AXA S.A.

Okay. Well, good afternoon.

Welcome, everybody, here in London for AXA's Full Year '19 Earnings Presentation. Welcome to those of you in London, who've just come through the rain.

Welcome to those of you who are dry and not just walked through the rain, who are with us on the webcast. Good to have you here.

For those of you on the phone as well. If you'd like to ask questions later on in the Q&A, of course, we'll give priority to the people here in the room, but we will take questions also from the webcast.

Just follow the instructions you're given, and we'll happy to -- I'll be happy to have your questions as well. I'd like to welcome, of course, we have with us Thomas Buberl, our CEO; Etienne Bouas-Laurent, our new CFO.

As well, we have some of the management team. We have Jacques de Peretti from France; and Antimo Perretta from Europe, Alban de Mailly Nesle, who is -- I always get the wrong.

Alban here as the group Chief Risk Officer -- Investment Officer, I'm sorry, Alban; and we have Scott Gunter, and that's a lot easier for me to pronounce. And Scott, welcome, who's the very new CEO of AXA XL.

All of them will be available to answer your questions later on. And without any further ado, Thomas, I hand over to you.

Thank you, Andrew, and good afternoon to all of you. Very glad to see you here in London.

When we look at the highlights of the results, they follow the rule 5 5 7 -- 5 5 5 5 7. 5% growth on the top line and across the board, 5% underlying earnings per share, reflecting a very strong technical result, 5% more solvency in an environment that is characterized by continuously low interest rates.

And certainly, in our case, characterized by a significant portfolio shift between AXA Equitable and AXA XL, and the proposal to the shareholders will be plus 7% on dividend, leading to a dividend of €1.43 per share. Let's have a look into the details.

The first 5%. Plus 5% revenue, leading us to €104 billion revenue.

What is interesting to see is that this is happening across the board. So all geographies are contributing to it, and I would like to draw your attention to the bottom right, which is AXA XL, we'll come back to that again, plus 10% is really a very good growth number, certainly resulting from strong price increases.

But if you go to the top part, you will certainly see that this strong growth was again, driven by the preferred segments that we are focusing on commercial lines, where we are today, the biggest in the world. Health and protection, where we are also very strong on a global level.

But what is nice to see that all the geographies, France, Europe, Asia and the international markets have really contributed to it, and I'm sure we will talk later on with Jacques with Antimo about more details. When we look at the earnings, as I mentioned earlier, that's the second 5%, we will see 5% earnings growth in terms of underlying earnings per share, which results in a 6% adjusted earnings per share growth, resulting in a proposed 7% growth of the dividend per share to €1.43 at a payout ratio that is slightly higher than last year, but very stable at a 52% level.

If we look at the priorities that we had, and they were clearly 3 priorities that we mentioned here as of last year. One was the selling down of AXA Equitable.

You see here on the left-hand side that we used to have a year gross to 51%, we now have 0. So AXA Equitable is now in the hands of American shareholders and free in the market.

Second priority was clearly reducing our temporary increased leverage. A year ago, we had -- we were at 32%.

Now we are at 29%, with a very clear way downwards to the target range of 25% to 28%. You've already seen the announcements of the necessary means in order to get there.

And certainly, on Equitable, we have achieved 100% of what we wanted to achieve because we issued, at the time, at $20 and never sold below $20, always above. What is also interesting is to look at the solvency.

The solvency has continuously increased now, and there is still some upside in it. You remember, we are still going -- putting XL into the internal model of AXA, which will hopefully happen this year.

But what is also interesting not only to see that there is a 5 percentage point increase, we are also, for the first time, not relying on equivalents anymore. This was certainly a discussion that we used to have a lot on the U.S.

business. Today, it is a solvency that is pure without any equivalents.

When we look at AXA XL. AXA XL, as I said earlier, has had a very strong growth.

So 10% growth of the premium, very much based on a strong and disciplined re-underwriting of the book, and this is very different across the different lines. So you have some lines, for example, in the U.S., liability, which is at 22%.

