Politics, Not Markets, Makes Banking Unstable

Is fractional-reserve banking inherently unstable? Over at National Review, Edwin Burton, a visiting economics professor at the University of Virginia, argues that the “mismatch between the maturity of the source of funds and the maturity of the use of funds” creates run-prone system. His argument, however, overlooks important contributions to monetary economics. Fragile banking systems usually result from political design, rather than economic necessity.

The fractional-reserve fragility hypothesis makes a prediction that is largely falsified by banking history. Inherent fragility implies runs should be largely random. But this isn’t so. Runs usually occur at banks where depositors have a rational basis to question the health of the balance sheet. The most recent round of failures, including Silicon Valley Bank and Signature Bank, didn’t happen at sound institutions beset by bad luck and depositor hysteria. They happened at unsound institutions with foolish capital structures.

Public policy, not market forces, makes banking unstable. US banking history is case in point. Contrary to the popular impression of unregulated “cowboy capitalism” in the republic’s early years, banking has always been heavily controlled by the state. Two especially costly restrictions imposed on banks were (a) limits on note issue based on government bond holdings and (b) limits on branching. The first rendered the money supply inelastic to the needs of trade. Money demand shocks needlessly threatened balance-sheet integrity. The second overexposed banks to location-specific risk. Shocks to agriculture and industry, sectors largely underwritten by nearby banks, could have been absorbed had banks been permitted to branch. It also meant that notes traded at a discount when circulating far from the issuing bank, reflecting the cost of redemption, and that money demand was somewhat less stable, since note-traders would gather notes up in one location, transport the notes back to the issuing bank, and then present them all at once for redemption.

Today, banking lacks a crucial supporting institution: multiple liability. Historically robust systems, such as those in Canada, Scotland, and Sweden, did not permit all banks to incorporate on a limited-liability basis. Some combination of double, triple, and unlimited liability was the norm. Bank shareholders were thus liable for their personal wealth, not merely their investment in the bank, should the bank prove unable to meet depositor claims. This greatly improved bank owners’ and managers’ incentives for moderating risk.

Burton prefers a form of 100-percent reserves to the current system. He rightly notes this would not eliminate banking, meaning financial intermediation. Banks could perfectly align time deposits with loans to eliminate duration risk. (Repayment risk would remain, of course). But he overlooks the significant transaction costs of this system. Making loans out of pooled depositor funds is much less costly for the bank, which makes for thicker and more-responsive capital markets. Most of what’s wrong with banking is political in nature; why hamper it even further with government “fixes” that are likely to come with a host of unintended costs?

Our financial system is a mess. It was tolerable, at best, until 2008, and has gone downhill since then. First-best reforms are likely off the table, and reasonable people can disagree about acceptable policy in a highly imperfect world. Nevertheless, it’s important we get the diagnosis correct. We won’t be able to make marginal improvements if we fundamentally misidentify the problems with banking. 

Burton’s critique of fractional reserves contains much more heat than light. It repeats several theoretically and historically unfounded claims about how banking systems work. Banking reform should rank high on our list of policy priorities, but moving away from fractional reserves shouldn’t be a part of the conversation.