Measuring the Costs of “Too Big To Fail”

by Mark Roe

CAMBRIDGE – The idea that some banks are “too big to fail” has emerged from the obscurity of regulatory and academic debate into the broader public discourse on finance. Bloomberg News started the most recent public discussion, criticizing the benefit that such banks receive – a benefit that a study released by the International Monetary Fund has shown to be quite large.

Illustration by Pedro Molina

Bankers’ lobbyists and representatives dismissed the Bloomberg editorial for citing a single study, and for relying on rating agencies’ rankings for the big banks, which showed that several would have to pay more for their long-term funding if financial markets didn’t expect government support in case of trouble.

n fact, though, there are about ten recent studies, not just one, concerning the benefit that too-big-to-fail banks receive from the government. Nearly every study points in the same direction: a large boost in the too-big-to-fail subsidy during and after the financial crisis, making it cheaper for big banks to borrow.

But a recent research report released by Goldman Sachs argues the contrary – and deserves to be taken more seriously than previous efforts to dismiss the problem. The report concludes that, over time, big banks’ advantage in long-term funding costs relative to smaller banks has been one-third of one percentage point; that this advantage is small; that it has narrowed recently (and may be reversing); that it comes from the big banks’ efficiency and their bonds’ liquidity; and that historically it has been mostly small banks, not big ones, that have failed.

Goldman is surely right that the United States has historically propped up small banks that it would not allow to fail. The savings-and-loan crisis of the 1980’s was a $100 billion problem. The banks that failed during the Great Depression of the 1930’s were small banks. More perniciously, because small banks, unlike big ones, paid for most of the costs of their failure through an insurance fund, small banks were given local monopolies, for fear that vigorous competition would lead too many to fail. Customers suffered because competition was weak.

The Goldman report seems to be aimed at a recent proposal by Senators Sherrod Brown and David Vitter to increase big banks’ capital requirements sharply, while exempting small banks (presumably to get the small banks’ political support, or out of nostalgia for local banking, or because the senators believe that small banks present no regulatory problems). Because small banks have historically been failure-prone, and have failed en masse, the Brown-Vitter proposal sidesteps a major problem – call it “too-many-to-fail.”

But, although Goldman is right to widen the focus to include small banks’ weakness, doing so undermines the logic underlying its view that the narrowing of the long-term funding-cost advantage implies that America’s banking problems are over. Goldman is right that small banks fail and have been protected by government insurance and bailouts; but that means that the big banks’ funding advantage over small banks needs to be added to whatever advantage the small banks themselves have.

Likewise, if big banks’ funding advantage is really not so large now as it was in the past, that narrowing may mean that others understand what the Goldman report is saying: small banks fail, too, and the government or the insurance fund bails them out. A narrowing gap in funding costs could just mean that financial markets recognize this.

Moreover, by measuring the too-big-to-fail subsidy on the basis of banks’ long-term bonds, the Goldman report misses much. It is a big bank’s short-term debt that sinks it in a crisis, and it is this debt that is bailed out first. Big banks, not small banks, are the major players in the market for short-term debt, which makes their bonds riskier than small banks’ bonds. So, if the market prices the big and small banks’ long-term debt similarly, even though the big banks’ debt is riskier, something must be giving the big banks’ riskier debt a boost.

US regulators are strongly hinting that they will allow long-term debt to default in a bank failure, while affirming that they will find a way to bail out short-term debt. If financial markets view them as likely to follow through, the too-big-to-fail boost may apply more to the big bank’s short-term debt than to any banks’ long-term debt.

And is it right to say that the amount involved – one-third of one percentage point in annual interest savings on long-term debt – is small? Big banks use so much debt nowadays and so much less equity – the ratio is typically ten to one – that a small financing advantage amounts to a large fraction of a bank’s profit. Depending on how much long-term debt a bank uses, a financing advantage of one-third of a percentage point can readily amount to 10% or more of its annual profit. That is not small change by anyone’s definition.

Finally, if we focus on the too-big-to-fail subsidy at any particular time, we miss too much. As the economy improves, failure is less likely. The banks may remain too big to fail, but, with their chance of failing lower in a more buoyant economy, the immediate subsidy declines. And the largest cost of major bank failures is not the subsidy, or the cost of a bailout. It stems from the economic havoc caused by too many financial institutions weakening and unexpectedly failing simultaneously, cutting back on lending, and degrading economic activity overall. Mass financial failure is costly, even if no one, big or small, is bailed out.

See the original article >>