Debunking the Myth That Bigger Banks Are More Efficient and Necessary

A very good op ed by Thomas Hoenig in the New York Times, “Too Big to Succeed” provides a solid recap of why the business of reining in the too big too fail banks is crucial. It isn’t simply that this is yet another version of “Mission Accomplished”; the bailouts actually made industry concentration worse, as Hoenig indicates.

Mike Konczal sheds more light on how the rescues helped the biggest banks at the expense of their smaller bretheren. However, what astonishes me is this drive by ad hominem by Matt Yglesias that Konczal links to in his post.

While Yglesias often has sound observations, I believe in not going beyond the limits of one’s knowledge, and Yglesias is out of his depth here. Yglesias is effectively saying smaller banks are less competitive, which is bunk. Every study ever done of banking efficiency in the US over the last 20 years shows that once a certain size threshold is reached (studies vary in their conclusions due to variations in sample and methodology as to where that is, but $25 billion is) banks show an increasing cost curve, meaning they are less, not more, competitive.

So why has there been consolidation in the banking industry? Why would banks get bigger to become less efficient? Simple. It’s the bad incentives and the usual principal-agent problem. Banks are not run to suit the interests of shareholders but of top management (and in big dealer firms, the various producers, who as we describe in ECONNED, can hold the company leadership hostage).

CEO pay is highly correlated with the asset size of banks. Big banks don’t out-compete smaller banks; they simply buy them up. The acquiring bank managers get a pay increase; the managers of the purchased banks get a windfall, since the deal will trigger payouts under their golden parachutes. It’s a win for everyone except the shareholders of the acquiring bank. Typically, the pitch for these deals is cost savings, which is fallacious. The increasing cost curve suggests at least three factors are at work:

1. Many of the cost savings pushed through in the merger could have been achieved by each organization separately. The merger was an excuse for rather than necessary for cost reduction

2. Planned merger savings may not be achieved. For instance, many banks find they can’t close as many branches as they thought, or if they try, they lose customers.

3. The visible cost reductions (which operate within lines of business) are offset by diseconomies of scope (more management layers, more coordination, and of course, those more costly top executives)

Now there is yet another factor that is often overlooked, that smaller banks can do better on the revenue side. Customers will often accept slightly lower rates on deposits (in the days when deposits actually paid meaningful interest) or pay more in loans to deal with a smaller bank if they perceive they get better service. Smaller banks also have flatter hierarchies, and are thus more able to include local, qualitative information in their lending decisions, which has the potential to lead to better outcomes.

Moreover, banks now enjoy large ongoing subsidies thanks to the TBTF status. As Bank of England officer Andrew Haldane tells us:

Table 2 looks at this average ratings difference for a sample of banks and building societies in the UK, and among a sample of global banks, between 2007 and 2009. Two features are striking. First, standalone ratings are materially below support ratings, by between 1.5 and 4 notches over the sample for UK and global banks. In other words, rating agencies explicitly factor in material government support to banks.

Second, this ratings difference has increased over the sample, averaging over one notch in 2007 but over three notches by 2009. In other words, actions by government during the crisis have increased the value of government support to the banks. This should come as no surprise, given the scale of intervention. Indeed, there is evidence of an up-only escalator of state support to banks dating back over the past century.

Table 3 takes the same data and divides the sample of UK banks and building societies into “large” and “small” institutions. Unsurprisingly, the average rating difference is consistently higher for large than for small banks. The average ratings difference for large banks is up to 5 notches, for small banks up to 3 notches. This is pretty tangible evidence of a second recurring phenomenon in the financial system – the “too big to fail” problem. It is possible to go one step further and translate these average ratings differences into a monetary measure of the implied fiscal subsidy to banks. This is done by mapping from ratings to the yields paid on banks’ bonds;6 and by then scaling the yield difference by the value of each banks’ ratings-sensitive liabilities.7 The resulting money amount is an estimate of the reduction in banks’ funding costs which arises from the perceived government subsidy.

Table 4 shows the estimated value of that subsidy for the same sample of UK and global banks, again between 2007 and 2009. For UK banks, the average annual subsidy for the top five banks over these years was over £50 billion – roughly equal to UK banks’ annual profits prior to the crisis. At the height of the crisis, the subsidy was larger still. For the sample of global banks, the average annual subsidy for the top five banks was just less than $60 billion per year. These are not small sums.

There are other spurious arguments, that we “need” big sprawling international banks. There are some activities that do require international scale, such as payment and cash management platforms for big multinationals, and acting as a dealer in OTC debt markets. But overly large, overly internationalized banks is precisely what gives them leverage over nation-states. We have a very peculiar regulator arrangement in which an international bank is nominally regulated in each of the countries in which it has licensed operations, but the capital actually (in theory) sits in the home country. So if the German regulator of a US bank is unhappy with the books of the German operation, the German sub has to get more capital from the mother ship. This has the effect of making the home country responsible for overseas operations that its regulators do not supervise.

One way to make the TBTF banks somewhat less influential is to put them on an “every tub on its own bottom” footing and require banks in foreign countries to be separately capitalized. That would have the effect of compartmentalizing the operations of each bank further than they are now. That would also mean that a German regulator could shut down the German operations of a US bank and not need much coordination with the US regulators. The resulting greater national compartmentalization would presumably make resolution less difficult (bankruptcies are national; thus forcing international banks to limit all but the most important elements of international integration would be a step forward).

The large banks would of course complain bitterly, since this sort of change, which would have to be implemented over a period of years, would increase their operational costs and require them to hold more equity or shrink their balance sheets (they would presumably not be able to “share” capital as readily across operations). But with a bloated, heavily subsidized, and value destroying financial system, measures to increase safety will inevitably reduce efficiency and shrink the industry. These outcomes would be features, not bugs, and should be celebrated.


Originally published at naked capitalism and reproduced here with permission.