Esta tesis trata de ser una contribución a la literatura en predicción con modelos no lineales en diversas vías. En el primer capítulo, se extienden a un contexto multivariante los modelos STAR para investigar las relaciones no lineales que aparecen entre la producción de la economía americana y el Composite index of Leading Indicators (CLI). Usando la estimación máximo verosímil, se extienden al contexto VAR los test de linealidad, selección de modelos y adecuación de modelos. Se encuentra evidencia empírica en favor del uso de modelos logísticos para identificar los estados de la economía. En el segundo capítulo, se encuentra que el filtro óptimo para convertir los datos del CLI en una probabilidad de recesión es la combinación de un modelo Markov-switching y de una especificación no-paramétrica. Usando este filtro, se muestra como una tasa de crecimiento del indicador igual a cero no es nada en sí misma sino que se debe usar el filtro apropiado para que, teniendo en cuenta el estado de la economía, convierta este dato en una apropiada probabilidad de recesión. Finalmente, en el tercer capítulo se estudian relaciones de cointegración en las cuales, incluso aunque el atractor a largo plazo se asuma lineal, la fuerza con la que los errores de cointegración se desvanecen depende del estado del ciclo económico con un modelo Markoz-switching. Se muestra como este hecho conecta con la idea de tendencias estocásticas comunes no lineales. Así mismo, se estudia la repercusión sobre las funciones de respuesta al impulso y la descomposición de la varianza. / This dissertation, is an attempt to contribute to the literature on nonlinear forecasting in several ways. In the first chapter, I extend to a multiple equation framework the STAR models in order to investigate the nonlinear interactions between US output (GDP) and the Composite index of Leading Indicators (CLI). Using maximum likelihood as the base for estimation, I extend to the VAR case linearity tests, model selection tests, and model adequacy tests. Additionally, I find empirical evidence in favor of using logistic models for capturing the business cycles phases. In the second chapter, I find that the optimal filter to convert the CLI releases into a probability of recession is a combination of the forecasts from a Markov-switching model and a nonparametric specification. Using this filter, I show how a release of zero rate of growth for the CLI must be interpreted carefully since it implies different probabilities of recession depending on the actual state of the economy. Finally, in the third chapter, I study cointegrating relationships such that, even though the long-run attractor is linear, the strength with which the equilibrium errors vanishes is assumed to follow a Markov-switching dynamics. I show that this assumption connect with the idea of nonlinear common stochastic trends. In addition, I study its implication for asymmetric impulse responses and variance decomposition.

In this paper, focus is on the global vector autoregressive (GVAR) model. Its attractiveness stems from an ability to incorporate global interdependencies when modeling local economies. The model is based on a collection of local models, which in general are estimated as regular VAR models. This paper examines alternative specifications of the local models by estimating them as regime-switching VAR models, where transition probabilities between different states are studied using both constant and time-varying settings. The results show that regime-switching models are appealing as they yield inferences about the states of the economy, but these inferences are not guaranteed to be reasonable from an economic point of view. Furthermore, the global solution of the model is in some cases non-stationary when local models are regime-switching. The conclusion is that the regime-switching alternatives, while theoretically reasonable, are sensitive to the exact specification used. At the same time, the issue of specifying the regime-switching models in such a way that they perform adequately speaks in favor of the simpler, yet functional, basic GVAR model.

