April 10, 2015
Erik P. Gilje, University of Pennsylvania
Gilje empirically tests whether firms engage in risk-shifting. Contrary to what risk-shifting theory predicts, he finds that firms reduce investment risk when they approach distress. To identify the effect of distress on risk-taking, Gilje uses a natural experiment with exogenous changes to leverage. Risk reduction is most prevalent among firms that have shorter maturity debt and bank debt. The researcher also finds that risk reduction occurs in firms with tighter bank loan financial covenants. These findings suggest that debt composition and financial covenants serve as important mechanisms to mitigate debt-equity agency conflicts, such as risk-shifting, that are not explicitly contracted on.
Xavier Giroud, MIT and NBER, and Holger Mueller, New York University and NBER
Giroud and Mueller argue that firms' balance sheets were instrumental in the propagation of shocks during the Great Recession. Using establishment-level data, the researchers show that firms that tightened their debt capacity in the run-up ("high-leverage firms") exhibit a significantly larger decline in employment in response to household demand shocks than firms that freed up debt capacity ("low-leverage firms"). In fact, all of the job losses associated with falling house prices during the Great Recession are concentrated among establishments of high-leverage firms. At the county level, the authors find that counties with a larger fraction of establishments belonging to high-leverage firms exhibit a significantly larger decline in employment in response to household demand shocks. Thus, firms' balance sheets also matter for aggregate employment.
Manuel Adelino, Duke University; Antoinette Schoar, MIT and NBER; and Felipe Severino, Dartmouth College
Adelino, Schoar, and Severino provide a novel interpretation of the debt dynamics leading up to the financial crisis of 2007. Earlier research suggests that distortions in the supply of mortgage credit, in particular a decoupling of credit flow from income growth, may have been responsible for the rise in house prices and the subsequent collapse of the housing market. By focusing on individual transactions rather than zip codes, authors show that the relationship between individual mortgage size and income growth during the housing boom never decoupled and was strongly positive, in line with previous periods (and independent of how income is measured). Zip codes with large house price increases experienced significant changes in the composition of buyers, i.e. home buyers had increasingly higher income than the average residents in an area. Instead, there was a significant expansion of credit along the extensive margin due to the increase in the pace of home buying. The fact that a larger fraction of people were holding recent mortgages and were thus close to their maximum debt capacity likely contributed to the build-up of systemic risk prior to the crisis. Compared to prior years, middle and high income borrowers (not the poor), as well as those with high fico scores, made up a much larger share of delinquencies in the crisis. These results are most consistent with an expectations based view of the financial crisis where both home buyers and lenders were buying into an unfolding housing bubble.
Craig Doidge and Alexander Dyck, University of Toronto; Hamed Mahmudi, University of Oklahoma; and Aazam Virani, University of Arizona
Can institutional investors generate sufficient power through collective action to drive improvements in governance? Doidge, Dyck, Mahmudi, and Virani use proprietary data on the private communications of a coalition of Canadian institutional investors and find that its private engagements influence firms adoption of shareholder democracy measures, say-on-pay advisory votes, improve compensation structure and disclosure, and influence CEO incentive intensity. Spillovers from engaged firms to non-engaged firms through board interlocks and informal regulation through definition and dissemination of performance relative to best practices, suggest a broader impact. This form of activism is both a substitute and complement to other interventions to address governance concerns.
Claire Célérier, University of Zurich, and Boris Vallée, Harvard University
Célérier and Vallée empirically test the hypothesis that relatively high returns to talent explain the wage premium for working in finance. The researchers exploit a special feature of the French educational system to build a precise measure of talent that they match with compensation data on graduates of elite French institutions. Using this measure, the authors show that wage returns to talent are three times higher in the finance industry than in the rest of the economy. This greater sensitivity to talent almost fully explains the level of the finance wage premium, its evolution since the 1980s, and, at the individual level, the pay increase workers obtain when joining the finance industry. Finally, returns to talent correlate with the share of variable compensation.
Harrison Hong, Princeton University and NBER, and Inessa Liskovich, Princeton University
Three reasons are often cited for why corporate social responsibility is valuable: product quality signalling, delegated giving, and the halo effect. Previous tests focus on consumers and cannot easily separate these channels because consumers are affected by all three. Hong and Liskovich focus on prosecutors, who are only susceptible to the halo effect. Using prosecutions of the Foreign Corrupt Practices Act (FCPA), the researchers find that more socially responsible firms pay $2.5 million or almost 50% of the median fine less for bribery. The authors use the FCPA's unexpected increase in enforcement to address reverse causality, rule out political donations as a confounding factor and text-mine case files for prosecutorial sentiment.
Jose Azar and Isabel K. Tecu, Charles River Associates, and Martin C. Schmalz, University of Michigan
Many natural competitors are jointly held by a small set of large diversified institutional investors. In the U.S. airline industry, taking common ownership into account implies increases in market concentration that are 10 times larger than what is "presumed likely to enhance market power" by antitrust authorities. Azar, Schmalz, and Tecu use within-route variation over time to identify a positive effect of common ownership on ticket prices. A panel-IV strategy that exploits BlackRock's acquisition of Barclays Global Investors confirms these results. The researchers conclude that a hidden social cost reduced product market competition accompanies the private benefits of diversification and good governance.
Atif Mian, Princeton University and NBER, and Amir Sufi, University of Chicago and NBER
Academic research, government inquiries, and press accounts show extensive mortgage fraud during the housing boom of the mid-2000s. Mian and Sufi explore a particular type of mortgage fraud: the overstatement of income on mortgage applications. They define "income overstatement" in a zip code as the growth in income reported on home-purchase mortgage applications minus the average IRS-reported income growth from 2002 to 2005. Income overstatement is highest in low credit score, low income zip codes that Mian and Sufi (2009) show experience the strongest mortgage credit growth from 2002 to 2005. These same zip codes with high income overstatement are plagued with mortgage fraud according to independent measures. Income overstatement in a zip code is associated with poor performance during the mortgage credit boom, and terrible economic and financial economic outcomes after the boom including high default rates, negative income growth, and increased poverty and unemployment. From 1991 to 2007, the zip code-level correlation between IRS-reported income growth and growth in income reported on mortgage applications is always positive with one exception: the correlation goes to zero in the non-GSE market during the 2002 to 2005 period. Income reported on mortgage applications should not be used as true income in low credit score zip codes from 2002 to 2005.