Legal & General Group Plc (LGEN) Earnings Call Transcript & Summary

Okay, and thank you very much for joining today's Q&A session on IFRS 17. I hope you've all had the opportunity to watch the video or to look through the slides, both of which are on our website. My name is Ed Houghton. I'm the Group Strategy and IR Director. I'm joined today by Jeff Davies, Group CFO; and Richard Crookes, IFRS 17 Accounting Lead. Can I remind analysts please to raise your virtual hand since I can see some of you have done if you'd like to ask a question. I'll then invite you to turn on your camera and to unmute your line. We need to allow you to do this from our end. So please bear in mind that it may take a moment or 2 while we click the necessary buttons. [Operator Instructions]

The first question is from Andy Sinclair from Bank of America.

Thanks for the time today and thanks very much for all the details you provided. I could see first set of tracks of the U.K. life insurers. And so 2 for me, please. Firstly, I think you've been pretty clear that nothing in the real world is changing, but then you did give some color on IFRS distributable reserves in the presentation. But I just really wondered if you could put some more numbers around that, just to give comfort that we're not going to run into any issues of distributable earnings and that becoming a binding constraint. That's my first question. And my second question, I've asked a few other companies similar questions as well. I was just looking at the difference between own funds and Solvency II world and IFRS shareholders' equity plus CSM and I guess, risk adjustment as well. I think oil funds are about GBP 17.4 billion at half year, but that includes various qualifying debt. I think book value plus CSL and risk adjustment, I get to about 17.5 to 18.5 roughly billion from the numbers you put out today. Just wondered if you can give me a kind of a waterfall between the 2 of them because I don't think you get too much goodwill, but just if you can walk through that.

Yes. Cheers, Andy. Thanks. Yes, as we said, we have material distributable reserves at both the LGAS main insurance entity and at the group level. So we see no concerns at all about that, obviously, with the additional statement from the Board about dividend payments. It's in the many billions. So we have no concerns at all about it. So -- but we're very happy with that. And we've been conscious of it from the start and looking at what some of the dynamics would be and being comfortable throughout. And Richard have been one of the people monitoring that. Yes, in terms of the analysis, we do show the Tier 1 owned funds against equity plus CSM because there is some of that analysis there. Obviously, you get some slight differences in risk margins and -- versus transitionals and risk adjustments. Discount rates are slightly different in terms of actually the fundamental spread is more prudent than the 41 basis points we have under IFRS 17 being, for example. And so there are a few differences amongst that. But broadly, there are different ways of looking at the same balance sheet. And so yes, we think that's a good way to look at it, as you say, shareholder equity plus CSM and what you see as the addition. We don't have any sort of material goodwill at all in us.

Great. And the next question is from Greig Paterson.

Can you hear and see me, everybody?

We can hear you.

Can't see my hands and face?

No, we can't, which is a disappointment, Greig.

Sorry, jokes aside. Just 2 questions. In terms of the operating profit, I mean, I've obviously just had a quick scan through your presentation. Am I correct in understanding that there'll be no mark-to-market impacts going through the operating profits or either below the line or straight to LCI? And the second point is -- second question is, excluding the movement in the CSM, am I correct that there will be a supplementary income statements look very, very similar to what we saw before in terms of new business in-force contribution, operating assumption? And I see even you're referring to MSG. So by definition, that still has to exist. So basically, the income statement for an investor will look pretty similar, except for the CSM movement. Is that a fair statement?

