Abstract:

An important problem in risk analysis is the distinction between aleatoric (probabilistic) uncertainty and epistemic uncertainty (resulting from a lack of precision or lack of information). The article describes the problem of making an investment decision based on the analysis of the profitability and risk of a tangible investment, in a situation where some input parameters of the decision model are presented as probability distributions and some as possibility distributions, in other words, in a situation of hybrid data. The risk analysis for such a defined problem was performed using a method combining Monte Carlo simulation with a method for performing arithmetic operations on dependent fuzzy numbers. In the article, the latter was implemented using the nonlinear programming method. As a result of this type of calculation, a fuzzy random variable was obtained, the interpretation of which may be difficult for business practitioners. Based on a critical review of existing risk measures defined for the use of hybrid data, a new method for investment decision-making was proposed. The practical part presents the investment decision-making process on the example of a tangible investment in the metallurgical sector.