av justering. Detta gör dessa test väldigt enkla jämfört med existerande test som inte har denna egenskap. Testens asymptotiska fördelningar härleds och simuleringsresultat tillhandahålls, vilka visar att testen har goda egenskaper jämfört med andra populära test. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/545339
- author
- Westerlund, Joakim LU
- supervisor
- opponent
-
- Professor Banerjee, Anindya, Economics Department, European University Institute
- organization
- publishing date
- 2005
- type
- Thesis
- publication status
- published
- subject
- keywords
- Nationalekonomi, ekonometri, ekonomisk teori, ekonomiska system, ekonomisk politik, Cross-Sectional Dependence, Common factor restriction, Feldstein-Horioka Puzzle, Fisher Hypothesis, economic policy, economic theory, economic systems, Economics, econometrics, Structural Break., Sieve Bootstrap, Information Criteria, Residual-Based Cointegration Test, Panel Cointegration, Monte Carlo Simulation, Model Selection, International R&D Spillovers, International Health Care Expenditures
- in
- Lund Economic Studies
- volume
- no. 129
- pages
- 193 pages
- publisher
- Department of Economics, Lund University
- defense location
- EC3:210, Holger Crafoords Ekonomicentrum
- defense date
- 2005-11-12 11:15:00
- ISSN
- 0460-0029
- language
- English
- LU publication?
- yes
- id
- 19e63da7-678f-4e17-b65b-0c965e4c8992 (old id 545339)
- date added to LUP
- 2016-04-01 17:01:19
- date last changed
- 2019-05-21 16:51:55
@phdthesis{19e63da7-678f-4e17-b65b-0c965e4c8992,
abstract = {{This thesis develops new techniques for analyzing cointegrated relationships in panel data. The first chapter is introductory while the remaining six contain the main contributions.<br/><br>
<br/><br>
The second chapter is concerned with the estimation of a cointegrated panel relation with endogenous regressors, in which case the least squares estimator is biased unless it is conditioned on the lags and leads of the differences of the regressors. The problem is how to choose the appropriate number of lags and leads. This issue is illuminated by examining the performance of several information criteria that facilitate a data dependent choice. The Monte Carlo evidence suggests that the criterion with the best performance also leads to the best performing estimator.<br/><br>
<br/><br>
Although cointegration is usually considered to be the most natural choice of null hypothesis, most existing panel tests are based on the null hypothesis of no cointegration. The third chapter therefore develops a new test with cointegration as the null. Asymptotic properties of the test are derived and verified in small samples via Monte Carlo simulations, and implementation is illustrated through an application of the test to international R&D spillovers.<br/><br>
<br/><br>
Relationships that span extensive periods of time are prone to structural breaks. Yet, there is presently no test that is general enough to allow for such breaks. The fourth chapter takes a step in this general direction by proposing a test that allows for multiple structural breaks in the cointegration relation. Asymptotic distribution of the test is derived and critical values are provided to permit accurate testing even in small samples, which is verified using Monte Carlo simulations. An application of the test to the Feldstein-Horioka puzzle is also provided.<br/><br>
<br/><br>
Empirical evidence suggests that the Fisher hypothesis does not hold, which seems at odds with many theoretical models. The fifth chapter argues that these results can be attributed to the low power of conventional time series tests and that the use of panel data can generate more powerful tests. For this purpose, two panel cointegration tests are developed that allow for cross-sectional dependence, and are shown to be more powerful than existing tests. The empirical results suggest that, based on the new tests, the Fisher hypothesis cannot be rejected.<br/><br>
<br/><br>
In the sixth chapter, four new error correction based tests for the null hypothesis of no cointegration are proposed. These tests are less restrictive than most existing tests and are therefore more widely applicable, which implies that they are also expected to be more powerful. This is illustrated via simulations. The empirical application shows evidence of cointegration between health care expenditures and GDP.<br/><br>
<br/><br>
The seventh chapter develops two panel cointegration tests that allow for very general forms of serial correlation structures without the need for any kind of adjustment. This makes them very simple in comparison to existing tests, which do not share this invariance property. Asymptotic distributions are derived and Monte Carlo evidence suggests that the new tests compare favorably with several other popular tests.}},
author = {{Westerlund, Joakim}},
issn = {{0460-0029}},
keywords = {{Nationalekonomi; ekonometri; ekonomisk teori; ekonomiska system; ekonomisk politik; Cross-Sectional Dependence; Common factor restriction; Feldstein-Horioka Puzzle; Fisher Hypothesis; economic policy; economic theory; economic systems; Economics; econometrics; Structural Break.; Sieve Bootstrap; Information Criteria; Residual-Based Cointegration Test; Panel Cointegration; Monte Carlo Simulation; Model Selection; International R&D Spillovers; International Health Care Expenditures}},
language = {{eng}},
publisher = {{Department of Economics, Lund University}},
school = {{Lund University}},
series = {{Lund Economic Studies}},
title = {{Essays on Panel Cointegration}},
volume = {{no. 129}},
year = {{2005}},
}