The cautious approach to fixing banks will not work

By Martin Wolf
Published: June 30 2009 20:03 Financial Times

With one bound the banks are free, or so it seems. Already, the panic of the autumn of 2008 is fading. The period within which lessons can be learnt and changes made is closing. Yet without radical changes, another crisis is certain. It may not even be that long delayed.

In a recent speech, governor Elizabeth Duke of the Federal Reserve told an anecdote from just after the failure of Lehman Brothers last September. Ben Bernanke, chairman of the Federal Reserve, was asked: “Well, what if we don’t do anything?” To which he replied: “There will be no economy on Monday.” Instead, all institutions deemed systemically significant were saved, by shifting almost all of the risk on to taxpayers.

“Never again” might be too much to ask. But “not for a generation” is essential. Governments cannot afford an early repeat, financially, politically, perhaps morally: the lives of so many cannot soon be sacrificed to the whims of a foolish few.

Yet what has emerged after the crisis is, as I argued last week , an even worse financial system than the one with which we began. The survivors are an oligopoly of “too-big-and-interconnected-to-fail” financial behemoths. They are the winners not because they are necessarily the best businesses, but because they are the best supported. It takes no imagination to realise what these institutions might now do, given the incentives for risk-taking.

So what is to be done? The characteristic, but futile, response is to move the regulatory deckchairs on the deck of the Titanic. Recent proposals from the US Treasury fall partly into this category. But the financial system had to be rescued from its own mismanagement of risk. This is not going to be changed by external supervision. It is going to be changed only by fixing incentives.

The starting point has to be with “too big to fail”. We need a credible system for winding up even huge financial institutions. The most attractive proposals are for “good banks”, in which unsecured creditors become shareholders. That would be easier if, as President Barack Obama has proposed, and Mervyn King, governor of the Bank of England, has argued, a regulated institution has to produce a plan for an orderly wind-down of its activities.

Yet bank failures are like buses: you do not see one for hours and then a fleet arrives together. The authorities cannot make a credible promise that they would be prepared to put all affected institutions through bankruptcy in a systemic crisis. This would be a recipe for still-greater panics. “Too big and interconnected to fail” is a reality. It is so, because, as Andrew Haldane of the Bank of England pointed out in a recent speech, the financial system is an increasingly tight network.*

My colleague John Kay has argued that the right response is to create “narrow banks”, which are perfectly safe, leaving the rest of the financial system to go on its merry way, subject to a then-plausible threat of bankruptcy. I find this idea both attractive and unpersuasive. The attraction seems evident. It is unpersuasive in part because it is so hard to agree on what narrow banks should do. It is also unpersuasive because the narrower the banks are made to be, the more vital is the role of the rest of the financial system and so the less plausible it is that governments would let it collapse.

If institutions are too big and interconnected to fail, and no neat structural solution can be identified, alternatives must be found: much higher capital requirements and greater attention to liquidity are the obvious ones. At present, big financial institutions operate with next to no capital: in the US, the median leverage ratio of commercial banks was 35 to 1 in 2007; in Europe, it was 45 to 1 (see chart). As I noted last week, this makes it rational for shareholders to “go for broke”, with the results we have seen. Allowing institutions to be operated in the interests of shareholders, who supply just 3 per cent of their loanable funds, is insane. Trying to align the interests of management with those of shareholders is then even crazier. With their current capital structure, big financial institutions are a licence to gamble taxpayers’ money.


So how much capital makes sense for systemically significant institutions? “Much more than today” is the answer. Moreover, the required capital must also not be risk-weighted on the basis of banks’ models, which are not to be trusted. Shareholders’ funds should make up a minimum of 10 per cent of capital. In the US, it used to be far higher.

Higher capital is, in addition, a good way to internalise the negative “externalities” – more precisely, risks – created by one institution for the entire system. Ideally, therefore, the required capital should be correlated with the systemic significance of institutions, as the excellent new annual report from the Bank for International Settlements argues. Moreover, the requirement should be set against all activities, on the basis of fully consolidated accounts.

Within a far better capitalised financial system, it would also be relatively easy to operate a “macroprudential” regime, with the required capital rising during booms and falling during busts. Again, the bigger the stake of shareholders, the less one would worry if the rewards of managers were aligned with them. Even so, regulators have to have some sort of control on the incentives of management, as long as taxpayers bear residual risk.

Two difficulties remain: the transition; and regulatory arbitrage.

On the former, a demand for much higher capital ratios today would imperil the recovery. The answer is a lengthy transition, perhaps of as much as a decade. On the latter, it is evident that the so-called “shadow banking” system cannot be allowed to operate outside capital constraints if entities within it are likely to be systemically significant, as proved to be the case for money market funds. Moreover, capital ratios would have to be imposed by all significant countries. But the US is powerful enough to force movement in that direction by insisting that any foreign bank operating within it must be appropriately capitalised.

In sum, deleveraging is the right starting point for a healthier financial system. This would work best if we also eliminated today’s huge fiscal incentives for borrowing.

It is cautious incrementalism, not radicalism, that is now the risky option. Where should such radicalism start? The answer is clear: it is the incentives, stupid.