Yes. Okay. Yes, our profit, they will be very little in terms of mark-to-market there. Certainly, the changes in terms of insurance business will be very much an unwind and expected on our excess assets and then the unwind from the amortized cost assets that we've allocated, which effectively backed the CSM. And so we won't get that. We then get the expected unwind, if you like, from the prudence in the discount rate, the sort of 41 basis points. And so the rest would fall to investment bearings to the extent it exists, but we have looked to take as much of that out of the result as possible, for example, through the protection business going through OCI in terms of rates impact and by allocating some of the assets, the back, the CSM for the annuity business to amortize cost as well. So there should be a lot less noise from the insurance business in respect of mark-to-market certainly. In terms of the statement, I mean to some extent, that's true. We're into the depths of disclosures and what's required, but there will be a breakdown. But it will be -- a lot of it will be centered around sort of CSM, if you like, a new business metric. What are you adding from CSM? What is the unwind from CSM? Then what is your interest rate accretion to that. What have you earned on your assets and then that will give you your overall operating result. But there will be -- and we will also help walk people through what we saw before and what are you seeing now certainly in the early stages. That could be one of the things we would do in May, for example, to try and make that a bit easier.

Till have all those constructs that it had that I mentioned that it had before pre-CSM, and then there will be the CSM adjustments to those old?

You'll be able to see experience variance. You obviously see CSM unwind. So most of them will be there, but you won't have the net release from operations doesn't exist in the same way. That is effectively will your CSM, to some extent, and what's been thrown off the in-force comes from that?

All right. So there will be differences in the impact?

Yes, absolutely. But as I say, we'll try and walk people through that as we give much broader disclosures going forward.

Okay. Thanks, Greig. And the next question is from Ashik Musaddi. Go ahead, Ashik. Yes, we'll come back to Ashik.

We'll come back after the next one.

Yes. So let's go ahead, please, and open the line to Nasib. Thank you.

Can you guys hear me?

Yes. Yes, all good.

Two questions. So I guess the first one is on the EPS target being greater than the EPS previously. I see in the release, you've said the GBP 10 billion of PRT would give you 6% to 7% CAGR on UK PRT operating earnings. So -- and you've changed the DPS growth to 5% as well. So it seems like EPS is still growing greater than DPS. And is the growth actually bigger because you kind of rebased your starting point lower by 20% to 25%. So that's the first question. Second question on Slide 14. Just on the prudence on IFRS 17, the 41 bps. Is that just credit risk? Or does that include risk adjustment, CSM as well? Because 41 bps seems a little bit high. And Jeff, you mentioned that the fundamental spread is higher, which is about -- around about 50 bps. So if I compare kind of the prudence on Solvency II, that will be higher on a like-for-like basis, right? So we could kind of put 50 bps next to that on a Solvency II basis. Is that correct?

Yes. Yes, so I'm not quite sure. I'll say apart from you're right or not, I mean EPS -- yes, so EPS growth in DTS. We have -- we tried to give the example that if we write sort of our ambition levels, when we talk about GBP 10 billion of PRTs after what we've talked about and equally on a self-sustaining portfolio, then you would get profit growth of 6% to 7%, broadly across the insurance business. Clearly, there's upside to that if we are to capitalize on the very large PRT market that's out there, anything we write on top of that would give us an increase in profit growth. And then, of course, any assumption changes. If there are longevity releases in the future, those will also add to CSM and give us greater profit growth. So that's a sort of a clean number. Everything happens as expected. You write this business with a sort of a 9% of CSM and risk adjustment added on GBP 10 billion, and that then it runs off at 8%. So just trying to give the math for people to build something. But yes, there are -- well, upsides and you can argue downsides, but there's definitely upsides to that if we write more volume or there are longevity releases, et cetera, that come through over these. You're spot on. I've been arguing the points. 41 basis points does seem quite prudent. It's supposed to be a best estimate view of cost of default and downgrade but it is quite a big number. Our own experience is about 1 basis point for defaults, as you know. So that's where roughly 20 basis points of that would unwind into up profit and then the next 20 basis points, if experience is in line with what we've had, would unwind to investment bearings. And so we're looking about 200 million of profits each year rolling out of that 41 basis points on top of CSM unwind and risk adjustment unwind.

Thanks, Nasib. And next question is from Andrew Crean.

Okay. Can you hear and see me?

Yes. All good, Andrew.

