Dynamic stochastic general equilibrium (DSGE) models are the common workhorse of modern macroeconomic theory. Whereas story-telling and policy analysis were in the forefront of applications since its inception, the forecasting perspective of DSGE models is only recently topical. In this study, we perform a post-mortem analysis of the predictive power of DSGE models in the case of Austria's Great Recession in 2009. For this purpose, eight DSGE models with different characteristics (small and large models; closed and open economy models; one and two-country models) were used. The initial hypothesis was that DSGE models are inferior in ex-ante forecasting a crisis. Surprisingly however, it turned out that not all but those models which implemented features of the causes of the global financial crisis (like financial frictions or interbank credit flows) could not only detect the turning point of the Austrian business cycle early in 2008 but they also succeeded in forecasting the following severe recession in 2009. In comparison, non-DSGE methods like the ex-ante forecast with the Global Economic (Macro) Model of Oxford Economics and WIFO's expert forecasts performed comparable or better than most DSGE models in the crisis.

Essay 1: Estimation of DSGE Models with Uninformative Priors DSGE models are typically estimated using Bayesian methods, but because prior information may be lacking, a number of papers have developed methods for estimation with less informative priors (diffuse priors). This paper takes this development one step further and suggests a method that allows full information maximum likelihood (FIML) estimation of a medium-sized DSGE model. FIML estimation is equivalent to placing uninformative priors on all parameters. Inference is performed using stochastic simulation techniques. The results reveal that all parameters are identifiable and several parameter estimates differ from previous estimates that were based on more informative priors. These differences are analyzed. Essay 2: A DSGE Model with Labor Hoarding Applied to the US Labor Market In the US, some relatively stable patterns can be observed with respect to employment, production and productivity. An increase in production is followed by an increase in employment with lags of one or two quarters. Productivity leads both production and employment, especially employment. I show that it is possible to replicate this empirical pattern in a model with only one demand-side shock and labor hoarding. I assume that firms have organizational capital that depreciates if workers are utilized to a high degree in current production. When demand increases, firms can increase utilization, but over time, they have to hire more workers and reduce utilization to restore organizational capital. The risk shock turns out to be very dominant and explains virtually all of the dynamics. Essay 3: Demand Shocks and Labor Hoarding: Matching Micro Data In Swedish firm-level data, output is more volatile than employment, and in response to demand shocks, employment follows output with a one- to two-year lag. To explain these observations, we use a model with labor hoarding in which firms can change production by changing the utilization rate of their employees. Matching the impulse response functions, we find that labor hoarding in combination with increasing returns to scale in production and a very high price stickiness can explain the empirical pattern very well. Increasing returns to scale implies a larger percentage change in output than in employment. Price stickiness amplifies volatility in output because the price has a dampening effect on demand changes. Both of these explain the delayed reaction in employment in response to output changes.

Essay 1: Expectation-Driven House Prices, Debt Default and Inflation Dynamics We contribute to the literature on dynamic stochastic general equilibrium (DSGE) models with housing collateral by including shocks to house price expectations. We also incorporate endogenous mortgage defaults that are rarely included in DSGE models with housing collateral. We use this model to study the effects of variations in house price expectations on macroeconomic dynamics and their implications for monetary policy. Model simulations show that an increase in expected future house prices leads to a decline in mortgage default rate and interest rates on household and business loans, whereas it leads to an increase in house prices, housing demand, household debt, business debt, bank leverage ratio and economic activity. In contrast to previous studies, we find that inflation is low during a house price boom. Finally, we show that monetary policy that takes into account household credit growth reduces the volatility of output and dampens a rise in housing demand, household debt and bank leverage ratio that enhances financial stability. However, a central bank that reacts to household credit growth increases the volatility of inflation. / Essay 2: House Price Expectations, Boom-Bust Cycles and Implications for Monetary Policy This essay examines the role of household expectations about future house prices and their implications for boom-bust cycles and monetary policy. Our findings are as follows. First, waves of optimism and pessimism about future house prices generate boom-bust cycles in house prices, financial activities (household debt, business debt, bank leverage, interest rates on household and business loans) and the real economy (housing demand, consumption, employment, investment and output). Second, we find that inflation declines during a house price boom and increases during a house price burst. Third, we find that monetary policy that reacts to household credit growth reduces the magnitude of boom-bust cycles and improves household welfare. Fourth, we find that the case for taking into account household credit growth becomes stronger in an economy in which the bank capital to asset ratio requirement is low, interest rates on loans and deposits adjust immediately to changes in the policy rate, or the household sector is highly indebted. / Essay 3: Credit Disruptions and the Spillover Effects between the Household and Business Sectors This essay examines the effects of credit supply disruptions in a New Keynesian DSGE model with housing collateral and working capital channels. A tightening of business credit conditions creates negative spillovers from the business sector to the household sector through labor income and housing collateral channels. A tightening of household credit conditions has negative spillover effects on the business sector via the housing collateral channel. We find that spillovers are more sensitive to changes in leverage where the shock occurs. A negative business credit shock creates upward pressure on inflation, whereas a negative household credit shock creates downward pressure on inflation. The working capital channel magnifies the response of inflation to a business credit shock, whereas it dampens the response of inflation to a household credit shock.

