Análise da influência do sentimento, da regulação e da auditoria nas estimativas de perdas com créditos ofertados pelos bancos

Abstract

This thesis aimed to investigate the influences of Sentiment, the regulatory environment, and independent auditing on exposure to banking credit risk. The literature establishes that economic factors and cycles affect banks' credit risks, considering the interactions between surplus and deficit agents intermediated by the financial system. However, this research is based on the premise that, beyond these factors, Sentiment, regulation, and auditing play fundamental roles in the relationship between banks' exposure to credit risk levels. We defined Sentiment as irrational expectations regarding asset behavior that are not based on traditional economic principles. While previous studies have analyzed the relationship between managerial Sentiment and credit risk, this thesis advances the discussion by examining the impact of investor sentiment, both at the market and firm levels, on banking credit exposure, an aspect still underexplored in the literature. The regulatory environment is assessed through capital solvency and indicators such as the CAMELS model (Capital Adequacy, Asset Quality, Management Quality, Earnings, Liquidity, and Sensitivity to Market Risk), as well as regulatory and prudential capital and the adoption of expected credit losses based on IFRS 9. Auditing is analyzed based on quality attributes such as auditor rotation, the experience of partners and firms, audit fees, and the composition of the audit committee. Credit risk exposure was measured using incurred losses (CMN, according to Resolution Nº. 2,682, 1999) and expected losses (IFRS 9/CPC 48 model). The analysis compared groups of banks that adopted IFRS 9 early with those that had not yet implemented the model. The sample included both listed and non-listed financial institutions on B3, with available data from 2010 to 2023, justified by the adoption of IFRS in Brazil. The first hypothesis (H1) analyzed the relationship between Sentiment and banks' exposure to credit risk. The results indicate that investor sentiment, both at the market and firm levels, positively influences the levels of incurred and expected losses. During periods of optimism, banks tend to recognize higher losses, reflecting greater discretion in the estimates made by managers and accountants. The second hypothesis (H2) examined the influence of the regulatory environment on credit risk, and it was found that the CAMELS model and banks' productive efficiency mitigate risk exposure. At the same time, efficient capital and leverage management, measured by the Basel Index and the Capital and Leverage Ratio, is associated with risk reduction. However, expected losses measured by IFRS 9 exceed incurred losses, particularly in larger banks, which exhibit greater exposure due to the size of their portfolios. The third hypothesis (H3) explored the relationship between independent auditing and credit risk, and it was found that auditor rotation, the experience of partners and firms, audit fees, and the composition of audit committees positively influence risk exposure. The presence of women on committees was found to be relevant for strengthening audit governance and increasing scrutiny in auditing procedures. Finally, the thesis presents relevant contributions for regulators, managers, and auditors. The practices related to Sentiment, regulation, and auditing, combined with predictive loss estimates, promote a conservative approach to credit risk measurement, which is essential for mitigating future crises and ensuring financial stability for institutions.