Construction and Bayesian estimation of DSGE models for the Euro area [Elektronische Ressource] / vorgelegt von Ernest Pytlarczyk

Informations

• Publié par: universitat_hamburg
• Publié le: 01 janvier 2007
• Nombre de lectures: 18
• Langue: English
• Poids de l'ouvrage: 4 Mo

Extrait

Construction and Bayesian Estimation of DSGE Models for the
Euro Area
DISSERTATION
zur Erlangung des akademischen Grades
eines Doktors
der Wirtschafts- und Sozialwissenschaften
(Dr. rer. pol.)
des Departments Wirtschaftswissenschaften
der Universität Hamburg
vorgelegt von
MSc Ernest Pytlarczyk
aus Warschau
Hamburg, den 13. Dezember 2006Mitglieder der Promotionskommission:
Vorsitzender: Prof. Dr. Georg Hasenkamp
Erstgutachter: Prof. Dr. Bernd Lucke
Zweitgutachter: Prof. Dr. Michael Funke
Das wissenschaftliche Gespräch fand am 16. Mai 2007 statt.
iACKNOWLEDGEMENTS
ManypeoplehavehelpedmeinvariouswayswhileIwasworkingonthisthesis. Greatestthanks
are due to Bernd Lucke, my supervisor, for his comments, helpful advice, guidance and continuous
support. Special thanks go to Matthias Paustian for cooperation on the joint research project and
suggestions on earlier drafts of my thesis. I am also grateful to Heinz Herrmann, the head of the
Bundesbank research centre, for giving me the opportunity to conduct part of my research at the
Bundesbank. I would also like to thank Michel Juillard and Stephane Adjemian, both affiliated
withCEPREMAP as wellas WolfgangLemkeand MichaelKrause from the Bundesbank and Keith
Küster from the ECB for discussing my research and subsequent working papers.
I also greatly benefited from comments and discussions of seminar participants and colleagues
at the University of Hamburg, Deutsche Bundesbank, the University of Frankfurt, CEPREMAP
in Paris, Georgetown University in Washington DC, the National Bank of Poland, Institut für
Wirtschaftsforschung Halle and the Centre for European Integration Studies in Bonn.
Finally, I would like to express my thanks to my friends and companions at the Universities
of Hamburg and Frankfurt, in particular Beatriz, Timo, Atilim, Olaf, Stefan, Jacopo, Omar, and
Theofanis who all took time to provide helpful comments on various draft chapters and articles.
Financial support through a doctoral scholarship from the DAAD is greatly acknowledged.
On a personal level I owe gratitude to my wife, my parents and my grandparents for their
steady and kind encouragement.
iiTo my wife
Niniejszą pracę dedykuję mojej żonie.
iiiContents
1 Introduction 1
1.1 Objective of the study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Summary of the chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 Bayesian estimation of DSGE models 4
2.1 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Basic concepts of Bayesian analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 Estimation procedure: an overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 DSGE framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4.1 Model solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4.2 Setting up a state space framework . . . . . . . . . . . . . . . . . . . . . . . 13
2.4.3 Using the model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.5 Prior distributions in DSGE model estimation . . . . . . . . . . . . . . . . . . . . . 16
2.6 Computation of the data likelihood . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.6.1 Problem of filtering and prediction . . . . . . . . . . . . . . . . . . . . . . . 17
2.6.2 Prediction, and likelihood of linear Gaussian models . . . . . . . . 18
2.7 Approximations of the posterior distribution . . . . . . . . . . . . . . . . . . . . . . 20
2.7.1 Posterior simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.7.2 Numerical optimization of the posterior . . . . . . . . . . . . . . . . . . . . 27
2.8 Model evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.8.1 Model-data fit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.8.2 Robustness of the results and identification issues . . . . . . . . . . . . . . . 32
3 Sticky contracts or sticky information? Evidence from an estimated Euro area
DSGE model 35
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 Outline of the model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.1 Households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.2 Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2.3 Market clearing, monetary policy and exogenous processes . . . . . . . . . . 40
3.3 Staggered wage and price setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.3.1 Calvo set up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.3.2 Sticky information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.4 Equilibrium and linearized equations . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.4.1 Log-linearized versions of sticky information price and wage equations . . . 45
3.5 Empirical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.5.1 The data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.5.2 Calibrated parameters and priors . . . . . . . . . . . . . . . . . . . . . . . . 47
3.6 Results for the baseline model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.6.1 Empirical properties of the seminal closed-economy DSGE model . . . . . . 48
iv3.6.2 Model comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.6.3 Extension I: Increasing the truncation point . . . . . . . . . . . . . . . . . . 57
3.6.4 II: Departing from exponential decay . . . . . . . . . . . . . . . . 61
3.7 Alternative model comparison method - models of heterogenous price and wage setters 62
3.7.1 The nested Calvo - sticky information model . . . . . . . . . . . . . . . . . 63
3.7.2 Hybrid sticky information with backward looking agents . . . . . . . . . . . 65
3.7.3 Estimation results for heterogenous agents models . . . . . . . . . . . . . . 65
3.8 Overall discussion of model dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4 Transmission of economic fluctuations in an estimated two-region DSGE model
for the Euro area 90
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.2 Related literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.3 Theoretical Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.3.1 Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
4.3.2 Households and preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
4.3.3 Relative prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4.3.4 Fiscal authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4.3.5 Market clearing conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4.3.6 Monetary authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
4.4 Macroeconomic equilibrium and model solution . . . . . . . . . . . . . . . . . . . . 107
4.4.1 Model in stationary variables . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.4.2 Solution method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4.5 Data consideration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
4.6 Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.6.1 Evaluation of the posterior . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.6.2 Calibrated parameters and specification of the priors . . . . . . . . . . . . . 118
4.7 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4.7.1 Posterior estimates of the parameters . . . . . . . . . . . . . . . . . . . . . . 120
4.7.2 Modification I: Estimation based on detrended data . . . . . . . . . . . . . 125
4.7.3 Mo II: Excluding the nominal exchange rate from the set of observ-
ables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.7.4 Empirical performance of the model . . . . . . . . . . . . . . . . . . . . . . 130
4.7.5 Role of frictions and open economy mechanisms . . . . . . . . . . . . . . . . 131
4.8 International spillover effects within the Euro area . . . . . . . . . . . . . . . . . . 132
4.8.1 Variance decomposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
4.8.2 Evidence from reduced-form models . . . . . . . . . . . . . . . . . . . . . . 136
4.8.3 Impulse responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
4.9 Out-of-sample forecasting experiment . . . . . . . . . . . . . . . . . . . . . . . . . . 144
4.9.1 Design of the experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
4.9.2 Empirical results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
4.10 Conclusions and direction of further research . . . . . . . . . . . . . . . . . . . . . 149
A Steady state 206
A.1 The two-region model . . . . . . . . . . . . . . . . . . . .