Can anyone be expected to manage a modern bank?

The latest approach to the problem of regulating systemically important financial institutions (Sifis) can't be criticised for complexity: Bernie Sanders, a veteran US senator, has introduced a bill simply requiring that every designated Sifi (and any other institutions regarded as too big to fail) should be broken up by the US government, on the basis that no institution should be allowed to grow to the point where its collapse alone could threaten systemic stability.

While the bill's very unlikely to pass, it has a certain elegance from an engineering point of view: any Sifi is, by definition, a single point of failure for the financial system, and you shouldn't have single points of failure in critically important systems. But a related argument is that financial institutions of Sifi scale are not only too big to fail, they're simply too big to manage.

This is not an isolated view. HSBC's chief executive, Stuart Gulliver, complained earlier this year that it was "not reasonable" to hold him responsible for the actions of all 275,000 of the bank's employees (the context being the Swiss authorities' investigation of HSBC money-laundering). Bank of America faced accusations of being too big to manage effectively at its shareholders' meeting in May. And even in 2012, the St Louis Fed was already citing examples such as the London Whale to argue that the largest banking groups are simply unmanageable in their current form.

The St Louis Fed argued that the position was particularly bad at banks because boards of non-expert directors are out of their depths managing complex international financial institutions. But it's also possible - and plausible - that a large multinational bank is simply too complex for any person to manage successfully, just as the first generation of US military supersonic aircraft were so lethal to their pilots (26% of whom could be expected to die in accidents) because human reflexes and senses were no longer fast and accurate enough to control them directly; they were, in fact, borderline unflyable. Only the advent of fly-by-wire technology, in which a computer intermediated between the pilot and the airframe, reduced this unacceptable death rate. And it is not easy to see how a parallel technological leap would render large banks manageable again.

The imperatives of air combat meant that it was not feasible to simply revert to older, slower, simpler aircraft. But this option is available for the financial sector. It may even be advisable, given the growing evidence for diseconomies of scale at larger size, and the harm to economic stability produced by an overdeveloped financial sector.

And every new management failure in the sector will simply strengthen the argument for using force to create a smaller, simpler banking industry; one which might once again be manageable by human beings, as the current one apparently is not.