The Price of Crisis Prevention

This estimate may look astonishingly high, but it is relatively conservative, as it assumes that only one-quarter of the initial shock will result in permanently lower output. According to the true doomsayers, who believe that most, if not all, of the shock will have a permanent impact on output, the total loss could be two or three times higher.

One year of world GDP amounts to $60 trillion, which corresponds to about five centuries of official development assistance or, to be even more concrete, 10 billion classrooms in African villages. Of course, this is no direct cost to public budgets (the total cost of bank rescue packages is much lower), but this lost output is the cost that matters most when considering how to reduce the frequency of crises.

Assume that, absent adequate preventive measures, a crisis costing one year of world GDP occurs every 50 years (a rough but not unreasonable assumption). It would then be rational for the world’s citizens to pay an insurance premium, provided its cost remains below 2% of GDP (100%/50).

A simple way to reduce the frequency of crises is to require banks to rely more on equity and less on debt so that they can incur more losses without going bankrupt – a measure that is currently being considered at the global level. Thanks to reports just released by the Financial Stability Board and the Basel Committee – one on the long-term implications of requiring higher capital-to-asset ratios, and one on the transitory effects of introducing them – we know more now about the likely impact of such regulation.

The first report finds that, starting from the current low level of bank capitalization, a one-percentage-point increase in capital ratios would permanently reduce the frequency of crises by one-third, while increasing interest rates by some 13 basis points (banks would need to charge more because it costs them more to raise capital than to issue debt). In other words, the price of losing one year of income every 75 years instead of every 50 years would lead banks to increase the rate on a loan from 4% to 4.13%. Such an insignificant increase would at most lead a few bank customers to turn to alternative sources of finance, most likely with no discernible effect on GDP.

It is stunning to find that a regulation can do so much good at such a small cost – much smaller than in many other fields where public policy imposes economically costly safety requirements. Think, for example, of the environment or public health, where the precautionary principle guides decision-making.

So much for the long term. The hotly debated question nowadays is whether the transition to higher mandatory reserves would involve excessive short-term costs (as banks will likely increase lending spreads and reduce credit volumes). Subjecting banks, some of which are still ailing, to new regulation may lead them to curtail lending even more, thereby further weakening the pace of economic recovery. Sound judgment is needed regarding the speed and timing of regulatory tightening.

The second report finds that a one-percentage-point increase in bank capital ratios, if introduced gradually over four years, would lower GDP by about 0.2%. Given that increases by three percentage points are frequently mentioned, the total effect could be 0.6%.

But uncertainties abound. The report oddly finds that raising the target capital ratio would have a significantly greater adverse effect in the United States than in the eurozone, despite the latter’s more pronounced reliance on bank-based financing. Moreover, the report assumes that monetary policy can offset part of the shock, which may not be true where near-zero interest rates already prevail – or in the euro area where the effort may vary across countries while monetary policy is uniform. So the impact of new regulations on countries where banks are significantly under-capitalized could easily be four or five times larger than the headline figure – say, in the vicinity of a percentage point at a four-year horizon.

This may still look small, but it is hardly a trivial sum in view of the short-term growth outlook in the developed world. At a time when growth is too slow to reduce massive unemployment, every decimal matters.

To lower growth by this magnitude at a time when the private sector has not yet completed its deleveraging cycle – and governments are starting their own – is to risk a prolonged period of near-stagnation, which could turn crisis-induced unemployment into structural unemployment. Furthermore, tighter credit standards over a prolonged period are likely to fall disproportionately on cash-poor, fast-growing companies, with consequences for innovation, productivity, and ultimately growth potential.

None of this means that banks should be granted a regulatory holiday and forget the need to recapitalize. But it does suggest, first, that timing matters. Policymakers should be wary of imposing a regulatory shock simultaneously with a fiscal shock. For this reason, enacting new regulatory standards now while setting distant deadlines is a sensible strategy.

Second, the existence and magnitude of transition costs means that not everything that reduces the probability of financial crises is worth undertaking. For politicians obsessed by current difficulties, whether the next crisis occurs in 50 or 75 years may seem secondary. As a result, regulatory reform must be designed in a way that minimizes short-term costs.

Higher capital ratios and liquidity ratios are only one of several approaches to making the financial system safer. Other measures – say, capital insurance, or reform of the boundaries within the financial industry, À la Paul Volcker – are worth considering.

There is no doubt that the long-term price of insuring against crises is worth paying. But that doesn’t mean that reform should not be as cost-effective as possible.

Jean Pisani-Ferry is Director of Bruegel, the European think tank.