Good. That's amazing, I can actually work with technology. Two things. Firstly, historically, you've hedged the IFRS balance sheet, which is unique. Now that you've got a steadier IFRS position, will you switch to hedging the Solvency II balance sheet and therefore, putting an underpin to your 220% to 225% coverage ratio? And secondly, I'm still struggling to understand why you want to grow your earnings faster than your dividends, given the fact you've got a much steadier profile of earnings. I think you're saying over the last few years, you've generated about 0.5 billion more capital, surplus and dividends. Is it not time to increase the payout ratio not reduce it?

So yes, good questions, Andrew. Hedging, well, it's interesting, actually. And we are now looking at what you've talked about. The movement to IFRS 17 definitely makes our ALM easier and the hedging easier. Because what we effectively do, you get 2 best estimates that are much more alike on an IFRS 17 and a Solvency II basis. So we can do our cash flow match and do our best estimate liability matching. Then we have excess assets and we can decide what we do with that. The fact that we have allocated quite a lot of the ones back in the annuity CSM to amortize cost means that doesn't give any noise in the investment variance. And therefore, we can have those to offset movements in rates for the solvency ratio. And so we will be looking as we embed all of this to see if there is something we can do, which could involve lengthening some of those with some more assets outside the best estimate liability, which could potentially take some of the rate sensitivity out to the solvency ratio. But we obviously -- at the moment, we're bedding in that moving from IFRS 4 to IFRS 17, and it's something we're investigating to see would we want to do more of that length and its likely using assets, which would slightly reduce that solvency sensitivity that you talk about in the ratio. It is, of course, as we've always said, a noneconomic hedging that you are doing, and so we would want to trade off what are the costs and benefits of that, but it's definitely something we're investigating. And as we do more thinking on it, we will obviously update, and you'll see it coming through in our sensitivities. The earnings less -- greater than dividend. Well, we think it's a good thing to grow the earnings, of course. And we have grown the book value consistently. We see that as an indication of good quality earnings. I mean as much as anything, it's to show we also manage the investment variance and our good quality result, which has been -- had earnings growing bigger than dividend. We do want to reinvest to some extent, whether that is capital or liquidity, whilst growing the book value, gives us reducing leverage and gives us options around that. I mean in terms of restating it this time, obviously, we didn't want to use an IFRS education session to completely change all of our targets. We did update on dividend to some extent. But we didn't want to restate every item. So in terms of earnings, greater than dividend, but we do see value in quality of earnings, growing book value and having some optionality around that. And as we've said before, we'll continue to review where we think the capital policy takes us and what we should be doing, but we're happy with the 5% growth in dividend at this stage and using our earnings to potentially deploy against the PRT demand that is out there and invest on an ongoing basis into LGC and other growth opportunities.

Okay. So reducing payout ratios overtime?

Potentially, yes. If you're growing your earnings, yes, that's right. Sorry, I was thinking at the [indiscernible]. Yes.

Okay. And we try and go back to Ashik, perhaps. Ashik, hopefully, we can hear you if you unmute your line this time.

Can you hear me?

Yes.

That's great. So I just have a couple of questions. So sorry, first of all, for clarification first. You mentioned that the divisional earnings are going to go down by 20% to 25%. So is that only LGRI and LGI? Is it the total divisional earnings? And what about the final operating earnings? Is it possible to get a bit of a view on what is the final operating earnings? Does it mean that if it's all divisional earnings, then does that mean that final operating earnings is going to go down more than that? So that's the first one. And just related to that is any sense of how much of this is only coming from annuities would be helpful? Second thing is now, I guess, one of the reasons why your earnings are going down is because of growth because probably you will have now negative new business trend. So what happens to your earnings growth if, let's say, you do not grow at all? You just keep your book flat, whichever let's say, GBP 3 billion, GBP 4 billion of annuities to maintain a flat book. What happens to your earnings growth? I'm just trying to get a bit of understanding is what is the steady state earnings growth or earnings number, if you don't grow at all? Just keep a flat, steady-state business.

