More to follow but see below for a partial transcript of the Treasury Committee hearing yesterday, with Harriett Baldwin quizzing Jon Cunliffe and others about the effect of Covid on life insurers.
Does Sir Jon agree with Sir John Vickers “when he says that, er, the Matching Adjustment is more of a mask than a cushion“?
A lot of waffle from Sir Jon. He says that ‘market liquidity risks’ are not suffered if assets are held to maturity, which is correct, then says “if you were to price the assets that insurance companies hold on their balance sheets at market prices, you would be picking up how liquid the assets were, whether you could sell them in stress etc”, which is clearly false, but at least confirms that he thinks it is a cushion.
Baldwin complains that she is out of time, and Stride (Chair) closes with the remark that he senses “a slight frustration there, and you might have valued a little more time to probe”
I think on that basis we might write to the panel after this session, and if we do if I can ask the members of the panel to respond very promptly to any letter we might send on the issue of insurance and stress testing.
Clearly more to follow. Stay tuned.
—Treasury Committee Hearing 20 May 2020
Baldwin 15:22:40 … why the Bank has chosen this moment not to publish the 2019 stress test for insurers.?
Cunliffe: there are two separate things here … [first is about business continuity]
Baldwin: I really want to talk about the publication of the stress test …
Cunliffe: the decision not to publish the stress test I think was unconnected with the first issue [business continuity]. The issue I think on the stress test was, this was the first time that the stress test had included life insurers [15:23:28] as well as general insurers. We’ve run a general insurance test for a couple of years beforehand, and it was to some extent an exploratory test, and when the results came back it was clear that in order to get a clear enough message for publication it was going to require quite a lot of work to actually get back to the companies and probe how they had done the tests and asset it. And that’s not unusual we had the same thing when we brought in the stress test for banks. That then happened at the start of the crisis, and the decision was that the resources of the company and the Bank would be better spent by actually doing specific stress tests around what we’d already seen in the markets and around what was likely or what could happen [15:24:28] to the asset side of insurers balance sheets.
So there have been subjecting to the tests. Their solvency ratios are pretty strong, and on [in] the light of that there was a statement in the IFS [?] that we had subjected them to individual tests, which have a really big downgrade of credit rating in the economy and whether their asset side can deal with that. But the actual tests that we have done, I think it would have been, er, to get it to the point of publication I think would have been difficult and I don’t think it would have told us very much about ..
Baldwin: .. but don’t, don’t you think Sir Jon that it looks as though you’re trying to hide something by not publishing them now and when will they be published?
Cunliffe: er, I’m not sure .. I don’t know what we’ll do about the 2019 test as we go through this crisis, because it may not be very relevant to publish it later on. But as I say it would need considerable effort and work to get it to a point where it could be a reliable guide, particularly on the life insurance side. But on the other side [GI?] we have the solvency ratios of the companies and they are pretty robust. [15:25:53]
Baldwin: So can you say on the record then that the stress test would not show that any companies are failing?
Cunliffe: I don’t think that stress tests show that any companies are failing. [But] some of the evidence needs a lot more work to get a clear message that we could publish.
Baldwin: Governor I think you were trying to come in on this.
Bailey: … for general insurance the stress test was not a pandemic, but natural disasters such as flood, earthquake .. so it wouldn’t tell you a lot about the situation we are in at the moment, on the GI side. I think the point you raised in respect of your constituents [about business continuity] is a very important one, I’ll leave the FCA to one side. On the prudential side I think there is a very important question there which is that the prudential capital framework is obviously calibrated to the risks that are written, to the policies that are written. Of course the issue, and that is that the FCA [had?] its court case .. for many policies the risk that is written is clear, for some it is not as clear as it needs to be and that’s why there’s a court case. If I could just .. on that life insurance side, because that’s [mutters] again for a pandemic style risk stress test you would have to run a completely different stress test. [15:27:19] It is relevant for two reasons. One is credit risk, as Jon was saying, where Solvency II does already, it’s what’s called a 1 in 200 – a 95.5% confidence limit, and when the so-called quantative indicators were put together, I do remember the members of this Committee complained quite vocally a few years ago. They are calibrated around the 1930s, the credit events of the 1930s. So that’s the basis of Solvency II. You can also stress that. The second thing I would say is finally on stress tests of life insurers, and this is really a sad thing, is that you would have to look at the assumptions around the restructuring of the mortality models, because an annuity business, which is what they are, obviously Covid, I’m afraid, is going to have an impact there, but it has the [potential?] to go the other way. So all …
Baldwin: OK ..
