Розробка моделі прогнозування раціональності фінансових рішень в умовах діджиталізації фінансових ринків

The object of this study is to predict the rationality of financial decisions in the context of digitalization of financial markets. In the context of digitalization of financial markets, about ¼ of financial decisions turn out to be irrational for financial market participants. Under these conditions, the problem is the inability of financial market participants to predict the rationality of financial decisions. The devised multi-vector model for predicting the rationality of financial decisions in the context of digitalization of financial markets makes it possible to evaluate key indicators of decision-making efficiency and minimize risks. It was found that the use of the adaptive Adam optimization algorithm provides a reduction in the average forecasting error by 18.7 % compared to conventional methods, such as gradient descent. The use of a utility function with a correction parameter β made it possible to smooth out market fluctuations, reducing the deviation of predicted values from actual values by an average of 12.3 %. The conducted scenario modeling using the Monte Carlo method demonstrated that under conditions of high market volatility, the accuracy of forecasts remains stable and exceeds 85 %. Testing the model on the example of five Ukrainian financial companies (Moneyveo, LeoGaming Pay, Ukrfinzhytlo, European Microfinance Alliance, Smart Pay) over the period 2018–2023 showed that the level of irrational financial decisions decreased on average from 24.6 % to 15.2 %, which is equivalent to saving financial resources in the amount of UAH 37.8 million per company. This indicates the significant potential of the model in improving the quality of financial management and ensuring sustainable development of financial markets. The practical value of the devised multi-vector predictive model of financial decisions rationality relates to its ability to optimize the process of assessing risks and investment returns, taking into account the multifactorial nature of the modern market environment. The results could become a tool for strategic planning and assessing investment attractiveness