Bursting The Bubble

Based on research by Brian Akins, Lynn Li, Jeff Ng and Tjomme Rusticus

• Low levels of competition in the banking industry make it likelier that banks will engage in riskier activities — and, ultimately, collapse.
• States with lower levels of banking industry competition saw greater housing price inflation during the 2007-2008 financial crisis.
• In states with higher levels of competition, housing price fluctuations were mitigated.

When banks become “too big to fail,” does their size empower them to take risks smaller banks wouldn’t? It’s easy for Monday-morning quarterbacks to blame the economic collapse of 2007-2008 on bloated financial institutions who lacked rivals to keep them on their toes. But research co-authored by Brian Akins, a professor at Rice Business, applies classic economic measures of competition to determine whether oversized banks and limited competition played a key role in the building — and bursting — of the housing bubble.

Akins’ research does, in fact, link a lack of competition in the banking industry with riskier bank investments, a higher level of regulatory intervention, and, ultimately, bank failure. It also associates low levels of banking industry competition with housing price inflation and steeper housing price declines during the 2007-2008 financial crisis.

While the results may seem intuitive — especially in hindsight — several economic theories actually predict the opposite. For example, the charter value hypothesis presumes that large banks have correspondingly large charter values, and hence have more to lose in case of failure. The theory predicts that larger banks are therefore motivated to counter the potentially massive potential for failure by engaging in lower-risk activities. An increase in competition, the theory holds, would lower the bank’s charter value and free it to engage in riskier investments.

But this theory’s detractors argue that it ignores the effect of bank competition on borrowers’ behavior. According to a counter theory, as the lending market becomes more concentrated, banks use their market power to charge higher loan rates, leading to an increase in their interest margin. Higher loan rates increase the probability of bankruptcy for borrowers — and the increased likelihood that they’ll default on their loans reduces bank stability.

Akins and his colleagues set out to conduct a comprehensive examination of the relationship between banking industry competition and financial stability in the United States — specifically, the impact of competition on individual bank outcomes and on the housing and mortgage market within states.

Their first round of analysis demonstrated that just before the 2008 crisis, more competition was associated with lower interest margins, a less risky portfolio of assets, lower profitability and lower liquidity. During the crisis, more competition was associated with less risk-taking, and, consequently, a lower likelihood of enforcement actions and bank closure.

In a second layer of analysis, the researchers examined the relationship between competition and changes in housing prices before and during the crisis. They found that greater competition led to higher mortgage rejection rates, especially for the highest-risk mortgages. More competition, the authors concluded, seemed to have a disciplinary effect that mitigated the housing price inflation before the crisis and the deflation afterward.

So the Monday-morning quarterbacks are right — and Akins’ study provides the data to prove it. A lack of competition, these results suggest, fueled the hyperinflation of housing prices, their subsequent collapse, and the devastating effects for the U.S. financial system and economy that still linger a decade later.

Brian Akins is an assistant professor of accounting at Jones Graduate School of Business at Rice University.

To learn more, please see: Akins, B., Li, L., Ng, J., & Rusticus, T.O. (2014). Bank competition and financial stability: Evidence from the financial crisis. Journal of Financial and Quantitative Analysis, 51(1), 1-28.