Esta tesis, titulada “Ensayos en Macroeconometría dinámica” se compone de tres ensayos y estudia dinámicas macroeconómicas desde una perspectiva empírica. El primer capítulo, titulado “Choques Tecnológicos y Horas Trabajadas: Nueva Evidencia de un Modelo Estructural de Factores” estudia los efectos de los choques tecnológicos sobre las horas trabajadas. Hasta ahora, el análisis se ha llevado a cabo utilizando exclusivamente los modelos estructurales vectoriales autoregresivos (SVAR) y los resultados obtenidos varían en función de la especificación de las horas trabajadas. En niveles, las horas trabajadas aumentan; expresadas en tasas de crecimiento, las horas caen. Aquí un enfoque diferente se toma. Los efectos sobre las horas son estimados usando un modelo estructural de factores dinámicos. El análisis se realiza con un conjunto de datos que contiene 102 series trimestrales de EE.UU. durante el período 1959Q1-2007Q4. El resultado principal es que un choque tecnológico positivo aumenta las horas trabajadas en el mediano y largo plazo aunque no tiene efecto en el impacto. Este hallazgo contrasta con el resultado obtenido en los modelos SVAR con horas expresadas en tasas de crecimiento ya que la respuesta es negativa. La diferencia es atribuible al hecho de que el choque tecnológico es no fundamental, lo que implica que un modelo VAR con un número finito de retardos no se puede utilizar para recuperar el choque tecnológico. El segundo capítulo, titulado “Los Determinantes de Expansiones sin Crédito” estudia las características y en particular los factores determinantes de expansiones sin crédito. Después de documentar algunos hechos estilizados de expansiones sin crédito en las economías de mercados emergentes, este análisis emplea modelos de panel probit para analizar los determinantes de las expansiones sin crédito. Nuestras principales conclusiones son las siguientes: en primer lugar, nuestro análisis de frecuencias confirma hallazgos anteriores de que expansiones sin crédito no son raros. Por otra parte, se muestra que la frecuencia de expansiones sin crédito se duplica después de una crisis bancaria o de divisas. En segundo lugar, los resultados de modelos de panel probit estimados sugieren que expansiones sin crédito suelen ser precedidos por importantes descensos en la actividad económica y la tensión financiera, en particular si el endeudamiento del sector privado es alto y el país depende de las entradas de capital extranjero. Por último, nos encontramos con que la predicción de probabilidad de una recuperación del crédito, en los Estados de Europa de este que son miembros de la UE durante los próximos años varía según los países, pero es relativamente alta en los países bálticos. Por último, el tercer capítulo, titulado “Transmisión de la Política Monetaria durante las Crisis Financieras: Una perspectiva empírica'” contesta a la pregunta de si la transmisión de la política monetaria en EE.UU. ha sido diferente durante las crisis financieras de los últimos cuarenta años. En particular, se analizan los efectos de una expansión de la política monetaria en tiempos de tensión financiera alta y en tiempos "normales". Como la pregunta a mano supone un entorno no lineal, el análisis se lleva a cabo mediante la introducción de un modelo de factores de transición suave (STFM). En este modelo, la transición entre estados (“normales” y las crisis financieras) depende de un índice de condiciones financieras que resume la información de los mercados financieros. Utilizando un conjunto de datos trimestrales durante el período 1970Q1 2009Q2 que contiene 108 EE.UU. series temporales macroeconómicas y financieras encuentro que una expansión monetaria tiene efectos más fuertes y más persistentes en las variables macroeconómicas durante las crisis financieras que. Las diferencias en los efectos entre los regímenes parecen originarse en la no linealidad en el canal de crédito. / This dissertation, titled “Essays in Dynamic Macroeconometrics’’ is comprised of three essays and analyzes macroeconomic dynamics from an empirical perspective. The first chapter, titled “Technology Shocks and Hours Worked: New Evidence from a Structural Factor Model’’ studies the effects of technology shocks on hours worked. So far, the analysis has been exclusively conducted using Structural Vector Autoregression (SVAR) models and the results obtained differ strongly depending on the specification for hours worked. In levels, the hours worked increase; in growth rates, the hours fall. Here a different approach is taken. The effects on hours are estimated using a structural dynamic factor model. The analysis is performed with a data set containing 102 quarterly U.S. macroeconomic time series over the period 1959Q1-2007Q4. In line with former VAR analysis, the technology shock is identified assuming that it is the only shock having a permanent effect on the level of labor productivity. The main result is that a positive technology shock increases hours worked in the medium and long run while having no effect on impact. The finding is in contrast with that obtained in SVAR models with hours in growth rates since there the response is negative. The difference is attributable to the fact that the technology shock is nonfundamental for the growth rates of labor productivity and hours, implying that a VAR model with a finite number of lags cannot be used to recover the technology shock. The second chapter of this dissertation, titled “The Determinants of Credit-less Recoveries” is written together with my co-author Martin Bijsterbosch and aims to shed light on the characteristics and particularly the determinants of credit-less recoveries. After building a dataset and documenting some stylized facts of credit-less recoveries in emerging market economies, this analysis uses panel probit models to study key determinants of credit-less recoveries. Our main findings are the following: first, our frequency analysis confirms earlier findings that credit-less recoveries are not at all rare events. Moreover, our analysis shows that the frequency of credit-less recoveries doubles after a banking or currency crisis. Second, results from estimated panel probit models suggest that credit-less recoveries are typically preceded by large declines in economic activity and financial stress, in particular if private sector indebtedness is high and the country is reliant on foreign capital inflows. Finally, we find that the predicted probability of a credit-less recovery in central and eastern European EU Member States during the coming years varies across countries, but is relatively high in the Baltic States. Finally, the third chapter, titled “Monetary Policy Transmission during Financial Crises: An Empirical Approach’’ aims to answer the question of whether the transmission of monetary policy in the United States has been different during the financial crises of the last forty years. In particular, I analyze the effects of a monetary policy expansion, i.e., a decrease in the Federal Funds rate, in times of high financial stress and in good or “normal” times. As the question at hand demands a non-linear environment, the analysis is carried out by introducing a Smooth Transition Factor Model (STFM). In this model the transition between states (“normal” times and financial crises) depends on a financial condition index summarizing information from financial markets. The STFM is estimated using Bayesian MCMC methods. Employing a quarterly dataset over the period 1970Q1-2009Q2 containing 108 U.S. macroeconomic and financial time series I find that a monetary expansion has stronger and more persistent effects on macroeconomic variables such as output, consumption and investment during financial crises than during “normal” times. Differences in effects among the regimes seem to originate from non-linearities in the credit channel.

