After the recession hit and stock values plummeted, in October 2008 the Financial Accounting Standards Board loosened accounting standards for banks to report assets that fluctuated wildly in value amid the market uncertainty.
Xu Jiang is a professor at Duke University's Fuqua School of Business who studies the economic consequences of accounting standards. Working with Pingyang Gao of the University of Chicago's Booth School of Business, Jiang built an economic model that found relaxing accounting rules can prevent bank runs that are based on panic rather than a bank's fundamental health.
Jiang discusses his work in this Fuqua Q&A.
You found that giving banks some flexibility in reporting their assets can avoid bank runs - mass withdrawals that threaten a bank's existence. But there are different kinds of bank runs. Why is that distinction important?
Some bank runs occur because the bank is not being managed properly. That's just the market weeding out the bad banks. But there are also bank runs that result from avoidable panic, when creditors worry that other creditors may also be withdrawing their funds, and fear they will get nothing if they are too late. Banks that are fundamentally sound but not able to withstand a certain number of creditors running are particularly vulnerable to panic-based runs.
The financial crisis forced many banks to sell assets at heavy discounts. They were required to mark down the value of those assets on their balance sheet. The banks argued this mark-down made them look weak and made creditors more eager to run based on panic. We show this concern is at least partially justified in our study.
Giving banks some flexibility in reporting their assets provides the opportunity for those vulnerable banks to appear as strong as the best banks. The stronger pool alleviates creditors' fear that others may be running and thus alleviates the chances of a panic-based run.
What can banks and the FASB take away from this?
Banks are special because they are particularly vulnerable to panic-based runs that are detrimental to social welfare. This vulnerability is due to their business model. Banks typically finance long-term assets (their loans) with short-term liabilities (their deposits). So if all their creditors suddenly demand their money, banks may not be able to satisfy their requests since their loans have not yet paid off. It may not be a bad idea for standard setters such as FASB to give more flexibility in reporting their assets to alleviate those runs.
There's obviously a risk of fraud when institutions are given this kind of leeway. What makes it acceptable in this case?
There is a risk that really bad banks - those with such weak fundamentals that customers should be looking to withdraw funds - can use the flexibility to pretend they are as strong as good banks. In our study we do not find that banks should be given infinite reporting flexibility, as the risk of fraud would be too high. But some degree of flexibility will allow vulnerable banks to avoid panic-based runs without allowing bad banks to appear as strong as good banks. In our model, as in all analytical models, we cannot be very specific about what degree of reporting flexibility is optimal. What we can say is that the more severe the asset-liability mismatch of the balance sheet, the more reporting flexibility should be given.