La problématique de cette thèse se résume à deux questions liées aux économies en transition. La première est pourquoi quelques pays convergent vers les critères d'accession à la zone Euro, tandis que d'autres sont toujours loin de ces critères de stabilité. La deuxième question est comment a changé la structure de l'économie et la politique monétaire pendant la transition. Je réponds à ces questions en analysant des modèles dynamiques et stochastiques d'équilibre général (DSGE) en utilisant des méthodes économétriques Bayésiennes. Les techniques d'évaluation habituelles ont été étendues pour considérer des changements structurels de l'économie. Dans le premier chapitre, on a présenté la méthode d’estimation Bayésienne des modèles DSGE linéaires. Dans le deuxième chapitre, on construit un modèle DSGE incorporant quelques caractéristiques des économies en transition et l'ai évalué en utilisant la méthode Bayésienne. Enfin, dans le troisième chapitre, on estime un modèle intégrant une date de rupture structurelle dans les paramètres et de l’heteroskedasticité des chocs. / In this thesis, I have considered two questions related to transition economies in Central Europe. The first is why some countries converge toward the Euro area accession criteria, while others are still far from the stability criteria. The second question is how did the structure of the economy and the monetary policy change during the transition. I answer to these questions by analysing dynamic stochastic general equilibrium (DSGE) models using Bayesian econometric methods. I have extended the usual estimation techniques in order to consider structural changes in the economy. In the first chapter, I introduce the general methodology of Bayesian estimation of linear DSGE models. In the second chapter, I have built a DSGE model incorporating some features of the transition economies and have estimated it using the Bayesian method. Finally, in the third chapter, I have estimated a model considering a structural change date in parameters and heteroskedasticity of shocks.

Šiame darbe buvo nagrinėjamas dinaminis stochastinis stacionarus modelis aprašytas I.Carabenciov, I.Ermolaev, Ch.Freedman, M.Juillard, O.Kamenik, D.Korshunov, D.Laxton „A Small Quarterly Projection Model of the US Economy“ straipsnyje. Duomenis pateikė „Euromonitor International“ įmonė. Modelyje naudojami keturi Jungtinių Valstijų ekonomikos rodikliai: realus bendras vidaus produktas, nedarbo lygis, infliacija ir federalinių fondų palūkanų norma. Rodikliai stebimi 1994 m. І ketv. – 2009 m. ІІ ketv. laikotarpiu. Darbo tikslas – pakartoti „A Small Quarterly Projection Model of the US Economy“ straipsnio rezultatus. Buvo pasirinkti du programavimo paketai – Dynare ir Winbugs. Modelis buvo suprogramuotas dvejomis skirtingomis programomis, kurios remiasi Bajeso metodologija. Po to, gauti rezultatai buvo lyginami su straipsnyje pateiktais rezultatais. Atlikus visus skaičiavimus buvo gauti tokie rezultatai: su Dynare puikiai pavyko pakartoti modelio rezultatus. Su Winbugs programa gauti rezultatai nepilnai sutapo su straipsnyje pateiktais rezultatais. Iš to galima buvo padaryti išvada, kad Dynare programa labiau tinka dinaminių stochastinių stacionarių modelių vertinimui. Šio darbo rezultatai bus labai naudingi tyrimo planuotojams bei vykdytojams. Modelio rezultatai bus įtraukti į globalųjų modelį, kuris apjungs dar kelių valstybių modelius. / In the research paper a dynamic stochastic general equilibrium described in I.Carabenciov, I.Ermolaev, Ch.Freedman, M.Juillard, O.Kamenik, D.Korshunov, D.Laxton article „A Small Quarterly Projection Model of the US Economy“ was analyzed. The data for analysis was provided by “Euromonitor International“ company. The benchmark model has only four variables: real gross domestic product (GDP), unemployment rate, consumer price index, federal funds rate. The model is estimated over sample period from 1994QI till 2009QII. The aim of the research was to reiterate the results given in the article „A Small Quarterly Projection Model of the US Economy“. For this aim two programming packages were chosen – Dynare and Winbugs. The model was programmed using two different programs, both based on Bayesian methodology. Afterward the results were compared with results presented in the article. After all the calculations were done, the results were following: model was successfully repeated with Dynare. Although the results obtained with Winbugs program differed from the results given in the article. It therefore concluded that Dynare program is more suitable for the assessment of stationary stochastic dynamic models.