Bailey: So all of that needs to be considered, as is the position of your constituents by the way.
Baldwin: I appreciate that and I am very short of time, and I just want to put on record that I think from the point of view of Parliament, we feel that we would like to know whether the prudential policies that you have put in place, and that you have been tasked to put in place are actually resulting in insurance firms that are sufficiently capitalised, and so I just urge you to be transparent on that. And then can I come on to a question to Elizabeth if I may, along the lines of the quotes from Sir John Vickers, saying that if the Bank is pausing stress tests for insurance companies as the result of Covid 19, then those same firms shouldn’t be able to pay out dividends. So is it appropriate for Legal and General to be making its £750m pound dividend payout at present.
Stheeman [15:29:10] I should just say that the question of insurance stress tests was something that would have been discussed on the PRC, the Prudential Regulation Committee, which I am not an external member of. So my colleagues, external members of the PRC, together with – the Governor obviously chairs both PRC and the FPC – [the Governor] would obviously have discussed as well as the other deputy governors who are members. If I could just talk about what happened to banks, and that was discussed in the hearing earlier, is that this is something which was discussed with banks, obviously the FPC doesn’t get involved in individual bank discussions, again that would be the PRC, and given the work being done on capital, and how capital would develop under certain scenarios based on what the MPC had done, there was the view that it was better for banks not to pay dividends and that’s what the scenarios are based on, that there are no dividends paid from 2019 and 2020.
Baldwin: [15:30:20] Ok so perhaps I could come to Sir Jon on the insurance question in particular and the Legal and General question.
Cunliffe: Yes, so as Andrew said, insurance companies are stressed against a 1 in 200 stress, which you could think of as a single letter downgrade, a whole single letter downgrade across half of a diversified portfolio which is pretty similar to what happened in I think 1932 is the worst year of the Great Depression, and their capital is I think resilient to a stress of that sort. And the decision that the PRA took was not that all insurance companies should be able to pay dividends but that it should be a case by case approach for the insurance industry as opposed to a more across the board approach that banks decided to adopt, with some encouragement from the Bank of England. And the reason for that was that one needed to go into the position of every company to see how exposed it was and whether it had the resilience. Now I don’t want to go into individual … I can’t discuss individual companies ..
Baldwin: OK, what about generically then, Sir Jon, the question of the Matching Adjustment, do you agree with Sir John Vickers when he says that, er, the Matching Adjustment is more of a mask than a cushion.
Cunliffe: [15:31:56] Erm, [pauses] I spent a long period of my life negotiating the Matching Adjustment in Brussells – I’m pleased to see that it’s alive and well. I think the Matching Adjustment simply reflects the fact that insurance companies hold their assets to maturity. So while they suffer credit risk, credit events hurt them, market liquidity risks are not suffered, and therefore if you were to price the assets that insurance companies hold on their balance sheets at market prices, you would be picking up how liquid the assets were, whether you could sell them in stress etc., whereas given that the insurance companies are trying to match liabilities they hold those assets to the end of their life, and therefore that’s a credit risk not a liquidity mark to market risk, and the Matching Adjustment was designed to reflect that.
Baldwin: [15:32:52] OK. I’m out of time I’m afraid.
Stride: Harriet thank you, and I sense a slight frustration there, and you might have valued a little more time to probe, and I think on that basis we might write to the panel after this session, and if we do if I can ask the members of the panel to respond very promptly to any letter we might send on the issue of insurance and stress testing.
Bailey [15:33:15] That would be fine, by the way, and obviously as Jon said, Sam Woods is also very much the appropriate person to get involved in that.