| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.62 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
No modelo colectivo considera-se que é a carteira no seu conjunto que gera
os sinistros. Para além de se conhecer o montante do prémio a pagar para o conjunto
das apólices de uma carteira de seguros, é também importante saber qual o prémio a
pagar por cada apólice em particular, tendo em conta as características do risco a
segurar. Com esse objectivo, foi desenvolvido o presente trabalho, que se compõe de
três fases.
Na primeira parte desenvolve-se uma variante vectorial do modelo colectivo
com prazo fixo, que permite estimar os quantis para a distribuição da soma S dos
montantes dos sinistros, assim como a variância dos estimadores desses quantis, sem
se pressupor a hipótese de homocedasticidade como era usual em estudos anteriores.
O total dos prémios deve ser o quantil para a probabilidade q da distribuição de S,
caso se queira que a probabilidade de perda seja inferior a 1-q. Mostra-se ainda a
razoabilidade do princípio do desvio padrão no cálculo dos prémios.
Na segunda parte utiliza-se uma decomposição aleatória da carteira em m
sub-carteiras, juntamente com o método dos Mínimos Quadrados, para realizar o
rateio desse total pelas apólices que constituem a carteira.
Desenvolve-se por fim uma aplicação prática deste modelo a uma carteira do
ramo Multirriscos Habitação, para a qual se constrói uma tarifa, que se revelou
fortemente competitiva.
In the collective model, the portfolio, as a whole, generates the claims. It's important to know the amount of premium to pay for the portfolio, but is also important to know the premium to be paid for each one of the policies, taking into account the characteristics of the risks to ensine. This Thesis has been divided into three main steps and next I will explain its objectives. In a first step, a vectorial version of the collective model with a fíxed delay is developed, and allows the estimation of the amounts for the distribution of the sum S of the claim amounts, as well as the variance of the estimators of these amounts, without the restriction of homocedasticity used in previous studies. The total of the premiums must be the quantile for probability q of the distribution of S, or, if it is wanted, the probability of loss to be inferior tol-q. We found a justification to use the standard deviation principie for the premium calculation. In the second step, a random decomposition of the portfolio in m sub-portfolios, as well as the method of the minimum squares are used to share the total of the premiums between the policies of the portfolio. Lastly our results are applied to a Fire and Allied Risks portfolio. The tariff thus obtained is highly competitive.
In the collective model, the portfolio, as a whole, generates the claims. It's important to know the amount of premium to pay for the portfolio, but is also important to know the premium to be paid for each one of the policies, taking into account the characteristics of the risks to ensine. This Thesis has been divided into three main steps and next I will explain its objectives. In a first step, a vectorial version of the collective model with a fíxed delay is developed, and allows the estimation of the amounts for the distribution of the sum S of the claim amounts, as well as the variance of the estimators of these amounts, without the restriction of homocedasticity used in previous studies. The total of the premiums must be the quantile for probability q of the distribution of S, or, if it is wanted, the probability of loss to be inferior tol-q. We found a justification to use the standard deviation principie for the premium calculation. In the second step, a random decomposition of the portfolio in m sub-portfolios, as well as the method of the minimum squares are used to share the total of the premiums between the policies of the portfolio. Lastly our results are applied to a Fire and Allied Risks portfolio. The tariff thus obtained is highly competitive.
Description
Mestrado em Actuariado e Gestão de Riscos Financeiros
Keywords
prémio rateio versão vectorial do modelo colectivo tarifa estimação quantis Multirriscos Habitação share collective vectorial model tariff quantile Fire and Allied Risks risk
Pedagogical Context
Citation
Caravina, Maria Teresa Palos (1998). “Cálculo de prémios e seu rateio”. Dissertação de Mestrado. Universidade Técnica de Lisboa. Instituto Superior de Economia e Gestão
Publisher
Instituto Superior de Economia e Gestão
