| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.47 MB | Adobe PDF |
Authors
Abstract(s)
Para as empresas de seguros, é indiscutível a importância da constituição de provisões adequadas
para fazer face a pagamentos futuros de indemnizações decorrentes das responsabilidades com os
contratos de seguro. Se, por um lado, a constituição de provisões insuficientes pode levar as empresas a
entrar em insolvência, a constituição de provisões em excesso afeta a sua rentabilidade e pode, ainda,
levar ao aumento da tarifação, comprometendo, deste modo, a competitividade de mercado.
A nova Norma Internacional de Relato Financeiro, IFRS 17, divide as responsabilidades decorrentes
dos contratos de seguro em duas provisões designadas de LIC (Liability for Incurred Claims) e LRC
(Liability for Remaining Coverage). A LIC compreende os sinistros RBNS (Reported But Not Settled)
e os sinistros IBNR (Incurred But Not Reported). Já a LRC refere-se a sinistros CBNI (Covered But Not
Incurred). A LIC apresenta um peso elevado no balaço das empresas de seguros, assumindo um papel
de extrema primazia, com destaque para o ramo Não Vida.
Uma vez que, aquando da sua ocorrência, o custo total de um sinistro assume um valor
desconhecido, existe a forte necessidade de estimação. Com base nos dados históricos, as empresas de
seguros realizam análises casuísticas (por sinistro, caso a caso) e análises agregadas para sinistros com
características semelhantes. A análise casuística revela-se mais adequada na presença de sinistros
atípicos e a análise agregada quando existe um elevado número de sinistros, permitindo estimar custos
associados a sinistros IBNR com recurso a matrizes de run-off.
O presente trabalho aborda dois métodos clássicos de estimação agregada: o método de Chain
Ladder e o método de Bornhuetter-Ferguson. Para cada método é proposto um modelo e dado um
exemplo de aplicação. Por último, é discutida a validação dos pressupostos assumidos nos modelos
propostos, sendo apresentadas as limitações inerentes a este processo.
For insurance companies, the importance of setting aside adequate provisions to cover future liabilities arising from insurance contracts is indisputable. If, on the one hand, setting aside insufficient provisions can lead companies into insolvency, on the other hand, setting aside excessive provisions affects their profitability and can also lead to increased pricing, thus compromising market competitiveness. The new International Financial Reporting Standard, IFRS 17, sets the liabilities arising from insurance contracts in two provisions called LIC (Liability for Incurred Claims) and LRC (Liability for Remaining Coverage). LIC includes RBNS (Reported But Not Settled) and IBNR (Incurred But Not Reported) claims. LRC refers to CBNI (Covered But Not Incurred) claims. LIC has a high weight in the balance sheet of insurance companies, assuming an important role, especially in the Non-Life business. Since the total cost of an accident is unknown at the time of its occurrence, there is a strong need for estimation. Based on historical data, insurance companies carry out case-by-case analysis and aggregate analysis for claims with similar characteristics. The case-by-case analysis is more appropriate in the presence of atypical claims and the aggregate analysis when there are a large number of claims, estimating also costs associated with IBNR claims, using run-off matrices. The present work addresses two classic aggregate estimation methods: the Chain Ladder method and the Bornhuetter-Ferguson method. A model is proposed for each method and an application example is given. Finally, the validation of the assumptions made in the proposed models is discussed, and the limitations inherent in this process are presented.
For insurance companies, the importance of setting aside adequate provisions to cover future liabilities arising from insurance contracts is indisputable. If, on the one hand, setting aside insufficient provisions can lead companies into insolvency, on the other hand, setting aside excessive provisions affects their profitability and can also lead to increased pricing, thus compromising market competitiveness. The new International Financial Reporting Standard, IFRS 17, sets the liabilities arising from insurance contracts in two provisions called LIC (Liability for Incurred Claims) and LRC (Liability for Remaining Coverage). LIC includes RBNS (Reported But Not Settled) and IBNR (Incurred But Not Reported) claims. LRC refers to CBNI (Covered But Not Incurred) claims. LIC has a high weight in the balance sheet of insurance companies, assuming an important role, especially in the Non-Life business. Since the total cost of an accident is unknown at the time of its occurrence, there is a strong need for estimation. Based on historical data, insurance companies carry out case-by-case analysis and aggregate analysis for claims with similar characteristics. The case-by-case analysis is more appropriate in the presence of atypical claims and the aggregate analysis when there are a large number of claims, estimating also costs associated with IBNR claims, using run-off matrices. The present work addresses two classic aggregate estimation methods: the Chain Ladder method and the Bornhuetter-Ferguson method. A model is proposed for each method and an application example is given. Finally, the validation of the assumptions made in the proposed models is discussed, and the limitations inherent in this process are presented.
Description
Trabalho de Projeto de Mestrado, Matemática Aplicada à Economia e Gestão, 2024, Universidade de Lisboa, Faculdade de Ciências
Keywords
Provisões para sinistros Atividade seguradora Teses mestrado - 2024