This thesis suggests a Bayesian vector autoregressive (VAR) model which allows for explicit parametrization of the unconditional mean for data measured at different frequencies, without the need to aggregate data to the lowest common frequency. Using a normal prior for the steady-state and a normal-inverse Wishart prior for the dynamics and error covariance, a Gibbs sampler is proposed to sample the posterior distribution. A forecast study is performed using monthly and quarterly data for the US macroeconomy between 1964 and 2008. The proposed model is compared to a steady-state Bayesian VAR model estimated on data aggregated to quarterly frequency and a quarterly least squares VAR with standard parametrization. Forecasts are evaluated using root mean squared errors and the log-determinant of the forecast error covariance matrix. The results indicate that the inclusion of monthly data improves the accuracy of quarterly forecasts of monthly variables for horizons up to a year. For quarterly variables the one and two quarter forecasts are improved when using monthly data.

The Swedish households’ debt relative to income has increased for some time now, with the Riksbanks’ executive board expressing its concern for the risk it brings. It has been debated whether or not to take the high indebtedness into account when setting the policy rate. There is at the same time no consensus about the relationship between the repo rate and household debt. This study aims to examine the effect of a change in the repo rate on household debt-to-income ratio, using a VAR-model. The result is that a 1 percentage point shock to the repo rate for one quarter will have a negative impact on the household debt-to-income ratio by 1.75 percentage points after about 8 quarters. However this may not decrease the risk associated with the debt due to higher unemployment.

Submitted by Diogo Saraiva (diogosaraivarj@gmail.com) on 2016-06-21T18:22:50Z No. of bitstreams: 1 Tese_Diogo_Saraiva.pdf: 1438960 bytes, checksum: 75bf0c000613e5b32c7480ba2c0a1f9b (MD5) / Approved for entry into archive by GILSON ROCHA MIRANDA (gilson.miranda@fgv.br) on 2016-06-30T14:47:59Z (GMT) No. of bitstreams: 1 Tese_Diogo_Saraiva.pdf: 1438960 bytes, checksum: 75bf0c000613e5b32c7480ba2c0a1f9b (MD5) / Approved for entry into archive by Maria Almeida (maria.socorro@fgv.br) on 2016-07-13T12:51:22Z (GMT) No. of bitstreams: 1 Tese_Diogo_Saraiva.pdf: 1438960 bytes, checksum: 75bf0c000613e5b32c7480ba2c0a1f9b (MD5) / Made available in DSpace on 2016-07-13T12:51:41Z (GMT). No. of bitstreams: 1 Tese_Diogo_Saraiva.pdf: 1438960 bytes, checksum: 75bf0c000613e5b32c7480ba2c0a1f9b (MD5) Previous issue date: 2015-11-27 / The knowledge of the current state of the economy is crucial for policy makers, economists and analysts. However, a key economic variable, the gross domestic product (GDP), are typically colected on a quartely basis and released with substancial delays by the national statistical agencies. The first aim of this paper is to use a dynamic factor model to forecast the current russian GDP, using a set of timely monthly information. This approach can cope with the typical data flow problems of non-synchronous releases, mixed frequency and the curse of dimensionality. Given that Russian economy is largely dependent on the commodity market, our second motivation relates to study the effects of innovations in the russian macroeconomic fundamentals on commodity price predictability. We identify these innovations through a news index which summarizes deviations of offical data releases from the expectations generated by the DFM and perform a forecasting exercise comparing the performance of different models.

