Use of determined models in analysis of the borrower of the borrower

Abstract

Purpose: At present, there is a need for precise instruments for assessing the credibility of the lending borrower, allowing you to monitor the dynamics of financial performance of its activities, not only from the historical period, but also to predict them for the entire planned period of lending even at the stage of making a loan decision. Stochastic statistical models used for internal credit rating are inherent in certain shortcomings and limitations.
In particular, the analysis of the influence of credit decisions and factors of the local business environment on the borrower's risk level causes considerable difficulties. We proposed to combine the strengths of deterministic and stochastic models, assigning the first function of assessing the changes in financial reporting indicators as a result of the credit decision and leaving the function of assessing the probability of default for the second. The results of calculations based on deterministic models will be input parameters of stochastic functions.

Discussion: The shortcomings of the existing stochastic models used in the rating evaluation of borrowers are highlighted, a system of criteria is developed that identifies the need to use deterministic models to analyze the factors of their external and internal environment.
Results: deterministic factor models are constructed, allowing for rapid assessment of the consequences of credit decisions