Dependence Relationships Modelling In Financial Appraisal Under Uncertainty
Free (open access)
C. R. García & F. Amador
The modelling of dependence relationships is one of the most difficult processes in simulation models. Dependence relationships are especially important in financial problems under uncertainty conditions. Unfortunately, in complex situations we cannot easily assume mathematical functions -technical relationships- for relating cash flow components, like revenues or costs. In an attempt to approach this problem, fuzzy logic is used for designing and modelling \“fuzzy” relationships between selected financial parameters. A fuzzy inference engine has now been included in the PRAPPIS v2.0 software for project appraisal under risky and uncertainty conditions. This engine can assess relationships using a set of basic fuzzy rules that control their behaviour in a given context. Fuzzy relationships are easily defined by the user or decision maker guided by a natural language system. Mizumoto’s approach is used to obtain fuzzy relationship results and standard Monte Carlo simulation is applied to assess the financial project. Statistical distributions of many financial parameters are available in PRAPPIS with or without dependence relationships. In this paper, different results are shown corresponding to different statistical distributions of financial parameters using a first set of fuzzy inference rules called here \“neutral” in comparison to standard Monte Carlo simulation assuming independent statistical distributions. In an agricultural environment, which is uncertain, we have chosen some controlled relations between two selected parameters such as, for example, part-time hand labour in harvesting and production. Results show that the variance of the statistical distributions in the project results is reduced if fizzy relationships are used.