Thesis advisor: Zhijie Xiao / This dissertation contains three chapters in theoretical Macroeconometrics and applied Macroeconometrics. This first chapter addresses the issues related to the estimation, testing and computation of ordered structural breaks in multivariate linear regressions. Unlike common breaks, ordered structural breaks are those breaks that are related across equations but not necessarily occurring at the same dates. A likelihood ratio test assuming normal errors is proposed in this chapter in order to detect the ordered structural breaks in multivariate linear regressions. The estimation of ordered structural breaks uses quasi-maximum likelihood and adopts the efficient algorithm of Bai and Perron (2003). I also provide results about the consistency and rate of convergence when searching for ordered structural breaks. Finally, these methods are applied to one empirical example: the mean growth rate of output in three European countries and United States. This second chapter focuses on the parameter stability of dynamic stochastic general equilibrium (DSGE) models. To this end, I solve and estimate a representative New Keynesian model using both linear and nonlinear methods. I first examine how nonlinearities affect the parameter stability of the New Keynesian model. The results show that parameter instabilities still exist even using nonlinear solutions, and also highlight differences between two nonlinear solution methods: perturbation method and projection method. In addition, I propose a sequential procedure for searching for multiple structural breaks in nonlinear models, and apply it to the New Keynesian model. Two common structural breaks among these estimated parameters are identified for all the five solutions considered in this chapter. One structural break is in the early 1970s, while another one locates around the middle 1990s. In the third chapter, we investigate changes in long run productivity growth in the United States. In particular, we approach productivity growth from a sectoral perspective, and decompose the whole economy into two broad sectors: investment goods-producing sector and consumption goods-producing sector. Although the evidence of changes in the aggregate productivity growth is far from obvious at conventional test size, we find evidence of structural breaks in the sectoral productivity growth using both growth accounting and DSGE model based measures. There are two structural breaks in investment goods-producing sector using growth accounting measures, which indicates that the era of investment and productivity boom in the middle 1990s may have ended before the Great Recession. In addition, our results show there is one structural break in consumption goods-producing sector around the 1970s and attribute the aggregate productivity slowdown at that time to consumption goods-producing sector. These results are broadly consistent with Ireland and Schuh (2008). Our results offer up with a modestly pessimistic outlook on future productivity growth and, therefore, potential output. / Thesis (PhD) — Boston College, 2013. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Economics.

The Dynamic Stochastic General Equilibrium (DSGE) model, which is based on the New Consensus Macroeconomics (NCM) theoretical framework, has become the workhorse of macroeconomic analysis in academia, research institutes and monetary authorities since the 1980s. The dominating popularity of the DSGE type of models can be witnessed by their extensive use by central banks, such as the Bank of England (BoE), the European Central Bank (ECB), the Federal Reserve (FED) and other central banks. One of the most important and attractive advantages of the DSGE model is its compatibility with a variety of micro- and macro- economic foundations, including short-run nominal rigidities in the goods and services markets, heterogeneities in production, monetary policy and a rich set of exogenous shocks; not that there are no problems with these aspects of the DSGE model as discussed in this thesis. Although a lot of efforts have been made in DSGE modelling in industrialized economies, literature of DSGE modelling in emerging market economies is still at an early stage. The DSGE models especially designed for the economic and social features of these economies are hard to find. In this thesis, we develop a new DSGE model with special consideration of the economic and social features of emerging market economies, and account for some of the DSGE problems. The major development and innovation of this thesis is the heterogeneities not only on the supply side but also in terms of households. Additionally, the housing market and real estate assets are explicitly introduced into our model. Thirdly, we introduce a financial sector into our final model. In this sector, financial frictions are included and entrepreneurs are no longer riskless. Financial intermediates take deposits from households and then lend them to entrepreneurs at an interest rate, which is higherthanthedepositrate. Armed with these developments and improvements,the complete model in this thesis is expected to produce better empirical results and thereby more accurate explanation of economic movements in emerging market economies. Based on these models and data samples, we are able to make empirical analysis on the target economies, namely Brazil, China and India. In conclusion, the models developed in this thesis, based essentially on the DSGE type, can be the pioneer dynamic macroeconomic models for emerging market economies such as Brazil, India, and China. Based on these models, we conduct empirical analyses on data from China, Brazil, and India. We use the Bayesian estimation methodology to identify parameters in our model. The empirical results of these newly developed models show a good coherence with our theoretical hypotheses. Additionally, the performance of these models is consistent with the observed samples and the stylized facts in Brazil, China and India in terms of economic features, such as standard deviations of important economic variables including GDP and fixed asset investment. The results are promising, indicating that our DSGE type of model successfully captures the major economic features and dynamics in these countries with improved accuracy and explanatory power.

