Fannie Mae and Freddie Mac were probably the world’s most heavily supervised financial institutions, subject to a specialist agency, the Office of Federal Housing Enterprise Oversight.  The Office employed 236 people at the time of its last annual report.

OFHEO did not fail because it was understaffed, or because it was not well informed about Fannie Mae’s activities, but because it lacked authority.  The entire staff earned less in aggregate than the remuneration of Franklin Raines, the aggressive CEO who masterminded Fannie’s expansion.  And Mr Raines was good at some parts of his job:  an experienced Washington insider, he knew how to see of political interference in his business.

Like Martin Wolf, I yearn for a world in which regulators would moderate the inherent instability of the modern financial system.  But my yearning is tempered by modest expectations of what regulation can achieve.  Martin Wolf’s realism, which I share, acknowledges that public expectations are much higher and politicians will claim to respond to these expectations.  But the politicians will fail.   The next financial crisis will be different in origin and the rules which will be introduced to close the doors of today’s empty stables will prove irrelevant.

It is easy to assert that the solution to any market failure is better regulation.  If regulators were all knowing and all powerful;  if they were wiser than the chief executives of great financial institutions, but nevertheless willing to do the job for a small fraction of the remuneration awarded to such executives;  if they understood what was happening in the dealing rooms of Citigroup, Merrill or Lehman better than Chuck Prince, Stan O’Neal, or Dick Fuld, then banking regulation could protect us against financial instability. 

But such a world does not exist.  The reason market economies outperformed planned economies is not that businessmen are smarter than civil servants – sometimes they are, sometimes they are not.  But no-one has enough information or foresight to understand the ever changing environment, and so the market’s messy processes of experiment and correction yields better results than a regulator’s analysis.

In an imperfect world, the simple rules that Martin seeks have unanticipated and counterproductive consequences.  As with the reserve requirements imposed under the Basle agreements.  Good banking practice was translated into a regulatory obligation.  Reserve ratios were transformed from an internal discipline of prudent management to an external burden to be evaded when possible.    Because these rules distinguished different asset categories, they opened the doors to regulatory arbitrage, fuelling the explosion of securitisation which is the root of current problems. Capital requirements proved ineffective in preventing banking failures.  Northern Rock was setting out its plans to distribute surplus capital only weeks before it collapsed through shortage of liquidity.  And as soon as crisis struck, capital requirements proved counterproductive, forcing banks to constrain good lending to meet regulatory obligations.  The proposed solution – of course – is further refinement of the regulations – to legislate against SIVs, to supervise the categorisation imposed by rating agencies, and to introduce counter-cyclical reserve requirements.

In our debate, Martin used a forceful metaphor to describe the impact of the development of financial conglomerates –a utility is attached to a casino.  The utility is the payments system which enables individuals and non-financial companies to go about their everyday business confident that they can make and receive payments, and lend and borrow to finance normal transactions.  That activity needs to be protected from the consequences of the booms and busts which are an inevitable concomitant of securities trading in volatile markets.

There are two routes to this result.  One is to separate the utility from the casino.  Narrow banking prevents conglomerate institutions from relying on the assets of their unsophisticated customers as collateral for their highly sophisticated trading.  Another approach regulates the casino sufficiently heavily to ensure that failure there cannot jeopardise the utility.  This latter outcome is not feasible and to come close to achieving it would end financial innovation.

The financial services industry will successfully resist both the ring-fencing of everyday banking and the meaningful regulation of trading operations. Martin and I both recognise that in the next crisis, as in this, the taxpayer will step in to fund the casino in order to protect the utility.