Should banks be nationalized to protect shareholders?

What I mean is that banks are risk-magnifiers. When they lose money, credit to the whole economy gets choked off, thus causing recession. Banks are critical hubs in a network economy.

Put it this way. In the 2008 financial crisis, the US's biggest financial institutions lost between them less than $150bn. But during the tech crash of 2000-03 investors in US stocks lost over $5 trillion. The former led to a great depression, the latter to only the mildest of downturns. Why the difference? One big reason is that losses are easier to bear if they are spread across millions of (mostly unleveraged) people, but cause real trouble if they are concentrated in a few leveraged strategically important institutions.

One reason why non-financial stocks have fallen recently is that investors fear a repeat of 2008 - a fear which is all the greater because banks are so opaque. Yes, the bosses of Deutsche and Credit Suisse claim that they are sound - but nobody believes bosses these days. As Nicholas Taleb said, bankers are "not conservative, just phenomenally skilled at self-deception by burying the possibility of a large, devastating loss under the rug."

I suspect the CAPM has got things backwards. It says that banks fall a lot when the general market falls because they are, in effect, a geared play upon the general market. But sometimes, the market falls because banks fall.

Which leads me to the case for nationalization. This wouldn't prevent banks losing money: these are inevitable sometimes because of complexity, bounded rationality and limited knowledge. However, when banks are nationalized, their losses would create only a very minor problem for the public finances as governments borrow money to recapitalize them*. That needn't generate the fears of a credit crunch or financial crisis that we've seen recently. In this sense, nationalization would act as a circuit-breaker, preventing blow-ups at banks from damaging the rest of the economy. (Given that countries are exposed to financial crises overseas, the full benefit of this requires that banks be nationalized in all countries).

You might reply that the same effect could be achieved by demanding that banks were better capitalized, as Anat Admati and Martin Hellwig have argued: calls for 100% reserve banking are to a large extent just an extreme version of this.

However, the former would require massive share issues, which would themselves hurt stock markets. And the transition to the latter - as even its advocates acknowledge - would be complex: in fact, Frances has argued that it would kill off commercial banking. Nationalizing banks would be simpler.

You might object that doing so would impose losses upon shareholders, and the adverse wealth effects would depress demand. I'm not sure. By reducing the chances of future financial crises, the risk premium on non-financial stocks should fall, causing their prices to rise. And to the extent that banks have a positive net present value at all, their transfer to the public sector represents not a loss of wealth but a mere transfer: what bank shareholders lose, the tax-payer gains. The only wealth loss would come if banks are worse-managed in the public sector than they would be in the private - and that's a low bar. Net, there might well be a positive wealth effect.

My point here is, however, a broader one. One fact illustrates it. During the golden age of social democracy - from 1947 to 1973 - UK real total equity returns averaged 5.1% per year. If we take the fall of the Berlin wall in 1989 as its starting point, they have returned 4.9% per year in the "neoliberal" era. This alerts us to a possibility - that perhaps some social democratic policies are in the interests not just of workers but of shareholders too. Maybe the beneficiaries of neoliberalism are fewer than one might imagine.

* Because losses are most likely to happen when the economy is depressed, such borrowing would be done when demand for gilts is high and borrowing costs low. In fact, in such depressed conditions there might well be a case for quantitative easing, whereby the Bank of England buys the bond issues directly.