Negli ultimi dieci anni siamo stati testimoni di una delle più grandi crisi che il mondo ha visto. Il lavoro dei macroeconomisti è diventato più attivo, nell'urgenza di trovare la via d'uscita, molti degli strumenti applicati per la professione di economista sono stati rispolverati ed aggiornati per le nuove esigenze della crisi economica. Tra gli strumenti per la ricerca economica c'è lo modello dinamico stocastico di equilibrio generale (DSGE). Questa tesi è composta da quattro capitoli che coinvolgono l'intermediazione e/o politiche condotte dai governi o banche centrali finanziarie. I primi tre capitoli partono sul modello DSGE mentre l' ultimo su un modello macroeconomico principale-agente. Il primo (scritto come una rassegna delle principali indagini in DSGE) trata dei cicli di credito, di acceleratori finanziari, del mercato immobiliare, del settore bancario, dell'assunzione dei rischi e delle politiche monetarie. Il secondo analizza l'impatto delle variazioni tra il rapporto di leva e le riserve necessarie all'interno, che al giorno d'oggi regolano alcune delle politiche popolari. Il terzo capitolo incorpora un mercato dei prestiti interbancari per l'analisi degli shock di rischio generato nel settore bancario e come si sviluppa l'economia. Il quarto invece è un'indagine che si scosta dal modello macroeconomico principale-agente; comprende un governo attivo con le tasse e sussidi di disoccupazione. / During the last decade we were witness of one of the biggest crises that the world has seen. The job of the macroeconomists became more active, in the urgency for finding the way out; many of the tools applied for the economist profession were dusted off and updated for the new needs of the economic crisis. Among the tools for economic research there is the Dynamic Stochastic General Equilibrium Model (DSGE). This dissertation consists of four chapters involved in financial intermediation and/or policies conducted by the governments or central banks. The first three chapters depart from the DSGE model while the last is a macroeconomic agent based model. The first, written as a review of the main investigations in the DSGE, covers several fields as credit cycles, financial accelerator, housing market, banking sector, risk taking and monetary policies. The second chapter analyses the impacts of changes in the leverage ratio and the required reserves within, some of the popular regulation policies nowadays. The third chapter incorporates an interbank lending market for the analysis of risk shocks generated in the banking sector and how it is spread to the economy. The fourth chapter is an investigation that departs from the macroeconomic agent based model; it incorporates an active government with taxes and unemployment subsidies.

