The growth performance of the South African economy over the past two and a half decades has been disappointing. The economy has not reached the high growth rates of the 1960s, which is desperately needed to alleviate poverty in the country. While the sources of growth have been a subject of much debate, recently the notion that the Real Exchange Rate (RER) level of a country matters for growth has attracted attention. While it is generally expected that the value of the currency should not remain constant and that the exchange rate fluctuates over time, in the long-term, it is expected to converge to an equilibrium level South Africa follows an inflation targeting framework and a free floating exchange rate regime. The exchange rate has been highly volatile since the abolishment of the dual exchange rate system in 1995. This implies that there were periods of overvaluation and undervaluation from the equilibrium level; in other words the rand experienced times of misalignment. In the event of misalignments, the RER is moved to levels which make it difficult for an economy to sustain international competitiveness over the long-run, and this is harmful to growth rates in the economy. This is especially true for countries, such as South Africa, which is heavily dependent on exports. The RER is therefore very powerful and has been argued to be the cause of loss of competitiveness and growth slowdowns. This study investigates this notion for South Africa. The main aim of this study is therefore to investigate the effects of RER misalignment on economic growth in South Africa. This implies that the study aims to determine the level of RER equilibrium, the misalignment in the real value of the rand, and how this misalignment has affected economic growth in the country. The Behavioural Equilibrium Exchange Rate (BEER) approach is followed to determine the Equilibrium Exchange Rate (EER), which allows for the use of fundamental macroeconomic variables to determine the real equilibrium level of the rand. Identified fundamental variables, which are the main drivers of the current RER in South Africa, include GDP per capita, trade openness, terms of trade, gross fixed capital formation and the real interest rate differential. A Vector Error-Correction Mechanism (VECM) is used in the estimation of the Real Equilibrium Exchange Rate (REER). Misalignment is calculated as the difference between the actual and the equilibrium real exchange rate. It is found that during the period under investigation (1985 to 2011) there have been substantial misalignments in the RER of the rand, though the currency was mostly overvalued. It is also shown that the rand does revert to its equilibrium level over time. The least square method is used to determine the effect of this RER misalignment on economic growth. Additional variables such as the initial level of GDP per capita, trade openness, terms of trade as well as gross fixed capital formation, are included in the growth specification. Trade reforms emphasise the importance of export-led growth in a commodity-rich economy, such as South Africa. The results indicate that the RER misalignment has a positive coefficient; this implies that a misalignment in the rand has not necessarily been harmful to economic growth. Therefore, it can be concluded that in the case of South Africa, misalignment is generally stimulating growth, but more so when the currency is undervalued. The results therefore show that the RER should be kept at competitive levels in order to boost economic growth in the country. The results also show support for the strategy of export-led growth in South Africa. / Thesis (MCom (Economics))--North-West University, Potchefstroom Campus, 2013.

This study analyzes validity of the relative purchasing power parity (PPP) and the uncovered interest rate parity (IRP) theories for the dollar/euro exchange rate. The period of analysis is from 1990 to 2003. The dollar/euro exchange rate represents the currencies of a country, the USA, and a region, the Euro Area. The basic data needed for this study are the dollar/euro exchange rate, and the inflation and the interest rates for the USA and the Euro Area. Since the Euro Area was officially formed on January 1st, 1999, we had difficulty in finding the data for the Euro Area. For the lacking Euro Area data, synthetic values are created by using the individual data of Euro Area countries. These synthetic values are treated as the equivalents of the actual values and are used in the parity implied dollar/euro exchange rate calculations. The parity implied dollar/euro exchange rates are compared with the actual dollar/euro exchange rates. Our results indicate that the parity implied dollar/euro exchange rates are statistically significantly different from the actual dollar/euro exchange rates. In other words, both the PPP and the IRP theories do not hold for the dollar/euro exchange rate.

The recent empirical literature has shown that the dynamic generating mechanism of macroeconomic variables can be asymmetric. Inspiring from these empirical results, this thesis uses a class of nonlinear models called smooth transition autoregressive models to investigate possible asymmetric dynamics in inflation and nominal exchange rate series of Turkey. Estimation results imply that variables under consideration contain strong nonlinearities and these can be modeled by STAR models.

