BEHAVIOURAL DETERMINANTS OF INVESTMENT DECISIONS IN THE FINANCIAL CONTROLLING SYSTEM OF REAL SECTOR ENTERPRISES

The research subject is the impact of cognitive biases on the efficiency of investment decisions within the real economy and how these biases can be integrated into financial control systems. The research methodology takes a comprehensive approach, combining regression analysis to identify statistical patterns with the Mamdani fuzzy inference system to model non-linear interactions between behavioural determinants. The empirical basis comprises 67 investment projects of Ukrainian enterprises for the period 2022–2024 and data collected via a structured questionnaire containing 45 questions designed to diagnose cognitive biases. The research objective is twofold: first, to identify the dominant behavioural determinants of investment decisions based on empirical analysis; and second, to develop tools for considering these determinants in the financial control systems of enterprises in the real sector. The research results demonstrate that cognitive biases systematically reduce the efficiency of investment projects. The Regression Model illustrates the detrimental effects of the sunk cost fallacy, overconfidence bias, the anchoring effect and herding behaviour on the discrepancy between the actual and planned NPV, accounting for 54.7% of the variation in the dependent variable. The most influential factors are the sunk cost fallacy (β = -1.40, p < 0.001) and the overconfidence bias (β = -1.05, p = 0.002). The developed Mamdani Model, which incorporates nine expert rules, enables the formalisation of nonlinear synergistic effects between biases and the identification of critical risk zones when combining high values of determinants. The correlation between the Mamdani Model and the regression results is 0.89 (p < 0.001), which confirms the validity of the fuzzy approach. The practical significance lies in creating operational tools for behavioural control with linguistic rules for diagnosing cognitive distortions when making investment decisions. These results can be used by the financial services departments of enterprises to improve control systems and reduce the impact of psychological factors on the quality of investment decisions.