I have so far dwelt on the positive aspects of PS 31/18. It is a landmark paper by the PRA in that for the first time it settles, with great authority and a wealth of cogent reasoning, how life insurers should correctly value a portfolio of simple European put options. Other non-insurance institutions have valued them correctly for a long time, but never mind that, it is an important breakthrough notwithstanding. In his letter of 10 December, David Rule stresses the importance of not valuing options in a way that assumes future house price growth (and by implication any asset growth) in excess of the risk-free rate.
But all is not well.
In its wisdom, Mr Rule and the PRA seem to have decided that insurers do not have to value guarantees correctly for business written before the introduction of Solvency II.
… we are not taking forward the proposal to apply an equivalent of the Effective Value Test as a single approach to determining the illiquidity premium in the pre-Solvency II Individual Capital Adequacy Standards regime for the purposes of the transitional measure on technical provisions.
This is almost impenetrable, but I think it means this. The ‘effective value test’ is simply a test to ensure a firms are valuing the no negative equity guarantee correctly, and the ‘illiquidity premium’ is the element of growth over risk free that firms were hitherto baking into their valuation. So Rule is saying that firms did not have to value guarantees correctly under the old ICAS regime, therefore they don’t have to value them correctly for any elements of their book written under that regime.
So we could have the odd scenario where there are two guarantees on the books, of identical size and nature, one written in December 2015, the other written January 2016, hence under different regimes. The second guarantee absolutely must be valued correctly (or rather, according to the new valuation rule), as though the fate of the universe depended on it. But the first (pre-Solvency) guarantee doesn’t have to be valued correctly. Or rather, the correct valuation doesn’t have to be a ‘single approach’ to the valuation. There could be more than one value, you see.
Did anyone ask whether this made any sense? The word ‘correct’ rather suggests there is only one thing that could be correct. How could answer A be correct, but different and conflicting value B be absolutely correct as well? It also sits uneasily with the cogent reasoning of the policy paper, showing how the proposed approaches by the industry and the Institute are not correct in any way.
Perhaps the PRA will object that firms don’t have to value some guarantees correctly. But what does that mean? The value is what it is, and the PRA have carefully shown how to arrive at it. We applaud their reasoning. What they appear to be suggesting is that while the value is what it is, and nothing else, a firm can write down or represent the value as something different.
But how’s that? It’s as though after long and careful thought the PRA had added up all the costs in the accounting statement and found that they correctly summed up to £1,000 (say), whereas the firm had incorrectly thought the cost was only £50. But then the PRA says ‘OK lads, £1,000 is a bit too much, so while between ourselves we agree that £1,000 is the correct cost, you can still write down £50 in the published accounts. No one will notice’.1
But isn’t that, er, a bit dishonest? Remember we are not talking capital requirements here, which have always been a bit finger in the air, to be honest. We are talking about book values. You know, the values that investors take seriously on the assumption that they are, er, correct. Moreover it’s not £1,000, the total misvaluation approaches £1bn or more, if the Prof and I are correct. Big number, no?
Makes no sense to me.
Rule finishes off saying that the PRA strategic goal is ‘… avoid any risk that the credibility of this important element of the prudential framework is undermined’. Good luck with that, as they say.