AXA SA

PAR:CS

Earnings Call Transcript 2022-Q3

[Anu Venkataraman, Head, IR]

Good afternoon, and welcome to the AXA Investor event. We'll share with you today 2 presentations. The first one is a short one, focus on our 9 months activity indicators that we published after the market closed yesterday.

Our Group CFO, Alban de Mailly, will walk you through the highlights of the 9-month report, after which we would be happy to take your questions. We'll then pause for a short break, after which, Alban; and Grégoire de Montchalin, who is our Group Chief Accounting and Reporting Officer, will walk you through our presentation on the impact of the new accounting standards.

So thank you very much for spending the next few hours with us here today. We really appreciate it. And with that, I'll turn over to Alban.

[Alban de Mailly Nesle, Group CFO]

So thank you, Anu. Good afternoon to all of you. It's good to have you here in London. So as we have the presentation on IFRS 17, we also prepared a few slides for the 9 months and that should make the discussion a bit more lively than usual on our side.

So let's start with the key highlights for 9 months. On revenues, so we stand at 38 billion. That's plus 2%. And we're happy both with the growth, but also the fact that it's very much aligned with our strategy of growing in certain lines and reducing our exposure in others. And obviously, I'll come back to that in a second.

The second very important aspect, obviously, given the context is about pricing and the fact that we see a very good pricing momentum everywhere at XL. But obviously, that has been so for quite a few years now, but also now in Europe, both in Commercial lines and in Retail lines, and that will support our margins going forward.

The third message, but that's not a surprise, is the fact that we have a robust balance sheet with a high level of solvency with a ratio that stands at 225%. And with also a good asset mix of quality, which is the result of years of prudent asset allocation and asset selection.

The fourth item is obviously, also on Ian and we're pleased to show that Ian cost $400 million -- EUR 400 million net of reinsurance and growth of tax. That's a 0.7% market share. We estimate that the cost of the hurricane for the whole industry should be around EUR 60 billion. And that shows very clearly the efforts that we've made over the last few years to drastically reduce the exposure to nat cat that we have at XL. So I think it's a good result overall.

And last, in terms of capital management, so we have completed the 1 billion share buyback that we had announced in August. And you also saw that we announced the acquisition of the Spanish insurance subsidiaries of Groupe Crédit Mutuel and you saw with the main multiples that we are very much in line with what we told you as we wanted to compare any acquisition to the benefit of a share buyback. And that kind of acquisition is exactly what we want to do, both in terms of price but also in terms of what it brings to us in terms of diversification, notably of our distribution locally.

So now if we move to revenues, so the main messages here are there are 2 of them. One is it's very much in line with what you saw half year. And second, good pricing momentum, as I said. So if I take them one by one by line of business, so P&C grew by 3%. We had very good growth in Commercial line insurance, led notably by France and Europe that are up 7% and 8%, respectively.

XL on the insurance side is up 1%. That's very much in line with what we showed also half year with a clear focus on profitability, but nevertheless, still benefiting from significant price increases, and we see that on the next slide. So that's for Commercial lines.

On the Retail lines, good momentum as well. So it's plus 4%, France and Europe. It's 2% in Motor, 4% in non-motor and generally supported by momentum in prices. And that's a good signal for what is to come in the next quarters because clearly, inflation is picking up and is probably there for a few quarters again.

The last item on P&C is XL Re. And on this, revenues are down 20%, but again, no surprise. That's very much the reflection of the reduction in nat cat exposure that we have told you about. So that's P&C.

On the Life side, we had good growth in protection 3%. That comes mainly from Switzerland and Asia. On unit-linked, we are down by 12%, very much the same story again at a half year. It's mainly the fact that in France, there was a large corporate contract last year and no equivalent this year. But what matters to us on unit linked is the net inflows and we are still very positive at 1.6 billion.

On general account, we are down, and that's driven by the traditional GA savings for which revenues are down 16%, mainly. There's also a corporate contract on that part that we had in '21 and that we don't have in '22. That's for Life.

Then we have Health. Health is growing significantly, 14%. That's mainly driven by the international business booked by AXA France, but you will see in the documents that we published yesterday that we have growth in Health, in all geographies and both Retail and Group. That's a very satisfactory outcome for those 9 months.

And finally, Asset Management. Asset Management revenues are up 2% in a difficult environment. But what's good is the net inflows standing at EUR 18 billion, exclusively from third parties that went for EUR 5 billion to our AXA IMS platform, for another EUR 5 billion to our AXA IM Core and to -- for EUR 8 billion to our Asian joint ventures. So overall, as I said, a good mix of growth in the lines that we favor.

So moving to pricing. So the first thing to say on this slide is that we are showing you here slightly different things when it comes to XL and when it comes to the other entities. When it comes to XLs, the numbers here are pricing. Price increases on renewals. So directly comparable.

When we show the numbers for Commercial lines, AXA XL or retail business, it is price effect, i.e., by how much do price increases contribute to increase in revenues. That underestimates the real price increases by circa 2%. So we will work in the future to get this, the renewal and renewal for the next quarters. But it's important to signal that here on this slide, if you compare the same business in 2 different periods. For Commercial lines, AXA XL and Retail, that's probably underestimated.

So that being said, on XL, both insurance and reinsurance 9 months we see a significant price increase still. XL insurance plus 8%. Still significantly above loss trend. And what matters also to us is to see that in the lines that are most sensitive to inflation, such as property, price increases hold up quite well. Property insurance at XL, price increases are 13%, so significantly above inflation.

When it comes to Commercial lines XL, so in our GI entities in France and Europe, it's the same story where you see the -- an acceleration in price increases and notably, again, in the -- in Property lines. So Property in Q3 discrete in Europe, prices were up 9%. So you do see the momentum that we had told you about linked to the indexation on local indices that protects us against inflation.

And last, in Personal lines. So here, the numbers are a bit lower, but you need to keep in mind that notably in Motor, we have or had the benefits of lower frequency. Now we are at a point where we see still good levels of frequency compared to 2019, but we also see a bit more inflation, hence, the need for more price increases in Europe. On this, you have several situations. You have clearly U.K. and Spain that were among the first to feel the need for price increases, given local inflation. At the other end of the spectrum and that's true both for retail and Commercial lines, you have Switzerland. Where you simply don't see inflation at all. And in France, France, you now see what I was telling you about, which is more need now for price increases. You don't see it yet for Q3.

You will see it for Q4 and even more for next year. So overall, we remain very disciplined on pricing and we see an acceleration in momentum.

If we move to solvency, so our Solvency 2 ratio stands at 225%. It's obviously a very high level. It's slightly down compared to half year. If I take the various items one by one, the operating return gave us 1 point of solvency instead of the usual 2 that we have by quarter. The difference is the Hurricane Ian.

On the market impact, so we have the benefit of higher interest rates, 3 points. Some of you might think that it's lower than expected. Keep in mind that obviously interest rates went up significantly in euro and in dollars, but much less so in Swiss franc and Japanese yen. So if you take the weighted average of all this that leads you to the plus 3 points that we had this quarter. Equity markets were down and therefore, we lost 2 points, and implied volatility on interest rates, was very high at the end of Q3, and that cost us 5 points of Solvency.

And last, debt redemption. So you saw that we bought back around 600 million of subordinated debt this quarter and we refinanced that through senior debt. And therefore, the net is a reduction of 2 points of our solvency ratio.

Two more things to stay on solvency, the first one is given the higher interest rates, you can expect a lower sensitivity to interest rates in our solvency. And the second, there is a change in interpretation by the regulatory community in Europe on subordinated debt that sits not at the top holdco level, but at local level. And so the change is such that you lose the benefits of local debt in your consolidated solvency. So that would be in force at the end of the year. For us, the impact relates to sub debt that we had at AXA XL level that we found in the -- at the moment of the acquisition. That's roughly 2 points of solvency that we will lose at full year.

And then last slide before we move to your questions, it's on Ian. So as I said at the beginning, for the time being, we estimate that Ian will cost 60 billion to the industry which makes it the second costliest hurricane ever, Katrina being the largest, obviously, in current dollar terms. What you see on this slide is that the cost for us is, as I said, 400 million in round numbers. Net of reinsurance, gross of tax, a 0.7% market share.

Irma 5 years ago, we had a 2.6% market share. So we reduced our market share by 70% and that's what we have been telling you of what we were doing on our portfolio at XL Insurance and XL Reinsurance, notably this year. So as I said at the beginning, it's a good illustration of the efforts that we've made.

So what does it mean for our cat load for this year? At half year, we were slightly above half of our annual cat flow. With Ian and obviously, depending on what we'll see at Q4, but most probably, we'll be above our annual cat load. But we believe that we will be able to mitigate most of that deviation, thanks to good news that we have on other fronts such as better investment income or better frequency.

And the last message on nat cat is about 2023. Several things. The first one is on XL Re. XL Re's strategy will be next year like this year to have even better balance between cat and other business lines. What it means is that they will further reduce their cat exposure in '23 compared to '22.

The second thing is when you look at our insurance business and its exposure to nat cat. Now a few things to take into account. One, when you look at the effect of inflation, the simple fact that you have inflation for a given return period, your cat load increases with the amount of inflation and so does any deviation. Second, there is clearly a hard market now on reinsurance for U.S. periods.

And there are tensions probably not yet to a hard market in Europe, but there are tensions also in Europe, which means that capacity will be more expensive.

So if I combine the 2, it means that on our insurance business, our cat exposure will grow mostly because of inflation. But overall, the aim that we have for next year is to compensate, thanks to XL Re's reduction in nat cat exposure, the increase that we will have on the insurance side. So net-net, we aim for a constant cat exposure in '23 at group level compared to '22. Obviously, we need and we'll tell you about this in February when we have finalized the renewals in terms of reinsurance, but I think that's an important message for '23 as well.

So overall, for me, the key takeaways, good growth overall, notably in our preferred lines of business, good momentum in pricing. And on nat cat, I think we have demonstrated through those numbers, the efforts that we have accomplished over the last 2 to 3 years.

And with that, I think we can move to your questions.

[Anu Venkataraman, Head, IR]

We will move to Q&A. I'm going to point to the person, so the person with the mic can reach you. I request that you focus this Q&A session on 9 months. So maybe we can start with Ashik at the back.

[Ashik Musaddi, Morgan Stanley]

This is Ashik Musaddi from Morgan Stanley. Just a couple of questions I have is, first of all, is it possible to get some color on the lapsed momentum, especially in France and Italy, given that interest rates have increased significantly? I mean are you seeing any pressure on the lapses on the back book? And secondly is any color on the -- I mean, clearly, the pricing momentum in Personal lines is getting better. I guess, 9 months, you showed 3%, 4%, but the discrete quarter, you should say like 5%. But how does it compare to versus inflation? Are there any particular geographies where you would say that inflation is still not yet covered where we might see some impact on the combined ratio?

[Alban de Mailly Nesle, Group CFO]

Well, thank you for your 2 questions. On lapses, for the time being, when you look at our main portfolio, which is the one you mentioned, which is AXA France, what you see on the general account is 2 things. One, we lost a couple of corporate contracts that we were not unhappy to lose, to put it that way. And second, when you look at lapses on the retail side, they're not lapses, they are transferred to unit-linked. And then when you look at real lapses on retail, you don't see anything.