Yes. Okay. Yes, obviously, the only businesses that are impacted are the insurance businesses. So specifically, it's the annuities and the protection businesses within the Retail division and obviously, the PRT business, LGRI. So those are the only ones impacted. Therefore, if you run through their numbers to the op profit by divisions, you get the 20% to 25%. If you then look at the bottom line of profit, you obviously get a slightly higher number than that depending on the relative impacts of the group costs, et cetera, and just get a small difference between that. And so that's just slightly higher. That's all there is. You can drop it in on what we've had over the last sort of 3 years, if you like. And in terms of flat, I think the simple answer is it would be flat. So if we keep it flat, it'd be flat because then, we will effectively be -- the mass we've tried to give to help people in their models is the CSM, you add interest of about 3%, and the CSM amortizes at about 8%. So that would be net about 5%. And if you were -- so that would run down at about 5% of the CSM. But it would be -- if you're keeping the business flat, your investment margins, et cetera, would be pretty similar. So you might get a slight reduction if you were doing at that level. But clearly, what we're out to do is to add to the CSM. So even if we're writing GBP 7 million, GBP 8 million, GBP 9 billion, GBP 10 billion, then we're going to be growing the business. And we'll try to give again some numbers, so people can model that. You add an easy because it's round numbers. You had GBP 10 billion of new business with 9% of CSM and risk adjustment that runs off at about -- into profit of about 8% per annum. And so that gives you a growing book, plus you get the investment margin, which increases on top of that. And so that's where you get to the sort of 6% to 7%. So there's enough pieces in there for people to model an annuity book. I'm sorry, because you did mention in the first part, it is dominated by the annuities. 80% of the CSM risk adjustment is from the annuity book. And so that dominates really the dynamics of it. Clearly, the protection book is in there. It's interesting. It's got a material amount in the billions of CSM, but the annuity book absolutely dominates quite clearly for insurance business.

And just one thing on this again. So it's fair to say that it's not that new business strain is adding any drop in earnings. It's nothing to do with new business strain going forward as well. Does your growth rate, the 6% to 7% changes, if you do more growth or less growth?

Yes.

Because in Solvency II, the way it works is if you grow faster, then your total capital generation drops, it's counterintuitive, but it drops because of the new business strain. So does that same dynamic work here as well?

No, it's the opposite. So the more business you write, the more CSM you add, assuming it's all profitable, obviously. The more CSM, you add, which just means your profits grow in the following period. It's very predictable. It's not volatile. You can see what happens. It's all math, how much are you adding, how much is running off. You all know the duration of an annuity book. You add some interest to that and we make the yield on the underlying assets, which is an additional source of profit. So it's very straightforward. And you can work out the maths to grow the book. So there's no strain elements involved as long as you operate in a loss-making business.

Thanks, Ashik. Let's go please to Larissa next.

Two questions, both pertaining to Slide 13, please. I recognize that this is illustrative, so not science. But if we can look at the component parts, on your existing CSM stock, if we roughly eyeball it, it looks like it decreases by about 3-odd percent a year. And if the question is how much of next year's profit has already locked in from the existing book, is that a reasonable rough way to look at it?

Sorry, almost all of next year's profits is locked in from the existing book because the new business profit is just at most half a year's runoff on the CSM. So [ 1/24 ], let's say, making it up of what you write in the year. So it's all locked in.

Thank you. Because here, it appears to be going down, which seemed wrong. I guess we can basically assume that year-on-year will stay flat if we route your new business. Is that correct?

Yes. Well, actually, we do show on a previous slide that the runoff does slightly increase as a percentage. So actually, the contribution from the in-force, especially over the first few years, is pretty flat in your earnings, which is obviously a nice feature to have of the business.

So effectively, like I say, we take a 20% to 25% [ knock ] next year, but then effectively, the profits are locked in for the next 12 years, is that a reasonable way to think about it?

Yes, that's right, as long as experience, obviously tells that.

And then if we -- hopefully, fingers crossed. And then if we take new business profit, can we -- would it be a reasonable expectation to take VNB divided by 12 and add that to the previous year's profit?

