Break Up the Banks
From the April 5, 2010, issue of NR.


Arnold Kling

Turn now to the question of efficiency: Is bigger better for consumers? Bankers speak mystically about the “financial supermarket” and claim that there are tremendous economies of scope in financial services, meaning that a consumer benefits from being able to have a checking account and a stock portfolio at the same large firm. But in practice, whatever benefits might be derived from such a supermarket are probably more than offset by the diseconomies of managing such a complex entity.

Another unsound argument is that large banks are needed to finance large multinational firms. If large international firms require big capital investments, these can be obtained by issuing securities or by loan syndication, in which the risk of borrowing is spread across several banks. The existence of large non-bank firms does not imply the need for similarly gigantic banks.

There are economies of scale, but small banks can take advantage of them, too. For instance, a small bank can join an ATM network or contract with a third party to develop Internet services. It does not have to build such systems from scratch, and we do not need big banks to make them possible.

Which brings us to the question of systemic risk. Regulation can, of course, make systemic risk worse: The U.S. banking crisis of the 1930s was exacerbated by the fact that banks could not start new branches across state lines or, in many cases, even within the same state. This led to poor diversification of regional risk. The regulation in question was admittedly poor, but we need not return to the banking system of the 1930s to achieve a reduction in the size of America’s largest banks.

Some point out that the Canadian banking system performed relatively well during the financial crisis, noting that Canada’s assets are concentrated in just five large banks. This is offered as evidence that large banks are conducive to financial stability. But while Canada’s big banks have a big share of the country’s assets, they still are much smaller than America’s largest banks: Bank of America and JP Morgan Chase are three or four times the size of the Royal Bank of Canada, Canada’s largest. And while its banking marketplace is dominated by five big players, Canada’s population is less than one-seventh that of the United States; even if we concede that Canada is served well by five large banks, the equivalent in the United States would be 35 large banks. In 2008, total assets of the U.S. banking system were about $10 trillion, with the top five bank holding companies in possession of $6 trillion. If the entire $10 trillion had been divided evenly among 35 banks, none would have accounted for more than $300 billion in assets; all of our banks would have been smaller than the fifth-largest Canadian bank.

Overall, there is little evidence that really big banks are necessary to a sound financial system. The financial crisis demonstrated that they are not sufficient for a sound financial system. And it is possible that without very large banks the system actually would be more robust. Certainly, the failure of any one bank would be less traumatic if the size of that bank were small relative to the overall market.