AI shows how governance practices in SMEs have real implications for Credit Ratings and access to Funding

We all agree that small and medium-sized enterprises (SMEs) play a crucial role in the economy, and effective governance practices are instrumental in their success. At INBONIS RATING we wanted to go further and show scientifically and using AI models the significance of implementing proper governance practices in small and medium-sized enterprises (SMEs) and its impact on the financial capital of the company, specifically focusing on credit ratings and access to diversified and suitable funding sources.

Thanks to our comprehensive database that includes over 500 European SMEs rated by INBONIS RATING and contains a wealth of information we gathered a valuable resource for advanced AI models, with over 50,000 data points, hundreds of rating drivers analysed, and more than 50 qualitative questions recorded for each SME.

How we used AI

Technically, to analyse the correlation between environmental, social, and governance (ESG) practices and credit ratings, ESG scores were developed for each of the 500+ rating reports using Natural Language Processing (NLP) models. NLP, a subset of Artificial Intelligence (AI), focuses on processing and understanding human-written text. In this study, three models built upon Google’s Bidirectional Encoder Representations from Transformers (BERT) model were utilized: FinBert, FinBert-ESG, and FinBert-ESG-9. Credit Rating Reports, which provided detailed information on companies and their methodologies, were analysed using these models.

To obtain the scores, over 500 Credit Rating Reports with detailed information on companies and their working methodologies where studied. Each sentence of the report was run through the models and scored based on the prevalence of each subject and the sentiment of the sentence itself. E.g. a company with a high mention of Climate Change but a high negative sentiment would score poorly in this subsection. To score each company the scores of either positive or negative sentences were averaged to obtain a final score for each of the identified sectors.

The key findings are striking

Our findings indicate a strong correlation between governance practices, as defined in this analysis, and credit ratings, consequently influencing the availability of financial capital for the company. Analysis of the rating distribution among the 500+ sample revealed that companies with lower governance scores obtained credit ratings ranging from CCC- to BB-, while those with higher governance scores achieved ratings from BB- to BBB, representing an average improvement of four notches.

This analysis further underscores the importance of SMEs having a credit rating, as it highlights the limited visibility of good governance practices through simple quantitative scores alone. Instead, it emphasizes the value of more comprehensive evaluations conducted by credit rating agencies (CRAs) such as Inbonis Rating, which delve deeper into assessing and understanding the intricacies of a company’s governance practices.

Israel es el COO de INBONIS y es responsable del desarrollo de estrategias analíticas multiplataforma. Con más de 20 años de experiencia en banca internacional, ha trabajado en BBVA durante 18 años, los últimos 12 en Estados Unidos, asumiendo diferentes responsabilidades, como la gestión del riesgo crediticio de carteras minoristas o los procesos de detección de fraude para diferentes sistemas de pago. Tiene una amplia experiencia en el desarrollo de productos y canales, gestión de riesgos y equipos de desarrollo analítico en Estados Unidos. También ha formado parte de los consejos de entidades de tecnología financiera (Fintech) como Upturn y de entidades sin ánimo de lucro como la Casa Ronald McDonald. Posee un título en Economía de la Universidad de Alcalá, está certificado como "Gestor de Riesgos Financieros" por el Instituto GAAP y ha completado su educación superior en la Universidad Northwestern en Evanston, Illinois.