In this thesis we analyze learning mechanisms applied to a variety of macroeconomic models. In the first chapter, we present and discuss the advantages and limitations of estimating Dynamic Stochastic General Equilibrium (DSGE) models added with learning, thus suppressing the central assumption of rational expectations. First, we introduce the reader on how learning can be inserted in those models, starting from the discussion of where and how the rational expectations operator is substituted by the learning mechanism. We then present several additional learning setups related to the information set available to agents considered by the literature, which affect directly the dynamics of the final model. Last, we estimate three different models to assess the advantages of learning in our artificially generated data and real data for Brazil. In the second chapter, we algebraically show the limitations of learning and propose two flexible methods to deal with the parameter instability in data. The first of these methods is closely related to the DSGE-VAR methodology, which we call Learning DSGE-VAR, and the second, which departs even further from the DSGE model, which we call Learning Minimum State Variable, or LMSV. Finally, in the third chapter we provide evidences that the supposedly moderate improvements found in the previous chapters have more to do with the nature of the model at hand than to the learning method itself. To do so, we simulate problems using a time-varying structure similar to the one presented in chapter 1 and evaluate the likelihood improvements with different learning mechanisms. We then provide empirical evidences of learning in reduced form models to forecast inflation, interest rates and output gap for the Brazilian economy, using ad-hoc reduced form models commonly used by practitioners. / Nesta tese analisamos os instrumentos de aprendizado (Learning) aplicados a uma variedade de modelos macroeconômicos. Em nosso primeiro capítulo, apresentamos e discutimos as vantagens e limitações de se estimar modelos dinâmicos e estocásticos de equilíbrio geral (DSGE) acrescidos de um mecanismo de aprendizado, ou seja, abandonando-se a hipótese de expectativas racionais, tão cara a estes modelos. Em primeiro lugar, mostramos como esse mecanismo pode ser introduzido nesses modelos, começando pela discussão de onde e como o operador de expectativas racionais é substituído pelo operador de aprendizado. Em seguida apresentamos configurações alternativas em relação ao conjunto de informações disponível aos agentes dentro do mecanismo de aprendizado, que afeta diretamente a dinâmica do modelo final a ser estimado. Por fim, estimamos três modelos usando nosso mecanismo de aprendizado, aplicando-o a dados artificiais e reais para a economia brasileira. No segundo capítulo, mostramos algebricamente as limitações do mecanismo de aprendizado em modelos DSGE e propomos dois métodos mais flexíveis para lidar com a instabilidade dos parâmetros nos dados. O primeiro desses métodos é intimamente ligado à literatura de DSGEVAR, e que chamamos de Learning DSGE-VAR, enquanto o segundo método, que se afasta ainda mais do modelo DSGE, ao qual chamamos de LMSV. No terceiro capítulo, provemos evidências de que os ganhos supostamente moderados de nosso modelo de aprendizado apresentados nos dois primeiros capítulos têm mais a ver com a natureza dos modelos estimados do que com o método de aprendizado utilizado. Para tal, simulamos dois grupos de dados usando uma estrutura econômica que varia no tempo, semelhante àquela estudada no primeiro capítulo, e estimamos os modelos utilizando diferentes mecanismos de aprendizado. Por fim, fornecemos evidências empíricas de aprendizado em modelos de forma reduzida para projetar inflação, taxas de juros e hiato do produto para a economia brasileira, através de modelos ad-hoc comumente utilizado por econometristas.

