Markov-switching models : empirical applications using classical and Bayesian inference

Detalhes bibliográficos
Ano de defesa: 2019
Autor(a) principal: Mendes, Fernando Henrique de Paula e Silva
Orientador(a): Caldeira, João Frois
Banca de defesa: Não Informado pela instituição
Tipo de documento: Tese
Tipo de acesso: Acesso aberto
Idioma: eng
Instituição de defesa: Não Informado pela instituição
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
Palavras-chave em Inglês:
Link de acesso: http://hdl.handle.net/10183/202145
Resumo: In this thesis, we present three empirical applications on finance and macroeconomics. The general modeling framework in all chapters is based on extensions of the Markov-switching model. And the statistical methodology is divided into two distinct areas; Classical and Bayesian inference.1 In the first one, we test for the presence of duration dependence in the Brazilian business cycle. The main results indicated that as the recession ages, the probability of a transition into an expansion increases (positive duration dependence in recessions). On the other hand, as the expansions ages, the probability of a transition into a recession decreases (negative duration dependence in expansions). In the second paper, we extend the research concerned with the evaluation of alternative volatility modeling and forecasting methods for Bitcoin log-returns. The in-sample estimates suggest evidence of long memory in the data series. When performing one-day ahead Value-at-Risk (VaR), our results outperform all standard single-regime GARCH models considered in the study. Finally, in the third paper, we capture different regimes in Bitcoin volatility returns and test the mean-reversion hypothesis for multi-period returns. In general, we found evidence of mean-aversion for different holding returns. We also confirmed this result for alternative specifications and also carrying the analysis for sub-sample periods.