Financial dollarization, defined as the substantial presence of foreign currency denominated assets and liabilities in the balance sheets of the main sectors of an economy, is a widespread phenomenon among developing economies, especially in Latin America and Turkey. Since financial dollarization often causes financial fragility and limits the effectiveness of monetary policy, the causes and consequences of it and dedollarization strategies have been placed at the forefront of policy debates especially in developing countries. The purpose of this study is to analyse the determinants of corporate sector asset and liability dollarization in ten Latin American countries (Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Mexico, Peru, Uruguay and Venezuela) and Turkey for the period 1990-2001. To this end, this study considers the effects of monetary policy stance (exchange rate flexibility and adoption of a de facto inflation targeting regime), institutional structure (governance) and macroeconomic stance variables (volatilities of inflation and real effective exchange rates) on financial dollarization. The results based on panel data estimations suggest that high and volatile inflation and depreciation of domestic currency induce a switch to dollar denominated assets and liabilities. Furthermore, exchange rate regime flexibility appears to reduce liability dollarization and encourage asset dollarization. Finally, the empirical analysis supports the hypothesis that adoption of inflation targeting regime and strengthening the institutional structure are significant in decreasing the level of financial dollarization.

I.1 Motivation Exchange rates are a key issue in international economics and politics. While the determinants of exchange rates have been extensively studied in previous works, this dissertation contributes to the literature by deriving exchange rate expectations from stock market (ADR) data and analyzing their determinants. This exercise is done for three cases where one has to resort to exchange rate expectations since the national exchange rate is either manipulated by the central bank (the first paper in Chapter II), fixed in pegged exchange rate regimes (the second paper in Chapter III), or not existent as the considered country is part of a currency union and therefore has no national currency (the third paper in Chapter IV). The first paper presented in Chapter II analyzes exchange rate expectations for the case of China in the period 1998-2009 in order to test standard exchange rate theories. American officials repeatedly accused China of systematically undervaluing its currency against the U.S. dollar , which produces political tensions between both countries. A recent climax in this dispute was reached on September 28, 2010, when the House of Representatives passed the Currency Reform for Fair Trade Act, which would allow the imposition of import duties for countries with undervalued currencies, namely China. Although this bill did not pass the Senate, Chinese officials clearly opposed the bill arguing against significant undervaluation of the yuan and in favor of political opportunism of U.S. officials. As the assessments of a fair exchange rate significantly differ among officials of both countries, the Chinese-American exchange rate dispute continues. Measuring the development of market determined exchange rate expectations may help to find a compromise in this international political dispute and knowing the determinants of these expectations may help to identify macroeconomic policies necessary to influence future exchange rates. The second paper presented in Chapter III investigates the development of exchange rate expectations and their determinants for the currency crisis episodes in Argentina (2001-2002), Malaysia (1998-1999), and Venezuela (1994-1996 and 2003-2007). Large devaluations of Southeast Asian and Latin American currencies were to be observed during the currency crises in the 1990’s and at the beginning of the last decade. Due to an appreciation of foreign currency denominated debt, capital withdrawals, and bank runs, for example, currency crises typically lead to significant output losses in the affected economies (Hutchison and Noy, 2002). Avoiding currency crisis outbreaks has therefore become one of the major policy goals in many developing countries, which may explain the rapid accumulation of foreign exchange reserves aimed to fend off speculative attacks in these countries. The costs of this currency crisis prevention policy are however often overseen. Since foreign exchange reserves are typically invested in U.S. Treasuries, they yield a relatively low return compared to the high cost of domestic capital in these countries. Moreover, foreign exchange reserves may lose in value as the domestic currency appreciates against the U.S. dollar (Rodrik, 2006). An alternative way to avoid the outbreak of currency crises may be to regularly adjust the official exchange rate (typically managed by the domestic central bank) to levels in line with market expectations. Knowing market-based exchange rate expectations and their determinants may therefore be a cheaper way to avoid currency crises than holding excess amounts of foreign exchange reserves. The third paper presented in Chapter IV uses daily ADR data to analyze the determinants of the risk of withdrawals from the Economic and Monetary Union (EMU) for the five vulnerable member countries Greece, Ireland, Italy, Portugal, and Spain