In this thesis we analyze learning mechanisms applied to a variety of macroeconomic models. In the first chapter, we present and discuss the advantages and limitations of estimating Dynamic Stochastic General Equilibrium (DSGE) models added with learning, thus suppressing the central assumption of rational expectations. First, we introduce the reader on how learning can be inserted in those models, starting from the discussion of where and how the rational expectations operator is substituted by the learning mechanism. We then present several additional learning setups related to the information set available to agents considered by the literature, which affect directly the dynamics of the final model. Last, we estimate three different models to assess the advantages of learning in our artificially generated data and real data for Brazil. In the second chapter, we algebraically show the limitations of learning and propose two flexible methods to deal with the parameter instability in data. The first of these methods is closely related to the DSGE-VAR methodology, which we call Learning DSGE-VAR, and the second, which departs even further from the DSGE model, which we call Learning Minimum State Variable, or LMSV. Finally, in the third chapter we provide evidences that the supposedly moderate improvements found in the previous chapters have more to do with the nature of the model at hand than to the learning method itself. To do so, we simulate problems using a time-varying structure similar to the one presented in chapter 1 and evaluate the likelihood improvements with different learning mechanisms. We then provide empirical evidences of learning in reduced form models to forecast inflation, interest rates and output gap for the Brazilian economy, using ad-hoc reduced form models commonly used by practitioners. / Nesta tese analisamos os instrumentos de aprendizado (Learning) aplicados a uma variedade de modelos macroeconômicos. Em nosso primeiro capítulo, apresentamos e discutimos as vantagens e limitações de se estimar modelos dinâmicos e estocásticos de equilíbrio geral (DSGE) acrescidos de um mecanismo de aprendizado, ou seja, abandonando-se a hipótese de expectativas racionais, tão cara a estes modelos. Em primeiro lugar, mostramos como esse mecanismo pode ser introduzido nesses modelos, começando pela discussão de onde e como o operador de expectativas racionais é substituído pelo operador de aprendizado. Em seguida apresentamos configurações alternativas em relação ao conjunto de informações disponível aos agentes dentro do mecanismo de aprendizado, que afeta diretamente a dinâmica do modelo final a ser estimado. Por fim, estimamos três modelos usando nosso mecanismo de aprendizado, aplicando-o a dados artificiais e reais para a economia brasileira. No segundo capítulo, mostramos algebricamente as limitações do mecanismo de aprendizado em modelos DSGE e propomos dois métodos mais flexíveis para lidar com a instabilidade dos parâmetros nos dados. O primeiro desses métodos é intimamente ligado à literatura de DSGEVAR, e que chamamos de Learning DSGE-VAR, enquanto o segundo método, que se afasta ainda mais do modelo DSGE, ao qual chamamos de LMSV. No terceiro capítulo, provemos evidências de que os ganhos supostamente moderados de nosso modelo de aprendizado apresentados nos dois primeiros capítulos têm mais a ver com a natureza dos modelos estimados do que com o método de aprendizado utilizado. Para tal, simulamos dois grupos de dados usando uma estrutura econômica que varia no tempo, semelhante àquela estudada no primeiro capítulo, e estimamos os modelos utilizando diferentes mecanismos de aprendizado. Por fim, fornecemos evidências empíricas de aprendizado em modelos de forma reduzida para projetar inflação, taxas de juros e hiato do produto para a economia brasileira, através de modelos ad-hoc comumente utilizado por econometristas.