So there's no change for the time being compared to '21 or the other years. Now on your second question on price increases, I would say, in most of our -- and your question was on retail specifically, I would say in most geographies, we're fine. I would put as an exception in the U.K. where you see that there is a good momentum, but you need more of it. And then, as I mentioned, in France, that's a tipping point, where before that, we could offset inflation through the track frequency. Now it's -- you need price increases.

There is no danger to our combined ratio in France, to be clear. That's -- but I wanted to highlight that tipping point. The other way to put it, you may remember that in the half year, I told you that compared to 2019, we had 7% of better frequency and 5% higher severity or inflation. Now those numbers would be 9% on higher severity and 8% on better frequency for the whole of Europe and France. So you see that inflation or severity, whatever you call it, is now slightly above the frequency benefit.

[Anu Venkataraman, Head, IR]

Okay, maybe Dom, in the middle?

[Dominic O'Mahony]

Dominic O'Mahony, BNP Paribas. A couple of questions on nat cat, if that's all right. The first is given the very significant shift out of reduction in exposure, should we be thinking about recalibrating our expectation for the combined ratio stack to the budget, and in particular, should we think that if the cat budget reduces, should the attrition will be going up or actually, are we actually thinking that there's a permanent sort of change up or down in the profitability of that business? Second is on the 0.7% share, which is great, is that representative of your ongoing North America hurricane exposure whether that you'd particularly underweight or you'd overweight Florida? So if there's a nasty one in a few weeks' time, would that be a good proxy? And then last question on health. It continues to grow very strongly. I wonder if you could just help us unpick the sustained drivers of health. So market growth versus market share growth in particular.

[Alban de Mailly Nesle, Group CFO]

On nat cat, as I said, we want to maintain our cat load in euro terms flat. And so what this means, given that, that we have price increases is that in percentage terms, cat load should diminish. On the attritional, no impact. I mean we don't -- I mean, we're taking measures through price increases to offset the impact of inflation, but there is no reason why attritional loss ratio should be impacted positively or negatively as such.

On the 0.7% market share, I think -- I mean, obviously, there can be some variations in the U.S. in our reinsurance business is exposed to Retail line writers and Commercial line writers. So I think it's probably the same in all places. But we are also in Commercial lines. So depending on where the hurricane would go, if it's more of an industrial area or retail one or it would have slight differences, but I don't think that would be massive.

Where you're right is that we significantly decreased our exposure to Florida, specifically, and more generally to the Gulf of Mexico over the last years. And on Health, several aspects. We've gained market share when we reviewed that overall. We gained sort of 0.5% market share over the last years. So there is growth there. The second effect is you have inflation on health, which has been there for some time between 3% and 5%.

In May jurisdiction. So there is that aspect. And then there is the fact that in France, in particular, there's specific development of their international business, but you shouldn't expect that growth rate going forward.

[Anu Venkataraman, Head, IR]

Andrew?

[Andrew Crean, Autonomous]

This is Andrew Crean of Autonomous. Three questions. What was the gross loss from Ian? Secondly, what is the nat cat low for '23 as a percentage of premiums? And thirdly, AIG showed a substantial reserving exercise for D&O yesterday, you're very.

I think you're a market leader in D&O. Should we be concerned?

[Alban de Mailly Nesle, Group CFO]

So on the gross loss for Ian, it's 670 million -- I'm sorry, it's dollars, but these days, it doesn't make much of a difference. The nat cat load, so that's always the same. I gave you a bit of information and obviously you want more, so I am not able today because we have not completed everything to tell you, but the nat cat load in percentage terms would be. I can tell you what the intention is in euro terms and that's what I said a minute ago, but not more than this at this stage. And the third one on D&O, so you're right, D&O, we are probably the leader in the U.S., which means that we look at this book very, very, very often.

And we are not concerned at all by the amount of reserves that we hold on that book. And you obviously have in mind that someone else looked at it when we booked in ADC, and it was not an issue. If you needed some third-party view on this. So we are comfortable because, again, we are a leader and we want to make sure that it's a profitable book, but you had that implicit second opinion by the reinsurer that provided the ADC.

[Anu Venkataraman, Head, IR]

Let's go back here. Will Hawkins.

[William Hawkins, KBW]

William Hawkins from KBW. Alban, you've given good guidance about the nat cat impact at the group level. would you also go so far as to say that you think you can continue to grow earnings in XL this year as you were kind of talking about at the beginning of the year? So just nat cat exposure specifically XL. And then secondly, could you just give us a little bit more color on these kind of small and midsized events that have tipped you above after Hurricane Ian?

A little bit about what are they? I think you're specifically talking about nat cat, but is there any man-made that you're also having in mind? And is there any risk here that you're talking about large losses, but really, this is just another form of claims inflation hitting you in the back door, things that would have been quite small are actually adding up?

[Alban de Mailly Nesle, Group CFO]

XL earnings, you saw the 668 million of earnings at XL that we had at half year and the underlying trend. And that trend is still the same for the second half, but there will be the impact of Ian. So on a normalized basis, yes, XL would contribute to growth in earnings in '22. On the small and medium-sized events, what I had in mind were the -- some -- so there's no man made into this. It's purely some hail events or drought or flood that we had mainly in Europe, but to which -- and in France in particular, but to which XL is also exposed through its insurance business notably.

[Anu Venkataraman, Head, IR]

Farooq.

[Farooq Hanif, JPMorgan]

Farooq Hanif from JPMorgan. Just to clarify on the popular nat cat topic a little bit more. Could you put -- maybe put it another way, how much of your nat cat budget roughly has been consumed by these other events in the second half, apart from Ian? Maybe that's a good way for us to think about it so that we can work out that delta. Because of course, you've had those events that you talked about, but it would be just interesting to know.

And then secondly, on your vol impact in Solvency II, so I remember you said previously that if vol remained elevated, there will still be a negative impact. Can you just talk about what happens at vol remains where it is or goes down? What -- how should we think about that modeling terms?

[Alban de Mailly Nesle, Group CFO]

So on the first one, if you allow me, I will answer the other way around. When I showed the Solvency II numbers, you saw that operating return was in line, except for Ian. So that gives you an idea about the rest of the natural events. And on volatility, it's very much a mark-to-market. If volatility goes down again, we will regain those points of solvency that we lost.

[Anu Venkataraman, Head, IR]

Michael?

[Michael Huttner, Berenberg]

A, the cat load in euros and maybe you can specify by business; b, the pricing you'd expect now coming up to set inflation in motor in France and Germany; third, I think Hanover today released on COVID. Are you going to release the rest of what you've got, which I think is 400 million-or-something? And then the last one is the Spanish deal is fantastic. You say as an example for others is that like a pipeline you can kind of, I don't know, hit that?

[Alban de Mailly Nesle, Group CFO]

So on cat load. This year, our cat load in euros were around 1.9 billion for the whole group. And that's the way we think about it, as you saw, I mean, they are offsetting impacts and that's the way we construct our cat budget more than saying this one should grow specifically on this -- that's the way we think about it. Second, pricing in Motor. I think it depends really country by country.

As I said, what you see probably in the U.K., for instance, is around 7%. It's the same in Spain. It's 0 in Switzerland. It should be around 3 or 4 discrete in France, either at the end of this year or beginning of next.

The third question on your -- the release of COVID provision. We -- as you know, we do our reserve review twice a year. So we will be able to tell you that when -- I mean, in February, fundamentally. And the last one -- I'm sorry, I can't read my own writing. Yes, the pipeline, that's the word what I wrote, sorry.

So we want to be extremely selective. There aren't many potential transactions on the market in general, but we have very strict criteria. So even if we would like to do more in line with Spanish one, don't expect the regular flow.

[Anu Venkataraman, Head, IR]

Thomas?

[Thomas Fossard, HSBC]

Thomas Fossard from HSBC. Two questions. The first one on the 0.4 Ian losses, can you split it between primary and Re, please, if you can? And the second question would be on the Commercial lines. Can you talk about the persistency rate because we've seen some initiatives in the market from corporates to build up captives, but it seems to be that there is a kind of pushback on the price increase or indexations?

So any early sign that actually there is a -- could be a persistency issue?

[Alban de Mailly Nesle, Group CFO]

So the 0.4 billion split is put half, half between insurance and reinsurance. On persistency, so you're absolutely right, and that's a phenomenon that we've seen for XL customers, not for the GI. It doesn't impact SMEs or midsized corporates. What we've seen is a number of corporate customers that reduce their cover -- reduce the cover they buy. So they might accept a higher attachment point with lower capacity or the -- or they create a captive.

And that's something that we've seen probably since the beginning of the year. And that's one of the reasons why overall at XL, you don't see a lot of revenue increase despite price increases because one of the offsetting factors is the exposure and the volumes that we use because of that.

[Andy Sinclair, Bank of America]

Andy Sinclair from Bank of America. A couple left for me. Firstly, just on Commercial lines pricing renewals within XL, just looking at the 7.9% for 9 months versus 9.5% for H1, looks like a bit of a slowdown in Q3 stand-alone. Just wondered if you can give us any color there. And secondly, just on managing nat cat exposure going forward. Just your thoughts for purchased reinsurance program coming into 2023.

[Alban de Mailly Nesle, Group CFO]

Okay. So the first one, what do we see at XL, there is one line of business, which is U.S. professional or D&O put simply, where you see some price reduction. And you've seen that over the last 2 quarters, and that is obviously a significant one as we discussed a moment ago, but an extremely profitable one with one way in which we can afford to have lower prices. When you exclude that line, what you see in terms of price increases at XL is plus 11% for 9 months and plus 9% for Q3 discrete that you still see that in all other lines, you have significant price increases even when you look at Q3 only.

On nat cat purchase reinsurance, so that's something that, obviously, we are negotiating these days and until the end of December. Overall, as I said, we want to maintain our cat load on a net basis. Part of it will be achieved by a reduction at XL Re. As far as reinsurance purchases are concerned. As much as possible, we want to stick to our current attachment points, but obviously, they will need to reflect the fact that everything being equal, there is inflation. And therefore, if you insure a building, that building is worth more the cat, it was more. And so for a given written period, the attachment point in euros will be higher.

[Anu Venkataraman, Head, IR]

Peter, at the back?

[Peter Eliot, Kepler Cheuvreux]

First one, if I can just come back on the earnings at XL. I'm probably missing something obvious, but you mentioned 668 million for the half year just above the sort of nat cat load at that point. So the full year, we're looking probably at being just above the annualized amount. So I'm just wondering why we can't double that. You seem to suggest that, yes, you are in line with 1.2 on a normalized basis, but not on a reported basis.

I just -- sorry, I just wanted to clarify that. And second point, it may be too early, but I'm just wondering what you're seeing, if anything, in terms of the pressure on people spending sort of discretionary spending as the cost of living that goes up. Is there any sign that people are being a little bit more disconcerning on what they might purchase?

[Alban de Mailly Nesle, Group CFO]

I don't understand that question. What are you referring to in particular?

[Peter Eliot, Kepler Cheuvreux]

Just generally in terms of where people have a choice over buying insurance on the Retail lines and people being -- yes. Third question, sorry. At the half year, obviously, you gave an estimate of the Ukraine cost. You haven't mentioned anything today, so I assume nothing's changed, but can we assume that your thinking on that hasn't change at all.

[Alban de Mailly Nesle, Group CFO]

So taking them in the reverse order. In Ukraine, we will -- for Ukraine, we will revisit that at the end of the year. That's part of our reserve review taking into account. You remember that we built different scenarios to assess what better thing it could be. We need to see the probability that each of those scenarios developed and we'll just if needed, are reserved for Ukraine. On Retail lines, no. For attending, we don't see a shift to lower guarantees, price reductions and so on.

