Modelos de volatilidade estatística

• Autor(a) principal: Ishizawa, Danilo Kenji
• Data de Publicação: 2008
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFSCAR
• Texto Completo: https://repositorio.ufscar.br/handle/ufscar/4525
• topic: Análise de séries temporais; Análise de séries temporais heterocedásticas; Função de auto-correlação (a.c.f.); função de auto-correlação parcial (p.a.c.f.); Modelo ARCH; Modelo GARCH; Modelo EGARCH; ARCH model, auto correlation function (a.c.f.); EGARCH model, GARCH model; Partial auto correlation function (p.a.c.f.); CIENCIAS EXATAS E DA TERRA::PROBABILIDADE E ESTATISTICA::ESTATISTICA

In the financial market usually notices are taken of the shares sequentially over the time in order to characterize them a time series. However, the major interest is to forecast the behavior of these shares. Motivated by this fact, a lot of models were created based on the past information considering constant averages and variance over time. Although, in financial series a feature often presented is called volatility, which can be noticed by the variance to vary in time. In order to catch this characteristic were developed the models of the family GARCH, that model the conditional variance through known information. These models were well used and have passed by many formulation modifications to be able to catch different effects, such as the effect leverage EGARCH. Thus, the goal is to estimate volatility patterns obeying the specifications of the family GARCH verifying which ones of them describe better the data inside and outside the sample.