The chancellor has announced his plan to "electrify" the proposed ring-fence between retail and investment arms of British banks — which George calls the "Jurassic Park solution" — ensuring that if a bank tries to breach the ring-fence, it will be separated fully.

Much of the criticism of the ring-fence, as well as the Chancellor's defence of it, stems from its effects on financial stability. That's obviously important, but there's a bigger reason why such a move is long over-due, and that's the state backing of retail banks.

This backing isn't even a question of the too big to fail subsidies which hand around £34bn to the biggest banks. Instead, it's the effect of the financial services compensation scheme.

That's the government body which protects up to £85,000 of individuals' deposits with accredited banks. Since it was formed, it has paid out over £26bn, mostly in the aftermath of the financial crisis, to customers of retail banks which went bust.

And that's good! The FSCS is necessary in a world in which customers can't be expected to judge the financial health of a bank when deciding where to keep their money, and even more necessary given that a bank account is largely deemed a prerequisite of living a normal life in the UK today (hence the concern over the lack of basic bank accounts). But, at least on a cursory analysis, the FSCS also reduces the cost of capital for banks, because they don't have to compensate customers for the risk that they will lose all their deposits.

For a bank that's not failing, though, the FSCS is an in-kind subsidy, and it makes sense to limit that distortion. That's the reason for a ring fence: it ensures that the government is only subsidising the consumer banking sector, rather than the entire sector at once.

The stability arguments are important; and the ring-fence does indeed lessen the downside of a casino bank going under (although the amount it will lead to broader stability of the financial system is more questionable). But even without them, there'd be a strong prima facie reason for some kind of limit to the amount consumer deposits can be leveraged.