Essays on employment dynamics and asset prices

Date
2013
DOI
Authors
Siemer, Michael
Version
OA Version
Citation
Abstract
This dissertation consists of three essays that examine the importance of financial constraints and uncertainty for employment outcomes and asset prices. The first chapter documents that the recession of 2007-2009 has been characterized by: (1) a large drop in employment concentrated in small firms, (2) an unprecedented decline in the number of firms, and (3) a slow recovery. I study if these observations can be reconciled in a theoretical model. I develop a heterogeneous firm model with labor adjustment cost, endogenous firm entry, and financial constraints that generates these key facts. The model further predicts that a large financial shock results in a long-lasting recession when the model is calibrated to match the firm size distribution. The second chapter provides empirical evidence that financial constraints affected employment growth most strongly in young small firms during the 2007-2009 recession. I use a confidential firm-level employment data from the Bureau of Labor Statistics and financial data from Compustat to construct a sector level external financial dependence measure. I find that in the period of 2007-2009, small and young firms in sectors with high external finance dependence exhibited lower employment growth than those in low external finance dependent sectors. The effect of external finance dependence on employment growth in small and young firms is largely driven by firm entry and exit. For large firms I do not find a significant difference between sectors of difference external financial dependence. The third chapter studies the asset pricing implications of learning about economic uncertainty in an endowment economy. This work is motived by the widely held belief at the beginning of the 2007-2009 financial crisis that the economy might be at the onset of another Great Depression. Economic uncertainty in the model takes the form of rare events that are not directly observed. We show that when agents employ Bayesian learning rules, learning endogenously introduces a time-varying risk premium. The theoretical model is able to match several empirical asset-pricing facts including the risk premium and the predictability of returns.
Description
License