Esta tese é composta de três ensaios que avaliam os impactos de choques monetários sobre a economia brasileira. Sob um arcabouço de economia fechada, no primeiro ensaio, intitulado \"Choques antecipados de política monetária: uma investigação para a economia brasileira\", apresentamos o conceito de news shock em política monetária, ao mesmo tempo em que apontamos que tais choques não são tão relevantes para economia nacional quando outros choques antecipados são considerados. No segundo ensaio, intitulado \"Mudanças exógenas e antecipadas na meta de inflação e impactos sobre a economia brasileira\", estudamos como alterações nas percepções dos agentes econômicos em relação à meta de inflação futura do Banco Central explicam flutuações no produto, na inflação e também nas taxas longas de juros. No terceiro ensaio, intitulado \"Forward Guidance do Fed e impactos sobre países emergentes: o caso brasileiro\", avaliamos como sinalizações antecipadas da política monetária norte-americana geram impactos sobre as variáveis endógenas domésticas, em um arcabouço no qual o Brasil é a pequena economia aberta e também tem choque na meta de inflação / This thesis consists of three papers assessing the impact of monetary shocks on the Brazilian economy. Under a closed economy framework, in the first paper, entitled \"Anticipated monetary policy shocks: an investigation for the Brazilian economy\", we present the concept of news shock in monetary policy, at the same time that we point out that such shocks are not as relevant for national economy when other anticipated shocks are considered. The second paper \"Exogenous and anticipated changes to the inflation target and effects on the Brazilian economy\" assesses how changes to perceptions of economic agents regarding future changes to the inflation target explain changes in output, inflation and long-term market interest rates. The third paper, \"Forward Guidance by the Fed and effects on emerging economies: the case of Brazil\" assesses how anticipated changes to U.S. monetary policy affect the domestic endogenous variables, applying a small open economy framework. Brazil is the SOE and also has news shocks at the inflation target

Il y a un consensus que la sévérité des chocs sur les marchés pétroliers tend à diminuer, ainsi que la dépendance des économies développées vis-à-vis de ces chocs. Les pays développés sont généralement les importateurs d'énergie et l'effet des chocs pétroliers sur les pays exportateurs de pétrole peut être différent, surtout s’il s’agit des pays dont la grande partie de l’exportation est le pétrole ou les produits pétroliers. En outre, l'orientation sur l'exportation des matières premières peut modifier la performance relative des modèles économétriques qui sont généralement utilisés pour les prévisions. La thèse étudie et développe des modèles de l'analyse structurelle et de la prévision à court terme d'une économie exportatrice de pétrole où les données russes sont utilisées pour toutes les applications empiriques. Le premier chapitre est consacré à la construction d'un modèle DSGE pour un pays exportateur de matières premières. Le modèle DSGE est estimé par des méthodes bayésiennes. Nous constatons qu'en dépit de l'impact important sur le PIB des chocs pétroliers, les cycles économiques en Russie sont essentiellement d'origine intérieure. Le deuxième chapitre examine comment les méthodes bayésiennes peuvent être appliquées aux prévisions à l'aide d'un modèle BVAR. Le troisième chapitre applique ces techniques et compare la performance d'un groupe de modèles non structurels (univariés et multivariés) pour prévoir un ensemble d'indicateurs macroéconomiques russes. Dans le quatrième chapitre, les prévisions se sont concentrées sur les modèles structurels multivariés (DSGE) et non structurels (BVAR). Le cinquième chapitre quantifie l'effet de différents types de chocs pétroliers sur plusieurs variables macroéconomiques russes. / It is generally agreed that the severity of oil markets shocks tends to decrease as does dependence of developed economies on those shocks. Developed countries are generally energy importers, and the effect of oil market shocks on oil-exporting countries may be different, especially if energy represents a large percentage of the country’s exports. In addition, the focus on commodity exports may change the relative forecasting performance of econometric models that are generally used for forecasting. This thesis studies and develops models for structural analysis and short-term forecasting of an oil-exporting economy using Russian data for all empirical applications. The first chapter is devoted to a construction of a DSGE model for a country with commodity exports. The DSGE model is estimated by Bayesian methods We find that despite a strong impact of commodity export shocks on GDP, the business cycles in Russia are mostly domestically based.. The second chapter discusses how the Bayesian methods may be applied for forecasting with a BVAR model. The third chapter applies these techniques and compares the performance of a group of non-structural models – univariate and multivariate – for forecasting a set of Russian macroeconomic indicators. In the fourth chapter, the forecasting focuses on multivariate structural (DSGE) and non-structural BVAR models. The fifth chapter quantifies the effect of different types of oil market shocks on several Russian macroeconomic variables.

