A carregar...
Publicação
Backtesting a parâmetros de risco de crédito
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.61 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A realização do exercício do Backtesting aos parâmetros de risco utilizados no modelo de cálculo das perdas por imparidade, constitui um requisito regulamentar imposto inicialmente na publicação da norma IAS 39 Financial Instruments: Recognition and Measurement.
Tratando-se de um tema que, não só apresenta uma elevada relevância em termos da rendibilidade das instituições financeiras, como uma complexidade inerente, levou-me à sua exploração numa vertente mais académica.
Assim, este Projeto tem como objetivo desenvolver e implementar um exercício que visa testar se existem diferenças estatisticamente significativas entre as probabilidades estimadas por um modelo de cálculo de parâmetros já existente, que tem por base um universo histórico de 7 anos, e as probabilidades observadas no último ano.
Para os casos em que foram verificadas diferenças estatisticamente significativas, irá ser sugerido e implementado um algoritmo de medidas corretivas a essa curva das probabilidades estimadas utilizada no modelo de imparidade. Desta forma, pretende-se aproximar os valores das perdas estimadas às perdas efetivamente incorridas.
The performance of the Backtesting exercise on the risk parameters used in the Impairment Loss Calculation Model, is a regulatory requirement imposed initially in the statement of IAS 39 Financial Instruments: Recognition and Measurement. Because this is a topic issue that has high relevance in terms of profitability of financial institutions and also an inherent complexity, I was led to develop it as an academical exploitation. Thus, this Project's target is to develop and implement an exercise that aims to test if there are statistically significant differences between the probabilities estimated by an already existing parameter calculation model, based on a historical universe with 7 years, and the observed probabilities on the previous year. For the instances in which statistically significant differences were verified, an algorithm of corrective measures will be suggested and implemented to the estimated probabilities curve. On the other hand, it is also used in the calculation model of the impairment losses. In this way, it is intended to approach the values of the estimated losses to the actual losses incurred.
The performance of the Backtesting exercise on the risk parameters used in the Impairment Loss Calculation Model, is a regulatory requirement imposed initially in the statement of IAS 39 Financial Instruments: Recognition and Measurement. Because this is a topic issue that has high relevance in terms of profitability of financial institutions and also an inherent complexity, I was led to develop it as an academical exploitation. Thus, this Project's target is to develop and implement an exercise that aims to test if there are statistically significant differences between the probabilities estimated by an already existing parameter calculation model, based on a historical universe with 7 years, and the observed probabilities on the previous year. For the instances in which statistically significant differences were verified, an algorithm of corrective measures will be suggested and implemented to the estimated probabilities curve. On the other hand, it is also used in the calculation model of the impairment losses. In this way, it is intended to approach the values of the estimated losses to the actual losses incurred.
Descrição
Mestrado em Mathematical Finance
Palavras-chave
modelo de cálculo das perdas por imparidade parâmetros de risco Backtesting testes de hipóteses impairment model risk parameters hypothesis tests
Contexto Educativo
Citação
Bravo, Joana Margarida Gomes (2017). "Backtesting a parâmetros de risco de crédito". Dissertação de Mestrado, Universidade de Lisboa. Instituto Superior de Economia e Gestão.
Editora
Instituto Superior de Economia e Gestão