As JM Keynes observed, if you owe the bank £100 you have a problem. But if you owe £1m, the bank has a problem.
Now, with the turmoil in the financial markets, it is clear that if a bank owes a bank billions of pounds we all have a problem. Bailing out failing banks and investment houses has apparently become necessary if governments are to protect citizens.
As the US struggles to get a rescue package working and governments worldwide consider what changes to make to financial regulation, it may seem an inauspicious moment for the Department of Health to launch its consultation on a regime for unsustainable NHS providers.
There are parallels between the NHS and finance markets when it comes to institutional failure: no one would think it acceptable that the innocent should suffer. But how to protect them while not undermining the ultimate market punishment for failure - "exit" from the market? There are ways round this moral hazard: bail out the bank, write off the debts with taxpayers' money, but target the punishment by sacking the management, for example.
The DH's proposed solution for a failing provider makes more transparent and formalised what has in many ways been a process the NHS has used for many years. For example, the secretary of state has always had the power to remove members of a hospital board. And a traditional approach to the need to reorganise local services has been merger of trusts.
Interestingly, too, the DH has rejected importing the insolvency scheme used in the private sector which focuses on financial failure and is not easily translated into dealing with other sorts of failure such as problems with quality of care (see first chart). It seems that when things get bad enough in the finance markets existing rules have to be bypassed.
The new failure regime represents a new, more clearly spelled-out back stop - focused on protecting patients and NHS staff. How often this insurance policy will have to be called on is made clear in the DH's associated policy impact assessment. The net benefits of the failure regime, estimated to be around£10.3bn over 20 years (see second chart) assumes every year 2.1 per cent of trusts would be classed as failing (see third chart). This represents up to six organisations each year.
It is worth every trust board reading the DH's proposals very closely.
And Keynes' last words (literally) on a life grappling with this sort of stuff? "I should have drunk more champagne."