Uma crítica ao cálculo da PCLD e à divulgação dos riscos de créditos em instituições financeiras: comparativo IFRS e COSIF (BRGAAP)

Abstract

The topic of this research it concerns matters of transparency and the measurement of information related to credit risk, presented in the consolidated financial statements and in the risk management report of Brazilian financial institutions. This linked with the changes in accountancy rules, from the convergence to IASB reports and the transparency requirement from Pilar 3, of the New Basel Agreement (Basel II). Part of this study was to continue Martins and Dantas´ (2014) research that aimed to identify which of the two financial disclosure patterns adopted in Sistema Financeiro Nacional (SFN) – IFRS or COSIF – was the best for the disclosure requirements about credit risk in Basel Agreement. Moreover, to quantify the impacts with the adoption of the Brazilian rule Resolução CMN nº 2682 (COSIF) and IAS 39 (IFRS) in the profit/result of institutions, considering that the criteria for the provision for doubtful accounts (PCLD) is different, for COSIF – expected loss, and IFRS – incurred loss. For this comparisons, it was analyzed the disclosed information about credit risk in annual financial statements and in the risk management report (additional report not included in Martins and Dantas´ (2014) research), concerning 2010 to 2014, followed by two disclosure models from the ten biggest consolidated banks in Brazil, through total asset criteria. Initially, it was made a bibliographic revision, to offer a general view of the subject. Afterwards, it was applied content analysis, with qualitative and quantitative data evaluation techniques. It was calculated the credit risk disclosure (DRC) for the sample institutions in the five periods, though the assignment of scores 0 and 1, for each one of the eight disclosure requirements in Pilar 3 Basel II, in the financial statements in IFRS and COSIF patterns. The information about the risk management report was added to it for attending the COSIF pattern. Panel data was used to evaluate differences in measurement. Differently from Martins and Dantas´ (2014) research, the COSIF pattern better attended to disclosure requirements, indicating that the publication of Circular BACEN nº 3477/09 and its later replacement for 3678/13 (this rule made the disclosure of the risk management report mandatory) reinforced the enhancement in disclosure. It was not possible to identify a significant statistical association that showed impacts in profit/result.