for the period 2007-2009. The subprime lending crisis has triggered significant financial turmoil in the EMU. Banking systems were destabilized and the governments of Greece, Ireland, and Portugal had to be bailed out. Reasserting national authority over monetary policy may help domestic policymakers to address the problems caused by banking and sovereign debt crises or an overvalued euro at national discretion. While the abandonment of fixed exchange rate regimes has so far been analyzed for countries with national currencies, the financial vulnerabilities in the EMU offer a new case to study the possibility of withdrawals from a monetary union. Although a country’s membership in the EMU is typically considered irreversible, many authors agree that sovereign states can choose to leave the EMU (Cohen, 1993; Scott, 1998; Buiter, 1999; Eichengreen, 2007). The new Treaty of Lisbon now includes a provision outlining voluntary withdrawal from the Union, which may cause the members to re-think the pros and cons of remaining in the EMU. Although the European Central Bank (ECB) has implemented measures meant to support the banking sectors and governments in the EMU, autonomous national central banks would probably pursue more expansionary monetary policies. By analyzing the determinants of exchange rate expectations in the monetary union one may therefore analyze the drivers of the risk of withdrawal from the EMU. I.2 Deriving exchange rate expectations from prices of American Depositary Receipts Measuring movements in exchange rate expectations is a relatively easy task for currencies in which a liquid and free forward exchange market exists. For the cases considered in this dissertation, however, the forward exchange market either produces bad forecasts or does not even exist. For the case of China, the yuan/U.S. dollar forward exchange rate is most likely managed by the Chinese central bank in the course of its foreign exchange market intervention policies, which hampers its ability to provide good signals for the future spot market exchange rate (see, for example, Wang, 2010). For the considered member countries of the EMU, no national currencies exist and consequently forward exchange rates cannot be used. For the case of the currency crisis episodes studied in this dissertation, one could use regression-based forecasting models that employ data on macroeconomic variables in order to produce currency crisis signals (see, among others, Eichengreen et al., 1995; Frankel and Rose, 1996; Kaminsky et al., 1998; Kaminsky and Reinhart, 1999; Karmann et al., 2002). The drawback of these approaches is the nature of macroeconomic data used, which enables one to create only monthly or quarterly crisis signals based on backward-looking data. In this dissertation I use stock market data to derive exchange rate expectations, which has several advantages compared to existing approaches. First of all, the prices of the considered stocks are most probably not manipulated by central bank interventions since these stocks are traded in the United States, which enables the derivation of exchange rate expectations formed under free market conditions (also for China). The used stock market data is available for the considered EMU member countries, which facilitates the analysis of the risk of withdrawals from the EMU. Moreover, stock market data is forward-looking and available on a daily basis, which enables the derivation of more accurate and up-to-date currency crisis signals for the considered crisis episodes. In order to derive exchange rate expectations I use data on a particular type of stock called American Depositary Receipt (ADR). An ADR is a financial instrument for foreign companies to list their shares at stock exchanges in the Unites States. An ADR represents the ownership of a specific number of underlying shares of a company in the home market on which the ADR is written. While the underlying stock is denominated in the currency and traded at the stock exchange of the home market, the ADR is denominated in U.S. dollars and traded at a U.S. stock exchange. Since both types of stocks of the same company generate identical cash flows and incorporate equivalent rights and dividend claims, cross-border arbitrage implies that the ADR and its underlying stock have the same price when adjusted for the current exchange rate. When capital controls or ownership restrictions are implemented, cross-border arbitrage is not possible and the law of one price is not binding. In such an environment, information efficiency suggests that the relative prices of ADRs and their underlying stocks – which only differ with respect to the currency they are denominated in – will signal exchange rate expectations of stock market investors. Using data on relative prices (or returns) of ADRs and their underlying stocks and the current exchange rate I can calculate measures for exchange rate expectations of stock market investors. Although the papers presented in this dissertation differ with respect to the considered companies, countries, and time periods, each paper uses the same kind of data and a similar methodology to derive exchange rate expectations – relative prices or returns of ADRs and their corresponding underlying stocks. In each paper I use a panel regression framework in order to analyze the determinants of exchange rate expectations. Each of the included papers focuses on a distinct facet of exchange rate expectations. The first paper focuses on standard