Rating of Non-Financial Corporate Borrowers Based on Open-Source Data

Purpose: This study is aimed at developing a rating model methodology, which allows for rating assessment of non-financial companies based solely on publicly available information.
Design/methodology/approach: This study follows a philosophy of positivism, applies deductive approach, it utilizes a survey strategy and employs closed-end questions in the assessment questionnaire as it performs quantitative assessments of responses of categorized and numeric nature.
Findings: The developed rating model showed good detection accuracy (97%) for performing companies and acceptable detection accuracy (78%) for defaulted companies. The rating comparison showed that the model ratings on average deviate by not more than three notches from the ratings assigned by the rating agencies, which is deemed acceptable. Also, there are certain limitations and further directions for development, laid out in the study.