The Unassuming Economist

From a new IMF working paper:

“The risk shifting incentive identified by Jensen and Meckling (1976) can induce excessive
risk taking by banks in a competitive environment (Hellmann, Murdock, and Stiglitz (2000)).
This paper tests this risk shifting hypothesis of competition in the U.S. mortgage market
between 2000 and 2005. Our study exploits a natural exogenous variation of local house
price volatility in the cross section of U.S. cities and counties, one of the most important
sources of risk for mortgage returns. This paper finds that banks in high-competition markets
lowered their lending standards (e.g., raising the loan-to-income ratio and acceptance rate) in anticipation of high house price volatility while those in low-competition markets did not, an indication consistent with the risk shifting hypothesis.

This paper also examines the real economic consequences of this risk taking pattern through
the credit supply channel. In particular, it studies the change in local employment in
non-financail sectors at the beginning of the Great Recession. We find that between 2007 and
2009 non-financial sector employment in high competition markets lost 1.5 percent for one
standard deviation increase in local house price volatility, while this relationship was
insignificant for low-competition markets. This exercise identifies a credit-supply channel, in
addition to the demand channel shown in Mian and Sufi (2014), that contributed to the rise in non-financial sector unemployment during the Great Recession.

The analysis in this study shows the importance of banks’ risk taking incentive due to
competition prior to the recent crisis. It helps deepen the understanding of why the financial
sector had accumulated so much mortgage risk despite that an reverting house price would
lead to massive mortgage defaults (e.g., Palmer (2015)). When studying the impact on the
real economy such as non-financial sector employment, this risk taking pattern can also be
used to identify the credit supply channel of bank lending. This analysis offers a possible
strategy to disentangle the supply and demand effects of bank lending on real economic
activities.”