Bankable Frontier Weblog

The US currently has 7598 commercial banks, down from just under 10 000 a decade ago. According to the first quarter 2005 FDIC Quarterly Banking Profile, 94% of these banks today reported assets of less than $1 billion. The assets of all these small institutions together constituted just 14.8% of total banking assets, 4% lower than the share four years ago. Medium sized institutions (assets $1bn-10bn) were similarly squeezed, with their asset share also reducing by around 4% to 13.7%. The gainers were clearly the large banks, which now control almost three quarters of US bank assets (up from around two thirds four years ago). The dominance of big banks in the US system is growing. What is going on here?

Little bank, big bank—what’s the difference?

The underlying data on performance by size category tell an interesting story. The smallest banks, with assets below $100m, report a net interest margin of 4.14%, considerably higher (almost 70bp) than that of the largest category (more than $10 billion). This is generated by a higher yield on assets (44bp) and lower funding costs, presumably from retail deposit bases (30bp). They also report much lower bad debt charge-offs. At this stage, you might begin to wonder why big banks are growing so fast.

The data also show, however, that very small banks earn less non-interest income than very big banks (presumably to do with trading and wider activities at big banks), have to carry more capital (almost 12% versus 10%) and have higher expense ratios (cost-to- income of 70% versus 55%). The overall result is that, at the bottom line, large banks are much more profitable than very small banks, reporting a Return on Equity of 13.5% versus 8.8%.

These US trends towards greater concentration in banking appear to confirm recent empirical findings about economies of scale in finance. Economic theory has long made the case that large financial intermediaries were more efficient than small. This was due both to the greater risk diversification made possible by larger size, as well as cost efficiencies. However, empirical research in the Eighties and early Nineties found little or no evidence of this in practice. If anything, the scale economies were either very modest, or confined to certain size categories—medium sized banks—and petering out at greater sizes.

More recent research has become more sophisticated, for example, controlling for the effects of risk taking on cost. Now, empirical researchers find large economies of scale for all sizes of bank. This research is well summarized in the second section of a 2002 IMF Working Paper by Biagio Bossone and Jong-Kun Lee. Cost economies at larger bank sizes are further reinforced not only by new technology but also by reputation signaling effects—larger banks are seen to be less likely to fail, hence can pay less at the margin for their capital. Studies in a variety of other countries have confirmed that these scale effects are not limited to the US.

With all these powerful forces working against them, can small banks survive in the long run? A chapter in a recent book on The Future of Banking (Ed. Benton Gup) is quite sanguine.  The authors, Robert de Young and William Hunter, suggest that there can continue to be a dual equilibrium market structure—in which there are fewer but stronger small community banks; alongside fewer but larger big banks. Community banks will survive, they say, by exploiting these informational advantages at the ‘last mile’ to the full.

Emerging markets are not exempt from the same forces shaping their banking systems, which will drive towards greater concentration of assets in larger institutions. Microfinance institutions are the community bank equivalents in developing countries today. In the face of these prevailing winds, this evidence from the US suggests that MFIs need strategies to get big fast; or else prepare to defend their niched local advantages strongly against large potential newcomers.