• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Uma investigação sobre modelos de previsão da inflação brasileira

Bortoluzzo, Maurício Mesquita 11 April 2017 (has links)
Submitted by Aline Amarante (1146629@mackenzie.br) on 2017-07-24T21:10:19Z No. of bitstreams: 2 MAURÍCIO MESQUITA BORTOLUZZO.pdf: 2244426 bytes, checksum: 50c3ac77956611398f2f5b0e5e01416e (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Eliana Barboza (eliana.silva1@mackenzie.br) on 2017-08-07T13:25:58Z (GMT) No. of bitstreams: 2 MAURÍCIO MESQUITA BORTOLUZZO.pdf: 2244426 bytes, checksum: 50c3ac77956611398f2f5b0e5e01416e (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-08-07T13:25:58Z (GMT). No. of bitstreams: 2 MAURÍCIO MESQUITA BORTOLUZZO.pdf: 2244426 bytes, checksum: 50c3ac77956611398f2f5b0e5e01416e (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-04-11 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Fundo Mackenzie de Pesquisa / The present thesis performs a pseudo real time study on the predictability of Brazilian inflation, measured by the IPCA, using data from January 1995 to December 2015. The main objective of the study is to compare the predictive accuracy of multivariate models, containing macroeconomic information, against Naive models and against disaggregated data models of inflation, which in the recent literature have been successful in overcoming benchmark models for Brazilian inflation. We found evidence that most models with macroeconomic variables have predictive accuracy higher than the traditional benchmark of the literature, the autoregressive model of order 1. There is also evidence regarding the superiority of forecasts generated by the model with greater data disaggregation. In addition, the ranking of model forecasts changes when we change: the loss function, the forecasts horizons, and the time windows used for evaluations. / A presente tese realiza um estudo pseudo tempo real sobre a previsibilidade da inflação brasileira, medida pelo IPCA, utilizando dados de janeiro de 1995 a dezembro de 2015. O principal objetivo do estudo é comparar a acurácia preditiva de modelos multivariados, contendo informações macroeconômicas, contra modelos ingênuos e contra modelos de dados desagregados da inflação, que na literatura recente apresentaram sucesso em superar os modelos benchmarks. Foram encontradas evidências que a maioria dos modelos com variáveis macroeconômicas possuem acurácia preditiva superior ao benchmark tradicional da literatura, o modelo Autorregressivo de ordem 1 (AR(1)). Também há evidências quanto à superioridade de previsões geradas pelo modelo com maior desagregação de dados. Além disso, verifica-se que o ranqueamento das previsões dos modelos se altera quando se alteram: a função de perda, os horizontes de previsão e as janelas de tempo utilizadas para as avaliações.

Page generated in 0.1098 seconds