Episódios atípicos de fluxos de capitais e crises financeiras

• Ano de defesa: 2020
• Autor(a) principal: Suzuki, Fellipe
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Economia
• Palavras-chave em Português: Fluxos de capitais; Capital Flows; Crises financeiras; Financial Crises; Surtos de capital; Surges; CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::ECONOMIA INTERNACIONAL; CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::METODOS QUANTITATIVOS EM ECONOMIA; Fluxo de capitais
• Link de acesso: https://repositorio.ufu.br/handle/123456789/30641; http://doi.org/10.14393/ufu.di.2020.3916

The objective of this dissertation is to analyze the relationship between unusual capital flows movements, identified in the literature as surges, and financial crisis, under a theorical and empirical perspective. This scientific paper contributes to the literature compiling a group of small method innovations applied in a group of different works, these incremental innovations are: the use of seven surges identification methods, control to a post-crises bias, estimators that treat unobserved heterogeneity and the use of ROC curve as tool to evaluate the performance of the estimators and models. An Early Warning System (EWS) model, based on a multinomial probit model is estimated to test the relationship between surges in capital flows and financial crises for a sample of 161 advanced economies and emerging markets and developing countries for the period of 1970-2017. The results indicate that statistically significant relationships were found between atypical episodes in capital flows and bank crises for all seven methods of identifying atypical episodes used, for total sample the results vary from 6.7% (Surge 1) to 2.3% (Surge 7), the increase in probability in advanced countries is greater than that observed in the total sample (and consequently in emerging countries) ranging from 8.5% (Surge 2) to 5.1% (Surge 4); in the case of currency crises, atypical episodes (identified by Surge 4) increase the probability of currency crises by 2.7% for the total sample, 3.4% for advanced economies and 2.5% for emerging and developing economies ; no statistically significant relationships were identified between atypical episodes in capital flows and sovereign crises. Having the preexisting literature at basis and the empirical tests developed in this paper, it can be stated that are evidences of relationships between capital surges, banking and currency crisis, where capital surges increase the probability of occurrence of this crisis, however the baseline estimated was considerably smaller than other researches. The relationship between capital surges and sovereign crises could not be established.