exchange rate theories such as the relative purchasing power parity or the uncovered interest rate parity in order to analyze the factors that drive exchange rate expectations in general. The second paper studies the determinants of currency crisis expectations. The third paper analyzes the determinants of the risk of withdrawals from the EMU as expected by ADR market investors. I.3 Contribution to the literature This dissertation adds to two strands of the literature. First, it contributes to a literature that studies the determinants of exchange rates, currency crisis outbreaks, and risk of withdrawal from the EMU. The first paper (Chapter II) contributes to a vast literature on the determinants of exchange rates. An incomplete list of exchange rate determinants analyzed in the literature includes: labor productivity (Chinn, 2000; Cheung et al., 2007); inflation rates (Lothian and Taylor, 1996; Taylor et al., 2001); interest rates (Froot and Thaler, 1990; Chinn, 2006); overvaluation of the domestic currency (Glick and Rose, 1999; Corsetti et al., 2000); or export growth (Williamson, 1994; Isard, 2007). I study the impact of these macroeconomic fundamentals on ADR investors’ exchange rate expectations for China. China makes a good case to study standard exchange rate theories since the Chinese central bank manages the official yuan/U.S. dollar exchange rate, which therefore reacts much less to changes in macroeconomic fundamentals than is suggested by theory. Using ADR market data, I can test exchange rate theories for the Chinese peg/managed float regime under free market conditions. The second paper (Chapter III) contributes to a literature, which analyzes the determinants of currency crisis outbreaks (Eichengreen et al., 1995; Kaminsky and Reinhart, 1999; Karmann et al., 2002). Existing papers employ low-frequent and backward-looking macroeconomic data to forecast currency crises. This dissertation uses ADR market data to derive more accurate and up-to-date currency crisis signals on a daily basis. Moreover, the determinants of currency crisis expectations, such as banking or sovereign debt crisis risk, can be studied using daily market-based risk proxies. The third paper (Chapter IV) contributes to a literature on the sustainability of the EMU. Several papers discuss the possibility of withdrawal from the EMU (Cohen, 1993; Scott, 1998; Buiter, 1999; Eichengreen, 2007). I present empirical evidence that daily ADR market data reflects the risk that vulnerable member countries may leave the EMU and analyzes which determinants drive this withdrawal risk perceived by ADR investors. Second, this dissertation contributes to the literature on the pricing of ADRs. A common finding in the literature is that the outbreak of a currency crisis negatively affects the returns of U.S. dollar-denominated ADRs as the devaluation of the local currency depresses the dollar value of the underlying stock (see, for example, Bailey et al., 2000; Kim et al., 2000; Bin et al., 2004). Several papers find that the introduction of capital controls (typically meant to prevent a currency crisis outbreak) can lead to a permanent violation of the law of one price between ADRs and their underlying stocks since cross-border arbitrage cannot take place (Melvin, 2003; Levy Yeyati et al., 2004, 2009; Auguste et al., 2006). Arquette et al. (2008) and Burdekin and Redfern (2009) find that the price spreads of Chinese cross-listed stocks are significantly driven by market-traded forward exchange rates. This dissertation builds on these findings and uses the relative prices (or returns) of ADRs and their underlying stocks to derive exchange rate expectations. I present empirical evidence that ADR investors’ exchange rate expectations are driven by theory-based determinants of exchange rates, currency crisis outbreaks, or the risk of withdrawal from the EMU. This analysis therefore provides new insights into the price (return) determinants of ADRs. I.4 Main findings and policy implications The findings of this dissertation may broaden the understanding of exchanger rate expectations. The results of the first paper (Chapter II) suggest that stock market investors form their exchange rate expectations in accordance with standard exchange rate theories. Based on a monthly panel data set comprised of 22 ADR/underlying stock pairs and 52 H-share/underlying stock pairs from December 1998 to February 2009 I find that stock market investors expect more yuan appreciation against the U.S. dollar: if the yuan’s overvaluation decreases (the incentive of competitive devaluation); if the inflation differential vis-à-vis the United States falls (relative purchasing power parity); if the productivity growth in China accelerates relative to the United States (the Harrod-Balassa-Samuelson effect); if the Chinese interest rate differential vis-à-vis the United States decreases (uncovered interest rate parity); when Chinese domestic credit relative to GDP decreases (lower risk of a twin banking and currency crisis); or, if Chinese sovereign bond yields fall (lower risk of a twin sovereign debt and currency crisis), ceteris paribus. These findings suggest that the theoretical links between macroeconomic variables and exchange rates in most cases also apply to exchange rate expectations. In this way, the results support the validity of many exchange rate theories and substantiate the rationality of stock market investors’ expectations. This approach (based on stock prices formed