ABOUT ONE APPROACH TO THE FORMAL ASSESSMENT FOR THE QUALIFICATION OF COMMERCIAL BANK CREDIT ANALYSTS

Non-effective credit management can lead to significant losses and even bankruptcy of a bank. Therefore, in order to minimize possible losses, bank specialists should assess the creditworthiness of a potential borrower at a high level. This fact and the high competition of banks has led to the widespread use of scoring systems. At the same time, there is no universal scoring model, and the automated system cannot detect suspicious behavior or a borderline state of the client. Thus, to make a credit decision, it is advisable to rely not only on the results of the scoring system, but also on the opinion of qualified experts (credit analysts), which is especially important when the economic situation changes. Therefore, the formal scheme of assessing the qualifications of a credit analyst based on available a priori information is relevant. In accordance with the proposed approach to assessing credit histories, analysts use their testing based on previously identified results of the analysis of credit histories in banks. The tested candidates have to assign former bank borrowers to one of two classes – a trustworthy borrower who easily fulfills the terms of the loan, or an unreliable borrower who repaid the loan by force. If, according to the test results, the average risk of a credit analyst decisions is less than the decisions a priori risk that a bank manager makes only on the knowledge basis for the probabilities of the reliable and unreliable clients appearance, then the credit analyst is considered qualified. In addition, a scheme for selecting the most qualified credit analysts in the credit decision-making group based on the proposed formal conditions is provided. For practical application of the proposed approach, it is sufficient to assess the a priori probabilities of reliable and unreliable customers in the test sample of credit histories and conditional probabilities of errors made by a potential credit analyst in classifying customers of a bank with a known credit history. A model example illustrating the proposed approach is given.