Yes, well, we say in the presentation that new business, focusing on annuities that it's approximately 9% of premium, so GBP 900 million on the GBP 10 billion book. So that goes straight to your CSM and risk adjustment. And then as you say, the following year, let's call it 8%, 12 of that runs off into profit. So yes, you're adding that each period, and that's where you can see building up. And don't forget, you're also growing the book. So the investment return component, which is not included in that, also grows as you grow in the book. So if GBP 3 billion, GBP 4 billion of annuities run off, but you write GBP 10 billion, you've grown the book by GBP 6 billion, which out of GBP 80 billion, GBP 90 billion is 7%, 8%. So that's where you get more growth.

Next and currently, the last question is from Dom Mahony.

Hello, folks. Can you hear me?

Yes, perfect.

Thanks also from me for the presentation very helpful. I've got a couple of questions. So just on the 20% to 25% change. I mean that's against the last 3 years. And the last 3 years, your assumption changes. If I look at your profit breakdown, they're very big, something like 20% of the operating profit. So can you just give us a sense of what the 20% to 25% reduction would be if you structure all the assumption changes out of the numbers? So really just thinking about the new business dynamic, I'm guessing it's lower than the 20% to 25%. And then can I just also just clarify, I'm hoping this is part of the same question, so I've got my allowance of 2. Experience variances will still be recognized in the year. So if you have excess mortality in a year, am I right in saying that will all be recognized in the year, won't be rolled up into the same set? And then second question, in terms of the transition of the balance sheet, I see you've got about 1/3 of the CSM comes from the fair value approach. Can you give us a sense of whether you think applying the fair value approach versus the full retrospective or modified perspective. Does that change the output in terms of how much CSM and how much profit you're expecting to generate? Or is it really just a detail of methodology and I shouldn't worry too much about the transition approach?

No, good questions. Yes, as you say, over the last 3 years, the assumption changes. Longevity releases, et cetera, have been quite material. The easiest way to think about new business. I mean there's some small numbers there from retail protection in the tens of millions. But of course, the number we've always used is 2.5% to about 4% in a really good year of premium would have been profit for the annuity business. So if we were right in GBP 6 billion, GBP 7 billion, GBP 8 billion, GBP 9 billion, you're talking a couple of hundred million of profit only would have been from the new business if you take the assumption changes at. So you can see that there, that 2 out of almost GBP 1.5 billion, GBP 2 billion, whatever the number is, I mean even of the annuity business, where we were making GBP 1 billion plus on PRT business, and the GBP 200 million wasn't a huge number of that for just new business alone, and that was only PRT let alone the whole group profits. So yes, you're right on that. It's that sort of the percent. If you think about as a percentage of premium, we would always talk about 2.5% to max 3.5%, 4% of premium as profits under IFRS 4. And now we're saying, well, look, 9% is set up a CSM and risk adjustment for annuity business, which is much more in line with our Solvency II new business value, which makes sense that, that's where you get to. On your experience variance, yes, the actual variance within the year will obviously flow through in exactly the same way. It's only if you make an assumption change. Does it go to the CSM. So if we made a change to our view of future longevity improvements, that would be capitalized and put in the CSM and then spread over the next 12 years as we've been talking about as a sort of duration element. As opposed to before, it would have dropped through to profit. So yes, that's a difference there. That's your AMB. Yes, 30% on the fair value. I mean Richard could bore us for a long time, the merits of the different methodologies and what they mean. To one extent, I would say, I wouldn't worry about it too much because most of the fair value is pre-Solvency II. That's been an easy cutoff for people because it was very hard to get all the data and granularity that you needed pre-2016. And the reason I say that is, to put it in context, in the last 5 years, we've written GBP 40 billion plus of annuities. Whereas in the 5 years before that, we wrote about GBP 16 billion. So the vast majority of value has come from the fully retrospective with a modification, but that's just minor tweaks where it didn't quite qualify. So it's as if we had IFRS 17 from when we wrote the vast majority of our book. And so I think I would just guide you towards that in terms of materiality.