There's slightly -- I mean in some countries where we have a direct insurer, which is low cost and a traditional insurer. I mean, I'm thinking, for instance, of Belgium or France, you do see more of an increase in the direct insurer, which is low cost than in the traditional one, but it's not massive at this stage.

And in terms of XL earnings, what I mean is, I mean, it's still 400 million for XL. And you have, as I said earlier, also natural events. So that's what I mean is we have the trend. We have smaller natural events, but we also had Ian. We can probably compensate even at XL level a good part of Ian as a deviation, probably not all of it.

[Anu Venkataraman, Head, IR]

Iain?

[Iain Pearce, Aviva UK]

Just a quick question on nat cat exposure and particularly into 2023. When the nat cat exposure reduction started, you weren't thinking about how markets in the U.S. towers in particular. So clearly, the situation has changed quite a lot. So I'm just wondering why the strategy hasn't evolved.

Do you just not want this risk at all full stop? Is it the volatility you're not willing to take regardless of price? Or do you think things like climate change are sort of not being priced appropriately still? And then when we're thinking about the absolute exposure in euro terms remaining flat, so is the right way to think about that net of inflation, that's actually quite a big exposure reduction?

[Alban de Mailly Nesle, Group CFO]

On your second question, fundamentally, yes. if we manage to maintain that exposure in constant euros, it is a reduction when it is deflated by inflation. On your first question, I think -- and you know that the financial profile that we want to have is one of the company that grows regularly its earnings with natural volatility. So I agree with you, prices when you sell reinsurance are significantly better now than 2 years ago. That being said, there's still a lot of volatility. And as, in any case, we grow our exposure in our insurance businesses, we don't want to grow it in reinsurance. But XL as such, when they look at their own business, they want to have a more balanced portfolio as well. They were very much skewed towards nat cat when we bought them 4 years ago.

[Anu Venkataraman, Head, IR]

Will?

[Unidentified Analyst]

I guess the comments on the Personal Motor severity and inflation were helpful for thinking about forward trajectory on accident year combined ratios. I guess is there anything we need to consider on inflationary trends for Personal Motor reserves, therefore, on headline margin. A quick one. Is there any color you can give on reinvestment yields as of now versus H1? And finally, just to comment on your D&O, given what's happening to reinvestment yields, it sounds like you're happy with the starting point on margin there. I guess, shouldn't we be happy enough that there's a bit of pricing pressure there? What's the average duration on that portfolio that we need to consider?

[Alban de Mailly Nesle, Group CFO]

So if I start with the reinvestment yield, let me find it. Yes, that's the numbers for September. We reinvest in U.S. dollars at 5.7%. We reinvest in euros at 3.6% these days. In Japanese yen, it's 1.3% and in Swiss franc at 1.7%. So the weighted average is 3.1%. So that's the current -- the most recent reinvestment rates that we have. So when we look at D&O, if I understand your question is, are there price decreases because we write more and therefore have more investment income? Is that your question?

[Unidentified Analyst]

[Indiscernible]

[Alban de Mailly Nesle, Group CFO]

Yes, for the time being, that's not what we see in the market. For the time being, the market, we believe, is disciplined and that not we inject higher investment income, notably in dollars, into pricing. D&O is a very specific case where there had been very significant price increase over the last 3 years to a point where the business was really profitable and -- as I said, the fact that we reduced prices doesn't make it unprofitable far from it. But it's not linked to investment income as much as we see for the time being.

[Unidentified Analyst]

[Indiscernible]

[Alban de Mailly Nesle, Group CFO]

We would need to look, but I would probably say 5 or 6 years, something like this.

[Anu Venkataraman, Head, IR]

We'll get back to you.

[Alban de Mailly Nesle, Group CFO]

And on Motor reserves, we don't expect particular impacts on this. In -- I mean, it's a European business, right? We don't write Motor in XL. And when you talk of inflation, it's about bodily injuries. And we don't see inflation on bodily injuries going forward. And even in some countries, even in Spain, we have the [indiscernible] which sets very clearly how much it is. It has increased a little bit recently, but that's it.

Overall, we'll review our results for full year, and we'll see whether we need to increase our reserves or if like at half year. I'm talking about the whole group, we are satisfied with the overall amount of reserves we hold.

[Anu Venkataraman, Head, IR]

Charles?

[Unidentified Analyst]

Can you talk a little bit about Asia because it's not that long ago that Asia has seen a big opportunity and a big potential growth market. But obviously, the expense in Hong Kong and some has been very difficult in the recent past. Has there been any change of thinking or strategy in there?

[Alban de Mailly Nesle, Group CFO]

So Asia, as you would know, is, in fact, different realities from country to country. Japan, it's an aging country where you have growth and it hasn't changed. And we're very happy with our product mix and we grow and notably in protection. And we fund the right products, and we keep inventing products such as recently protection with unit linked with a significant success. So it's a growing business.

When you look at Hong Kong, it's a different reality. The Mainland Chinese visitors have not come back. The channel is open for Macau, but not for Hong Kong itself. So that's still the same as before. We have focused on Health in Hong Kong, but everybody has a competitive market. We're still growing and it's fine, but nevertheless, it's competitive. In China, on the P&C and Health side, it's on the health side, we have repositioned the business. We had a digital partnership which was not profitable. We stopped it. We moved to distributions with branches, and that's quite successful.

We significantly improved and increased production there. But on the Motor side, it's a difficult market, a highly regulated one, where every time you have some profitability because prices go up, there is price tension and profitability goes down again. So a difficult market. And on the Philippines, Thailand and Indonesia, it's well positioned. We are growing both generally, both on -- with our bank insurance partnerships and with the agent -- agency distribution that we have locally.

So we're happy with those 3 countries.

[Anu Venkataraman, Head, IR]

Any final questions? Thomas?

[Thomas Fossard, HSBC]

And just staying on the topic of Asia, we heard this morning about Accident & Health clients pick up in Taiwan, specifically on the retail side. Anything you may have exposure to?

[Alban de Mailly Nesle, Group CFO]

No. I mean as you know, we're not in Taiwan, so no.

[Anu Venkataraman, Head, IR]

Michael.

[Michael Huttner, Berenberg]

This really is fantastic to have this level of detail.

[Alban de Mailly Nesle, Group CFO]

Do you have any questions?

[Anu Venkataraman, Head, IR]

Thank you, Michael. Let's just move to the gentleman behind Thomas.

[Unidentified Analyst]

Anthony from Goldman Sachs. Just one question on life and savings. On the back book disposal, can you give some color on what's the market appetite given the volatile backdrop?

[Alban de Mailly Nesle, Group CFO]

So you're also giving me the opportunity to comment on recent developments, but again on the fact that we have closed the transaction on our Belgium book that we had announced. In terms of appetite, it's still the same. The professional buyers that we have in front of us are still very much interested in growing their books. So as you saw, we tackled already some countries like Belgium and Germany. Now we're moving to other countries such as France, Switzerland, Italy, but appetite is there. Nothing has changed.

[Anu Venkataraman, Head, IR]

There are no further questions, then we'll stop for a brief break before we reconvene in this room for IFRS 17.

[Alban de Mailly Nesle, Group CFO]

At 3:15, please.

[Anu Venkataraman, Head, IR]

At 3:15.

[Alban de Mailly Nesle, Group CFO]

Thank you very much.

[Break]

[Anu Venkataraman, Head, IR]

Well, welcome back to our IFRS 17 presentation. Before I turn over to Alban and to Grégoire, a few housekeeping items. So let me briefly summarize for you where we stand in the IFRS 17 implementation process and also highlight for you a few things to keep in mind as we go through this presentation and for our financial reporting in the coming months. To start with, as you know, we currently report and manage under IFRS 4. Our full year '22 results when they will be published in February of next year will be on IFRS 4 basis.

At the start of 2023, IFRS 17 and IFRS 9 will replace IFRS 4 and IAS 39. As a result, our formal reporting under IFRS 17 will not start until the beginning of next year. Therefore, any measurements and estimates that we shared throughout this presentation are preliminary and subject to change. They're also unaudited. While we're working with our auditors to implement the new accounting standards, we thought we'd share with you a few numbers in order to make this a more useful and, dare I say, lively interaction. We do believe that the numbers that we're sharing with you will be close enough to the final outcome. But nonetheless, please let me point you to the disclaimers on Slide 2, which you can expand on your devices for easier reading. And let me also point you to the calendar that we have shared on Slide 5. That will show you exactly which accounting standards our financial reports will be based on for the upcoming few quarters.

And with that, let me actually reintroduce you to our 2 speakers. So most of you are familiar with our group CFO, Alban de Mailly Nesle. Joining him to present the impact of the accounting standard is our Group Chief Accounting and Reporting Officer, Grégoire de Montchalin.

And with that, let me hand over Alban.

[Alban de Mailly Nesle, Group CFO]

Thank you, Anu. Hello, again, and I'm pleased to share the floor with Grégoire for this highly technical subject. So the first thing to say on IFRS 17 is that we believe it's good news for the industry in the sense that it will lead to more convergence between insurance companies on the way they book their reserves and, therefore, release their earnings. And I think we are in an industry which is sometimes difficult to understand from the outside. And therefore, this is a welcome change, I believe.

So for us, what are the main messages that we want to give to you today? The first one is on our underlying earnings power, and very importantly, it remains unaffected between IFRS 4 and the transition to IFRS 17. And you will see that a number of times in the presentation. There will be continuity in our earnings capacity.

To the same extent, there will be -- I mean our shareholders' equity will be roughly at the same -- or broadly at the same level under IFRS 17 as they are today under IFRS 4. I said broadly, you will see in the presentation that it's exactly the same amount, but as Anu said, there are still potential adjustments, and we want to be cautious along the whole presentation on the number that we show you. But there again, there is stability.

So that's on the results, and then there is the how, which is the reporting. And on the reporting, you will see, in fact, little changes on our P&C and on our Health business and obviously, no change in our Asset Management business because it's not affected by IFRS 17. And those lines of business, P&C and Health, represent the majority of our revenues, and therefore, there would be limited reporting impacts overall, and they will focus mostly on long-term business and their full life.

The fourth important message is that it is an accounting change and only an accounting change. So it does not have any impact on our cash generation, our capital management policy nor our strategy. And that's extremely important. It is just, so to speak, the way we report earnings at consolidated level and not, for instance, in statutory accounts locally, and it does not impact Solvency II either. So that's extremely important.

And so given that there is continuity in our level of earnings and given that capital and cash are not affected, we are in a nice position, which is that we are able today to say that our 2023 targets are reaffirmed. I mean the 4 ones that you know that are about earnings and our treasury.

So going into a bit more details and starting with our earnings. The -- let's look first at the results because obviously IFRS 17 is about reserves. And you will see in the whole presentation that we discussed IFRS 17 because that's where the main impact is and much less -- IFRS 9 has much less impact for us.

So what are the changes -- the main changes brought by IFRS 17? The first one, importantly, is that reserves will be booked on the best estimate basis. Second, they will be discounted. And therefore, when you say best estimate and discounted, they will be extremely similar, extremely close to Solvency II best estimate liabilities.

Second, there will be a reserve called risk adjustment that will come on top of reserves. On top of best estimate liabilities to cater for uncertainty on nonfinancial risks. We will discuss more at length risk adjustment later. And you will see that, in fact, it has little bearing on our earnings.

