Problems of Assessing Credit Risks of Structured Financial Instruments

Approaches to risk assessment of structured financial instruments, namely asset-backed securities, in modern financial engineering are insufficiently systematized. The reasons for this may include the variety of risks that a bank wishes to avoid by transferring management to other parties; inadequate attention to borrower selection; and a lack of consolidation on the formation of a unified methodology for assessing typical risks. The author presents for analysis three main types of risks faced by a bank in the process of securitization: prepayment risk, interest rate risk, and counterparty default risk. The assessments are based on time factors and events that, under certain conditions, can critically affect the cash flows of financial institutions. Prepayment risks are associated with the calculation of average time values before loan payments occur, however, a unified assessment methodology in banking practice is currently absent. The assessment of the degree of risk is accompanied by the calculation of the percentage increase to the base price with a constant yield. The increase reflects the risk of interest rate changes over a certain period. Particular attention is paid to the analysis of methods for assessing the counterparty default of a structured instrument. It is established that E. Altman's methodology under different environmental conditions can yield contradictory results. Therefore, in recent years, fuzzy-neural network models can be considered optimal models for assessing default risk. Based on O.A. Nedosekin's methodology, the degree of borrower bankruptcy risk is shown, relying on expert assessment methods. On one hand, the fuzzy-neural network model can give results close to the truth, on the other hand, information asymmetry can affect the accuracy of the result. In this light, the model requires further improvements based on the Mamdani criterion. Thus, the problem of assessing the credit risks of structured instruments remains relevant in modern scientific circles and requires the development of common ideas for the unification of methodologies proposed by modern science.