Dividing Financial Institutions into Utilities and Casinos?

One proposal for addressing too big to fail, or to systemically interconnected to fail, among financial institutions is to separate out the proprietary trading and other “casino” activities from the “utilities” business of commercial banking with the public.  In some ways (not all) it is a revival of the Glass-Steagall approach.  Paul Volker has urged such a policy, as have others.  The Obama administration has not so far shown any appetite for such it, preferring, in its Treasury blueprint for reform, to allow the functional interconnections within holding company structures, and identifying institutions that are regarded as too big or too systemically interconnected to fail and apply “regulation and last resort lending” to apply to them.

Something like the same debate is taking place in Britain, and the Financial Times’s Martin Wolf makes a comment on why one could see the functional separation desirable, but also why it is hard to do and hard to ensure that it actually reduces the systemic risk.  In response to a recent speech by Mervyn King of the Bank of England calling to separate out the “casinos” from the “utilities,” Wolf says:

it is evident why this distinction is appealing. If we define the utility parts of the financial system narrowly, as management of the payment system, it works like clockwork. It is in the management of risk (and the advice given to its clients) that the financial system fails. The limited liability businesses at the heart of our credit-based monetary system have a tendency to mismanage risk (and uncertainty), with devastating results.

However, he ultimately says that he is unpersuaded that a modernized form of Glass-Steagall can work as a structural solution to systemic risk:

Yet I remain unpersuaded that the structural solution – the separation of utility from casino finance – is workable, as I pointed out in a column on the “narrow banking” proposal of my colleague, John Kay. Indeed, Mr King himself is well aware of the difficulties.

First, the border between utility and casino banking is impossible to draw. For Mr Kay, the utility is the payment system and protection of deposits. This would leave all lending – including to households and businesses – inside the casino. For those in the US who hark back to the Glass-Steagall Act, the distinction is between commercial and riskier investment banking.

Mr Kay’s distinction is clear, but problematic. If we followed him, all risk management would become unregulated. It is inconceivable that governments would, or could, leave them so. If we moved back to a Glass-Steagall distinction (itself never accepted in continental Europe), we would need to draw a line. But where? Why would lending to households and business be good, but securitising those loans bad? Why would hedging be good, but speculating bad and how might one draw the line between them? Mr King counters that prudential regulation already draws such distinctions. I would respond that regulation has made a mess in doing so. Furthermore, these are not distinctions between businesses.

This is not to argue that there is no way of making finance safe. There is. But it would be far more radical: deposits would be 100 per cent reserve backed; and the liabilities of other investment vehicles would be adjusted for the market value of their assets at all times. Banking would disappear.

Short of such radicalism, we must approach the task in a more subtle manner. First, create a set of laws and institutions that make it possible to bankrupt any and all institutions, even in a crisis. Second, make financial institutions safer, with much higher capital requirements, against all activities. Third, prevent off-balance-sheet activities. Fourth, impose dynamic provisioning. Fifth, require huge cushions of contingent capital. Finally, cease to favour debt-finance, throughout the economy.

If we did all this, the world of finance would be duller and safer. It would still not have the reliability of jet engines. So long as we allow people to make leveraged bets on the future, breakdowns will occur. The division of finance into utility and casino cannot solve this problem. Only the end of leverage would do so. Do we want that? I doubt it.

I am still thinking through policy on this.  I am more or less persuaded that the Treasury view in the US represents a bad compromise that won’t  prevent the next crisis while stifling activities under regulation that might well turn out to have been both intrusive and yet mostly pointless.  Yet, while accepting Wolf’s criticisms of the casino-utility distinction, I question whether there is still not a role for a structural separation even if one recognizes that it does not solve all the problems of systemic interconnection of institutions via markets.  Constructive comments on the best approach to too big too fail and systemic risk welcomed.