And third, and that's really where the main change is because it's on the life side. We have the creation of a stock of future profits, which is the contractual service margin, both on the life side and on the life-like Health business that we have, so as far as we are concerned, Japan and Germany. Those are the main changes.

So what does it mean when you take the helicopter view? P&C earnings will therefore be a bit more sensitive to interest rates because, with movement in interest rates from one year to the other, the discount that you will apply to claims in the current year will vary. The second aspect is that CSM and the CSM release will be the main driver of earnings on the life side going forward. And you will see later in the presentation that there is not complete stability, but there is a good degree of stability in terms of amount of CSM released year after year. And that will improve the predictability of our earnings on the life side.

Health business, depending on the country and the nature of the business, it will be like P&C, so typically what we have in France, in the U.K. or Mexico of like life, as I said, Germany and Japan.

And finally, Asset Management and holding, by construction, are not affected by the change in accounting standards. And overall, when you look at our earnings capacity in aggregate for the whole group, stability, as I said, between IFRS 4 and IFRS 17.

So that's for the results. For the reporting side, what we want to highlight is the following. On the P&C side, in terms of reporting, you will see little change in the main KPIs. We will keep on reporting on the basis of gross earned premiums. And normally, you should see the industry move to that.

Some others use, under IFRS 4, net earned premiums as the denominator of the combined ratio. I believe they -- we should see some move to gross earned premiums and, therefore, make our combined ratios more and more comparable.

Second, you will still see, in terms of technical profitability, the current year loss ratio and the prior year. That will not change in terms of concept. You will see, as I said, the amounts themselves will be affected by the discount, but you will see currently and prior year loss ratio.

On the life side, the -- as I said before, the main driver of earnings is CSM. So that means that we will give you more details on the CSM: what we are thinking of is the new business CSM, the roll forward from the beginning of the year to the end of the year of the stock of CSM and also the fact that we will be able to give you some indication of how the stock of CSM will be released across the years in the future, again, giving you more predictability on our earnings. I take also that opportunity to say that more globally with Anu, we are looking to improve our disclosures. And we're happy to get your feedback on what you would like to see. And just a warning, not everything is possible. But on what you would like to see in our disclosure, that's the moment when as we move to IFRS 17.

Health, I want to come back to that, is a mix of P&C and life and Asset Management in terms of reporting will not be affected at all.

Earnings power, unaffected post transition. When we say earnings power, we mean underlying earnings at 17. The difference between the 2 that underlying earnings are after tax. But -- so unaffected.

Between underlying earnings and net income, you have a number of things that I will describe later on in the presentation. We don't expect that part of our P&L to be more volatile under IFRS 17 than it is to our balance sheet. Shareholders' equity will be stable at EUR 58 billion. That's the amount excluding OCI, and that's very important. And that's something that we were -- that we have worked for in the sense that -- and Grégoire will mention that there are a number of options that we have taken, allowed by IFRS 17.

And one of the principles was to have continuity as much as possible for capital. And as I said, there will be the CSM. It will be EUR 34 billion. So that gives you an idea of the amount of profit going forward, and you will see later in the presentation by how much that should be released every year. And the EUR 34 billion is gross of tax.

And so just to reiterate, no impact on capital and cash generation, same balance sheet strength in terms of Solvency II, and it's important because, as I said, group level and at local level, no impact on solvency. Therefore, no impact on capital management policy and, therefore, no impact on business strategy. And so key targets are reaffirmed.

And with that, I'm happy to leave the floor to Grégoire.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

So thank you. So it's, first of all, a real pleasure for me to be here with you all and to present IFRS 17. Indeed, I but also many teams at AXA have been all working for, I think, many years on this topic. So that's the day where we first presented to you. We are clearly at the footholds between IFRS 4 and IFRS 17.

And I think it's fair to say that it's definitely an exciting time for the company and, I think, most generally for the industry. And you know IFRS 17 started already long ago. We've come a long way. And I believe really the standard will bring some features that investors have been hoping for, for a long time.

So let me first remind you on top of everything that Alban already introduced that the accounting philosophies on which IFRS 17 was built is, in a sense, quite similar to the one of Solvency II. And we introduced to make sure that our financial statements reflect the economic reality of our business over the long term.

So he gave us some guiding principles. Our main goal with the introduction of the new standard, as Alban already said, was to maintain continuity, especially on both capital and earnings, and to limit any additional volatility that will come from the new standard. When I say continuity, indeed, you can get it from different angles. It's continuity of methodologies because, wherever it was possible, we reused the methodologies already developed for [72]. It's continuity in reporting as already evidenced by Alban, and its continuity -- and that's even more important, obviously, of our key financials.

So earnings and capital, meaning more possibly that we will have largely stable shareholders' equity and transition. And we also expect, as already mentioned, our earnings power to be unaffected in aggregate.

So moving on to the following slide. The first aspect to consider is that AXA's businesses will be largely accounted for under what is a simplified model under IFRS 17. So as you know, there are 3 measurement models under IFRS 17 that depend on the features of the business. So the short-term technical businesses will be accounted for under the simplified model, which we call the PAA model. The long-term technical business will be under the general model, which is called the BBA model, and the long-term participating businesses, including all the Unit-Linked business, by the way, will be under the third model, which is called the VFA model.

And important to note is that, as a result of our strategic decision a few years ago that you know very well to rebalance AXA towards technical risks, with our primary focus now being P&C, Health and Protection, we see that 65%, and that's what is shown on the slide, of our business will be under the simplified model or even fully unaffected in the case of Asset Management.

So on the contrary, there are the 2 other models: BBA and VFA, which are the ones on which there is the introduction of the main new mechanism brought by IFRS 17, which is the CSM mechanism. But these 2 businesses only represent -- BBA and VFA represents only 35% of our overall activities, as you see. So by and large, approximately 2/3 that's within the strategy of our business will be largely unaffected by the move to IFRS 17.

So let me now, with this slightly more busy slides, walk you through the main choices that we made to implement IFRS 17. The preliminary remark when making those choices and as you will see in the following 5 examples, our guiding principle has been to have -- well, first, a consistent approach across the group because, obviously, that's important for the consolidated accounts as well as, as already mentioned, ensuring continuity notably in our earnings power and limiting volatility.

So if I look quickly at the 5 items disclosed on the slide, starting with 1 of the main changes introduced by IFRS 17, which is the discounting from now on of all reserves. So we had the choice between the top-down or bottom-up approach when it comes to building the discount rate curve, as you know. And we intend to use the bottom-up approach, which is the one that starts from the risk-free rates and adds up a premium for illiquidity, very similar to Solvency II. And as a matter of fact, the curve that we will use will be very close to the one that you know under Solvency II. And by the way, this will bring also some efficiency because having this proximity will allow us to leverage similar inputs between the 2 frameworks obviously.

The second item, still linked to discount rates but relating to the changes in this contract. So as you know, the framework requires with us to be discounted at that market-consistent rate. So we must revalue our reserves at each closing date using the latest market rates. So this is -- this brings potential volatility, obviously. We elected to account for those changes through OCI.

And so for OCI, whenever -- we have the choice between having them through P&L or OCI. And let me zoom a second on the OCI under IFRS 17. So you're all familiar with the asset OCI that exists under IFRS 4. And as you know, one of the main novelties of IFRS 17 is that it brings a liability OCI, which is conceptually and de facto counterparts and the balance to the asset OCI, and that will result in a net OCI that, first, will be structurally much closer to 0 at the scale of the group and also much more stable or much less volatile than the OCI that we had under IFRS 4, which, especially this year, as you well know, was very volatile.

So coming back to the choice that we made on the discount rates having the changes through OCI. Let's just keep the idea that this is something that we obviously need to mitigate the volatility in the P&L. And this specific volatility will be ring-fenced to the balance sheet.

Next item is the risk adjustment. So as you know, it's essentially an additional reserve sitting on top of our prudent, best estimate reserves. We will compute this adjustment with a percentile-based approach, and I propose to come back with a bit more detail on it in one of the next slides.

Moving to the fourth item and this time on the asset side of the balance sheet. So for listed equities, we had, under IFRS 9 this time, an important choice on how to account for them. And we have chosen the fair value through OCI method that is offered by IFRS 9. So completely equities will continue to generate dividends, obviously, that will remain in the P&L as before. But the mark-to-market volatility that goes with the value of the equities, this one will now go only through the balance sheet and including, as you know, the fact that we will not get the realized gains and losses will not get recycled through the P&L.

But as a counterpart, this will be another element adding to the reduction of the volatility of our P&L, and that's the main reason why we made this choice. And finally, when it comes to the transition approach, so as you know, 3 approaches are possible under IFRS 17. And we have chosen the retrospective approach, so be it the full or the modified retrospective approaches, which has the 2 first of the 3 possible approaches, for about 80% of our businesses, meaning that we have adopted the third one, which is the fair value approach, only by exception and concretely only in order to limit the risk of having onerous contracts.

So overall, summarizing this a bit busy slide, our key accounting choices reflects, well, our guiding principles already announced on, one, ensuring continuity in earnings power; two, limiting the P&L volatility wherever possible; and three, ensuring the convergence to Solvency II, where it was feasible.

So I come now to detailing one of the key message of today, which is that our shareholders' equity will be -- are expected to remain broadly stable at around EUR 58 million. So here on the slide, you see the walk-through of our shareholders' equity, excluding OCI, from IFRS 4 to IFRS 17. And this is basically a 3-step process after a preliminary step, which is about eliminating the technical intangible assets, so meaning the DAC and the VBI, which are ultimately fully eliminated when we move to IFRS 17.

So once we've done that, the 3 steps are, first, that we revalue the liabilities on a best-estimate basis and by discounting all those technical reserves. Second, we create the risk adjustments. And finally, we create, again, as already announced, the CSM, which, as you know, consists of future profits appearing on the liability side, and that's the importance of our balance sheet and gradually released into our P&L.

So let me now zoom on the following slides into each of the -- those 3 steps. So the first item is about the revaluation of our reserves. And before we start on this slide, let me just remind you that reserving under IFRS 17 is, broadly speaking, very similar to what we do under Solvency II. We use the same model. We have very close technical parameters. And that includes, obviously, discounting all the reserves.

So here, I focus on commenting these discounts that's notably, from now on, will apply to all the P&C reserves. They were just a very small part of the P&C reserves that we have discounted until now. So that's probably one of the key changes as already highlighted by Alban. So as you know, discounting, well, first of all, and that's really important, has no impact on the lifetime profitability of any business which is discounted. But it's just about giving more economic value to the newly created reserves by discounting them, obviously. And then as long as we hold that reserve that was first discounted, there are basically 2 parallel accounting mechanisms. So the first accounting mechanism is that we unwind in the following years as long as we still hold the reserve, the discounts and we use the rate's locked-in inceptions, so the one that was used for the discounting in the first year, to calculate this unwind, which is the part that flows through P&L. And the second parallel mechanism is, as already explained, the fact that there will be an impact on the face value of the reserve of the volatility of interest rates around that rate that was locked in at inception, and this volatility will go through OCI, as I already mentioned, and will be ring-fenced to the balance sheet.

So this is the accounting mechanism for all our nonparticipating businesses. Just note that for the participating business, which is under the VFA model notably, there is a CSM mechanism. That's a slightly different mechanism, but that is basically the one that plays the same role to absorb also all the volatility linked to the movement in interest rates. So overall, as you see, well, the discounting mechanism is something which is fairly straightforward.

