At a recent economics conference I came across a fascinating paper by Eugene White, which discussed how incentives built into banking during the National Banking Era helped reduce risk taking. The paper changed my views more than anything else I’ve read in recent years. Here’s a few excerpts:
The Dodd-Frank Act of 2010 exemplifies this confusion. Few observers believe that the bill will provide a lasting reform of the American financial system, and many suspect that it will sow the seeds of the next financial crisis. By focusing on the regulation of choices made by borrowers, depositors, shareholders, and bankers, the Act repeats the mistakes made by previous reform legislation. Instead, reform should focus on changing the incentives that parties face to insure that they are correctly aligned to induce the development of less fragile institutions.
. . .
Perhaps, the most important but least heralded change in the ten years prior to the 2008 crisis was the shift by most major investment banks from partnerships to limited liability corporations.
What is notable about contemporary reform is that there is little effort to change the incentives that caused bank executives to take the big risks and a huge emphasis on regulating their choices. The implicit assumption seems to be that incentives and the assignment of liability plays only a small role so that choices must be regulated—that is, the market cannot be made to adequately discipline banks. Could such a market-based system be devised? In this paper, I offer evidence from the American National Banking Era (1864-1913) for the ability of incentives to successfully limit losses from bank failures.
. . .
An essential feature of the National Banking System was double liability for national banks, chartered by the federal government. Concerned that shareholders would not devote the time and resources to adequately monitor banks’ officers, the National Banking Act of 1864 imposed double liability on shareholders. Under this rule, if a bank failed, the receiver could order shareholders to pay an assessment up to the par value of the stock to compensate depositors. This regulation provided a key incentive to shareholders to control the risk-taking activities of bank management. The results of double liability are striking—banks were frequently and voluntarily closed when performing poorly but before they failed, as shareholders sought to avoid assessments. When they did not close a bank soon enough and it became insolvent, it was typically not deeply insolvent; and depositors received a substantial partial payout.
. . .
The two senior officer of the bank were the president who was paid a salary of $10,000 and the cashier who received a salary of $7,000. Some banks required a bond of the president; but Merchants National did not; however, all employees down to the messengers were required to post a bond. The cashier posted a bond of $30,000, while the messengers had to provide bonds of $5,000.
The shareholders delegated the task of monitoring the management to six outside directors (the president was a seventh director). According to the national bank examiner’s report, they were men of “good character and standing” who met twice a week to review the bank’s discounts and loans and examined the bank twice a year, having “all its operations laid before them.” The president and the directors all owned stock in the bank, the president 196 shares and the six directors 213, 50, 40, 20, 10 and 10 shares. Altogether, the outside directors owned 343 shares, which while it was not a large portion of the 30,000 shares, represented substantial potential individual losses if they were assessed their double liability in the event of the bank failing. For the directors owning, just 40 or 50 shares, carried a potential assessment of $4,000 or $5,000, very substantial sums in the late nineteenth century. For the president, a loss of $19,600 would have been equal to double one year’s salary. The cashier could have been assessed $6,200 or slightly less than one year’s salary; but then he had to provide a $30,000 bond. These large potential downside losses for the senior bank management (and even junior officials, considering the bonds posted) and the directors charged with monitoring them created substantial inducements to control and reduce risk taking.
It might be objected that supervision by the regulator, the Comptroller of the Currency was a more important factor in insuring the safety of national banks. However, supervision under the pre-1913 Comptroller was relatively light, relying on the market to discipline the banks.
Take the US banking system of 1864-1913, end branching regulation, add NGDP targeting, and we’d be fine. No need for Dodd-Frank.