Wikipedia: “A single point of failure is a part of a system that, if it fails, will stop the entire system from working”. Right. You try to avoid this in any decent system, building in redundancy at every critical point. The classic example is a chain, where any single link is a single point of failure. You can build in redundancy by, e.g., adding a second chain. So long as each chain on its own can support the weight, nothing bad will happen if one chain breaks (bad luck if both break, though).
Something like that principle of redundancy was meant to be part of the life insurance Part VII regime. You have the PRA which approves the default model of the insurer. The FCA is meant to look at the same thing from the point of view of the policyholder (the PRA being a prudential regulator, as I pointed out to the court here, see page 5). The Independent Expert provides further assurance with his/her ‘independent’ report. Finally, the court itself forms a judgment.
However, that principle seems to have been disastrously weakened in the light of the recent judgment by Justice Zacaroli.
Justice Zacaroli stated unequivocally that it was not for the court to ‘go behind’ the Solvency II regime
83. In my judgment, in agreement with the submissions of Mr Weitzman and Mr Moore, in considering whether Utmost satisfies the solvency criteria laid down by Solvency II, I must apply the regulatory regime as it exists, and it is not for me to go behind the requirements embodied in legislation (even if it were possible for me to reach a concluded view on Dr Buckner’s objections to matching adjustment, which it is not in the absence of hearing evidence from competing experts presented for crossexamination on their opinions).
My emphasis.
But this confuses challenging the regime directly, which the court clearly cannot do, and questioning the probability of default estimated by a firm’s model. A low probability of default is fundamental to the safety and security of policyholders, and the Part VII hearing primarily exists to protect the safety and security of policyholders. The firm’s model will of course have been recognised by the regulator, but that does not imply the probability is correct, and to question it is not directly to question the regime itself.
But sadly we are now in the position where a court, by this precedent, cannot question a capital model. The Independent Expert report is no help – I know of no such report where the expert has looked into the credit or market risk of the underlying assets, and questioned the risk estimate of the model itself. The PRA position is, as Weitzman said, that the regulator cannot ignore its own rules. And the FCA simply defer to the PRA.
MR VAN SANTE: In relation to the particular issue of matching adjustment, the FCA does essentially defer to the PRA as it is the PRA’s remit.
So we are down to a single point of failure, viz. some nameless actuary in the PRA insurance department subject to forces on every side, including firms phoning up the director of Insurance to complain about him or her, lobbying from the Treasury, shouting from senior people and so on. And, to be honest, a culture within the regulator that takes as its starting point whether the firm can withstand the regulatory treatment initially proposed.
As I said on Monday, “it’s like a fire department that issues the fire safety certificate, also answerable for the financial condition of the firms that made the cladding”.