Next item, so if you remember the roll forward of our shareholders' equity from IFRS 4 to IFRS 17. So the second step that I already introduced is the risk adjustment. So as I already said -- so the risk adjustment is a new liability that is created and required under IFRS 17. This is an extra reserve sitting on top of the pure best estimate liabilities and which is conceptually meant to cover nonfinancial risk. I think you know all that.

So as already said, there is no prescription under IFRS 17 as regards how to build this risk adjustment, which shows a percentage-based approach to build it. At AXA, as you know, whenever we said best estimate liabilities, we nevertheless make sure to remain on the prudent side in the way we set our best estimate liabilities. And accordingly, we believe that it's appropriate to set the confidence interval for the risk adjustment that will be centered on the 65th percentile. More precisely or more technically, we will give us a range between the 62.5 and the 67.5 percentiles, within which the risk adjustment will actually be set. And from a number perspective, and I'm sure that's important to you, this is expected to represent between EUR 3 billion and EUR 4 billion in our balance sheet for the risk adjustments going from the bottom to the upper end of the range.

So moving now to the third and last step, which is building the newly created CSM. So as you know, it will act as a reserve for future profits, and the CSM applies basically to all long-term Life & Savings contracts as well as the long-term life-like, if you want, Health contracts.

So you can see, on the right hand of the slide, the typical mechanism of creation of CSM. So it's nothing more than the premiums that we received, netted of expected claims and costs, to which we add an extra buffer, which is the risk adjustment that we just discussed. And all these cash flows are obviously discounted. And this leaves the future value attached to the premium that we receive that we account for as a liability and as the contractual service margin attached to that given contract and the premium we receive for it.

So now if -- so that's for one single contract. So now, if we look at it from a stock perspective, we will have, in the opening balance sheet, the total amount of CSM of around EUR 34 billion gross of tax, corresponding to the estimate of 27 net of tax. And this EUR 34 billion is going to be split approximately EUR 25 billion for Life & Savings and EUR 9 billion for Health.

So if now we look at the evolution over time of the CSM, we expect it, and that's the key message here, to grow sustainably. So it doesn't mean that it cannot decrease in a given year. And by the way, the CSM will play as a really good volatility absorber. It will absorb at first all the technical volatility and, for all the VFA business on top of the technical volatility, the financial volatility impacting the valuation of the reserves. It will be first absorbed by the CSM.

And then if it's low volatility, it will disappear by itself, or if the impact is conserved over time, this will be gradually released through the P&L. So that's why it's really important for us to monitor the stock of CSM over the long run. Obviously, we'll disclose details on this roll forward.

And the reason why we believe, on average over the long period, it will grow sustainably is that we'll have the contribution every year of the new business CSM, which will be measured on a [recent roll] basis. We would expect them exactly like you see under Solvency II, the financial variance to be, on average over time, positive, but the difference this financial variance will reflect real-world returns versus CSM, which is first built on a risk-neutral basis. And we expect that the sum of those positive elements will more than offset the release of CSM that will feed the P&L in each and every year. So Alban will comment further on these mechanisms and what it means for our Life & Savings earnings in the next part of the presentation. But before moving to that profitability part of the presentation, so it's time now to take stock of everything I just explained by having a look at the resulting balance sheet. So this is what is presented on this slide.

If I look first on the asset side, so the key points to highlight here are the elimination of EUR 18 billion of technical intangibles, so namely the DAC and the VBI. On the contrary, the pure goodwill is fully unchanged by the move to IFRS 17. So obviously, the goodwill remains as it is before and after the move to IFRS 17.

On all the invested assets, I think a really key and important message is that overall, there is very limited changes attached to the -- to all those assets. And as I believe, the classification remains broadly unchanged. So that means that we still have, as before, most of our assets measured at fair value through OCI. And you may note as well that we will account as before. Before, it was for almost all the real estate, and now it's for absolutely all the real estate.

The real estate will be classified as cost in our IFRS 17 and 9 financials. If I look now on the liability side. So we find, again, consistently with everything that we just discussed, well, further remeasurements, notably through discounting of the reserves. So that leads to around EUR 530 billion of best estimate reserves. You will note that this is going to include a very small amount for onerous contracts that will correspond to less than 0.5% of our total best estimate liabilities.

You see the introduction of the risk adjustment, so a bit less than EUR 4 billion. You see the CSM, as announced, around EUR 34 billion, and we'll have, under IFRS 17, the new OCI, as I promised, much closer to 0. So technically, in the opening balance sheet, we expect it to be at minus EUR 3 billion.

So all in all, having summed all those elements, both on the asset and then on the liability side, this ends up with EUR 50 billion of shareholders' equity, which is a stable amount as compared to what we had under IFRS 4.

So with that, we've now established the impacts of IFRS 17 on the balance sheet. That was the main focus of this part. Stable shareholders' equity and also a much more economic representation on the face of the balance sheet. So we see all this as very positive.

On this last slide, before -- just for the sake of completeness, on the balance sheet-related items, let me just focus on 2 ratios, the return on equity and the debt gearing, that will obviously remain key to understanding our financial profile going forward. So regarding the ROE, we have a simple message here, which is no change. So the definition will remain unchanged, moving from 4 to 17. We still have the underlying earnings, and as you know, many of our peers refer to the same concept as operating earnings, but that's exactly the same concept, divided by the average shareholder equity. And since the shareholders' equity is stable in this transition, it makes perfect sense that we reaffirm and maintain the targets, which is the one that you know, a range between 13% and 15%.

Looking now at the debt gearing. Here, slightly different because we will amend the definition moving to IFRS 17. What we basically do is add the CSM net of tax to the denominator. And we believe it makes sense because the CSM is also a financial resource that can be used to meet our future financial debt obligations. Hence, it makes sense to report it to the amount of debt.

So this will, by construction, mechanically translate into a lower gearing ratio as compared to the one that we used until now. You have, on the slide, the new range to which the previous one would now correspond. But important is that despite this pure change of measurement, this has no impact, again, on our net debt issuance capacity of EUR 1 billion per annum at constant debt gearing, which is the one that you very well know.

So now just before handing back to Alban, maybe just quickly wrap up this part. So the transition to IFRS 17 resulted in a new EUR 34 billion item, CSM, on the balance sheet. And as you already understood, we welcome this change as it really better reflects the economic reality of our business over time. And so we've made accounting choices to implement IFRS 17, and that serve 2 main objectives that I just reiterate: first, the continuity of our accounts and, in particular, of our shareholders' equity, as I covered it in this section; and second, with the objective to keep a low level of volatility for our underlying earnings just as before.

So thanks for your attention, and I will, of course, be available for questions later, but I hand over back to Alban.

[Alban de Mailly Nesle, Group CFO]

Thank you, Grégoire. I'll try to be as clear as Grégoire on the P&L as he was on the balance sheet and the accounting options that we took.

So starting with the profitability by line of business. So the global picture, as we said, is that overall, in aggregate, the -- our earnings power will be unaffected. But this is also true broadly by line of business.

So in P&C, what you will see, and I will come back in detail on this in the coming slides, is, obviously, an improved combined ratio simply because -- or thanks to the discount effect that you will have in any given year. That will be offset by lower investment income because you will have the same growth investment income as today, but you will have the negative unwind of the discount rate.

There will be volatility in P&C. And importantly, as Grégoire also mentioned, our best estimate reserves will keep on having some degree of prudence. I mean our approach to reserving will not change significantly. As such, you can still expect positive prior year developments in line with our long-term average. As we said, Life & Savings will be predictable, and Health would be a mix of P&C and life.

So a bit more details on the P&C side. Starting with what does not change. Again, as Grégoire said it, the accounting change that is reflected by IFRS 17 does not change the lifetime profitability of the business or any given contract. It just changes the pattern of recognition of the earnings of that business. Second, we will carry on with our approach, which is to be prudent, both on underwriting and reserving.

And third, as I said in the beginning, and we will continue analyzing our profitability with the combined ratio on a growth basis. So what will be the change brought by IFRS 17? The current year claims will be discounted at inception, i.e., at rates that reflect the then-prevailing rates. The mark-to-market along the years of those changes in rates will go through OCI, will not go to P&L, as Grégoire said. And the unwind of those rates will go to P&L in the following years. So you have those 3 movements: discount in the current year at the then-prevailing rates, move over time of the mark-to-market in OCI and then unwind in the following years at the inception rate.

An important aspect is the risk adjustment. The net effect of the risk adjustment is expected to be really marginal. Because as Grégoire said, it will be within the range but within a relatively narrow range. Part of it will be released every year with the reserves themselves, but we will recreate immediately a risk adjustment with the new claims of the new year. So don't expect, on the long run, risk adjustment to contribute to earnings. And that's pretty important.

When you look at the combined ratio, obviously, the combined ratio will look better under IFRS 17 than under IFRS 4 for obvious reasons, thanks to the discount. It will be also, as I said and Grégoire also, a bit more volatile given the changes in interest rates from 1 year to another. And we will need to learn to live with that collectively because earnings will be earnings, and we need to take into account that volatility that, once again, does not impact the overall profitability of the business.

The combined ratio will be better than today. Investment results will be lower than today because, again, you will have the investment income as it is today, but you will have the unwind of the reserve discount. And the way we see it -- I mean once it stabilizes, as you invest every year, your investment income should reflect the various years of investment that you have. But so should also the unwind of the reserve discount. So there's a parallel track for the investment income and the unwind.

If we invest regularly, which we do, the difference between the 2 is that the unwind reflects the rate at which it was discounted, which is a risk-neutral rate or risk-free rate, whereas the investment income, obviously, is a real-world rate.

If we move now to life. So as we said several times, the life of earnings will be driven fundamentally by the release of CSM. And as opposed to what we showed on the P&C side, it's a completely different world, and that's why we don't show a bridge between IFRS 4 and IFRS 17. It wouldn't make sense. So as Grégoire said, there will be some volatility in the CSM itself. But very importantly, the CSM release will be reasonably stable. And why is it stable? Simply because it is a function of long-term assumptions that do not vary significantly from 1 year to the other.

The CSM is created on a risk-neutral basis, but its unwind is a function of real-world assumptions, like in Solvency II, where you have that financial variance year after year. That's why here, we say that the range of release of the CSM is between 9% and 11%. It's not because the CSM as a stock -- sorry, what I mean is it's to offset the volatility of the CSM. In other words, when the CSM is down, as the amount of CSM released will be relatively stable, it means that the percentage of CSM released will be higher, close to the 11%. When the CSM is up conversely, the CSM released, again, is still reasonably stable, and the percentage released will be closer to 9%.

So that's -- and that's the percentages that we will have on the life side. That is extremely important because, again, we see IFRS 17 as a mechanism that brings a lot of predictability on the life side. And that's why we will be able to give you a good indication on how the CSM will be released in the future. Again, we're not saying that it's to the euro, but we're saying that it's -- it will be reasonably stable.

So that's the CSM, main driver of our profits on the life side. You will also have investment results on the other life businesses that are not under VFA, so the nonparticipating businesses and the shareholders' fund. And you will also have the technical adjustments for the small part of the life business, which is PA.

So if I sum up, it will be effectively the line which is the most affected by the change. But the good thing about it is the overall predictability of the release that you will see year after year. A word on new business value and new business margin. Obviously, with IFRS 17, you will have a new business, CSM, that we will report. But on top of that, we thought it would be useful for everyone to keep the concept globally of new business value for all our life businesses, even those that do not generate a CSM because they are not accounted for under this format. And that will therefore be comparable to the new business value of today.

