Everyone wants to ring-fence something, but they can’t agree on what: Vickers, Liikanen, Volcker.
In all proposals, the idea is to have bank capital separately allocated for some activity, and to prevent that capital from being exposed to any other activity. Some people want to lock the wild animals in a cage to keep them away from us; some people want to lock the tame animals in a cage to keep them safe from the dangerous world outside.
Vickers wants to ringfence retail banking, with the idea of trying to protect Main Street from the other more risky activities of banks. Liikanen wants to ringfence all trading activities (i.e. Wall Street), apparently in the hope of keeping the rest of the bank safe from them. And Volcker wants simply to ring fence the market-making aspect of trading, in order to separate it from so-called proprietary trading which exposes bank capital to price risk.
All three proposals represent attempts to come to regulatory grips with the dramatic changes in the nature of banking over the last 30 years or so, changes that were revealed to the world by the global financial crisis that began in August 2007 and continues to this day.
My own view is that we need to begin by thinking of banking more generally as dealing, and distinguish between money dealers who quote buy and sell prices for funds, and risk dealers who quote buy and sell prices for risk. I think both of these activities need backstop, not just the money dealers, but different kinds of backstop since funding liquidity is a different thing from market liquidity.
Further, in both cases, we need to distinguish between matched-book dealing and speculative dealing. That’s essentially what Volcker is trying to do, but probably I get to this view from a different chain of logic. The ideal of matched book is to have offsetting risk exposures that exactly net out; if you could really do this, you would not need any capital since you would be bearing no risk. But there is one kind of risk that does not net out, and that is liquidity risk which is systemic.
The larger the scale of the matched book position, the larger the liquidity risk, even if all other risks net out. Because of this, there is always an implicit liquidity put from matched book dealers, both money dealers and risk dealers, to the central bank. My view is that we should make that liquidity put explicit, and then argue about the details, including how much it should cost.
Speculative dealing is an entirely different animal, at least conceptually. It is true that liquidity risk is involved, again in both money and risk dealing, but price risk is the big thing, so this is where you want there to be capital requirements, or other ways of ensuring that the taxpayer is not providing implicit capitalization.
Where does that leave me in terms of the proposals on the table?
I don’t think it makes sense to try to ring fence either Main Street or Wall Street. Shadow banking brought them together—money market funding of capital market borrowing—and the collapse of shadow banking has torn them apart. We could of course simply adopt a regulatory structure that reads that experience as the verdict of history, and so strives to keep the two sides apart forever more. My concern is that maybe that is just piling on, rather than constructively trying to imagine an ongoing engagement between the two, a market based credit system that is shorn of the worst elements of the shadow banking system.
Europe is having its Glass-Steagall moment, and maybe they just have to go through it. But the US has been there and done that. I’m with Volcker that we need to try to distinguish matched-book market making from speculative position taking. The former involves liquidity risk, and requires liquidity backstop, which should be forthcoming. (Maybe we should actively try to concentrate it at central clearing counterparties?) The latter involves price risk, and requires capital backstop, which should be demanded by counterparties in the first place, and by regulators protecting the public purse in the second place.