Submitted by PPG Economia do desenvolvimento (economia-pg@pucrs.br) on 2018-02-28T17:55:09Z No. of bitstreams: 1 CARLOS_ MARCHIONATTI__DIS.pdf: 1459801 bytes, checksum: 85429d200a6a1593f861f6d9277c6bb5 (MD5) / Approved for entry into archive by Caroline Xavier (caroline.xavier@pucrs.br) on 2018-03-06T14:11:06Z (GMT) No. of bitstreams: 1 CARLOS_ MARCHIONATTI__DIS.pdf: 1459801 bytes, checksum: 85429d200a6a1593f861f6d9277c6bb5 (MD5) / Made available in DSpace on 2018-03-06T14:16:25Z (GMT). No. of bitstreams: 1 CARLOS_ MARCHIONATTI__DIS.pdf: 1459801 bytes, checksum: 85429d200a6a1593f861f6d9277c6bb5 (MD5) Previous issue date: 2017-12-22 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior - CAPES / Pol?tica fiscal est? em debate nos dias atuais. Seus impactos no crescimento do PIB, infla??o, juros e taxa real de c?mbio trouxeram informa??es pelos artigos de Alesina (2010) e Cavalcanti e Vereda (2010 e 2015). Este trabalho vista extender os choques de pol?tica fiscal via gastos do governo usados no DSGE de Cavalcanti e Vereda (2015) em tr?s diferentes n?veis: federal, estadual e municipal. SVAR para a economia brasileira apresentaram novos par?metros para as tr?s esferas do DSGE proposto. Os resultados mostraram que, apesar de haver um aumento tempor?rio no PIB, uma pol?tica fiscal expansionista via aumento dos gastos p?blicos acarreta maior infla??o, maiores juros, taxa de c?mbio real apreciada e inicia uma recess?o. / Fiscal Policy is on debate nowadays. Its impacts on GDP growth, inflation, interest and real exchange rate brought insights by the works of Alesina (2010) and Cavalcanti and Vereda (2010 and 2015). This work aims to extend the fiscal policy shocks via government spending used in Cavalcanti and Vereda?s (2015) DSGE model into different levels: federal, state and city levels. SVARs for the Brazilian economy presented new parameters for all the three levels of the DSGE model proposed. The results presented showed that although there is a temporary increase on GDP level, an expansionary fiscal policy via government spending leads to higher inflation, higher interest rates, appreciated real exchange rate and starts a recession.

O objetivo do presente trabalho é avaliar o impacto da política fiscal sobre as variáveis de economia aberta, incluindo a taxa de câmbio. Para tanto, faz-se uso de um modelo DSGE com setor externo para o Brasil, tendo por base Grith (2007). Essa abordagem apresenta vantagens significativas em relação à literatura existente, como: (i) a presença de uma autoridade fiscal; (ii) rigidez nominal de preços e salários, (iii) uma Regra de Taylor, condizente com o sistema de Metas de Inflação; e (iv) a possibilidade de avaliar o impacto de choques gerados no país estrangeiro – no caso, os Estados Unidos –, sobre a economia local. Os resultados do modelo estimado, com dados trimestrais entre 2000 e 2011, sugerem que, entre as tributações sobre consumo, salário, capital e gastos do governo, a política fiscal que mais surte efeito sobre as variáveis do setor externo é a última. Além disso, é a política monetária que provoca o maior efeito em magnitude sobre a taxa de câmbio. / The present work aims to evaluate the fiscal policy impact on the open economy variables, including the exchange rate. In order to do this, it is used an DSGE model with external sector for Brazil, having Grith (2007) as a basis. This approach has significant advantages compared to the existing literature, such as: (i) the presence of a fiscal authority; nominal rigidity of prices and wages; (iii) a Taylor Rule, consistent with a Inflation Targeting system; and (iv) the possibility to evaluate the impact of shocks generated in the foreign country - in this case, the United States - under the local economy. The results of the estimated model suggest that among consumption, wage, capital taxations and government expenditures, the fiscal policy that has the biggest effect on the external sector variables is the last one. Furthermore, the monetary policy causes the greatest effect on the exchange rate.