Cette thèse s'intéresse aux fondements de la croissance et des déficits budgétaires. Elle se compose de trois essais distincts. Le premier essai examine les déterminants économiques et politiques des déficits budgétaires. Le second essai s'intéresse à la transmission de l'inflation aux revenus et dépenses dans la zone euro. Le troisième essai analyse l'interaction entre les dépenses publiques et la croissance économique lors du développement économique d'un pays. Dans le premier chapitre, je présente une modélisation des biais de comportement du politicien avec un facteur reflétant la probabilité pour le gouvernement de rester au pouvoir. Si un gouvernement est incertain concernant ses perspectives de réélection, il aura tendance à négliger l'avenir davantage qu'il ne l'aurait fait autrement. Sur le plan empirique, ce papier contribue à formaliser la relation entre déficit budgétaire, comportement de l'homme politique et contexte économique. Dans le second chapitre, afin d'étudier la relation entre dépenses publiques/revenus et inflation en France, j'utilise l'analyse de cointegration et de causalité du modèle de Granger. Les résultats indiquent que les dépenses/recettes publiques et l'inflation sont fortement cointégrées et qu'il existe donc une relation d’équilibre à long terme entre ces variables. De plus, nous estimons le déficit budgétaire `a partir de l'inflation et d'autres variables explicatives telles que les élections, la croissance économique et le taux de croissance du chômage. En outre, ce chapitre souligne que les revenus (cotisations sociales, impôts directs nets) sont fortement et positivement corrélés `a l'inflation en Europe Occidentale. En revanche, les dépenses publiques sont faiblement mais positivement corrélées `a l'inflation. Dans le troisième chapitre, nous analysons l'influence de la composition des dépenses publiques sur le taux de croissance au cours du développement économique. Nous constatons une forte corrélation entre le développement du pays et l'utilité des citoyens au fil du temps. Les pays pauvres déboursent fortement en dépenses productives tandis que les pays riches ont une proportion plus élevée de dépenses improductives. Nous illustrons nos résultats avec un panel de données de 147 pays. En utilisant une estimation GMM en système sur panel dynamique, nous constatons que les salaires publics, les paiements d'intérêt, les subventions et les dépenses de consommation finale du gouvernement ne stimulent pas la croissance contrairement aux dépenses d'éducation et de santé. En outre, nous observons que la réaffectation des dépenses d'éducation est associée à une augmentation de la croissance. / This dissertation addresses the interconnections between growth and public finance. It is made of three distinct essays. The first essay investigates the economic and political determinants of budget deficits. The second essay focuses on the transmission of inflation to public finance in the euro zone. The third essay analyses the two-ways interaction between government spending and economic growth over the course of a country's economic development. In the first essay, I consider the modelling of politician behaviour bias which diverges from the typical citizen by a factor reflecting the probability for governments to stay in power. If a government is sure it will stay in power, it will discount the future in an optimal way, otherwise it will tend to discount the future more heavily. On the empirical side, our paper contributes to formalize the impact of politician behaviour and economic context on budget deficit. In the second essay, I analyze the relationship between highly granular public expenditures/revenues growth and inflation in France. I use the cointegration analysis and Granger Causality Model and find that public expenditures/revenues and inflation are cointegrated and thus there exists a long-term equilibrium relation between these variables. We forecast in detail budget deficit in France based on inflation and other explanatory variables such as elections, economic growth and unemployment growth rates. Also, this paper emphasizes that Eurozone governments' total revenues are highly and positively correlated to inflation in particular net social contribution and direct taxes while government expenses are lowly but positively correlated to inflation. In the third essay, I investigate the influence of public expenditures composition on countries growth performance along their economic development. We find citizen utility significantly evolves as the country develops and significantly changes the role and intervention of governments. Poorer countries rely heavily on productive spending while richer countries have a higher proportion of unproductive spending. We illustrate our findings with a data panel of $147$ low-, medium- and high-income countries covering the period $1970-2008$. Using dynamic panel GMM estimators, we show that public wages, interest payments, subsidies and government consumption are not growth-enhancing while spending on education and health positively impact growth. We observe that a reallocation involving an increase in education spending is associated with higher growth.