So we will use the CSM and the IFRS 17 framework to calculate the new business value, notably when it comes to discount rates. But you will have a new business value comparable to what it is today but which will highlight the new business CSM probably as its main component. And that new business value will be in line with what you have today.

The new business margin as a concept doesn't change. I just want to highlight one thing. We were one of the last to report new business margin as with APEs at the denominator. So we will use present value of expected premiums only going forward as the denominator, which will also put us more in line with what our competitors do.

On the Health side, as I said, it will be a mix of -- I mean Health today is a mix of short-term business and long-term business. So you will have, in our Health earnings, a mix of the life mechanisms and the release of CSM and everything that we have just seen notably in Asia, and that's what I said at the beginning, notably in Japan, a bit in Europe with Germany. And for the short-term business, you will have something which would be similar to P&C.

So that's my technical slide, and it is the -- how you move from underlying earnings to net income. And I said at the beginning that we don't expect that part of our P&L to be more volatile under IFRS 17 than it is under IFRS 4. Why? So there are several reasons.

The first one is the mark-to-market of all assets that back VFA business to put it that way, [indiscernible] business, all that mark-to-market will be absorbed by the CSM. And therefore, you won't see it in that part of our P&L, and it will be released over time.

Second, today, in that part of our P&L, we book the realized gains that we have on our equity portfolio, listed equity portfolio and impairments, if any. We took the option to have that mark-to-market on listed equities in OCI and no longer in P&L. So the only thing that you will see in the P&L and it will be in underlying earnings is the dividends that we receive from that portfolio of listed equities.

And that also gives me the opportunity to say that we manage 2022 to IFRS 4. And that's a good example. What do I mean by this? As we won't have capital gains on equities from 2023 on, we realize those gains this year, and we save the capital gains that we have on real estate for future years, obviously, because we want still to have in '23, the same sort of amount of capital gains overall. But in '22, it will be done with equities in '23 with real estate, which means that when we show to you the '22 P&L under IFRS 17, and we'll do that in May next year, you won't see the capital gains in equities because they don't exist under IFRS 17.

So don't -- it's a sort of a forewarning on the P&L of '22 under IFRS 17, it will be interesting. But don't look at it too long because, again, it's managed through IFRS 4, and that's what matters to us and not through IFRS 17.

So those are the 2 reasons why that part would be less volatile. There is a part which is a bit more volatile is that the amount of investment funds that will be accounted through P&L and not through OCI will be slightly greater under IFRS 17 than under IFRS 4.

So that's overall, the impact that you have, obviously, everything else, which is not financial, such as restructuring provisions and so on, do not change, and there will be the same concepts between -- in IFRS 4 and in IFRS 17.

So what are the conclusions in terms of profitability. More economic view because you will have reserves on the best estimate basis but with some degree of prudency, but you will have the discount impact immediately and greater predictability because on the life side, we'll give you the CSM release pattern. Obviously, I haven't mentioned it yet, but given the continuity that we have underlying earnings remain the basis of our dividend policy, no change to that. And last, net income is not expected to be more volatile for the reasons I tried to explain.

So I will, in conclusion, get back to what I said at the very beginning before taking your questions with Grégoire. Underlying earnings power not affected overall, true in aggregate, but true also broadly line by line. Shareholders' equity ex OCI, stable, limited change in reporting and mostly on the life side for participating businesses. No change to capital management, strategy, cash generation and therefore, same targets for driving forward '23 as we had in our plan.

And with this, Grégoire and I are happy to take your questions.

[Anu Venkataraman, Head, IR]

Let's start in the middle. Andy?

[Andy Sinclair, Bank of America]

Andy Sinclair from Bank of America. Three, please. Firstly, it was just looking at shareholders' equity plus CSM and comparing that to own funds in a Solvency II world, I would have probably thought intuitively, they would be fairly similar numbers given the similar methodologies, but it seems quite a big difference, and quite a lot higher under IFRS 17. Just wonder if you can bridge the gap for me. Secondly, it was just the mix of P&C results seems to move towards technical.

Just how will that change as interest rate and bond yields go higher, I suppose, intuitively, it would be more going towards the investment result, but actually does more go to the technical result because you got more discounting. Just how does that mix evolve? And thirdly, just on the unwind rate of the CSM. 9% to 11% seems quite a high unwind rate on the life CSM. Why is that? And what does that say about duration? And similarly, for health, it seems a bit lower 6% to 8%, why lower there?

[Alban de Mailly Nesle, Group CFO]

So can you say again your second question? I was not completely sure I understand it and the mix between technical and nontechnical?

[Andy Sinclair, Bank of America]

Yes. Just so you've shown that the mix moves towards technical away from investment income on the P&C result. I was just thinking as interest rates as bond yields go higher how does that mix evolve? I would have intuitively thought normally you're getting more on the investment result, but actually does higher discounting mean actually you're getting perhaps more of the technical result?

[Alban de Mailly Nesle, Group CFO]

Okay. So Grégoire, so just I take the last 2 questions between shareholders equity and on funds. And if I start with the CSM release, what we see with the 9% to 11% is exactly as you said about duration, is that broadly, the duration of our life business is 10 years, which is reality. Our life business is -- has a 10-year duration. When you look on the health side, that release would be closer to 5% and that's because the duration of our health business, the life like health business, is closer to 20 years.

Think of what we have in Japan. Those are lifelong contracts and therefore, with a much longer duration and therefore, with a release, which is only 5%. On the mix between technical and investment income, as interest rates increase immediately, you will see a benefit in terms of combined ratio and technical profitability. And you will see that coming over time with investment income. But there again, bear in mind that the investment income will grow, as you will have invested at a higher rate, but the unwind will also be done at a higher rate. That's what I tried to explain earlier on when I said that both should move in parallel. And Grégoire on the...?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Yes, on the reconciliation between the IFRS 17 shareholders' equity and the Solvency II on funds. So we'll disclose in the future the full reconciliation similar to the 1 you already know. But to give you a few tips, I think it's going to be a very similar reconciliation. The main things that are going to change as compared to the 1 you know now is so when it comes to eliminating the intangibles, you will eliminate just the amount of goodwill and not the whole intangibles that you used to eliminate. On the technical reserves, the best estimate liabilities themselves, they are really very close like a difference below 0.5% or something like that to the one that are under Solvency II.

The only thing that really changes is the amount of risk adjustment under IFRS 17 versus risk margin of Solvency II. And as you know, for prudential reasons, risk margin under Solvency II is calibrated at a quite high level. And so this makes the gap that you have to take into account in your reconciliation as well.

And the third item in the reconciliation this is the one that you mentioned. The CSM obviously is presented as a liability under IFRS 17. So to come to something similar to Solvency II, you have to add it up, but you have to add it up basically net of tax, which is more EUR 27 billion rather than the EUR 34 billion gross of tax that we mentioned.

[Alban de Mailly Nesle, Group CFO]

I realized I told you it was 5% release for health, it's 6% to 8% as we said in the presentation.

[Anu Venkataraman, Head, IR]

Michael?

[Michael Huttner, Berenberg]

On the combined ratio, I'm really confused. If you could talk really, really, really slowly like and I'm 60, so like to 65-year-old. Because you say there's more volatility, but I can't see where the volatility comes from. If you have a discount rate, discount rates kind of, okay, so it changes maybe year-to-year, but it's not -- I can't get being such a big item. So -- and we've been interested.

And also, I'd like to better if you could just repeat. So the unwind of the discount rate in the combined ratio, I understand goes through the investment income. But what are the change in discount rate, where does that reappear or disappear? No, disappear is a wrong word, where does that reappear. I'd be really grateful for that.

And then on the point about the discount rate and the combined ratio and the technical profit, does that mean that there will be a non-life CSM as well?

[Alban de Mailly Nesle, Group CFO]

Okay. So actually, just Grégoire you take the third question, and I'll take the first 2. So the combined ratio volatility. There would be some volatility, but compared to the life side where it's going to be very stable. I just -- we wanted to highlight that. But what do we mean concretely? The -- when you look at the amount of reserves that we book for any given year's claims, is probably a 2-year duration. You multiply that with the change in interest rates that you can have, turns our claims. And you will see the amount of earnings, pluses or minuses that you could get. I mean it's not huge volatility, I agree with you.

But compared to today, where investment income was stable, there will be a bit more. Now on your other question on the unwind. To put it simply, it's exactly like a bond. When I buy a bond today, I have, let's say, a 4% yield. If market moves to 2%, that goes to OCI, but I will still have my 4% in the next years. It's the same here for the discount on the claims.

I will have for '22 to '22 rates, that will be the rate at which I will discount the claims. Then any variation of that will go to OCI, but will not impact the P&L going forward because it will still be unwound at the rates that it will have in '22. Grégoire...

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

I mean potential non-life CSM, the simpler size, there will not be any CSM on non-life, all -- or almost absolutely all the Non-life business will be accounted for under the simplified method and the particularity of the simplified method is precisely not to have any CSM contrary to others.

[Anu Venkataraman, Head, IR]

Farooq?

[Farooq Hanif, JPMorgan]

Thank you very much for going through all of that. This is a very, very good presentation, I thought. Can you tell us the proportion of CSM that's VFA and BBA, is it 2/3, 1/3? Can you comment on the sensitivity of both equity and CSM, I think, maybe to interest rate spreads and equities roughly? I mean are there anything -- is there anything there that would be slightly different from what we used to in own funds that we should bear in mind?

And can you talk about the CSM impact on profit? So if we decided as analysts that we wanted to add the CSM back, which I think some of us may do, what would the impact on underlying earnings be if we took the life and savings and health CSM movement and added it back. How material would that be?

[Alban de Mailly Nesle, Group CFO]

If you -- I'd say, on the third question can you say that again?

[Farooq Hanif, JPMorgan]

If we use like an better value quasi and better value approach, and we added back the CSM to equity. We know the impact because you told us roughly. But what would the impact on underlying earnings be? Could you say that the CSM is going to grow every year? So there's obviously a -- if I took that out and added that back to earnings, my earnings will also go up? If you understand what I'm saying. So what will the movement in CSM be every year and potential earnings?

[Alban de Mailly Nesle, Group CFO]

Do you want to take those 3 questions? I can start with the third one and you will -- can compliment. No, I'm not trying to -- so the -- if I well understood your third question, and I'll start with that one. CSM stock will grow year after year. That's our projections because you create -- you fill the bath tub with new business CSM on a risk-neutral basis. But you unwind it and release it on the basis of real-world assumptions. So the CSM will grow even though you release 9% to 11% of it every year. So that's the part I understood in your question, but maybe I...

[Farooq Hanif, JPMorgan]

Sorry, if the CSM grew by 100% because new business and will -- and unwind -- investment return is much bigger than the release, EUR 100 million, then I would add 100% to my earnings if I had to remove the CSM as a concept.

[Alban de Mailly Nesle, Group CFO]

So even if you didn't have CSM, you wouldn't take today all earnings one shot in the current year. So there's not a big difference when you talk -- when you think about notably investment assumption that will be one of the primary movers of the CSM. Rates move, you don't see the impact today on our earnings because they will materialize slowly as we reinvest. I'm sorry. Okay. Okay. Grégoire, do you want to take the other 2 on the...

