Regulatory reform of the world's banks is becoming a painful thing to watch. The main object, never that clear to start with, is disappearing under a mass of suffocating detail. Perhaps policymakers should turn for guidance to an unlikely source - health and safety.

That is not a joke, though it sounds like one. Health and safety has become a byword in some quarters for petty bureaucracy. But in the two centuries since the Industrial Revolution, it has developed the kind of intellectual clarity that bank regulators are still groping for.

Consider two of its central concepts, hazard and risk. Hazard represents the amount of damage if a chemical plant, say, blows up or spews out poison. Risk represents the odds of that actually happening. Hazard must be identified, risk measured.

If we want chemicals - or indeed, banks - risk cannot be done away with. So the concept has developed of "tolerable risk".

That represents the amount of risk that society is prepared to live with. That varies enormously - large in the case of driving along freeways, vanishingly small for nuclear reactors. But it can be defined, if in a rough and ready way. The trick is then to get the risk below that level.

With all that in mind, consider the issue of macro-prudential supervision for the banks. This is much touted at present as the answer to the failings of micro-regulation.

The concept is not new, having been devised by the Bank for International Settlements in the 1970s. Nor has it been neglected since, with central banks round the world turning out 200-page financial stability reports at regular intervals.

But in the run-up to the crisis, those reports were generally ignored. This was not because they were irrelevant: the Bank of England's reports furnished a hair-raising account of the risks building up in the credit system. But there was no real attempt to draw conclusions or set priorities. So the system managed to identify the hazards, but not to measure the risks.

Before taking that further, it is worth considering the place for macro-prudential supervision in the overall scheme, - specifically in relation to micro-regulation.

According to some regulators, the tide of micro proposals risks becoming self-defeating. That view is set out in a new paper* by Clive Briault who knows whereof he speaks, having been the UK's chief regulator of retail banking at the time of the Northern Rock collapse. He is now self-employed.

The micro approach, he points out, now calls for more of everything: more capital, more liquidity, curbs on bonuses, control of hedge funds, private equity and derivatives, more accountability of non-executive directors, changes in accounting standards and so forth. Implicitly, it seems, the level of tolerable risk has dropped to zero. So any measures are justified.

But the longer the list, the more chance of the whole enterprise failing. First, if we make it too tough to run a bank - or to build a nuclear power plant - we must settle for alternatives which are not necessarily safer. Second, the more proposals are on the table, the more scope there is for world governments to bog themselves down bickering over detail.

So we need to think realistically about what level of risk is tolerable and pare down the list accordingly. That also means setting priorities. Mr Briault's favourite is liquidity - which, he says, was neglected by regulators and academics for two decades before the crisis.

The less scope there is for micro- regulation, the more weight falls on the macro side. But there is considerable lack of clarity on what that means - or whose job it is.

Read the rest at FT!