A falta de coordenação das políticas fiscal e monetária no Brasil, freqüentemente, tem sido apontada como motivo para os desequilíbrios macroeconômicos que a economia enfrentou nas últimas três décadas. No período mais recente, pós-metas de inflação, diversos autores apontaram a política fiscal ativa como fator restritivo ao desempenho da política monetária. Nesse caso, a autoridade fiscal desconsidera a interferência do lado fiscal no monetário, o que implica em uma menor eficácia da política monetária. Somente a partir de uma condução de política fiscal com um comportamento passivo a política monetária poderia ser mais eficiente. Para testar a hipótese de políticas ativas e/ou passivas, estimou-se, por meio do método Bayesiano, um modelo DSGE com rigidez de preços e concorrência monopolística para a economia brasileira, baseado em Woodford (2003). Neste modelo, o superávit primário e a taxa de juros nominal são os instrumentos de política econômica. As estimações apontaram para um regime no qual as políticas fiscal e monetária foram ativas durante o período de 2000I a 2002IV. Porém, para o período posterior, de 2003I a 2008IV, a política fiscal foi passiva e a monetária ativa. / This paper seeks identify whether the way of fiscal and monetary macroeconomic policies in Brazil, to that period after inflation targets, were active way or/and passive way. For that, it’s estimated, for Bayesian methods, a model DSGE with price rigidities and monopolistic competition, in which the primary surplus and the nominal interest rates are the tools economic policy available. The lack of coordination of policies in Brazil, usually, has been identified as the reason for the macroeconomic imbalances. So, many authors pointed out the active fiscal policy, as a factor limiting the efficient performance of monetary policy. However, the analysis that relation within the framework of DSGE models is still limited, especially in applications for the Brazilian economy. The estimates of the model pointed out for a system where policies were active during the 2000/1Q to 2002/4Q both of them, and the later period, 2003/1Q – 2008/4Q, the fiscal policy behaved themselves on passive way and the monetary policy was active way.

A utilização de modelos de equilíbrio-geral estocásticos e dinâmicos (DSGE) para o estudo detalhado das relações entre variáveis econômicas reais e nominais tem crescido substancialmente nos últimos anos. Avanços computacionais recentes contribuíram significativamente para este movimento, permitindo que a modelagem DSGE se torne cada vez mais precisa, superando técnicas menos restritivas de modelagem macroeconômica. Contudo, a estimação destes modelos, usualmente realizada através de métodos Bayesianos, apresenta problemas, como a alta dependência da distribuição a priori. A principal inovação desta tese é propor uma solução para estes problemas, ao apresentar e utilizar o método de datacloning para estimar uma versão simplificada do modelo DSGE de Gali e Monacelli (2005), com o objetivo de avaliar a condução da política monetária pelo Banco Central do Brasil (BCB). Os principais resultados encontrados indicam que o BCB segue uma política anti-inflacionária, reage ao produto e a variações cambiais, além de gerar uma trajetória suave para a taxa de juros ao longo do tempo. Foram encontrados indícios de que a alteração de estratégia do BCB a partir de 2010, com a introdução de uma série de medidas macroprudenciais, não configurou quebra na condução da política monetária. / The use of dynamic stochastic general equilibrium (DSGE) models for the detailed study of the relationship between real and nominal economic variables has grown substantially in recent years. Computational advances have contributed significantly to this movement, allowing DSGE modelling to become increasingly precise, surpassing less restrictive macroeconomic modelling techniques. However, the estimation of these models, usually performed with Bayesian methods, presents problems, such as high dependence on the prior distribution. The main innovation of this thesis is to propose a solution to these problems, presenting and using the data cloning method to estimate a simplified version of Gali and Monacelli (2005)’s DSGE model, in order to assess the conduct of monetary policy by the Central Bank of Brazil (BCB). The main findings of this thesis indicate that the BCB follows an anti-inflationary policy, responds to GDP and exchange rate changes, and chooses a smooth interest rate path over time. Evidence suggests that the change in BCB’s strategy from 2010 onwards, with the introduction of a series of macroprudential measures, is not a conclusive indication of a parameter break in its reaction function.

