© The Financial Times Ltd 2015 FT and 'Financial Times' are trademarks of The Financial Times Ltd.
April 27, 2010 8:54 pm
Americans are obsessed with the case launched by the Securities and Exchange Commission against Goldman Sachs. At last, many hope, wrong-doers will be punished. But this misses the point. The problem is more what is allowed than preventing what is not. This is not to deny that there was much fraudulent behaviour behind the financial crisis. As John Kenneth Galbraith wrote, the “bezzle – the stock of embezzlement” – always rises in good times, But the real catastrophe, as I argued last week, is the risk taken on by the gamblers working legally inside the machine.
The role of big institutions is obviously problematic: they are, at one and the same time, the house, the biggest players at the gambling tables, agents for the other players and, if all goes wrong, beneficiaries of limited liability and implicit and explicit government bail-outs. This is a guarantee of repeated catastrophe. Under the gold standard, the scale of bail-outs was constrained. In a fiat system, there is no such limit, until the value of money collapses.
So what is to be done? Let us start with the least one could do to make greater stability credible.
An obvious solution is to revert to a tightly regulated, oligopolistic banking system. This is the sort of system Canada has enjoyed. But it is stodgy. It is also inconsistent with globalisation. Access by residents to foreign finance and by domestic institutions to foreign risks makes such cartels inherently unstable.
The alternative, then, would be to seek to make the present relatively free-wheeling system safe. To do this, we would have to make institutions and connections among them more robust, improve the quality of information and motivate the players to be less careless. There would be seven main elements.
First, raise capital requirements. There are many imaginable states of the world in which institutions with current levels of leverage would be insolvent. Leverage ratios of 30 to one are crazy. Three to one looks far more sensible.
Second, institutions must also have substantial liabilities that can be converted into equity or treated just as if they were equity, in a bankruptcy procedure. This is a necessary condition for resolution of problem institutions. What happened after Lehman’s failure showed that normal bankruptcy procedures do not work for complex institutions.
Third, make capital requirements powerfully counter-cyclical.
Fourth, make sure that banks hold a large stock of assets that are easy to value by lenders of last resort.
Fifth, shift incentives within firms. The managers should receive bonuses in shares they cannot sell until years after they have left. If that reduces the supply of aggressive risk-takers, so be it. Procedures for claw-back must also apply to other employees. Otherwise, it is too easy to gain hugely from trading strategies with a high probability of blowing up.
Sixth, impose much higher capital and collateral requirements against trading in derivatives. All such activities should be moved on to exchanges. Yes, innovation would be slowed. When the costs of innovation are borne by others, that is good sense.
Seventh, radically improve the quality of information available. Particularly important is a change in payment of rating agencies. Since these provide a public good, they must be funded by a general levy.
Would this make the system better? Yes. But the financial system would remain a doomsday machine. There are three difficulties. First, there is no sound basis for deciding how much capital is enough. Second, as the Bank of England’s Andy Haldane notes, “tail risk within the financial system is not determined by God but by man”. It is profitable to take risks whose upside accrues to oneself and whose downside accrues to others. So the safer regulators try to make the system, the more risk it can take on. Finally, it is easy to create the desired risk via regulatory arbitrage. That is precisely what the “shadow banking system” did.
So what else might be tried? The answer is structural reform. Three proposals are on the table.
The first, from Paul Volcker, is banning of proprietary trading by insured institutions. If this could be done (which I doubt), it should be.
The second, from my colleague, John Kay, among others, is in favour of “narrow banking”, under which deposit-taking institutions would be safe but the rest of the system would be little regulated. I remain unpersuaded that government could ignore the way the credit-creation system as a whole works (or, rather, does not). Nobody planned to rescue money market funds. Yet, in the crisis, they were saved all the same.
The third proposal, put forward by Laurence Kotlikoff of Boston University in his thought-provoking new book, is “Limited Purpose Banking”.* I like this idea. In essence, it says that you cannot gamble with other people’s money, because, if you lose enough, the state will be forced to pay up. So, instead of having thinly capitalised entities taking risks on the lending side of the balance sheet while promising to redeem fixed obligations, financial institutions would become mutual funds. Risk would then be clearly and explicitly borne by households, who own all the equity, anyway. In this world, financial intermediaries would not pretend to be able to meet obligations that, in many states of the world, they simply cannot.
None of this deals fully with the huge issue of securing greater macroeconomic stability: but a less unstable financial system would surely help. None of this deals fully with the issue of international co-ordination: in an open world economy, without exchange controls, financial stability is hard indeed to secure in just one country. Finally, none of this deals with the obvious problem of transition from what we have to a sounder system.
Even so, this discussion makes two fundamental points. The first is that to make anything close to the present system less unsafe requires radical changes in the rules. Tighter supervision is not enough. Incentives must change fundamentally. The second is that a financial system in which intermediaries assume risks on their own books is inherently unstable. It is too likely that they will make the same mistakes together, thereby creating a panic and threatening the system, with devastating economic consequences. This is the Achilles heel of market economies. We have been warned. For battered high-income countries, it will be the flood next time.
* ‘Jimmy Stewart is Dead’, Wiley & Sons, 2010.
Read and post comments at Martin Wolf’s blog
Please don't cut articles from FT.com and redistribute by email or post to the web.