Compared to the previous twenty years, residential investments in the US appear more stable after the mid-1980s. Chapter 2 explores key hypotheses regarding the underlying causes. In particular, it uses estimated DSGE models to examine whether a more responsive interest rate policy stabilizes the housing market by keeping inflation in check. These estimations indeed found a policy that has become more responsive over time. Counter-factual analysis confirms that the change stabilizes inflation as well as nominal interest rate. It does not, however, find the change in policy to have stabilizing effect on real economic activity including housing investment. It finds that smaller TFP shocks make modest contributions, while the biggest contributing factor to the fall in the housing volatility is a reduction in the sensitivity of the investment to demand variations. Chapter 3 constructs a richly specified model for the housing market to examine the empirical relevance of various costs and frictions, including the investment adjustment cost, sticky construction costs, search frictions, and sluggish adjustment of house prices. Using the US national-level quarterly data from 1985 and 2007, we find that the gradual adjustment of house prices is the most important and irreplaceable feature of the model. The key to developing an optimization-based empirical housing model, therefore, is to provide a structural interpretation for the slow adjustment in house prices. Chapter 4 uses US national-level time series of residential investment, price index of new houses, consumption and interest rate to explore whether the US, as a nation, experienced a drop in the price elasticity of supply of new housing. Maximum likelihood estimations with a simple stock-and-flow model found a statistically significant drop of the elasticity from 10 to 2.2, when the quarterly data between 1971 and 2007 are split at 1985. A richer model with mechanisms of gradual adjustment also indicates such a reduction, when existing knowledge about the adjustment parameters is incorporated in the analysis. For the Federal Reserve, an inelastic supply can be a source of concern, because policy-driven demand in housing market is more likely to trigger undesirable swings in prices.

Compared to the previous twenty years, residential investments in the US appear more stable after the mid-1980s. Chapter 2 explores key hypotheses regarding the underlying causes. In particular, it uses estimated DSGE models to examine whether a more responsive interest rate policy stabilizes the housing market by keeping inflation in check. These estimations indeed found a policy that has become more responsive over time. Counter-factual analysis confirms that the change stabilizes inflation as well as nominal interest rate. It does not, however, find the change in policy to have stabilizing effect on real economic activity including housing investment. It finds that smaller TFP shocks make modest contributions, while the biggest contributing factor to the fall in the housing volatility is a reduction in the sensitivity of the investment to demand variations. Chapter 3 constructs a richly specified model for the housing market to examine the empirical relevance of various costs and frictions, including the investment adjustment cost, sticky construction costs, search frictions, and sluggish adjustment of house prices. Using the US national-level quarterly data from 1985 and 2007, we find that the gradual adjustment of house prices is the most important and irreplaceable feature of the model. The key to developing an optimization-based empirical housing model, therefore, is to provide a structural interpretation for the slow adjustment in house prices. Chapter 4 uses US national-level time series of residential investment, price index of new houses, consumption and interest rate to explore whether the US, as a nation, experienced a drop in the price elasticity of supply of new housing. Maximum likelihood estimations with a simple stock-and-flow model found a statistically significant drop of the elasticity from 10 to 2.2, when the quarterly data between 1971 and 2007 are split at 1985. A richer model with mechanisms of gradual adjustment also indicates such a reduction, when existing knowledge about the adjustment parameters is incorporated in the analysis. For the Federal Reserve, an inelastic supply can be a source of concern, because policy-driven demand in housing market is more likely to trigger undesirable swings in prices.