The queues outside Northern Rock branches in September 2007 were the first run on a British bank in over a century, and they happened partly because the safety net of the day was feeble: deposit insurance covered 100 per cent of only the first £2,000, then 90 per cent of the next £33,000. A saver with £40,000 stood to lose real money, so standing in the rain to withdraw it was rational. Nearly two decades of rebuilding later, a UK bank failure is designed to be something most customers notice only as a letter and a rebranded app.
Two mechanisms do the work. The first is the Financial Services Compensation Scheme, which guarantees deposits up to £85,000 per person, per authorised institution — £170,000 for a joint account, since each holder gets their own limit. The scheme is funded by a levy on the financial industry rather than by taxpayers, and if a deposit-taker is declared in default, the FSCS aims to return money to the majority of savers within seven days, automatically. There is no form to fill in: the failed bank's records are used to calculate what each customer is owed, and payment arrives by cheque or transfer, or the accounts are simply moved wholesale to a purchasing bank. Temporary high balances — the proceeds of a house sale, an inheritance, a redundancy payout — get extended cover of up to £1 million for six months, an acknowledgement that people cannot always keep life events under the threshold.
The second mechanism is meant to stop the FSCS ever being needed for a bank of any size. The Banking Act 2009 created a resolution regime run by the Bank of England, which plans in advance how every UK deposit-taker would be dismantled or transferred if it failed. The tools are blunt and fast: the Bank can sell the whole business to a rival, move deposits into a bridge bank it controls, or — for the largest firms — write down bondholders and convert their debt into equity, the process known as bail-in, so that investors rather than taxpayers absorb the losses. The point is to do all of it between Friday's market close and Monday's opening. That is exactly what happened in March 2023, when Silicon Valley Bank's UK arm was resolved over a weekend and sold to HSBC for £1. Customers, many of them tech firms with payrolls due, logged in on Monday and found their money where they had left it. Compare that with the drawn-out nationalisations of Northern Rock and Bradford & Bingley in 2007–08, and the machinery has visibly improved.
The small print that still bites
The protection is per authorised institution, and an institution is a banking licence, not a brand. First Direct sits on HSBC's licence; Halifax and Bank of Scotland share one; several supermarket and app-based brands have historically sat on other firms' authorisations. A saver holding £85,000 with each of two brands that share a licence is covered for £85,000 in total, not £170,000. The Bank of England publishes a register of which trading names belong to which licence, but few people read it, and surveys repeatedly find savers assuming each brand carries its own limit. Anyone with serious cash savings should check the register — or simply keep no more than £85,000 with any one banking group.
There are other gaps worth knowing. E-money firms and some payment apps are not banks: customer funds are "safeguarded" in ringfenced accounts rather than FSCS-protected, and if the firm collapses, getting money back can take months and may be reduced by administrators' fees. Cash Individual Savings Account transfers in flight, funds held with credit unions (which are covered) versus those in some Christmas savings clubs (which are not), and deposits at overseas banks passporting under a foreign scheme all follow different rules. The 2008-era lesson was that panic spreads when protection is partial or unclear. The modern regime has largely fixed the partial; the unclear remains the saver's own homework.
