Bank of England Governor Mervyn King used a speech in Edinburgh Tuesday to suggest that it may be prudent to split up the activities of investment and commercial banks in order to prevent another financial crisis and the bailouts that followed.

King said that it may be necessary to consider splitting up 'banks and separate riskier activities from more stable businesses like taking deposits'.

The move echoes the thoughts of former Fed Chairman Paul Volker, who has been saying the very same thing in the US.

Critics point out, however, that, in the UK, the likes of Bradford & Bingley and Northern Rock (which both failed), didn't engage in so-called 'riskier activities', and, in the US, the 99 commercial banks that have gone to the wall did not employ investment bankers!

Here's an excerpt of the Governor's speech:

'The United Kingdom faces two fundamental long-run challenges. First, to rebalance the economy, with more resources allocated to business investment and net exports and fewer to consumption. That is consistent with the need - now widely accepted - to eliminate the large structural fiscal deficit and to raise the national saving rate. It is part of a need for a wider rebalancing of domestic demand in the world economy away from those countries that borrowed and ran current account deficits towards those that lent and ran surpluses.

Second, both the structure and regulation of banking in the UK need reform. Banks increased both the size and leverage of their balance sheets to levels that threatened stability of the system as a whole. They remain extraordinarily dependent on the public sector for support. That was necessary in the immediate crisis, but is not sustainable in the medium term.

These two challenges are interrelated. In creating the crisis, imbalances in the world economy led to unusually low real interest rates and large net capital flows from the emerging market economies to the developed world. That provided the fuel which an inadequately designed regulatory system ignited to produce the financial firestorm that engulfed us all. If our response to the crisis focuses only on the symptoms rather than the underlying causes of the crisis, then we shall bequeath to future generations a serious risk of another crisis even worse than the one we have experienced.

Tonight I want to focus on the second of those challenges - reform of the structure and regulation of the banking system. Why were banks willing to take risks that proved so damaging both to themselves and the rest of the economy? One of the key reasons - mentioned by market participants in conversations before the crisis hit - is that the incentives to manage risk and to increase leverage were distorted by the implicit support or guarantee provided by government to creditors of banks that were seen as 'too important to fail'. Such banks could raise funding more cheaply and expand faster than other institutions. They had less incentive than others to guard against tail risk. Banks and their creditors knew that if they were sufficiently important to the economy or the rest of the financial system, and things went wrong, the government would always stand behind them. And they were right.

The sheer scale of support to the banking sector is breathtaking. In the UK, in the form of direct or guaranteed loans and equity investment, it is not far short of a trillion (that is, one thousand billion) pounds, close to two-thirds of the annual output of the entire economy. To paraphrase a great wartime leader, never in the field of financial endeavour has so much money been owed by so few to so many. And, one might add, so far with little real reform.

It is hard to see how the existence of institutions that are 'too important to fail' is consistent with their being in the private sector. Encouraging banks to take risks that result in large dividend and remuneration payouts when things go well, and losses for taxpayers when they don’t, distorts the allocation of resources and management of risk.

That is what economists mean by 'moral hazard'. The massive support extended to the banking sector around the world, while necessary to avert economic disaster, has created possibly the biggest moral hazard in history. The 'too important to fail' problem is too important to ignore.

There are only two ways in which the problem can - in logic - be solved. One is to accept that some institutions are 'too important to fail' and try to ensure that the probability of those institutions failing, and hence of the need for taxpayer support, is extremely low. The other is to find a way that institutions can fail without imposing unacceptable costs on the rest of society. Any solution must fall into one of those two categories. What does this mean in practice ?

Consider the first approach. To reduce the likelihood of failure, regulators can impose capital requirements on a wide range of financial institutions related to the risks they are taking. This is the current approach underpinned by the Basel regime. In essence, it makes banks build a buffer against adverse events. It has attractions but also problems. First, capital requirements reduce, but not eliminate, the need for taxpayers to provide catastrophe insurance. Second, the 'riskiness' of a bank’s activities and the liquidity of its funding can change suddenly and radically as market expectations shift. This means that what appeared to be an adequate capital or liquidity cushion one day appears wholly inadequate the next.

Indeed, the Achilles heel of the Basel regime is the assumption that there is a constant capital ratio which delivers the desired degree of stability of the banking system. After the experience of the past two years, and a decade or more in which capital ratios fell and leverage ratios rose to historically unprecedented levels, banks need more capital to persuade investors to fund them. A larger buffer gives new creditors greater comfort that their claims will be met in future, without resort to the public purse. And rather than pay out dividends or generous remuneration, banks should use earnings to build larger capital buffers. But how much larger ? We simply don’t know. A higher ratio is safer than a lower one, but any fixed ratio is bound to be arbitrary.

One way of dealing with this problem is to require banks to take out insurance in the form of 'contingent capital', that is capital in a form that automatically converts to common equity upon the trigger of a threshold that kicks in before a bank becomes insolvent. If a bank’s regulatory core tier 1 capital - common equity - remained above the threshold, nothing would happen. But if the regulatory core tier 1 capital fell below the threshold then the contingent capital would automatically be converted to common equity. When bank failures impose costs on the rest of the economy, it is reasonable to insist that banks themselves purchase a sufficient degree of insurance in the form of a large capital cushion available automatically before insolvency.