under free market conditions) provides an opportunity to test exchange rate theories in managed floating regimes, where the official exchange rate is manipulated by the central bank and does therefore not necessarily respond to changes in macroeconomic fundamentals. Moreover, I use a rolling regressions forecasting framework in order to evaluate the quality of exchange rate expectations. I find that exchange rate expectations drawn from the ADR and H-share market have a better ability to predict changes in the yuan/U.S. dollar exchange rate than the random walk or forward exchange rates, at least at forecast horizons longer than one year. The People’s Bank of China may take advantage of ADR and H-share based exchange rate expectations in order to determine possible misalignments of the yuan/U.S. dollar exchange rate. In this way, the Chinese central bank may improve the timing and intensity of foreign exchange market interventions meant to manipulate the yuan/U.S. dollar exchange rate. The second paper (Chapter III) focuses on the derivation and determination of currency crisis signals formed by ADR market investors. Using daily data on 17 ADR/underlying stock pairs for the capital control episodes in Argentina (2001-2002), Malaysia (1998-1999), and Venezuela (1994-1996 and 2003-2007) we find that ADR investors anticipate currency crises or realignments well before they actually occur. Policymakers could use ADR investors’ up-to-date assessment of the peg’s sustainability in order to identify currency crisis risk earlier and to take the necessary steps to realign an (unsustainable) peg rate before a crisis breaks out. In this way, they could prevent the outbreaks of damaging currency crises without holding excess amounts of costly foreign exchange reserves. Using panel regressions we find that ADR investors anticipate a higher currency crisis risk when export commodity prices fall, the currencies of trading partners depreciate, sovereign bonds yield spreads rise, and interest rate spreads increase. These findings suggest that ADR investors’ currency crisis expectations are based on currency crisis theories even on a daily basis underlining the validity of these theories. The third paper (Chapter IV) studies a particular form of currency crisis risk: the risk that vulnerable member countries could leave the EMU. I use a multifactor pricing model to test whether the financial vulnerability measures assumed to reflect the incentives of national governments to withdraw from the EMU (banking crisis risk, sovereign debt crisis risk, and overvaluation of the euro) are priced in ADR returns. Using daily data on 22 ADR/underlying stock pairs of Greece, Ireland, Italy, Portugal, and Spain in the period January 2007 to March 2009 I find that ADR investors perceive a higher risk of withdrawal (priced in ADR returns) when the risk of banking and sovereign debt crisis and the overvaluation of the euro increase. Policymakers could use ADR market data in order to assess the stability of the EMU. Higher correlations between ADR returns and currency crisis risk factors would suggest a higher risk of withdrawals from the EMU. In such a case, financial vulnerabilities may be addressed within the EMU in order to preserve the integrity of the eurozone. However, time will show how long the policymakers in the EMU will continue with the implementation of even more anti-crisis measures. Growing controversies on the ECB’s sovereign bond purchases and the bailouts for Greece, Ireland and Portugal cast doubt on the sustainability of the EMU in its current form.

-Exchange Rate Dynamics under Financial Market FrictionsThis paper extends Dornbusch's overshooting model by proposing "generalized interest parity condition", which assumes sluggish adjustment on the asset market. The exchange rate model under the generalized interest parity condition is able to reproduce the delayed overshooting of nominal exchange rates and the hump-shaped response to monetary shocks of both nominal and real exchange rates.-Electoral Cycle of Exchange Rate in KoreaThis paper empirically investigates the real exchange rate behavior around elections in Korea. We find that the real exchange rate depreciates more before the elections but there is no clear pattern found after the elections. Interestingly, this result is the opposite of the electoral cycle found in Latin American countries. To explain this results we should consider the difference between economic backgrounds of Korea and Latin American countries.-Exchange Rate Regime, Capital Market Openness and Monetary Policy; The Trilemma in KoreaThis paper tests the trilemma proposition by performing an empirical study of Korea. Korea has distinct periods of all combinations of exchange rate regime and capital market openness in trilemma: pegged exchange rate regime under capital controls, pegged exchange rate regime under free capital mobility, and floating exchange rate regime under free capital mobility. We check whether monetary autonomy exists in each of the three different combinations. We find that monetary autonomy existed over the periods with capital controls and the periods with floating exchange rate regime. For the periods with the pegged exchange rate regime and free capital mobility, monetary autonomy was limited. In addition, we identify that just before the financial crisis the government pursued autonomic monetary policy under pegged exchange rate regime and free capital mobility, thereby defying the trilemma.

