John Kay recently had a must-read piece Financial Times, now available ungated on his website. Kay makes an interesting point: the obsession with “too big to fail” financial institutions and “the lobbying power of incumbent companies” is leading the revamping of financial regulation after the crisis down a dead end alley. Not that new rules are not generously produced, but they would not provide for that “financial stability” that anybody is longing for. “Perhaps the most fundamental confusion in the evolution of financial services regulation,” Kay argues, “is the equation of financial stability with the survival of established institutions.”

He writes:

Lehman was run (badly) for the benefit of its senior employees rather than customers or shareholders. In a market economy, such organisations fail while rivals with better business models and management structures gain at their expense. That process of selection is the reason market economies have an impressive record of promoting efficiency and innovation. The problem revealed by the 2007-08 crisis was not that some financial services companies collapsed, but that there was no means of handling their failure without endangering the entire global financial system.

Still, would it not be better if proper supervision ensured that no financial institution could ever get into a mess like Northern Rock or Lehman – or Royal Bank of Scotland or Citigroup or AIG? No, it would not. Just replace “financial institution” with “fast-food outlet” or “supermarket” or “carmaker” in that sentence to see how peculiar is the suggestion.

Begin with practicality. It is hard enough to find people capable of running financial conglomerates – the fading reputation of Jamie Dimon, JPMorgan Chase chief executive, confirms my suspicion that managing these businesses is beyond the capacity of anyone. The search for a cadre of people employed on public-sector salaries to second guess executive decisions is a dream that could not survive even the briefest acquaintance with those who actually perform day-to-day supervisory tasks in regulatory agencies. They tick boxes because that is what they can do, and regulatory structures that are likely to be successful are structures that can be implemented by box tickers.

We have experience of structures in which management or regulatory committees in Moscow or Washington take the place of the market in determining the criteria by which a well-run organisation should be judged, and that experience is not encouraging. The truth is that in a constantly changing environment nobody really knows how organisations should best be run, and it is through trial and error that we find out.

In short, Kay maintains that “financial stability is best promoted by designing a system that is robust and resilient in the face of failure.” But what should then be the features of a “robust” financial system? Larry White has an insightful article in the last issue of the Cato Journal. He borrows the word “antifragility” from Nassim Taleb’s last book. White explains that “the wrong tactic for enhancing antifragility is to have a central authority impose uniform rules” and summarizes antifragility as, “Let a thousand flowers bloom, but do not artificially preserve even one of them.” “Banking is ecologically rational only when a standard heuristic of the rule of law is observed: the shareholders, creditors, and management who stand to absorb the upside gain must also absorb all the downside loss”. This is what Kay expresses with the terse formula “treat finance and fast food alike.”

An often-heard criticism of this viewpoint is that it isn’t quite clear how to move from A to B, from today’s crony financial capitalism, to something that may have a closer resemblance to a free market. And yet, I think Kay is quite right in maintaining that the very idea that we should aim to restore accountability, free enterprise, and the possibility of failure for big banks (aka the rule of law) is shared neither by regulators nor by public opinion. That’s a bigger problem than planning the transition.