Cette thèse présente trois essais en macroéconométrie appliquée. Leur dénominateur commun est l’emploi conjoint de méthodes non-linéaires et bayesiennes afin de rendre compte de cycles économiques. Le choix de ces méthodes s’appuie sur deux constats fondamentaux. Premièrement, la plupart des séries temporelles macroéconomiques et financières présentent de soudains changements dans leur comportement résultant d’évènements tels que les crises financières, les changements brutaux de politiques fiscales et monétaires, l’alternance de phases d’expansion et de récession, etc. La prise en compte de ces changements discontinus et occasionnels nécessite une modélisation non-linéaire, c’est-à-dire la conception de modèles dont les paramètres évoluent au cours du temps. Deuxièmement, l’analyse économétrique moderne des modèles à vecteur autorégressif (VAR) et des modèles dynamiques et stochastiques d’équilibre général (DSGE) soulève de nombreux problèmes auxquels peut répondre un cadre bayesien. Tout d’abord, les modèles DSGE correspondent à une représentation partielle et simplifiée de la réalité, cette dernière étant généralement trop compliquée pour être formalisée ou trop coûteuse en termes de ressources computationnelles ou intellectuelles. Cette mauvaise spécification, inhérente aux modèles DSGE, s’ajoute en général à une pénurie de données informatives nécessaires à l’obtention de réponses précises. Dans un cadre bayesien, le praticien introduit une information supplémentaire, une distribution a priori, qui rend l’inférence des paramètres du modèle plus accessible aux macroéconomistes. S’agissant des modèles DSGE, la distribution a priori, construite à partir d’informations microéconomiques telles que les élasticités agrégées ou les taux de croissance moyens des variables macroéconomiques à long terme, permet de déplacer la fonction de vraisemblance du modèle dans les régions économiquement interprétables de l’espace de paramètres. Ceci, en vue de parvenir à une interprétation raisonnable des paramètres structurels du modèle, rendant ainsi l’inférence beaucoup plus précise. [...] / This dissertation presents three essays in applied macroeconometrics. Their common denominator is the use of Bayesian and non-linear methods to study of business cycle fluctuations. The first chapter of this dissertation revisits the issue of whether business cycles with financial crises are different, in the euro area since 1999. To do so, I fit a vector autoregression in which equation coefficients and structural disturbance variances are allowed to change over time according to Markov-switching processes. I show that financial crises have been characterized by changes not only in the variances of structural shocks, but also in the predictable and systematic part of the financial sector. By predictable and systematic part of the financial sector, I mean equation coefficients that describe the financial behavior of the system. I then examine the role of financial sector in financial crises and standard business-cycle fluctuations. The evidence indicates that the relative importance of financial shocks (“non-systematic part”) is significantly higher in periods of financial distress than in non-distress periods, but the transmission of these shocks to the economy appears linear over time. Counterfactual analyses suggest that the systematic part of financial sector accounted for up to 2 and 4 percentage points of output growth drops during the downturn in 2001-2003 and the two recessions, respectively. The second chapter examines the quantitative sources of changes in the macroeconomic volatility of the euro area since 1985. To do so, I estimate a variety of large-scale Dynamic Stochastic General Equilibrium (DSGE) models in which structural disturbance variances are allowed to change according to a Markov-switching process. The empirical results show that the best-fit model is one in which all shock variances are allowed to switch between a low- and a high-volatility regime, where regime changes in the volatilities of structural shocks are synchronized. The highvolatility regime was in place during the pre-euro period, while the low-volatility regime has been prevailed since the euro introduction. Although the size of different types of shock differs between the two shock regimes, their relative importance remains unchanged. Neutral technology shocks and shocks to the marginal efficiency of investment are the dominant sources of business cycle fluctuations. Moreover, the decline in the variance of investment shocks coincide remarkably well with the development of the European financial market that has increased access to credit by firms and households, suggesting that investment shocks reflect shocks originating in the financial system. [...]