Τα πολυεπίπεδα νευρωνικά δίκτυα έχουν εφαρμοστεί στο παρελθόν με μεγάλη επιτυχία στην πρόβλεψη χρονοσειρών από το χώρο της οικονομίας. Στην πράξη όμως παρουσιάζουν διάφορα προβλήματα όπως: Εύρεση του βέλτιστου υποσυνόλου χαρακτηριστικών και χρησιμοποίησή τους σαν εισόδου. Εύρεση της βέλτιστης δομής (επίπεδα κρυφών νευρώνων, αριθμός κρυφών νευρώνων). Εύρεση των βέλτιστων τιμών των παραμέτρων του αλγορίθμου εκπαίδευσης των τεχνητών νευρωνικών δικτύων (παράμετρος μάθησης, παράμετρος ορμής κλπ.) Σκοπός της διπλωματικής αυτής εργασίας είναι η δημιουργίας μιας υβριδικής μεθόδου γενετικών αλγορίθμων και νευρωνικών δικτύων. Ο γενετικός αλγόριθμος θα είναι υπεύθυνος στην εύρεση του βέλτιστου συνδυασμού των παραπάνω προς αναζήτηση παραμέτρων του νευρωνικού δικτύου. Η υβριδική αυτή μέθοδος θα εφαρμοστεί στο πρόβλημα της πρόβλεψης του δείκτη ASE-20 του ελληνικού χρηματιστηρίου καθώς και στο πρόβλημα της πρόβλεψης της ισοτιμίας δολαρίου-Ευρώ. / In the present thesis we attempted to create a combination of genetic algorithms and neural networks. The proposed methodology was applied to the problem of predicting the exchange rate between EUR/USD and the Greek stock market ASE 20 index. The idea of combining these two techniques for the solving of the above mentioned problems emerged by their innate ability of finding solutions where traditional methods fail. On one hand, neural networks imitate the human brain procedures and on the other, genetic algorithms imitate the physical evolution process. In fact, both techniques copy some of nature’s functions. Artificial neural networks, through educating and generalizing manage to learn a problem and provide solutions to it. Genetic algorithms, through the evolution circle can overcome local minima or maxima and reach global ones. Due to their compact, parallel and distributed format and their ability of learning, neural networks make the solving of complicated problems possible, by dividing them in smaller projects, which are taken over by the neural networks according to their capabilities. Due to the advantages of neural networks and genetic algorithms we created a combination of them to predict the exchange rate between EUR/USD and the Greek stock market ASE 20 index.

Esta pesquisa investiga o comportamento não-linear do pass-through de taxa de câmbio na economia brasileira, durante o período de câmbio flutuante (2000-2015), a partir de um modelo de equilíbrio geral dinâmico estocástico com mudança de regime Markoviana (MS-DSGE). Para isso, utilizamos a metodologia proposta por Baele et al. (2015) e um modelo Novo-Keynesiano básico, sobre o qual incluímos novos elementos na curva de oferta agregada e uma nova equação para a dinâmica cambial. Encontramos evidências de existência de dois regimes distintos para o repasse cambial e para a variância dos choques sobre a inflação. No regime denominado de “Normal”, o pass-through de longo prazo é estimado em 0.0092 pontos percentuais para inflação, dado um choque cambial de 1%, contra 0.1302 pontos percentuais no regime de “Crise”. A superioridade do modelo MS-DSGE sobre o modelo com parâmetros fixos é constatada de acordo com diversos critérios comparativos. / This research investigates the non-linearity of exchange rate pass-through on the Brazilian economy during the floating exchange rate period (2000-2015) in a Markov-switching dynamic stochastic general equilibrium framework (MS-DSGE). We apply methods proposed by Baele et al. (2015) in a basic New Keynesian model, with the addition of new elements to the aggregate supply curve and a new equation for the exchange rate dynamics. We find evidence of two distinct regimes for the exchange rate pass-through and for the volatility of shocks to inflation. During the regime named “Normal”, the long run pass-through is estimated as 0.0092 percent points to inflation, given a 1% exchange rate shock, in contrast to 0.1302 percent points during the “Crisis” regime. The MS-DSGE model appears superior to the fixed parameters model according to various comparison criteria.