A falta de coordenação das políticas fiscal e monetária no Brasil, freqüentemente, tem sido apontada como motivo para os desequilíbrios macroeconômicos que a economia enfrentou nas últimas três décadas. No período mais recente, pós-metas de inflação, diversos autores apontaram a política fiscal ativa como fator restritivo ao desempenho da política monetária. Nesse caso, a autoridade fiscal desconsidera a interferência do lado fiscal no monetário, o que implica em uma menor eficácia da política monetária. Somente a partir de uma condução de política fiscal com um comportamento passivo a política monetária poderia ser mais eficiente. Para testar a hipótese de políticas ativas e/ou passivas, estimou-se, por meio do método Bayesiano, um modelo DSGE com rigidez de preços e concorrência monopolística para a economia brasileira, baseado em Woodford (2003). Neste modelo, o superávit primário e a taxa de juros nominal são os instrumentos de política econômica. As estimações apontaram para um regime no qual as políticas fiscal e monetária foram ativas durante o período de 2000I a 2002IV. Porém, para o período posterior, de 2003I a 2008IV, a política fiscal foi passiva e a monetária ativa. / This paper seeks identify whether the way of fiscal and monetary macroeconomic policies in Brazil, to that period after inflation targets, were active way or/and passive way. For that, it’s estimated, for Bayesian methods, a model DSGE with price rigidities and monopolistic competition, in which the primary surplus and the nominal interest rates are the tools economic policy available. The lack of coordination of policies in Brazil, usually, has been identified as the reason for the macroeconomic imbalances. So, many authors pointed out the active fiscal policy, as a factor limiting the efficient performance of monetary policy. However, the analysis that relation within the framework of DSGE models is still limited, especially in applications for the Brazilian economy. The estimates of the model pointed out for a system where policies were active during the 2000/1Q to 2002/4Q both of them, and the later period, 2003/1Q – 2008/4Q, the fiscal policy behaved themselves on passive way and the monetary policy was active way.

En s'appuyant sur les limites de l'économétrie mises en évidence dans les débats sur la politique monétaire depuis les années 1960, cette thèse s'attache à montrer le primat de la théorie sur l'empirie et que l'économétrie ne peut pas être décisive dans la remise en cause de la théorie. Nous adoptons une démarche basée sur des arguments épistémologiques pour montrer que ces débat dépassent le clivage théorie/empirie et intègrent une différence de vision quant à l'utilité d'un modèle empirique. Le programme de recherche de la Commission Cowles s'est constitué autour d'une articulation particulière de trois éléments fondamentaux. Un référentiel théorique issu de la Théorie Générale de Keynes, un modèle formel s'appuyant sur le relatif consensus autour du schéma IS-LM et des techniques économétriques pour estimer les paramètres de ce modèle. C'est la nature et le degré d'interdépendance entre les trois éléments ci-dessus qui sont remis en cause par les monétaristes et les tenants de la modélisation VAR. Alors que les keynésiens établissent une nette distinction entre le modèle théorique et le modèle estimé, pour les monétaristes cette distinction n'est pas claire et ne leur semble pas pertinente. Sims (1980) reproche aux modèles structurels de la Commission Cowles de comporter trop d'hypothèses théoriques non testées empiriquement. Il propose de soumettre les hypothèses d'exogénéité à des tests économétriques directs et précis. Toutefois, l'indétermination empirique de la causalité dans un modèle VAR, liée au problème de l'équivalence observationnelle (Basmann, 1965), impose l'adoption d'un schéma d'identification sur la base d'a priori théorique pour identifier les chocs de politique monétaire. Ceci constitue un cas extrême du problème de la sous-détermination de la théorie par les données soulevé par la thèse de Duhem-Quine (Duhem, 1906, Quine, 1951). De plus, Hoover (2009) note que l'analyse des réponses impulsionnelles dans un VAR fournit un bon exemple de ce que Cartwright (2007) qualifie de ‘'contrefactuel imposteur''. Le développement des Modèles à Correction d'Erreurs et des modèles VAR cointégrés a permis de renouveler l'analyse des propositions monétaristes. Toutefois, les liens entre les propositions de cointégration, les notions d'équilibre de long terme et de déséquilibre de court terme sont rarement interprétés dans le cadre d'un modèle théorique rigoureux et complètement spécifié. Pour Faust et Leeper (1994), l'identification d'un modèle par l'imposition de contraintes peut s'avérer non fructueuse lorsque la théorie économique n'établit pas de