• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.45.2021.tde-06052021-100559
Document
Author
Full name
Mateus Gonzalez de Freitas Pinto
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2021
Supervisor
Committee
Chiann, Chang (President)
Lopes, Silvia Regina Costa
Montoril, Michel Helcias
Title in English
Long memory in high frequency time series using wavelets and conditional volatility models
Keywords in English
Asset returns
FIGARCH
High frequency data
Intraday data
Long memory
Volatility
Wavelets
Abstract in English
The goal of this dissertation is to describe a methodology for modelling the volatility of high frequency financial data, considering its features and stylized facts. In order to account for the long-range dependence in conditional mean and conditional variance, ARFIMA and FI(E)GARCH models are used respectively, when observed. To account for the non-normality, skeweness and kurtosis, features observed in the the distribution of the log-returns in high frequency, the Skewed Student t and the Generalized Error Distribution (GED) are adopted for the innovation term of the aforementioned models. Wavelet shrinkage is used in a non-parametric identification and separation of the intraday jumps from the time series data. The application of this procedure is presented using real high frequency asset returns from the Brazilian Exchange and OTC, as well as exchange rates from cryptocurrencies traded in Crypto Exchanges.
Title in Portuguese
Memória longa em séries financeiras utilizando ondaletas e modelos de volatilidade condicional
Keywords in Portuguese
Dados de alta frequência
Dados intradiários
FIGARCH
Memória longa
Ondaletas
Retornos
Volatilidade
Abstract in Portuguese
O objetivo desta dissertação é descrever uma metodologia para modelagem da volatilidade de dados financeiros de alta frequência, considerando suas particularidades e fatos estilizados. Os modelos ARFIMA e FI(E)GARCH são utilizados para modelar a longa persistência das séries na média e na variância condicional, respectivamente, quando isto for observado. A fim de contemplar não-normalidade, assimetria e curtose são utilizadas as distribuições t de Student Assimétrica e Distribuição Generalizada de Erros (GED) para o termo de inovações dos modelos supracitados. A limiarização de ondaletas é utilizada para identificação e separação dos "jumps" intradiários de forma não-paramétrica. A aplicação deste procedimento é apresentada utilizando séries financeiras reais de retornos de ações em alta frequência para ativos negociados no mercado à vista na bolsa de valores brasileira, além de séries de taxas de câmbio de criptomoedas, comparando o modelo semiparamétrico proposto a uma abordagem tradicional sem remover os "jumps".
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2021-05-06
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2022. All rights reserved.