Three Essays in Banking and Labor Markets

No abstract available

This thesis spans two fields: banking and labor markets. The first essay contributes to the former field, while the second and third essays contribute to the latter. The research questions that are broadly relevant to Macroeconomics. The second and third essays are based on joint work with Diego Comin, Riccardo Franceschin and Antonella Trigari. In the first essay, I ask: how can we measure and disentangle market power on lending and deposit markets? What are the implications on the relationship between market power and financial stability? I revisit this old question by developing a new structural approach to the joint estimation of markups on lending rates and markdowns on deposit rates for all US depository institutions between 1992 and 2019. Markups (markdowns) are wedges between the observed price for the output (input) good and the price that would realize if the bank was a price taker on that market. Markups have been trending downwards over time, while markdowns have been increasing after the great recession and decreasing as recovery began. Bigger banks tend to exert more market power on lending markets, while smaller banks exert more power on deposit markets. However, markdowns play a larger role in the profitability of banks relative to markups. I show that Herfindahl-Hirschman Indices are positively correlated with markdowns on deposit rates, but negatively with markups on lending rates. I compute the Z-score and the O-score as measures of financial stability within US banks. I show that higher markups are associated with a lower bankruptcy probability. Instead, markdowns correlate positively with default probabilities. When considered jointly, markups and markdowns both correlate negatively with the probability of bankruptcy. These results show that the sources of market power are important in addressing this old question in the literature. In the second essay, we seek to explain the differences in the time series of unemployment we observe across Germany, France, Spain and Italy. We write a standard Diamond-Mortensen-Pissarides (DMP) labor market model with search and matching frictions and we use it to assess differences in labor markets across Germany, France, Spain and Italy. We simulate the model feeding in exogenous shocks to aggregate productivity and to the discount factor. We obtain three main results. We first confirm the finding in Hall (2017) that financial returns are correlated to unemployment with European data, possibly more than labor productivity. Second, we find that discount factors are a promising source of variation to explain fluctuations in unemployment. Finally, we observe that the extent to which the DMP model explains the four countries' variations of unemployment depends on labor market institutions, as captured by the calibration, and on the degree of wage rigidity. However, the timing of fluctuations of unemployment is different across countries. This cannot be explained by different discounts, as they happen to be somewhat contemporaneous across countries. Based on the results of the second essay, we attempt to characterize differences across labor market outcomes in Europe in the third essay. We do so by writing a labor market model where Fixed-Term Contracts (FTCs) and Open-Ended Contracts (OECs) can simultaneously arise in equilibrium. We write a labor market model à la Diamond-Mortensen-Pissarides that allows for heterogeneous match-specific productivities. A random productivity is drawn when a firm and a worker meet, but the parties only observe a noisy signal initially. As long as the match persists, parties may perfectly learn the productivity with a Calvo-style lottery. At the match and based on the information they observe, workers and productivities may decide to reject the match, sign a FTC or an OEC.