This dissertation consists of essays that study exchange rate pass-through, China’s de facto exchange rate regime, and China’s capital controls. The first essay studies exchange rate pass-through (ERPT) by using a set of data from ten countries including four advanced economies and six Asian emerging markets. The price indices used in this essay include consumer price, producer price, import price and export price indices. While most literature only include the import price index, this essay also puts emphasis on the export price index. It investigates the asymmetry in the ERPT between depreciation and appreciation of domestic currency by using a non-linear OLS model; meanwhile, the short-run and long-run effects of ERPT are also compared with each other. It also detects possible structural change in the ERPT and finds most structural change points are around the Great Recession and Asia financial crisis. Finally, a VAR model is developed to detect the impulse responses of prices to exchange rate shock. The second essay is about China’s exchange rate regime. It has changed a lot since the 2005 reform. It is interesting and important to investigate China’s de facto exchange rate regime with the most recent data. This essay follows Frankel and Wei’s (2008) method, by applying both the basic model and new model with the exchange market pressure (EMP) variable to currency basket for the Chinese yuan exchange rate. I select the US dollar, the Euro, the British pound, the Japanese yen, the Canadian dollar, the Australian dollar and the Russian ruble as component currencies of the basket, based on free floaters, GDP and trade volume. I also add results from a VAR model, considering the endogeneity issue, and the results are consistent with those of OLS. I find the weight of the US dollar declines dramatically and the variation of the Chinese yuan becomes much larger after 2015. This implies that China has been transferring its exchange rate regime from dollar pegged to free floating. The third essay investigates the effectiveness of China’s capital controls. In recent years, after 2014, China’s foreign reserves declined dramatically, from 4 trillion US dollars to 3 trillion US dollars. There was a huge amount of capital outflows from China during 2015 to 2016. This phenomenon lets us reconsider the question: Are China’s capital controls still effective? I will use five methods to measure the effectiveness of China’s capital controls, including de jure indicators, saving-investment correlation test, covered interest rate parity, real interest rate differentials and Edwards-Kahn model. The de jure indicators I use are from Fernández et al. (2016) and Chinn and Ito (2008). I compare China with the US, the UK and Japan in the saving-investment correlation test, and with the Eurozone and Japan in covered interest rate parity, real interest rate differentials and Edwards-Kahn model. Various results indicate that China’s capital controls are still effective.

This research analyzes the Brazilian exchange rate system and the exchange rate interventions after flexible exchange rate system implementation in 1999, and have like base the "fear of floating" approach, the exchange rate interventions theory and of the fact exchange rate classification. The evidences found for Brazil suggest that the exchange rate interventions illustrate objectives of the exchange rate debt reduction and do not follow the "fear of floating objectives". In the others empirical tests find a weak short run relation between exchange rate and domestic interest rate, exchange rate backups, domestic inflation and public debt in proportion to the GDP, and a strong short run relation between exchange rate and country risk, indicating that the movements of the exchange rate do not affect heavy the variables of the "fear of floating" and that do not believe that the country suffer from that problem. Confirming that idea, finds that of the fact exchange rate system is compatible with the jure exchange rate system, despite it find low exchange rate flexibility. / Esta pesquisa analisa o sistema cambial brasileiro e as intervenções cambiais após a implementação do sistema cambial flexível em 1999 e tem como base a abordagem do medo de flutuar , a teoria de intervenções cambiais e a classificação de sistemas cambiais de facto. As evidências encontradas para o Brasil sugerem que as intervenções cambiais ilustram objetivos de redução da dívida cambial e não seguem os objetivos da abordagem do medo de flutuar . Em outros testes empíricos encontra-se uma fraca relação de curto prazo entre taxa de câmbio e taxa de juros domésticos, reservas cambiais, inflação doméstica e dívida pública em proporção ao PIB, e uma forte relação de curto prazo entre taxa de câmbio e risco país, indicando que os movimentos da taxa de câmbio não afetam pesadamente as variáveis da abordagem do medo de flutuar e que, portanto, não há razões para crer que o país sofra desse problema. Confirmando essa idéia, encontra-se que o sistema cambial de facto é compatível com o sistema cambial de jure, apesar de encontrar-se uma baixa flexibilidade cambial. / Mestre em Economia