Comment on "Competition and Financial Stability"

Allen and Gale (2004, this issue of JMCB) provide a thought-provoking tour d'horizon. I'd call it "ambitious," except that what they label a "prelude to the development of a theoretical framework" is too honest to be ambitious: they steadfastly refuse to give a definitive answer to their fundamental question, "Does competition encourage or discourage adequate financial stability?"—in other words, "Is there a tradeoff between competition and financial stability or are the two goals mutually reinforcing?"

In addressing this question, Allen and Gale employ several different interpretations of the two goals: In some sections, "competition" means price-taking behavior. In others, they interpret competition to be imperfect competition with some degree of market power. By "instability" Allen and Gale sometimes simply mean "the possibility of default." Often instability means the choice of overly risky or undiversified investments. But in their most interesting case instability means a high degree of interdependence between bank defaults.

1. The Arguments

I will also use this review as an opportunity to consider the often uncomfortable relationship between economic theory and policy making. Policy makers want, and need, definitive answers to questions like Allen and Gale's fundamental question. Theorists spin answers in every possible direction. When we are lucky, empiricists sort through those answers to determine their relative importance. Nonetheless this process takes time and often fails to reach any definite conclusions. In the meanwhile, the theoretical arguments are all we have, and like Harry Truman, current policy makers sigh for a "one-handed economist."

Allen and Gale give them no comfort. They provide, not two, but many hands:¹

1. The invisible hand.

Allen and Gale (2003), following Prescott and Townsend (1984), show conditions under which the welfare theorems hold in economies with financial intermediaries. In terms of this work "competition" means Walrasian price taking and "financial instability" means violating a contract by declaring bankruptcy. In this sense, financial instability is only present in an incomplete market setting; in a complete contract setting, there is no sense to the notion of violating a contract because every possible reaction and consequence is included in the terms of the contract itself. Allen and Gale (2003) specify a class of models in which incomplete contracts are optimally chosen, provided the market for underlying Arrow securities is complete. Crucial to the structure is the assumption of no ex ante differential information; contracts are written and Arrow securities traded by symmetrically informed agents. In such a world, there is no reason for a government to intervene to stabilize a financial market.

Although this is a natural starting point for a discussion, the argument seems to me to be a straw man. Even though bankruptcy procedures are expensive, sometimes particular banks ought to fail; few regulators or policy makers would argue otherwise. The real costs of financial instability must be found elsewhere.

2. A neighborly hand.

Geographic diversification reduces risk. It may also reduce competition, depending on whether the geographically diverse banks have local monopolies or end up competing region by region. Nonetheless, since it is hard to see how consolidation could increase competition, score one for the tradeoff camp.

3. A hand in the cookie jar.

Limited liability and high leverage, possibly exacerbated by deposit insurance, by a winner-take-all objective function, or by some, though not all, market-share objective functions (or by options in executives' compensation packages?) tempt banks into overly risky positions. The lower the bank's charter value (for example due to increased competition), the greater the temptation. Allen and Gale's model of this is attractive in its simplicity and illustrates one classic form of the tradeoff between risk and competition. And yet, as they point out, the result is reversed if the potential for excessive risk taking is in the hands of the borrowers rather than the banks, or if the objective includes being the "last one standing."

4. The dead hand of the past.

Lock-in of demand, through fixed costs of switching banks, can cause otherwise competitive firms...