• 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.92.2005.tde-29032023-144541
Document
Author
Full name
Felipe Villarino Prieto
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2005
Supervisor
Committee
Paula, Gilberto Alvarenga (President)
Artes, Rinaldo
Elian, Silvia Nagib
Title in Portuguese
Precificação de seguros de automóvel
Keywords in Portuguese
Finanças
Seguro de automóveis
Abstract in Portuguese
O mercado segurador brasileiro e, especialmente, o seguro de automóveis é extremamente competitivo, o que obriga as seguradoras a realizarem uma tarifação correta e bem ajustada de acordo com o perfil de risco do segurado. Cada vez mais tem se buscado precificar de uma forma mais granularizada havendo um preço para cada segurado, conforme as características de cada um. Os modelos estatísticos vêm ganhando grande importância em seguros, na medida que eles se adaptam muito bem aos problemas e necessidades apresentadas. Neste trabalho, descrevemos alguns aspectos do seguro de automóveis e a partir de uma base de dados fornecida por uma seguradora brasileira, calculamos o preço individual a ser cobrado para suprir todos os prejuízos ocorridos na carteira. Utilizamos três técnicas combinadas para calcular a frequência e o custo médio de sinistros: árvores de decisão, regressão logística e modelos lineares generalizados. A árvore de decisão foi utilizada para identificar as possíveis interações entre as variáveis e trouxe um enorme ganho na qualidade de ajuste dos modelos. Os diagnósticos do modelo e análise de resíduos apresentados sugerem um bom ajuste dos modelos. Assim, conseguimos encontrar uma forma bastante eficaz de precificar o seguro de auto mediante o perfil de risco do segurado.
Title in English
Car insurance pricing
Keywords in English
Auto insurance
Finance
Abstract in English
The Brazilian Insurance market and more specifically car insurance is extremely competitive, which forces insurance companies to practice tight fees, well adjusted to the clients risk profile. There has been an increasing attempt to price insurance in a more customized way. Statistical models have gained great importance in as much as they fit very well the problems and needs at hand. In the present work, some aspects of car insurance are described, and from a database supplied by a Brazilian insurance company, the individual fee is calculated in order to cover all damage caused to the cast. Three combined techniques have been employed to calculate the frequency and the average cost of accidents: decision trees, logistic regression and generalized linear models. The decision tree was been used to identify possible interactions among the variables, and has brought an enormous gain to the goodness-of-fít. The model's diagnostics and further analyses of the remains presented suggest the model's fine tuning. Thus, highly efficient means of pricing car insurance according to the client's risk profile has been found.
 
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
2023-03-29
 
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 © 2001-2024. All rights reserved.