Thursday, June 25, 2009

Reform of regulation has to start by altering incentives

Posted in the Financial Times by Martin Wolf:

Proposals for reform of financial regulation are now everywhere. The most significant have come from the US, where President Barack Obama’s administration last week put forward a comprehensive, albeit timid, set of ideas. But will such proposals make the system less crisis-prone? My answer is, no. The reason for my pessimism is that the crisis has exacerbated the sector’s weaknesses. It is unlikely that envisaged reforms will offset this danger.

At the heart of the financial industry are highly leveraged businesses. Their central activity is creating and trading assets of uncertain value, while their liabilities are, as we have been reminded, guaranteed by the state. This is a licence to gamble with taxpayers’ money. The mystery is that crises erupt so rarely.

The place to start is with the core of modern capitalism: the limited liability, joint-stock company. Big commercial banks were among the most important products of the limited liability revolution. But banks are special sorts of businesses: for them, debt is more than a means of doing business; it is their business. Thus, limited liability is likely to have an exceptionally big impact on their behaviour.

Lucian Bebchuk and Holger Spamann of the Harvard Law School make the big point in an excellent recent paper.* Its focus is on the incentives affecting management. These are hugely important. Still more important, however, is why a limited liability bank, run in the interests of shareholders, is so risky.

In a highly leveraged limited liability business, shareholders will rationally take excessive risks, since they enjoy all the upside but their downside is capped: they cannot lose more than their equity stake, however much the bank loses. In contemporary banks, leverage of 30 to one is normal. Higher leverage is not rare. As the authors argue, “leveraged bank shareholders have an incentive to increase the volatility of bank assets”.

Think of two business models with the same expected returns: in one these returns are sure and steady; in the other the outcome consists of lengthy periods of high returns and the occasional catastrophic loss. Rational shareholders will prefer the latter. This is what one sees: high equity returns, by the standards of other established businesses, and occasional wipe-outs.

Profs Bebchuk and Spamann add that four features of the modern financial system make the situation even worse: first, the capital of banks is itself partly funded by debt; second, the role of bank holding companies may further increase the incentives of shareholders to underplay risk; third, managers are rewarded for aligning their interests with those of shareholders; and, fourth, some of the ways managers are rewarded – options, for example – are themselves a geared play on rewards to shareholders. So managers have an even bigger economic interest in “going for broke” or “betting the bank” than shareholders. As the paper notes, the fact that some managers lost a great deal of money does not demonstrate they were foolish to make these bets, since their upside was so huge.

A solution seems evident: let creditors lose. Rational creditors would then charge a premium for lending to higher-risk operations, leading to lower levels of leverage. One objection is that creditors may be ill-informed about the risks being run by banks they are lending to. But there is a more forceful objection: many creditors are protected by insurance backed by governments. Such insurance is motivated by the importance of financial institutions as sources of credit, on the asset side, and suppliers of money, on the liability side. As a result, creditors have little interest in the quality of a bank’s assets or in its strategy. They appear to have lent to a bank. In reality, they have lent to the state.

The big lesson of the current crisis is just how far such insurance may go in the case of institutions deemed too big or interconnected to fail. Big banks rarely get into trouble in isolation: they often make very similar errors; moreover, the failure of one impairs the actual (or perceived) solvency of others. Thus, creditors are most at risk in a systemic crisis. But a systemic crisis is precisely when governments feel compelled to come to the rescue, as they did at the end of last year.

According to the International Monetary Fund’s latest Global Financial Stability Report, support offered by the US, UK and eurozone central banks and governments has amounted to $9,000bn (€6,400bn, £5,500bn), of which $4,500bn are guarantees. The balance sheet of the state was put behind the banks. This does not mean creditors bear no risk at all. But their risk is attenuated.

The well-known solution is to regulate such insured institutions very tightly. But an enormous part of what banks did in the early part of this decade – the off-balance-sheet vehicles, the derivatives and the “shadow banking system” itself – was to find a way round regulation. The obvious question is whether it will be “different this time”. Sensible people must doubt it. Indeed, it must be particularly unlikely when the capitalisation of banks is so small. This is the time to go for broke.

In a speech delivered just last week, Mervyn King, governor of the Bank of England, made clear why finding a better approach matters so much: “The costs of this crisis are not to be measured simply in terms of its impact on public finances, the destruction of wealth and the number of jobs lost. They are also to be seen in the lost trust in the financial sector among other parts of our economy ... ‘My word is my bond’ are old words. ‘My word is my CDO-squared’ will never catch on.”

Such a crisis is not only the result of a rational response to incentives. Folly and ignorance play a part. Nor do I believe that bubbles and crises can be eliminated from capitalism. Yet it is hard to believe that the risks being run by huge institutions had nothing to do with incentives. The unpleasant truth is that, today, the incentive to behave in this risky way is, if anything, even bigger than it was before the crisis.

Regulatory reform cannot end with incentives. But it has to start from incentives. A business that is too big to fail cannot be run in the interests of shareholders, since it is no longer part of the market. Either it must be possible to close it down or it has to be run in a different way. It is as simple – and brutal – as that.

Public support of banks

* Regulating Bankers’ Pay, Harvard Law and Economics Discussion Paper No. 641, May 2009

No comments: