Real effects of the IFRS 9 adoption in banking

We investigate the nexus between the interest rates charged to bank borrowers and loan-to-value (LTV). Using a unique dataset on mortgage applications from a major European bank (2017-2020), we test whether reporting entities change resource allocation in response to a new accounting rule (real effect hypothesis) by exploiting the mandatory IFRS 9 adoption in 2018 as a laboratory setup. We first find evidence of a tightening of loan-to-value in retail banking, suggesting that its introduction increases the LTV-related costs compared to its predecessor (IAS 39). Second, we demonstrate that the staging classification drives the tightening in LTV. For bank borrowers with previous underperforming exposures, increasing LTVs might become proportionally more expensive. Finally, we show that underperforming borrowers are exposed to further LTV reductions when they apply for mortgages with higher maturity and higher climate-related risks. Our findings provide important implications since the extensive staging downgrades create the conditions for credit crunch phenomena.