Gestão de ativos e passivos em entidades fechadas de previdência complementar

Abstract

In this study a methodology of managing assets and liabilities in defined-benefit pension plans was developed. The study presents how Assets and Liabilities Management can be benefited from the use of Operations Research techniques. The goal of pension fund wasformulated as a problem of minimizing the probability of default. The Genetic Algorithms technique was used as a solution methodology and succeeded in finding satisfactory solutions to the mathematical problem formulated. The study also investigated how changes in policy contributions, regular reviews of portfolio composition, choices on theperiod for estimation of parameters and degree of capitalization of pension funds interfere in the way fund's assets are allocated. The general conclusion is that the Assets and Liabilities Management approach with the use of Genetic Algorithms is a good tool for selecting portfolios that minimize the probability of default in a pension fund.