Such contingent capital instruments are very different from instruments, such as subordinated debt, which banks have been permitted to count as capital under the Basel regime, but which do not provide a reliable capital buffer until too late - after the bank has failed. So at present the taxpayer is providing capital that is exposed to future losses rather than investors in debt-like capital instruments of a troubled bank. Whether investors would be willing to provide contingent capital on the required scale would depend upon the price, but at least the cost would fall where it belongs - on the banks.

So the first approach to the 'too important to fail' problem might be made to work with a requirement for contingent capital. It is worth a try. But it has three drawbacks. First, banks still have an incentive to take really big risks because the government would provide some back-stop catastrophe insurance. Second, experience has shown that it is difficult to assess risks of infrequent but high-impact events, and so it is dangerous to allow activities characterised by such risks to contaminate the essential - or utility - services that the banking sector provides to the wider economy. Both of these drawbacks mean that it is almost impossible to calculate how much contingent capital would be appropriate. And a third drawback is that the approach probably requires the extension of detailed regulation, and especially a special resolution regime, to all institutions deemed 'too important to fail' - or could become 'too important to fail' in the middle of a crisis - in order to prevent a bank failure leading to conventional insolvency procedures. It remains to be seen whether that scope would be easy to define beforehand. Failure to deal with the underlying problem draws regulators deeper into murky waters.

The second approach rejects the idea that some institutions should be allowed to become 'too important to fail'. Instead of asking who should perform what regulation, it asks why we regulate banks. It draws a clear distinction between different activities that banks undertake. The banking system provides two crucial services to the rest of the economy: providing companies and households a ready means by which they can make payments for goods and services and intermediating flows of savings to finance investment. Those are the utility aspects of banking where we all have a common interest in ensuring continuity of service. And for this reason they are quite different in nature from some of the riskier financial activities that banks undertake, such as proprietary trading.

In other industries we separate those functions that are utility in nature - and are regulated - from those that can safely be left to the discipline of the market. The second approach adapts those insights to the regulation of banking. At one end of the spectrum is the proposal for 'narrow banks', recently revived by John Kay, which would separate totally the provision of payments services from the creation of risky assets. In that way deposits are guaranteed. At the other is the proposal in the G30 report by Paul Volcker, former Chairman of the Federal Reserve, to separate proprietary trading from retail banking. The common element is the aim of restricting government guarantees to utility banking.

There are those who claim that such proposals are impractical. It is hard to see why. Existing prudential regulation makes distinctions between different types of banking activities when determining capital requirements. What does seem impractical, however, are the current arrangements. Anyone who proposed giving government guarantees to retail depositors and other creditors, and then suggested that such funding could be used to finance highly risky and speculative activities, would be thought rather unworldly. But that is where we now are.

It is important that banks in receipt of public support are not encouraged to try to earn their way out of that support by resuming the very activities that got them into trouble in the first place. The sheer creative imagination of the financial sector to think up new ways of taking risk will in the end, I believe, force us to confront the “too important to fail” question. The belief that appropriate regulation can ensure that speculative activities do not result in failures is a delusion.

Separation of activities does not resolve all misaligned incentives. Where private sector entities outside of the utility banking sector engage in a high degree of maturity transformation on a scale that could have consequences for the rest of the economy, the government would not want to stand aside when such an entity fails. That is the heart of the matter. Maturity transformation reduces the cost of finance to a wide range of risky activities, at least some of which are beneficial, but the implicit government guarantee means that the true cost of that maturity mismatch does not, as it should, fall on those who receive the benefits. The aim of policy should be to minimise or eliminate that subsidy. Separation of activities helps not hinders that objective, not least because it is the mixture of activities that reduces the robustness of the system. Although there are no simple answers, it is in our collective interest to reduce the dependence of so many households and businesses on so few institutions that engage in so many risky activities. The case for a serious review of how the banking industry is structured and regulated is strong.

By international standards UK banking is highly concentrated. There are four large UK banking groups. Of these four, two are largely in state ownership and their assets are a multiple of the assets of the next largest bank. As in the English Premier League, getting into the top four will not be easy for those outside it. But in both cases I hope greater competition will produce less rigidity in the composition of the top four.

Whichever approach to the 'too important to fail' problem is adopted, there is growing agreement that such financial institutions should, as I argued at the Mansion House in June, be made to plan for their own orderly wind down - to write their own will. I welcome that, but without separation of the utility from other components of banking it will be necessary to develop detailed resolution procedures for a very wide class of financial institutions. The options may turn out to be separation of activities, on the one hand, or ever increasingly detailed regulatory oversight, with the costs that that entails for innovation in, and the efficiency of, the financial system, on the other'.

Please use the 'E-Mail' button immediately under the article title to send this item to a friend.

Press here to comment on this story >>

Today in Life: I Swear To Tell The Truth... >>

TODAY

• And The Best Place To Find A New Job Is.......
• Top Firm Said Under Pressure To Break Itself Up
• Bank Of America Sued Over Theft Of Parrott

THIS WEEK

• And The Best Place To Work In The Global Financial Markets Is...
• The Business Trip From Hell - Parts I & II
• And The Best Place To Find A New Job Is.......