Research
Research
This study examines the effect of the expected credit loss model under IFRS 9 on the role of credit relationships in facilitating firm financing in Spain. We document that credit relationships have a positive effect on the growth of credit, but that this positive effect is significantly reduced after implementation of IFRS 9. In 2018, we estimate that the negative impact of IFRS 9 on firm credit relationships established with its main bank led to a reduction in bank lending to Spanish non-financial firms of 2.8% of their total outstanding loans, suggesting a sizeable effect on the availability of credit. We document that this effect is concentrated among borrowers with non-impaired credit positions with low credit quality that has been deteriorated. While this may help constrain evergreening practices at an early stage, before default risk becomes material, it may also weaken the function of credit relationships in supporting credit to low credit quality but viable firms.
This paper examines the impact of the Current Expected Credit Loss (CECL) model on the efficiency of equity investors in processing banks’ earnings announcements. A potential concern with CECL is that it increases the complexity of loan loss provisions. I examine whether this added complexity impairs investors’ ability to efficiently process earnings announcements. Using a difference-in-differences methodology, I find that CECL reduces investors’ processing efficiency during banks' earnings announcements. The effect is stronger when provisions are driven by the origination of new loans. This finding is consistent with the idea that CECL introduces a timing mismatch between loss and revenue recognition—provisions for expected losses are recognized at loan origination while interest income accrues over the life of the loan—making it harder for investors to interpret the valuation implications of provisions. Collectively, my results show that equity investors face higher processing costs when interpreting banks’ earnings announcements under CECL.