From its roots in agrarian democracy, American banking policy has been skewed to favor small institutions and a diffusion of economic power. Small banks thrived, with efficiency and profitability measures historically exceeding those of bigger banks.

But a cult of bigness is overtaking those notions. Blockbuster mergers dominate bankers' thoughts, and a few institutions with hundreds of billions in assets seem more and more to be calling the industry tune.

What gives?

The free market is speaking. The message may be a bit ambiguous, but old assumptions are being swept aside as stakes are re-calculated for a new kind of endgame. Gross revenues and market power matter more than ever.

Bank bigness is counterintuitive, if a generation's worth of studies on scale economies-the ability to lower costs by getting bigger-are to be believed. The studies were hard-pressed to find benefits in scale, at least in the way banking has traditionally been defined and at least as far as the biggest banks were concerned. (For example, scale may have worked to the benefit of a $100 million-asset bank over a $50 million bank, but not to a $1 billion bank over a $100 million bank.)

David B. Humphrey of Florida State University concluded in a 1994 paper with Allen N. Berger of the Federal Reserve Board that scale economies "at the very smallest banks ... allow average costs to fall with increases in bank size, but they account for less than 5% of costs. For larger banks, constant average costs or slight diseconomies prevail."

But Mr. Humphrey and Mr. Berger collaborated on another study indicating that megamergers significantly increased revenue production, and hence "profit-efficiency," especially for banks that did not perform well before merging.

Edward J. Kane, professor of finance at Boston College, sees no contradiction, just an "over-reaching of the data." Mr. Kane said most academic research has relied on the Federal Reserve Functional Cost Analysis, a fount of real bank data that long showed large banks to be at a unit-cost disadvantage. The problem was that not enough of the biggest banks participated in the functional cost program.

"There has never really been adequate data on the big banks," said Mr. Kane, adding that there must be good reasons for the high market multiples that target banks are fetching. (Another complicating factor, he pointed out, is the idea that some banks are "too big to fail" or "too big to discipline," which gets factored into stock prices.)

Many observers also assume that technology is changing the game and that a bank has to be big to play it well. One body of economic thought would question this. Nobel laureate Ronald Coase of the University of Chicago postulated that as communications technology drives transaction costs down, corporation size tends to decline-not increase.

Why has this not happened in such a technology-driven world as banking?

Hoover Institution research fellow David B. Henderson, who recently quoted Coase's theory in Fortune magazine, suggested that it applies only when "all else is equal." Because of past restrictions on bank expansion, Mr. Henderson said, "all else is not equal." With deregulation, U.S. institutions are beginning to find their way toward a market structure closer to that of Canadian banking.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.