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Yes. On the pivot of the CSM between VFA and BBA, so we gave us the split mobile line of business between life and health, so 25% and 9%, respectively. I don't have the exact figure, but on top of my head, it's a small amount of pure BBA, probably between 10% and 15%, but we don't -- we didn't disclose the exact figure for the time being. And a bit the same for the sensitivity to the CSM. We don't have sensitivities to disclose for the time being.

You are right that to a certain extent we should find things a bit similar to what we observe under Solvency II, but this is typically the kind of things which we find in the future.

[Anu Venkataraman, Head, IR]

Dom?

[Dom O'Mahony, BNP Paribas Exane]

Dom O'Mahony, BNP Paribas Exane. So just starting with the technical question, the illiquidity premium and the discount rate, how are you going to calculate this? And is it going to respond to credit spreads? If I've understood correctly, then the investment income is essentially spread less illiquidity premium and less credit loss, that will be your investment income in P&C. So if, for instance, we ended the year and spreads have blown out. Are you going to have a much higher illiquidity premium, increasing the discount rate? Or are we going to expect one per investment income to come? If you could help us on that one, that would be very helpful. You showed us that essentially the earnings don't move very much at a group level and even at a sort of product segment level. My guess, but it's only a guess, is that at a geography level, there may be a bit of movement.

And it is only a guess, but I'm guessing that long-term health, for instance, has a deferral the profit recognition and the Asia, for instance, would have a lower level of profitability today. Is that right? Or is that wrong? And there be other geography segments which look better, which are offsetting that?

And then the third question, right at the beginning you said -- you gave us a very helpful reminder that it doesn't change cash earnings, it doesn't change Solvency II. Could you maybe give us a little bit of color on where the main misalignments might be. So for instance, German [GAAP], I think it's well understood, it's got quite different drivers, both from IFRS and Solvency II. Are there other drivers of your remittance capacity that you'd highlight as sort of being a different essentially from IFRS 17?

[Alban de Mailly Nesle, Group CFO]

So Grégoire, I take last 2 and you take the first one on the illiquidity premium. So the drivers of remittance, the way we think about it is mainly the local solvency and how much we have created in terms of solvency locally and that we can -- what is the excess amount that we can pay as a dividend to [access], that doesn't change fundamentally. So obviously, there's a link to the earnings, notably on the P&C side. But fundamentally, it's a solvency creation. So I don't think there is any change on this that we should expect compared to today.

And on the profit recognition versus today, so you saw that at line of business group level, even at entity level, it will be very, very similar. And you won't see a massive change or even a significant change at product level.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

And on the interest rate curve that we'll use. So as I said, we'll use a curve, which is very close to Solvency II. We said very closely because it's rate plus an illiquidity premium, like in Solvency II. As you know, when you look into the details, you have 10 to 15 detailed parameters across with the number of currencies. So for that reason, is not perfect arrangement of the Solvency II. But as with Solvency II, yes, indeed, you have kind of volatility adjustment mechanism that makes that the spreads go up. This will be reflected in the yield curve. And this is the standard curve that we'll use on all the businesses across the board for the BBA and VFA the businesses, but also for the P&C. So indeed, if the rates -- if the spread increase and the interest rate curve goes up for that reason, this would be reflected in the discount.

[Anu Venkataraman, Head, IR]

Andrew?

[Andrew Crean, Autonomous]

It's Andrew from Bernstein Autonomous. Thanks God, we got back to embedded value. Again, three questions. One, on Slide 12. What -- I think you said 24%. Yes. 24% of the earnings are under the VFA method. Roughly speaking, how much of that is GA savings of that 24%? Secondly, could you go into a bit more detail on the onerous contracts? How much are they and where are they?

And then thirdly, if you were presenting all this based on current I mean the difference between your equity in December was, I think, [75 place, 55] so if you include the OCI, I'm assuming they're going to be much more narrow now because the IFRS 4 balance sheet has come down a lot. And I assume that the IFRS 17 balance sheet hasn't moved much.

[Alban de Mailly Nesle, Group CFO]

Do you want to take all 3?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Yes.

[Alban de Mailly Nesle, Group CFO]

I can probably start with the first 1 on the...

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Yes. The first 1 to [indiscernible] I'm not sure to have right away like that, the proportion of GA savings as part of the 24% of the VFA, but it's probably a majority part we could refine that next year.

[Alban de Mailly Nesle, Group CFO]

And to be sure, you have this and you have unit linked. Unit-linked is also a VFA business, to be clear. We'll come back to you on this you know that already.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

So onerous contracts, 2 things. First, as I said, we have a very small amount of reserves deemed onerous contracts in the opening balance sheet, around 0.5% of the -- a bit less even of the total liabilities that's concretely around EUR 2 billion. These are reserves for onerous contracts where we are just faced to the technical risk as regards any impact on the P&L. We believe there are reserves on a prudent basis from a technical standpoint. So we don't expect any volatility coming from those.

And also they have any potential new onerous contracts, our stance here is that we definitely want to continue to be very selective in the business we write and not write completely any onerous new business. From what we write currently, there is no new onerous business.

[Andrew Crean, Autonomous]

[Indiscernible].

[Alban de Mailly Nesle, Group CFO]

So when you write loss-making business, the loss goes directly to P&L. So yes, you would see that if we are writing any, yes.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

And on the equity, I think the answer was indeed in your question. Actually, at the end of -- on the 1st of Jan '22 or the 31st of December '21, there was a EUR 275 billion equity, including a bit more than EUR 20 billion of OCI. This reduced by EUR 19 billion at half -- the OCI under IFRS 4 reduced by EUR 19 billion at half year. This was in the figures that we disclosed and rates have continued to move up. So this OCI under IFRS 4 is definitely very volatile and the shareholders' equity, as we speak, at a minimum at the same level as the IFRS 17 ones. The good news, again, is that this volatility coming from the OCI will be much lower under IFRS 17.

[Anu Venkataraman, Head, IR]

Will?

[William Hawkins, KBW]

William Hawkins from KBW. Thank you also for an excellent presentation. It's very helpful. Why are you encouraging us to look at shareholders' funds, excluding OCI? I can see under the existing regime, there's a big problem of comparability between assets and liabilities, so we should be doing it. But the whole point about this change is that assets and liabilities move on to a basis that's more comparable. So I really don't understand why we're being encouraged to do that.

And then sort of secondly, it's sort of slightly small, but why is the OCI negative? Should I be inferring anything from that? And then secondly, please, do you have an allowance for limited historical experience? I'm assuming not because you haven't referred to it. And I'm not even sure where this idea came from other than the fact that 1 of your peers invented it a month ago. Clearly, you don't have one, is that implying that you're less well reserved than a company that does?

[Alban de Mailly Nesle, Group CFO]

So on the OCI, I guess it's a tradition at AXA. It's -- when we looked at our ROE in the past or when we look at gearing, we always excluded OCI because there is a degree of variability in it. As Grégoire just said, under IFRS 17, it will be much, much less volatile because the volatility coming from the asset side will be offset totally or in a great part by the same movement coming from the liability side because you also defend your liabilities. So we expect the OCI to be much less volatile than today.

That being said, that's not something that we consider because of that volatility because -- and because it doesn't say much about our business and our net assets. But that's the way we have looked at it historically the -- why is it negative? It was already close to 0 at half year, simply because rates had gone up. So it's just a reflection of the current interest rates applied to our assets and liability fundamentally plus the variation on equities as well. Equity was still positive at 1.5 billion at 30 June, but they could vary and that could become negative, and that could go to OCI. The third question on -- did I forget on this?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Just to add on this, that's really close to 0. But in the current balance sheet, it's minus EUR 3 billion, which we believe at the scale of our other assets is close to 0. And the fact that I mentioned it this way, there is notably a small mismatch, notably for the real estate, which is accounted for at cost. So you don't have the unrealized gains on the asset side of the balance sheet, while when you calculate the reserves on the liability side, you take into account this -- the corresponding value notably for the shareholder -- for the policyholder, sorry, and you have set of that within OCI. And so you create -- and so it was a bit too simplistic saying it's centered to 0.

It's probably in, real life, the OCI that we'll have under IFRS 17 centered around minus EUR 3 billion, EUR 4 billion, something like that as you see in the opening balance sheet. Then with the rates moving, we'll have a bit of volatility around that too early to say to give you more detail on that, but a much lower volatility than again, the one that we observed especially this year on the IFRS for OCI.

[Alban de Mailly Nesle, Group CFO]

And on your question on [indiscernible] what we know and what we don't know, and we have IBNRs for that and we'll keep on having IBNRs and it would be in our best estimate liabilities. I guess you have felt from what I said that we think that the risk adjustment is not a necessary reserve. It will not move much. It will not be released through a P&L, net-net, so we don't see the merit of having yet another amount of reserves separately from our best estimate liability.

[Anu Venkataraman, Head, IR]

Claudia?

[Claudia Gaspari, Barclays]

Claudia Gaspari from Barclays. So this pay [ph] is the result of your pro mix, I guess, and geographic footprint. To what extent is it the result of accounting choices you're making here now? And to what extent it's actually like physical incentive to maximize the CSM at inception and there's equally, very clearly some constraints. I mean the obvious ones that everyone mentioned are always financial leverage and, I guess, the tax deal but there's probably 1 million other considerations that come into play.

So if you could just elaborate a little bit on your thought process when determining this balance with the inception?

[Alban de Mailly Nesle, Group CFO]

With your questions. You don't have volatility if you don't have onerous contracts fundamentally because the CSM as such and the way it is released, provides for stability in earnings without having to do more. So the question is onerous contracts. And that's 1 of the key options that we have, the amount of reserves that were onerous contracts. But that was a tiny bit of our portfolio, what remains is what kind of what I said, 0.5%, even 0.4% of our reserves.

But fundamentally, the stability comes from the quality of the business and the fact that for a very, very large majority, it was not onerous contracts at all. And so what drove the CSM, honestly, it's apart from that part on fair value. It is very very similar to indiscernible] than other. You have the profits going forward. The discount curve that we built in a given way that Grégoire described.

And that gives you the CSM. I think where we acted was to limit the amount of under-risk contract with those. Grégoire any...

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

That's it. And just on your first question, I think I tried to explain the [indiscernible] that we made, which, to a large extent, were kind of obvious choices to us notably as we don't like volatility in the P&L, for instance. So we made the choices made the calculation. As a matter of fact, but to be honest, we really saw that as a kind of normal thing. This ended up with [little] of the shareholders' equity, also the stability of earnings.

And in that context, no need to voluntarily boost the CSM and this is actually not the way under which we work because on the contrary, we want to have the earnings that we believe represents a fair level of earnings for the group and continue to have the ability to have the CSM stock to grow sustainably over time.

[Anu Venkataraman, Head, IR]

Ashik?

[Ashik Musaddi, Morgan Stanley]

Ashik Musaddi from Morgan Stanley. So if I understand correctly, what you mentioned is CSM is basically based on market consistent discount rates. So does that mean that a real-world CSM is a much larger amount because you would have much higher returns on real estate, equities, some of the illiquid assets you would have. So that's one, if my understanding is correct.

Second thing is, when you mentioned 9% to 12% release of CSM, does that mean that half of that is actually coming from the 5% because it's market consistent. But the other half is actually the real world excess returns you'll be generating I mean, a concept which is similar to OCG, that you would have higher returns on equities real estate. So going back to the duration point, I guess you mentioned duration is 10-year, but that means maturity profile of 20 years. i.e., your CSM unwinds at 5%, but on top, we add 5%, which is the real-world returns. That's basically the ultimate CSM unwind.

