Nobody thinks that utility-operating companies – whether in transport, such as railways, in energy, such as electricity, or telephone or water – are too big to fail. If they lose enough money and go bust, then, if another company cannot be found to take over the franchise, the government steps in to take over the operations. They keep the capital and (most of) the workers to continue running the utility. No one would think that it would make any sense to rip up the railway lines, electricity pylons, or water pipes, to sell them for scrap, and to push the skilled workers into unemployment. Nor is there any worry about utilities being too-big-to-fail.
Why is banking different?
The key reason is that its capital is fungible and can be quickly redeployed, unlike the water pipes, power stations, etc., of the utilities. The real danger with banking is the bankers’ ability to “gamble for resurrection”. With limited downsides from failure but enjoying the spoils from success, it is in the interest of both bankers and their shareholders to take on more risk than is socially desirable, especially when their stake in the business has already been eroded by losses. By the time that the bank has clearly become technically insolvent, it may well have run up far greater losses than any ordinary utility could aspire to emulate.
One answer, of course, is regulation. But regulation, whether by risk-weighted ratio control or by structural ring-fencing, has severe limitations; Anglo-Irish and Northern Rock were both, in effect, ring-fenced retail banks. If regulation is taken sufficiently far to ensure safety, creating narrow banks, it not only destroys the key utility function of such banks, i.e. credit allocation, but also leads to a major shift of business to non-regulated intermediaries, which can worsen both the fragility and the pro-cyclicality.
Prompt corrective action is needed
What needs to be done is bring intervention by the authorities forward in time: prompt corrective action, well before the bank’s managers can really drive it into the ground. A problem is that the available signals for doing so are quite faulty. The accounting value of equity capital is only available after a lag that can be far too long for comfort in fast-moving markets, and can be subject to all kinds of accounting tricks. On the other hand, the market value of equity can be subject to (temporary) manipulation, or to market over-reactions or flash crashes.
An answer to this is a requirement for both of these signals to be flashing red before intervention is contemplated, and having the accounting signal at a much higher value, say 5%, than the market signal, say 2.5%. If that is still thought too draconian, one could make such signals the trigger for a “comply-or-explain” exercise, whereby the authorities either comply by intervening, or explain in public why there is no need to do so.
But is this not a transgression of the rights of private property?
Maybe so, but is not all regulation (and taxation) also such a transgression? If banking is special, it is because it is a utility with special risks, and special risks justify special responses. Moreover, the shareholder could be protected in a variety of ways, notably by having a claim to the value of any (subsequent) sale of the business (beyond the appropriate cost to the government of the intervention, and residual to other senior creditors).
In general, no systemically important financial intermediary should ever be liquidated, for exactly the same reason that no other systemically important utility is liquidated and broken up. But bankers, and their shareholders, should be allowed to fail, just as other utility companies fail. The difference between banks and other utilities is that, because of the fungibility of their assets, banking “failure” and associated intervention by the authorities should come much earlier in the process, well before bank equity capital has been reduced to near-zero.
The best response to bank fragility remains prompt corrective action. It was tried in the US in 2008, but it failed – partly because it relied on a faulty trigger mechanism. The concept was right, but the mechanisms need repair.