VoxEU Column Financial Markets

Basel 3: Supervising the credit giants with progressive capital ratios

Government guarantees for financial institutions that are "too big to fail" seem unavoidable, but such insurance against insolvency imposes major costs and may increase the risk of crisis. This column argues for the necessity of non-conventional regulatory solutions to make big banks pay for their implicit insurance. Progressive capital ratios may be one such solution.

Some are already beginning to talk about Basel 3. So we should not forget an important lesson from these past months of financial crisis – large banks will never be allowed to fail, even if the cost to taxpayers and possible future market distortions are extensive. Capital coefficients that consider this risk in order to "insure against insolvency” and that increase as banks' sizes expand are needed. In other words, I propose considering a progressive capital ratio system.

Like a tsunami, the financial crisis is changing the landscape of the banking system. A distinctive aspect is that increasingly larger financial conglomerates seem to emerge as a necessary effect of states that have avoided systemic crises. The phenomenon is particularly apparent in the US, where the financial system is taking on distinctive features.

Before the crisis, the US financial industry was distinguished by two characteristics: a stronger development of investment company divisions and non-banking finance in general compared to commercial bank divisions, with both divisions closely and complexly entwined. In this financial system, a crisis in one, specific, limited sector – subprime mortgages – rapidly contaminated all other industries. Among the effects was a series of intermediary bankruptcies, with an increasingly widespread and massive government intervention occurring simultaneously – and not just in the United States – linked by at least one common element: to save the large intermediaries that were at immediate or potential risk of bankruptcy. At the same time, large non-banking intermediaries were allowed to begin to collect retail savings. The prospect therefore seems to be that of a market dominated by large conglomerates, confirming the character of national strategy, and hence the relative government protection.

Thus, in 2008, a sort of unwritten law returned, which is both explicit and shared, and which has historically characterised public intervention in the banking sector: large banks never fail. It is true that, in the past few months, we have seen and will continue to see various versions of this general rule, which could be varied in terms of the characteristics of the saved intermediary, the conditions of the government performing the bailout, and the urgent or potential nature of the bailout. But one common denominator remains: if the failure of a single bank makes a systemic crisis probable, this bank must be bailed out by the government. This probability often depends on the size of the bank itself, which is why the “too big to fail” rule was created. Protecting a public aim – stability – could therefore justify bailing out a large private enterprise. Unfortunately, the "too big to fail" rule risks being a worse remedy than the original ill.

The price of "too big to fail"

A similar rule, if supported over time, could stimulate managers to increase the size of banks, considering nothing else, because this creates a kind of insurance against insolvency. Since an optimal size of banking activity does not exist, trying to expand activity volumes at all costs implicates efficiency risks. But there are also risks in terms of stability, for at least two reasons. In the first place, managers who know they are insured against insolvency may increase their tendency to take risks, thus increasing the probability that insolvency will actually occur. In the second place, the risk of capture increases: a financial system in which very few large intermediaries dominate the scene could create chain reaction mechanisms between banks, politicians, and oversight authorities. The potential toxicity of insurance against insolvency favouring large banks is further demonstrated by the speed and the strength with which large enterprises in other industries – read: automobiles – are asking for the same treatment.

The too big to fail rule therefore represents an authentic paradox – in the presence of risk of systemic crisis, the rule must be temporarily adopted to prevent the crisis from actually worsening, but if it is not credibly revoked, it pushes the system towards an unstable and inefficient oligopoly.

But can the too big to fail rule be revoked after the financial crisis ends? It may be difficult since the law has never been written, but it is always imminent in oversight procedure. But if government insurance against the insolvency of large banks is difficult to abolish, the government could try making them pay. Defining the correct plan for such a distinctive kind of insurance would not be simple, but the current situation could certainly be improved upon. Now, large banks could reasonably believe that they are covered by a subsequent implicit insurance in case of an insolvency risk. Banks do not pay a corresponding initial premium, however, as the prudent regulation plan does not mention bank size, at least from this point of view.

Today, with Basel 2, the distinctiveness of large banks generally emerges only if they want it to – in the choice of models to measure “customised” risk – and certainly not in reference to insurance against insolvency.

But if during the crisis large banks were verified as having more of a chance of being subject to a public bailout after the fact, then there should be more checks after the fact as well. An oversight discipline that increases as bank sizes expand could be designed in many ways, more or less in line with the principles of a market economy.

On one hand, there are rather invasive procedures such the possibility of consistently present public oversight inside the largest banking organisations, in the area of governing boards, among others. It would not be new as it has happened in Italy and Spain. But it would be a clear jump backwards.

New ideas are needed: capital coefficients that, equal to other risks, consider the “irresponsibility risks” that could increase bank sizes out of proportions, for example. Thus, non-proportional capital coefficients which are progressive in relation to the activity volume are needed, possibly divided into size groups. In order to “put a price” on government insurance in favour of credit giants, non-conventional regulatory solutions need to be explored, striving to remain in line with the foundation of a market with prudent regulation. It is not simple, but it would be better than pretending that the problem did not exist.