distinction claire entre les dynamiques de court et de long terme. Faust et Whiteman (1997) relèvent l'absence d'un critère d'arbitrage dans ces démarches en présence de conflit entre le principe théorique et l'ajustement aux données, sinon une subordination de la théorie à l'économétrie. Parallèlement au problème de l'identification, la critique de Lucas (1976) constitue la seconde critique fondamentale à laquelle se heurtent les modèles économétriques. Lucas (1980, 1986) adopte une nouvelle posture épistémologique en considérant le modèle théorique comme une ‘'fiction'' et non plus comme un ensemble de propositions sur le comportement d'une économie réelle. Il défend l'idée d'une explication du cycle en termes de discipline de l'équilibre (Lucas, 1977). Les modèles DSGE, qui constituant les modèles de base de la Nouvelle Synthèse, sont fortement influencés par la méthodologie lucasienne et s'inscrivent dans la continuité des modèles RBC (Taouil, 2011). Benati et Surico (2009) ont établi la supériorité des DSGE par rapport aux VAR structurels (SVAR). Cet échec des SVAR est la conséquence directe des restrictions inter-équations imposées par l'hypothèse des anticipations rationnelles, tel que cela a été initialement soulevé par la critique de Sargent (1979). / The purpose of this thesis is to show the primacy of the theory over the empirics and prove that econometrics cannot be decisive to question the theory. For this, we rely on the limits of econometrics highlighted in discussions of monetary policy since the 1960s. We adopt an approach based on epistemological arguments to show that these debates go beyond the cleavage theory/empirics and that they integrate a difference of vision as to the usefulness of an empirical model. The research program of the Cowles Commission was formed around a particular articulation of three fundamental elements: a theoretical repository of Keynes' General Theory, a formal model based on the relative consensus on the IS-LM diagram and econometric techniques to estimate the parameters of this model. It is the nature and the degree of interdependence between these three elements that are contested by the monetarists and supporters of the VAR modeling. While Keynesians make a clear distinction between the theoretical model and the estimated model, this distinction is not clear and does not seem relevant to the monetarists. Sims (1980) criticizes the structural models of the Cowles Commission for including too many theoretical hypotheses empirically untested. He proposes to review the exogeneity assumptions through direct and specific econometric tests. However, the empirical indeterminacy of causality in a VAR, linked to the problem of observational equivalence (Basmann, 1965), requires the adoption of an identification scheme on the basis of a theoretical a priori to identify the monetary policy shocks. This is an extreme case of the problem of under-determination of theory by data raised by the Duhem-Quine thesis (Duhem 1906, Quine, 1951). Furthermore, Hoover (2009) notes that the impulse response analysis in a VAR provides a good example of what Cartwright (2007) calls “counterfactual impostor”. The development of the Error Correction Models and cointegrated VAR models has renewed the analysis of monetarist proposals. However, the links between the proposals for cointégration, the notions of long-term equilibrium and short term disequilibrium are rarely interpreted in the context of a rigorous and fully specified theoretical model. According to Faust and Leeper (1994), the identification of a model by imposing constraints may not be fruitful when economic theory does not clearly distinguish the short-term and long-term dynamics. Faust and Whiteman (1997) note the absence of an arbitration criterion in these approaches apparent in the presence of conflict between the theoretical principle and the adjustment to the data; otherwise subordination of the theory to the econometrics. Alongside the issue of identification, the Lucas critique (1976) is the second fundamental criticism facing the use of econometric models. Lucas (1980, 1986) adopts a new epistemological posture considering the theoretical model as a 'fiction' and not as a set of proposals on the behavior of a real economy. He supports the idea of explaining the cycle in terms of discipline of equilibrium (Lucas, 1977). The DSGE models, that constitute the fundamental models of the New Synthesis theory, are strongly influenced by Lucas' methodology and are a continuity of the RBC models (Taouil, 2011). Benati and Surico (2009) demonstrated the superiority of a DSGE model with respect to a structural VAR (SVAR). This failure is a direct consequence of inter-equation restrictions imposed by the rational expectations hypothesis, initially raised by Sargent's critics (1979).