The third question is, when we think about real-world unwind, is it the realized unwind? Or is it the assumed unwind, i.e. would you assume that equities will give us 5% real estate will give us 7% in that unwind of CSM? Or is it the actual realized, which is based on dividend rental income, et cetera? Yes, 3 questions.

Greg.

[Alban de Mailly Nesle, Group CFO]

So just Grégoire, I take the first 2 and you take the third one?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Yes.

[Alban de Mailly Nesle, Group CFO]

So on the market consistent. The short answer to your first question is, yes, under real-world set of assumptions, it would be larger. It will sort of a deterministic way rather than a stochastic risk neutral approach. And that's why I said earlier that when you move from the CSM at the beginning of the year to CSM at the end of the year, obviously, you have the new business, CSM, but the unwind takes -- I mean, will reveal the real-world assumptions exactly as you said, whereas it was booked on a risk-neutral basis.

Now the 9% to 11% is the unwind applied to the stock, that's the amount that will be released in a given year. In addition to that, you have the financial variance that you see as explained earlier when I was there on some other aspects. And Grégoire, on the third question?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Yes. I think you have to make the difference between 2 aspects. The stock of CSM is always calculated on in-neutral basis. And then every year, you relief through P&L, a portion of that which for Life & Savings is on average for the group and it could vary [indiscernible] between 9% and 11% and for the [indiscernible] between 6% and 8%. The way this portion, which is released, but that's technically a kind of different question is measured through projecting what is called coverage unit.

And it's in the way we project coverage you need that we need to take into account real world long-term assumptions. And that's what make us say that the exact amount that is released in a given year also depends on real-world assumptions.

[Anu Venkataraman, Head, IR]

Thomas?

[Thomas Fossard, HSBC]

Thomas Fossard for the HSBC. Two questions. The first one, can I challenge your assumptions or I mean data that 80% of the Life business has been made on a retrospective approach because I thought that it was a super complex situation to gather all the data from historical contract, and that was probably not super easy to do. So I'm still surprised that actually you managed to reach to 80% on the life book.

And the second question would be, can you talk a bit about the operational gains from IFRS 17 because initially when it has been presented, it was a big project, then it was mentioned to be meant a big reshuffling of the financial function, breaking the silos between actuarial and accounting. So is there any material gains that you're getting to start earning through, now that actually the process is -- the investment process is over and you're going live?

[Alban de Mailly Nesle, Group CFO]

So I'll take the second question. You take the first, Grégoire. So the simple answer to your second question on operational gains is I don't see any. Simply because, yes, there is convergence between IFRS 17 and Solvency II, and that's a plus. And yes, actuaries and accountants talk to each other, and that's good. But when you look at the finance function, they also need to take care of their statutory P&L and balance sheet and they have not moved to IFRS 17. So you have different frameworks that you need to live with.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

On your question on the retrospective approach, I confirm it's actually 80% of our business, which has been accounted for using retrospective approaches. Now you're right from the technical standpoint. You have either the full or what is called the modified, which is basically a proxy and most of what we did is a proxy, notably as we used to disclose embedded value for many, many ideas. Now we have this data, which helped to make appropriate proxies.

[Unidentified Analyst]

So you spoke about convergence, and you said your numbers are broadly very similar before and after, which kind of implies it's a competition conversion, which kind of implies they were wrong. And that was just kind of taking -- I mean I'm overstating obviously. Where do you think, in particular, is it the mutuals who I mean -- just give us a feel for your thinking there?

And then the other question is, is the rating agencies. How are they -- so you heard from some of us that we're wondering whether CSM is equity or not equity or earnings or not earnings. How are the rating agencies thinking about this?

And the last point, which is probably because I'm still struggling with this, does is mean because the equities go through OCI, but real estate gains do not, but you're going to give up your equities and just invest in real estate?

[Alban de Mailly Nesle, Group CFO]

Okay. Anu is it okay if you take the question on rating agencies, and I think the other 2?

[Anu Venkataraman, Head, IR]

Sure. So Michael, the rating agencies have their own model, and you know S&P has been in a common period. So it remains to be seen how they treat the changes under IFRS 17, it would not be for us to speak for them.

[Alban de Mailly Nesle, Group CFO]

And so on convergence, far from me, the idea of saying that anyone was wrong before. Nevertheless, what you see in some countries with that on the Life side, some were used to booking NBV as a profit first year. And so that's a difference. And with IFRS 17, that's not possible any longer.

The second thing is on the P&C side. There is the concept of best estimate liability is probably tighter now than it was with IFRS 4, where you could have a bit more of leeway. Those are the 2 differences. I don't think it affects significantly the comparability with our traditional peers. It's more with others, notably on the Life side that I mentioned.

On equities, no, I mean the equities bring value. It's not recognized in P&L for the mark-to-market. It will still be recognized in the OCI. We don't plan to get rid of our equities. That being said, you know that today, our net exposure to equities is very limited.

[Anu Venkataraman, Head, IR]

Andrew?

[Andrew Crean, Autonomous]

A very quick one. I mean you give the embedded value of your Life business in your embedded value report. How does that compare with the CSM? Is it possible for us to make that comparison? Is it the same scope in terms of lifetime health so that we can see the degree to which your CSM is conservatively struck with a risk-neutral basis.

[Alban de Mailly Nesle, Group CFO]

So we can provide that reconciliation. I don't know if you have it, Grégoire?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

[Indiscernible]

[Alban de Mailly Nesle, Group CFO]

I think the -- just to be clear, there's a perimeter or scope aspects because embedded value would be slightly broader to the same extent I described on the new business that not only would you have the new business, CSM, you would have new business value coming from other Life and Health activities, which would not be accounted for under VFA or BBA but on the PA. So there is the question of scope that is...

[Andrew Crean, Autonomous]

CSM [indiscernible] scope...

[Alban de Mailly Nesle, Group CFO]

But the system is narrower scope because -- it's slightly narrower. It's not a big difference. Other items to have in mind when we reconcile the 2?

[Andrew Crean, Autonomous]

Those are different questions. What thing -- if you put the CSM onto a real-world basis, what is edge by?

[Alban de Mailly Nesle, Group CFO]

So I mean embedded value, it's not on a real world basis either because it's market neutral. So -- and it's a long time since we last looked at the deterministic embedded value, but I mean our mid value is not deterministic. It's not real world. It is market neutral.

[Anu Venkataraman, Head, IR]

Since there are no more questions in the room, we're going to go to questions that we received online. So we have 2 questions from the webcast. The first question is from James Shuck from Citi. How much of your EUR 4 billion risk adjustment relates to P&C? And how does this number compare to your previous margin over best estimate, i.e., excess P&C reserves under IFRS 4?

[Alban de Mailly Nesle, Group CFO]

You want to take it?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Yes. So the EUR 4 billion risk adjustment is, broadly speaking, splitted half-half P&C and Life. So that makes EUR 2 billion which is to be compared to an amount of excess reserves, which was slightly higher than that, I think, at the end of last year.

[Alban de Mailly Nesle, Group CFO]

Exactly, it was EUR 2.4 billion at the end of last year but it's reasonably similar at the size of the group.

[Operator]

And the second question is from Rob Heim from JPMorgan. To follow up on an earlier question, what are the main items that reconcile your Solvency II own funds of EUR 45 billion at full year 2021 against our total shareholders' equity plus CSM of EUR 85 billion, i.e., EUR 58 billion plus post-tax CSM of EUR 27 billion especially given that your best estimate liabilities are similar between IFRS 17 and Solvency II.

[Alban de Mailly Nesle, Group CFO]

There are several differences. The first 1 is goodwill because we still have goodwill in IFRS 17 balance sheet, and we don't have that under Solvency II. When you look at our own funds, you have obviously debt in both parts.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

But the EUR 45 billion to which we try to reconcile on the Solvency II side is with no debt. And so you have to deduct EUR 7 billion, I think, of debt that are still under IFRS 17. And then you have the difference between the risk adjustment and the risk margin, much higher [indiscernible] Solvency II that also drives the level of shareholders' equity and the Solvency II done. So if you add up the goodwill, the debt, the defense between risk adjustment and risk margin. And you take into account CSM, which is net of tax, I think you have all the elements to make that exact reconciliation that was targeted here.

[Anu Venkataraman, Head, IR]

Do we have any other questions online?

[Operator]

No more questions from the webcast.

[Anu Venkataraman, Head, IR]

Will?

[Unidentified Analyst]

It's a quick one, probably [indiscernible] on the risk adjustment, will we get that disclosure going forward, the split of P&C and Life? And will we get sensitivity to percentiles? Are you going to give us what percentage is based on? I'm wondering whether we'll get sensitivities on that percent?

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

So on the split, I think when we're on the -- I'm not sure where it will show up or not in our disclosures, but this is related to...

[Anu Venkataraman, Head, IR]

But that's something you want to see. Then obviously, we'll take that into account, and we will be soliciting your views and those of everybody else to populate our new [indiscernible] as Grégoire mentioned.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

But sensitivities to the escalator, we truly believe it does not make much sense because we really expect, as we said, this amount to be quite broadly stable over time.

[Anu Venkataraman, Head, IR]

Michael?

[Michael Huttner, Berenberg]

Just picking up on maybe Ashik's question, but probably I didn't understand his question. So the new business land, excluding the scoping, if we -- so under a CSM, I understand it, it's market consistency use risk-free plus illiquidity. Is that number different from the new business value we have now?

And on the other question where you kind of said you haven't done that calculation for a while. I think you did it in my memory 2014 or '15, where you showed the movement from embedded value to realistic assumption to something else. And the gap was, from memory, it was about EUR 30 billion, EUR 34 billion, something like that. And that was, if you like, unwind in profits. It was in 1 of those quite complicated live presentations. I try to get my head around, but clearly, I didn't do that very successfully. And I just wondered whether you could say that's, I think, in answer to 1 of the questions, the difference between CSM and realistic would be of that order of magnitude, which we had back then, you did publish it.

[Alban de Mailly Nesle, Group CFO]

So on those questions, if we leave aside the scope aspect, the new business value on the CSM scope would not be different -- will not be very different before and after. You may have some issues because there are some differences because the boundaries of contracts are not exactly the same under IFRS 17 as they are today for our new business value, but that will not impact significantly the amount.

On the move to real estate assumptions, honestly, that's not something that we have in mind to do. I think when we give the pattern of release of CSM going forward, that's, to some extent, already a good indication of what it will look like.

[Michael Huttner, Berenberg]

I remember, and this is really many years ago. You did that exercise of comparing market consistence with more realistic assumptions is to show the potential cash which could be unwound because that difference effectively does come through its real value. And so really, it gets us back into kind of the mini question, which is not that the CSM allows you to have more cash or less cash but really whether we can have more cash than you're publishing now?

[Alban de Mailly Nesle, Group CFO]

I mean, it's the same question as for Solvency II fundamentally. At what rate do you unwind your risk and your Solvency II capital fundamentally. And what is the gain between real world and market-neutral assumptions, which is also the question that we had before fundamentally. That's the same mechanisms under IFRS 17 and under Solvency II.

[Anu Venkataraman, Head, IR]

Okay. If there are no further questions, we will end the presentation. We really appreciate your time and attention. Thank you.

[Alban de Mailly Nesle, Group CFO]

Yes. Thank you very much.

[Grégoire de Montchalin, Group Chief Accounting and Reporting Officer]

Thank you.