Submitted by Vinicius Vale Pereira (viniciusvale@gmail.com) on 2016-03-01T18:04:06Z No. of bitstreams: 1 tese.v4.1.FINAL.pdf: 1143339 bytes, checksum: e3167f8ba5e577bce7f0c40605716c5d (MD5) / Rejected by Renata de Souza Nascimento (renata.souza@fgv.br), reason: Vinicius, boa noite Para que possamos aceitar seu trabalho deverá realizar os ajustes: O título deve estar em letras maiúsculas. A ficha catalográfica após a contra capa, na parte inferior. Não deve constar números romanos. As páginas anteriores não podem estar numeradas. Centralizar os títulos Agradecimentos, Resumo e Abstract. Após ajustes, submeter novamente o trabalho. Att on 2016-03-02T00:43:38Z (GMT) / Submitted by Vinicius Vale Pereira (viniciusvale@gmail.com) on 2016-03-02T13:49:51Z No. of bitstreams: 1 tese.v4.2.pdf: 1141368 bytes, checksum: 31f0646f0c15281795a6b96bd647ef42 (MD5) / Rejected by Renata de Souza Nascimento (renata.souza@fgv.br), reason: Vinicius, boa tarde Por gentileza, colocar o código CDU no canto do lado direito, da ficha catalográfica. Att on 2016-03-02T16:18:05Z (GMT) / Submitted by Vinicius Vale Pereira (viniciusvale@gmail.com) on 2016-03-02T16:48:39Z No. of bitstreams: 1 tese.v4.2.pdf: 1141278 bytes, checksum: 9f554d94d8c1ab631bb8127fadbbe43a (MD5) / Approved for entry into archive by Renata de Souza Nascimento (renata.souza@fgv.br) on 2016-03-02T17:07:51Z (GMT) No. of bitstreams: 1 tese.v4.2.pdf: 1141278 bytes, checksum: 9f554d94d8c1ab631bb8127fadbbe43a (MD5) / Made available in DSpace on 2016-03-02T17:27:52Z (GMT). No. of bitstreams: 1 tese.v4.2.pdf: 1141278 bytes, checksum: 9f554d94d8c1ab631bb8127fadbbe43a (MD5) Previous issue date: 2016-02-03 / O presente trabalho propõe um modelo de previsão simultânea de taxas de câmbio de vários países utilizando a abordagem GVAR e analisa a qualidade destas previsões. Para isso foram utilizados dados de 10 países ou regiões de taxa de câmbio, taxas de juros e nível de preços com frequência mensal entre 2003 e 2015. As previsões foram feitas utilizando janela móvel de 60 meses e avaliadas através da comparação dos erros quadráticos médios contra o benchmark padrão, o random walk, e dos testes de Pesaran e Timmermann e de Diebold e Mariano. Foram feitas previsões out-of-sample para horizontes de 1, 3, 12 e 18 meses. Os resultados mostram que o modelo proposto não consegue superar sistematicamente o random walk, contudo apresenta algum poder de previsão em alguns casos específicos / This paper proposes a model that simultaneously forecasts foreign exchange rate for several countries using the GVAR framework and analyzes the quality of these forecasts. For this purpose, data from 10 countries or regions regarding exchange rates, interest rates and price levels on a monthly basis between 2003 and 2015 was used. The forecasting was performed using a 60 months moving window and the evaluation of these was performed comparing the root mean square errors against the standard benchmark, the random walk, and by Pesaran-Timmermann and Diebold-Mariano Tests. Out-of-sample forecasts were estimated for horizons of 1, 3, 12 and 18 months. The results show that the model cannot systematically outperform the random walk, although it has some predictive power in some specific cases.

The thesis contains four essays, covering topics in the field of real-time macroeconometrics, forecasting and applied macroeconomics. In the first two chapters, I use recent techniques developed in the "nowcasting" literature in order to analyse and interpret the macroeconomic news flow. I use them either to assess current macroeconomic conditions, showing the importance of foreign indicators dealing with small open economies, or linking macroeconomic news to asset prices, through a model that help us interpret macroeconomic data and explaining the linkages between macro variables and financial indicators. In the third chapter, I analyse the link between macroeconomic data in real-time and the yield curve of interest rates, constructing a forecasting model which takes into account the peculiar characteristics of the macroeconomic data flow. In the last chapter, I present a Bayesian Vector Autoregression model built in order to analyse the last two crisis in the Eurozone (2008-09, and 2011-12) identifying their unique characteristics with respect to historical regularities, an issue of great importance from a policy perspective. / Doctorat en Sciences économiques et de gestion / info:eu-repo/semantics/nonPublished