A platform to produce hybrid models in order to manage risks and uncertainties on financial and actuarial organizations

摘要

Planning under uncertainty requires reliable tools and methods for dynamic financial analysis. A financial and actuarial problem is how to consider returns of investment forecasts and foresees based on uncertain incomes and liabilities. Lifetime and demographic studies focus on the population dynamics of a financial and actuarial model that has, among others, rates of mortality, withdrawal, disability and death that must be considered in costs assessments and cash flows. Thus , to adress asset and liability management (ALM) problem one must model stochastic liabilities based on stochastic assets to set more reliability to investment policies simulators. This articles analyses a combination of methods and techniques to better model uncertainty. Delphi technique, factor analysis and SD methods were used to identify factors, model causation between variables and to describe the structure and the behaviour of a financial and actuarial system. Stochastic processes and Monte Carlo simulation is used to manage probabilities distributions and to predict what future could be; and, in order to give to financial and actuarial analysts and actuaries a way to perform more realistic analysis of the solvency and liquidity, fuzzy logic and agent based modeling. The conclusion propose a framework where a combination of these techniques are possible and useful to integrate all these approaches and techniques and, on an open system perspective as a way to manage long term investment policy based on current liabilities, both stochastic in nature.