O presente trabalho de dissertação tem o objetivo de analisar o crescimento da dívida pública brasileira no período entre 1994 a 2004. Assim, inicialmente é realizada uma revisão macroeconômica sobre o tema, e posteriormente são expostas políticas fiscal, monetária e cambial adotadas pelo governo nos últimos anos e sua influência sobre o crescimento do déficit público. Finalmente, analisa-se os fatores condicionantes da dívida, bem como eventuais propostas de solução para o problema. / This dissertation analyzes the growth in Brazilian public debt from 1994 to 2004. It begins with a review of the macroeconomic theory on the subject, followed by a description of the fiscal, monetary, and exchange rate policies adopted by the government in different years, and what influence did these policies have on public deficit growth. In the end, it analyzes the factors that cause the debt, as well as some propositions to solve the problem.

Cette thèse a pour objectif d'étudier le lien entre les régimes de change, les mésalignements de change et les déséquilibres globaux. Elle est conduite dans le cadre d'un panel évolutif d'économies allant des pays en développement aux pays développés. Elle couvre trois principaux thèmes. Nous examinons d'abord les mésalignements de change dans la zone CFA et le projet d'union monétaire en Afrique de l'Ouest. Nous étudions ensuite les implications du choix par un pays d'un régime de change sur sa capacité de résilience aux déséquilibres externes. Enfin, nous analysons l'influence des mésalignements de change sur la persistance des déséquilibres globaux, ainsi que les interactions entre les déséquilibres macroéconomiques.Nous montrons que la monnaie ancre (l'euro) joue un rôle prédominant dans l'explication des mésalignements du franc CFA, toute chose égale par ailleurs, y compris les fondamentaux du franc CFA. Sur la base d'une nouvelle méthodologie basée sur la synchronisation des mésalignements que nous proposons, nous mettons en évidence qu'il existe des similarités entre les pays de l'UEMOA, le Ghana, la Gambie et la Sierra Leone dans le cadre d'une union monétaire. En revanche, les cycles ne sont pas suffisamment synchronisés entre ce bloc et le Nigeria qui reste structurellement plus proche des pays de la CEMAC.Concernant les pays d'Afrique subsaharienne, nous montrons que les régimes de change flexibles sont plus efficaces pour limiter déséquilibres externes et pour assurer la soutenabilité externe.Nous montrons que la persistance des déséquilibres de comptes courants est fortement et asymétriquement liés aux mésalignements de change dans les pays développés et que les déséquilibres macroéconomiques interagissent fortement à travers une relation causale. / This thesis aims to study the link between exchange rate regimes, exchange rate misalignments and external imbalances. Our research is conducted in the context of an evolving panel of economies ranging from developing countries to developed countries. Our investigation covers three main themes. We first we examine exchange rate misalignments in the CFA zone and the monetary union project in West Africa. We then we study the implications of the choices made by a country concerning exchange rate regimes with regard to their ability to be resilient to external imbalances in sub-Saharan African countries. Finally, we analyze the influence of exchange rate misalignments on the persistence of global imbalances, as well as the interactions between macroeconomic imbalances.We show that the anchor currency (the euro) plays a predominant role in explaining the CFA franc misalignments, all else being equal, including economics fundamentals of the CFA franc. Relying on a new methodology based on synchronization misalignments that we propose, we highlight that there are some similarities between the WAEMU countries, Ghana, Gambia and Sierra Leone. In contrast, the cycles of misalignments are not sufficiently synchronized between this block and Nigeria, which remains structurally closer to the CAEMC countries.With regard to sub-Saharan Africa, we show that the flexible exchange rate regimes are more effective in limiting external imbalances and to ensure external sustainability.We show that the persistence of current account imbalances is strongly and asymmetrically related to exchange rate misalignments in industrialized countries and that the macroeconomic imbalances strongly interact through a causal relationship.