The economic consequences of bank transparency for risk-taking decisions is unclear and prior empirical evidence is ambiguous. On the one hand, greater transparency helps establish market discipline and provides incentives for early corrective actions by managers. On the other hand, greater transparency is harmful for financial stability if the information conveyed to the market contributes to the erosion of trust among market participants. I study the introduction of IFRS 9 in the European banking sector in the financial years 2017 and 2018 to address this question. The new reporting regulation requires banks to adopt a more timely, and therefore more transparent, approach to the provisioning for loan losses. The new rules follow an expected loss approach and replace IAS 39's incurred loss model which standard-setters and regulators had criticized for its alleged opacity about potential loan losses ("too little, too late"). The expected accounting consequences are massive and, for example, KPMG predicts a one-time increase in loan-loss provisions by up to 50%. The introduction of the new accounting regulation is therefore a suitable laboratory to study the economic consequences of greater transparency about banks' credit risk. Specifically, I will focus on whether risk-taking decisions by managers (e.g., risk shifting) and risk perception by market participants changes around the adoption of IFRS 9. I will exploit intertemporal variation in the exact adoption dates as well as cross-sectional variation in the magnitude of the accounting change to address plausible endogeneity concerns.