Cash Management of a Company Using Neural Networks

Cash management is one of the most important indicators of a company´s activities and plays an important part in decision-making. Cash management or cash flow management are incomes and expenses for a certain period. The aim of cash management is to mobilize, check and plan a company´s financial resources, which is not easy. The aim of financial managers is to find an effective and flexible tool for improving the processes for the optimization of cash management. One such tool appears to be a system of artificial neural networks. These networks are very flexible and outperform other models, including linear regression models, in many ways. However, networks also have certain pitfalls, which include the sensitivity of the input data, the longer time associated with training the networks, as well as the lack of possibilities to define the architecture and other parameters of the network. This article attempts to predict the future development of various kinds of cash flows. In so doing, it hopes to identify a suitable neural network that is able to predict these cash flows. This process involved the generation of 1000 accidental artificial neural structures, of which the 5 most suitable were preserved. A sensitivity analysis was subsequently carried out. The results of the research show that neural networks can be effectively used to predict the cash flow developments within a company.