A new global governance was forged in the white heat of the financial crisis. The G7 gave way to the G20 (Eichengreen 2009). Leaders representing 80% of the world’s population met and were resolute in calling for a global policy response to the crisis. Governments opened the fiscal sluice gates, interest rates were slashed, the IMF was given additional resources, and the OECD finally got tough with European tax havens.
But as the ashes cool, senior officials are quoted one day in the international press saying that a global approach to banking regulation is necessary, and the next, saying “nobody is going to tell us how to regulate our banks.” Beneath the veil of global unity, different regional perspectives and national action are emerging. This is most clear in the area of international bank regulation.
Regional perspective: National action
In Europe, the emphasis has been on how to regulate financial markets to moderate future crises. Credit mistakes are made during the boom, not during the crash. Better regulation during the boom years (and perhaps better monetary policy too) could limit the amplitude of the bust. Moreover, in a crash, markets are not very discerning, and it is hard to strike the balance of who to save and who to leave to the wolves. Consequently, “Europeans” are focused on how to limit the crashes by introducing macro-prudential regulation such as counter-cyclical charges and minimum liquidity buffers.
In the US, perhaps reflecting a greater belief in markets and a stronger mistrust of regulation, the emphasis has been on finding market-friendly ways to contain the spillover of bank failure. US policy debates are occupied with concerns that banks should not be “too big to fail”; that private investors, not taxpayers, should hold “contingent capital” which carry implicit or explicit conversion into equity in a crash, and that improvements must be made to the functioning of “over-the-counter” markets through greater use of centralised trading, clearing, and settlement. These proposals are less about modifying capital requirements and more about prohibition and taxation and are micro-prudential in nature. Banks would not be allowed to do “risky” things (the “Volcker Plan”) and large balance sheets will be taxed to repay the bailout funds (“the Obama Levy”).
In the former crisis countries of Asia, where banking regulation followed new international banking rules more closely, the view is more sanguine. Officials from Indonesia, Thailand, Malaysia, Philippines, and Australia argue that because Asia largely escaped from the financial part of the crisis, it must actually be about poor supervision by their European and American peers. They argue that we should tinker less with regulation and instead deepen supervisory capacity.
The principal point of convergence between these perspectives is that the crisis was caused by “the others”, and “they” had better get their house in order. But regional perspectives and national action should be seen as inevitable as long as national taxpayers provide the bailouts. An EU bailout of Greece will further integration, not kill it.
Elsewhere, more local regulation may not be such a poor second best. The record of policymakers in the international regulation of banks is hardly a source for optimism. One global mistake is far worse than a few smaller ones, and small developing countries in particular need to be wary of international banking rules forged in developed countries with a view to expanding developed country banks into the emerging world.
Tackling “too big to fail” is less critical than it seems in the US
One other point of intersection between these regional perspectives concerns big banks. This has less to do with a convergence of regulatory thinking on bank size or function and more to do with the common need to play to the political gallery and raise taxes somewhere. Robin Hood, previously banished from the Kingdom by Prime Minister Thatcher and President Reagan, has returned. This desire to raise taxes from bankers sits well with the idea that policy should offset the incentive for banks to become systemically more important. I have long argued this point (Persaud 2008), but I have also come to realise that the notion that we should concentrate our efforts on making sure banks are not “too big to fail” is partly based on an illusion.
Bank balance sheets bloated by leverage are systemically dangerous, and regulatory or fiscal policy should address this through liquidity buffers and leverage ratios. But given how contagious crises are, it is likely that what is “too big to fail” is actually small. Any list of institutions that were “too big to fail” conjured up in 2006 would not have included Northern Rock, Bradford & Bingley, IKB, Bear Sterns or even Lehman Brothers. Banks lend to banks. While some are more illiquid than others, they are intrinsically illiquid institutions. It does not take a large failure to lead to panic. High-yielding deposits can fly out of the website almost as quickly as money market funds can withdraw. In a crisis almost everyone is “too big to fail”.
Moreover, we can have as large a boom and subsequent crash, with the same economic misery, in a world of only small banks. Some will recall the 1973-1975 Secondary Banking Crisis in the UK, in which 30 relatively small financial institutions had to be supported by the Bank of England following the preceding property boom. The 1973-1975 crisis rivals, and on some measures exceeds, the impact on the UK of the current financial crisis. It was one of the reasons that developed country regulators began to take a more benign view of large banks snapping up smaller ones. The fashionable argument of the day was that small, competitive, financial institutions were inefficient and had little “franchise value” and so they would under invest in their own longevity – by having less conservative lending practices – than larger, more profitable banks.
