Please use this identifier to cite or link to this item:
https://repositorio.ufba.br/handle/ri/15719
metadata.dc.type: | Artigo de Periódico |
Title: | A top–bottom price approach to understanding financial fluctuations |
Other Titles: | Physica A: Statistical Mechanics and its Applications |
Authors: | Castro, Miguel A. Rivera Miranda, José Garcia Vivas Borges, Ernesto Pinheiro Cajueiro, Daniel Oliveira Andrade, Roberto Fernandes Silva |
metadata.dc.creator: | Castro, Miguel A. Rivera Miranda, José Garcia Vivas Borges, Ernesto Pinheiro Cajueiro, Daniel Oliveira Andrade, Roberto Fernandes Silva |
Abstract: | The presence of sequences of top and bottom (TB) events in financial series is investigated for the purpose of characterizing such switching points. They clearly mark a change in the trend of rising or falling prices of assets to the opposite tendency, are of crucial importance for the players’ decision and also for the market stability. Previous attempts to characterize switching points have been based on the behavior of the volatility and on the definition of microtrends. The approach used herein is based on the smoothing of the original data with a Gaussian kernel. The events are identified by the magnitude of the difference of the extreme prices, by the time lag between the corresponding events (waiting time), and by the time interval between events with a minimal magnitude (return time). Results from the analysis of the inter day Dow Jones Industrial Average index (DJIA) from 1928 to 2011 are discussed. qq-Gaussian functions with power law tails are found to provide a very accurate description of a class of measures obtained from the series statistics. |
metadata.dc.rights: | Acesso Aberto |
URI: | http://repositorio.ufba.br/ri/handle/ri/15719 |
Issue Date: | 2012 |
Appears in Collections: | Artigo Publicado em Periódico (FIS) |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Miguel A. Rivera Castro.pdf | 496,99 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.