And sorry, Jeff, do you mind if I just clarify? On the -- that's very helpful. On the 20% to 25% sort of the reduction in op profit, I'm guessing -- my question is really how much of that is because assumption change is no longer drop to the op profit? So I'm guessing a big portion of the 20% to 25% reduction in op profit is actually just the assumption change is no longer...

Yes. Absolutely, yes. I mean it's multiples.

I mean, if you -- if we were to rebase the last 3 years, excluding assumption changes, could you give us a sense of how much you would expect your profit removed?

I mean, there is in each year, but it's at least half that number, I would say.

Half is the assumption change?

Yes, yes, at least. Some years, it wouldn't be -- some years more. It's never less than half off the top of my head.

Okay. Very good. That's really helpful.

We can see in the numbers, yes. I mean, we can see in the numbers and we can share, yes.

Okay. We have a repeat question from Andrew Crean.

Look, you keep on hinting that you might increase your BPA writings. And clearly, given the funding position of the annuity market, that is very possible. It doesn't have an impact immediately on your IFRS earnings, but it will impact your solvency capital generation, and that, in the end, pays the dividend. Could you be more clear as to what your strategy is on writing BPAs? Are you going to stick with your current targets? Or do you think you're going to open your shoulders, if you have the capacity so to do?

So I mean what we've been saying it is our current approach is I think with the greater market demand out there that everyone is aware of and talking about, it is probably easier, if you like, for us to hit our 8 to 10 ambition and possibly be at the top end of that or slightly over. And that would then allow us to still be self-sustaining, the things we've talked about, et cetera, though we have headroom within that over our ambition period. And as everyone noticed, our capital position is strong at the moment. So we wouldn't be concerned with going over the limit slightly on sort of self-sustaining. However, on top of that, there is undoubtedly a good number of jumbo type cases out there that we are all in conversations with and they're trying to work on what they'd like to do, we're all trying to work out the best way to take them on board. And so on top of that BAU 8 to 10-ish type numbers, that we may well do some one-off transactions, which might repeat for a few years. And as and when we get to those, we'll communicate appropriately around those, how we're thinking about it, how are we using reinsurance, what does it do into our capital, how we thought about it, and we would do that at the time. But clearly, we will look at those. What metrics are available? Can we deliver them? We have the assets to back them, et cetera, and we'll talk about at the time. So we're certainly not shutting off that we're going to do that. But I think that's more likely the scenario then that we just happen to write 12 by mistake as BAU, if you like. And so we'll continue to communicate around that. And as we have more conversations with schemes, I think the intentions of all parties will become clearer on that.

Okay. And you'd like us to look more IFRS or your Solvency II capital generation. What is your primary metric you want us to measure you by?

Solvency II is clearly the more sort of economic view of the world. It's certainly the constraint that everyone has looked at in terms of dividend paying ability. What is the -- we're in a nice position of 220, 225 at the moment where rates are on any day. But it's that capital generation, what are we doing that helps us to determine what is available, what does that mean? What's coming off the business? We now have an underpin of very stable accounting, but we've seen that, that could be very volatile. So Solvency II is generally what we've looked at and you've all looked at over the last few years.

Oliver Steel, I can see you've got your hand up. You haven't asked a question. So please go ahead.

Hello. I hope you can hear me -- you're hearing me. 50% of your LTIP is linked to IFRS earnings and has been, I think, over the last few years. And you're now telling us that all the assumption changes have basically been written back over. So how is your Board assessing the LTIPs that you've paid in recent years? And how is -- how are you planning to adjust the LTIP targets for 2020 to 2023 and '21 to '24?

Well, I didn't think we'd be talking about pay for that. So that's a good -- but we have considered that. I mean, we've obviously looked at what does it mean to EPS? What are the elements in there? I mean the biggest thing I would point you towards is, of course, the vast majority of our longevity releases were excluded from everything. We excluded them explicitly. They were excluded from what you looked at in terms of earnings they were excluded from all of our earnings discussions. And so that has been taken out of that. We always and will continue to have the very interesting discussions on Remco's of one-offs, releases, what does that look like? And we absolutely are considering all of that. So it is very relevant. And equally, where are we rebasing for EPS, et cetera, and we have considered that. So that will come out in the next -- we'll have to explain it further in the next report and accounts.