Crashes follow booms. Booms are fuelled by some new dawn – normally the arrival of new technology – that makes bankers feel that the world is a brighter place, risks have fallen, and that they are therefore justified in lending and leveraging more. This behaviour is even more acute in the modern age of “risk-sensitive” regulation than in the old-fashioned world of credit and concentration limits and lending rules of thumb. The systemic effect of having one large bank engaged in rapid lending growth is no different than having several small banks do so. It may even be easier to resolve a crisis with one large bank.
It is argued that the underestimate of risk by bankers is encouraged by the further belief that if it all goes “belly up” their institutions are too big to fail and their jobs would be safe. But if that latter belief dominated bankers’ thinking, they may not worry about their job but they would still worry about their savings. They would not wrap themselves up in the equity of their institutions and the leveraged products they are selling. But they did. The revealed preference of bank and banker behaviour is that they did not lend more because they thought they could get away from selling risk to others, or because they knew there would have to be some sort of bail out, but because they thought it was safe. They were more fools than knaves.
The hollowness of the risk-sensitive approach to regulation and an alternative
The main driver of excessive lending and leverage is a mistaken view of risks, by everyone and not just big banks. The riskiest institutions were not the largest. Some very large institutions, such as J. P. Morgan and HSBC, proved safer than others and did not seek or require state funding, while those that failed were relatively small: IKB, Bear Sterns, etc. There is an all too narrow dividing line between this argument and the one that got us into this mess. Big banks like the idea that regulation should care less about size of banks and more about their riskiness, and so they championed the “risk-sensitive” approach to bank regulation, not least because they had the bigger risk models and databases, so risk-sensitive regulation was more onerous for their smaller competitors.
A better argument for curbing bank size is the excessive influence of big banks on policy. The problem is that if booms are fuelled by an underestimation of risks, and regulation is made more sensitive to the estimation of risks, then the booms will be bigger and the busts deeper. This is the fallacy of the “risk-sensitive” approach to risk management, regulation, and auditing. This is a fairly fundamental point, but I have discussed it before and would like to highlight here how the solution to this issue relates to current proposals.
A financial system less sensitive to the error in the markets’ estimate of risk
What we are looking for is regulation that makes the financial system less sensitive to the error in the markets’ estimate of risk, not more so. There are two ways to do this.
- The first is to observe that this error is strongly correlated to the boom-bust cycle.
Booms have similar characteristics – strong growth in bank balance sheets and credit, a rise in leverage etc. The appearance of these would imply an increased probability of the market underestimating risk and so regulators should raise minimum capital requirements when they spot these trends. Counter-cyclical capital requirements fit in with this idea. A range of indicators would have to be used to calibrate the rise in the capital requirements and perhaps some discretion, but not too much. Market-efficiency zealots will complain that the market cannot make predictable mistakes, but there are many reasons why the market fails to correct its systemic error, the least rigorous being that booms are always founded on a good story of why “this time it is different”, one that takes in regulators and bankers. Recall the essays in central bank stability reports on how credit derivatives were bringing new stability to the financial sector?
- The second way to reduce the sensitivity of the financial system to the errors of estimating risk is to limit the flow of risks to institutions with a structural capacity for holding that risk, and not a statistical capacity.
That way when the risk modellers get it wrong we are in less trouble. Credit risk is best hedged through diversification across uncorrelated credits. Liquidity risk is best hedged through diversification across time. Market risk is best hedged through a combination of diversification across assets and time (having time to decide when to sell). In the past, risks of similar statistical magnitudes were considered fungible and simply flowed to whoever was prepared to pay for it. But while banks with short-term funding and many branches originating different loans have a deep capacity for holding credit risks, they have a limited capacity for holding market risks and little capacity for holding liquidity risk. Insurance companies and pension funds on the other hand have limited capacity for credit risk, but more for market and liquidity risks.
The capacity for risk is related to the maturity of funding and not what an institution is called. This idea resonates a little with Paul Volcker’s idea of banks severing their exposure to proprietary trading (market risks), hedge funds, (market risks) and private equity (liquidity risks). But we must be wary of arbitrary distinctions – some hedge funds behave like banks and some banks like insurance companies.
Although the response to the crisis is becoming more national and less global, current proposals to reform banking regulation have some promise. This is especially the case in the area of counter-cyclical measures being discussed by the Basel Committee and the Volker proposal to fragment financial institutions, though fragmentation should be along the lines of risk-capacity and not legal entities or names. We should also be careful, however tempting it may be, to expect too much from bashing big banks.
Eichengreen, Barry (2009), “The G20 and the crisis”, VoxEU.org, 2 March.
Persaud, Avinash (2008), “Why bank risk models failed”, VoxEU.org, 4 April.