I examine the economic consequences of the Federal Reserve's enhanced disclosure of supervisory stress test model between 2017 and 2025. Enhanced model disclosure substantially reduces uncertainty about stress test outcomes. I document heterogeneous stakeholder reactions to the enhanced disclosure: equity investors in affected banks anticipate net benefits, bond investors show a muted response, and uninsured depositors shift away from these banks. Consistent with these positive equity reactions, affected banks reduce capital ratios by increasing payouts. I find no evidence that affected banks reduce loan portfolio risk, consistent with the muted bond market response. The capital payout effect concentrates in banks with high disclosure exposure, consistent with model-specific knowledge driving the effects. Finally, shareholders’ positive valuation of model disclosure is driven primarily by higher expected payouts. Overall, the results speak to the current debate on supervisory model disclosure by highlighting uneven distributional outcomes: shareholders capture gains through higher payouts, while depositors and the real economy see no corresponding benefit. 

with Oliver Binz (ESMT Berlin) and Matthew A. Phillips (Northwestern) link 

→ 3rd Round Review and Resubmit at The Accounting Review 

We find that ASU 2016-13 (Current Expected Credit Loss Model) adoption by banks reduces managers’ ability to learn from the market, especially macroeconomic information. 

with Katja Kisseleva-Scherenberger (Frankfurt School of Finance and Management) and Per Olsson (ESMT Berlin) 

Using the full population of Swedish firms from 1998 to 2011, we find that firms’ accounting profitability is predictive of future performance, which combats the view that accounting numbers by young firms are not informative.