Okay. Good to hear that the exceptionals were excluded. That's actually quite a relief. And just going back to Andrew Crean's question then about what you see is more important between Solvency II capital generation versus IFRS earnings. I mean you're almost unique amongst the U.K. life companies is still using IFRS earnings and the LTIP. So are you going to change that going forward?

That's an interesting question. So we have brought in a Solvency II metric more quantitatively into our earnings management objectives, et cetera. And we'll continue to look at that for this very reason. We've been looking at what makes sense is obviously then the debate of what's the right one. Is it operating surplus generation? Net surplus generation? Is it capital budgets and usage? So yes, that's further discussion we have. We do have one -- explicit one in there, and we will continue to do that. But equally -- it's a bit like we said, earnings versus [ dividend ]. We quite like growing book value and showing that you've got good quality earnings. And so you can't completely ignore one. One you can say is accounting, one's a regulatory basis, and there's a balance, but it's more what's the constraint, what is management incentivized as done because that's the constraint on the business. And if it is more about surplus generation, then we will potentially look at that and decide whether we put more weight on it or not.

Okay. Let's go to Ashik again, please.

Just one question. I mean, is -- this risk adjustment, I mean, so if I understand correctly, is it fair to say that risk adjustment number is about, say, GBP 4 billion, GBP 5 billion? And what is the basis of coming to this risk adjustment number? Because -- I'm asking it because, I guess, given that you have done a lot of longevity transfers to reinsurers, why do we have such a big risk adjustment number at this point? We have assumed that it's a bit lower. So...

Yes. And actually, as you've asked, it's not explicitly in there anyway. I don't know if you've got your ruler out on one of the pages. It's a smaller number than that. It's let's call it GBP 2 billion to GBP 2.5 billion. Obviously, it's not all -- none of this is finalized yet. How do we get to that? It is -- it's not the same type of calculation of the risk margin. It's very much what is your claims experience, lapse experience in the 85th percentile. And if that's what you get throughout, how much extra would you need for that? What's the sort of reward that you need for that? What is the extra claims cost? And so that's where you get to the number. And whilst you say our new business has had a lot of reinsurance, which is true, of course, we haven't reinsured any of the individual annuities. We had a reasonable size back book. And so I think we are still, let's call it round number, 70% or so retained on longevity, might be slightly lower than that now after we eventually write all the 22 but it's certainly 65% still retained. And so you get a reasonable number just for pure longevity. And of course, we have mortality in the U.S., lapse experience on the protection books as well.

Okay. Just one more question I have is, I guess, this time I'm asking a bit too much. So your net obviously -- I know I'm at a risk of asking too much. So your net of tax CSM is about GBP 8 billion to GBP 8.5 billion. Is it possible to split this between what is the spread component and what is the technical margin component, a longevity component or say how much is the protection component? And what is the spread component?

So we do give that it's about 80% of it is annuities.

No. Within annuities, is it all spread? Or is there a longevity component of that?

No, no, no. There's no spread component, and that's in the discount rate. So that's in the -- the prudence of the discount rate is within your best estimate liability. So that's the 41 basis points. If you think of it very simply on day 1, you write annuity. You have a premium. You discounted your expected payouts at a prudent rate, which with the 41 bps of the yield, what's left is the CSM. So that's all -- that's just left over from the premium less the best estimate, less the risk adjustment, obviously.

I can see Richard suffering. So I think that we'll draw a line under things here. Greig, we can follow up with you off-line, if that's okay. Thank you very much for your questions today. I hope you found the video, slides and Q&A session helpful. Please do follow up with us in IR if you've got further questions. Thank you very much.

Yes. Thank you, everyone. Yes.