Abstract. We develop a simple quantitative model of a bank to study hidden loan losses under regulatory scrutiny. The model incorporates stochastic regulatory audit, enabling temporary concealment of losses. Calibration shows that banks typically hide about 0.8% of loan portfolios as defaulted loans. Following a contractionary aggregate shock, this rises sharply to 2.9% staying elevated for four years. We further show that hidden defaults heighten the risk of bank failure and increases its expected costs.

Abstract. We investigate the effect of earlier loan loss recognition on bank loan supply and stability. Although earlier loss recognition improves stability by strengthening the overall loss absorption capacity, it may amplify lending procyclicality. Under the expected provisioning approach, the bank recognizes the bulk of future losses after the deterioration of the aggregate state which constrains its lending capacity during bad times. The tax-deductibility of expected provisions further amplifies procyclicality. The calibration of our model predicts an economically significant increase in lending procyclicality under the new accounting standards of IFRS 9 and US GAAP which adopt the expected provision approach. 

Past presentations:  IFABS, Angers, 2019; The 27th Finance Forum, Madrid, 2019; 2nd Endless Summer Conference Financial Intermediation and Corporate Finance, Athens, 2019;   CERGE-EI, Brown Bag Seminar, Prague, 2019; The Contemporary Issues in Banking Conference, St Andrews, 2019; University of Zürich, Brown Bag Seminar, Zürich, 2020; 2nd Biennial Bank of Italy and Bocconi University conference, Rome, 2020, Bank of Portugal (2021), University of Bristol (2021)

Abstract. In a basic New Keynesian DSGE model with involuntary unemployment and inflation target shocks, we study the role of labor markets in the transmission of persistent monetary policy shocks that increase households' inflation expectations. The model predicts that labor market conditions play important role in the transmission channel of the persistent inflation target shock: quantitatively realistic labor market frictions reduce the expansionary effect of inflation target shock on output by around a half compared to that under the model without labor market frictions. Using a VAR analysis, we further provide empirical evidence consistent with the predictions of our theoretical model.

Past presentations:  EEA 2021;  AEA 2022