You have the excess casualty, which is at over 40%. So the hardening cycle is there.

We are making use of it. And when I look at the January numbers, I see that this trend is continuing.

So we will see this continuing in 2020, and we will also see the benefit of this earning through in 2020 and 2021, because this 10%, if you take on a net earned premium of €10 billion, you can make the math yourself, this is a very interesting figure. Second point was very much around the questions, which are related to social inflations, are the reserves adequate.

We had these questions plenty of times in this room. As you know, we have built a very significant reserve at the time of the opening balance sheet through the P-GAAP.

We, obviously, have a continuous look at this, and we can state today that the reserves are adequate and that we even have €0.8 billion, of additional annual unallocated reserves at the end of 2019. The result of XL is, for 2019, €0.5 billion, with a natural catastrophe fee experience that is beyond what we normally would expect.

I personally have a much higher ambition than €0.5 billion. And so we have taken 3 key decisions in order to get to a more profitable XL in 2020: #1 is we have increased the prices significantly, and we will continue these price increases, which we see, we are benefiting from a significant growth.

First decision; second decision, we have reduced our volatility, reduced our volatility, so 2 means. One is we have reduced our exposure in underwriting to natural catastrophes.

So for example, on the Reinsurance side, in natural catastrophes, we have reduced our exposure by 10%, which means also less premium and less profit or less losses in case there is a lot of natural catastrophes. Secondly, on the question around liability we've reduced the line sizing.

So in the past, we used to write much bigger net lines than we do today. Same logic again, less risk equals less premium, but also less losses.

And then thirdly, in this category of reducing risk, we have taken out more reinsurance to protect ourselves better against volatility. That was the second decision; the third decision was a new management.

We are passing now from a phase of integration which was a very successful phase, where it was very difficult to bring 2 large organizations together in a very short time, we have succeeded well on this today, AXA XL is really a new combination, works well. Greg Hendrick has done a fantastic job in really bringing those 2 worlds together.

In bridging, let's say, the cultural differences as well that used to exist that we are now passing into a new phase, which is around acceleration. The strategy will remain exactly the same.

We are focusing on increasing the profitability, on bidding the cost synergies that we have communicated, but we felt it's time to do this with a new man, and I'm very pleased that Scott Gunter today is with us, and I would like you, Scott, to stand up quickly, introduce yourself to the audience today. And maybe talk a little bit about what you have seen so far because today's your first day as active as CEO, but you have been on board for quite a few weeks already, preparing yourself for this new challenge.

And if you could stand in that box to talk to us quickly.

Yes. Thank you, Thomas, and I'm very excited to be part of the AXA group.

As Thomas mentioned, I've been on board for a couple of weeks and sort of studying it from afar, and I'm very excited. When you look at the AXA XL organization, we have a tremendous product breadth.

There's a lot of product capability across a number of geographies. What does that do for us?

That allows us to operate in marketplace and find where there's pricing that fits what we need to get. If the pricing is not there, we can turn the dial down on that business and go, "okay, we'll get that later and write more where the margin is."

So the breadth of product and the breadth of geography is a fantastic opportunity. Also impressed by our hidden gem.

We have a large risk consulting organization, over 400 engineers. And they not only do work for the underwriters, they do work for clients, for a fee.

So we're collecting non-risk revenue from that operation. I'm excited by that because that is a risk-taking business.

You always like non-risk fee business, for sure. And really, the third excites me is the combination of XL and AXA.

XL, great organization, lots of product geographic spread, leveraging the assets of AXA, the relationships in Europe, relationships in the U.K., right, the relationships in Asia. Just leveraging all of that and selling more product where we wrote property casualty business, and they'll bring those products to those brokers.

So you put those 3 together, very, very excited. Now, and what we're going to work on?

Thomas mentioned it, we're going to work on the margin, right? The pricing environment allows that to happen.

We're going to constantly look at our portfolio, again, where do we need to get pricing, pricing is right now is looking very good. We get a lot of products that are running ahead of trend, right?

That's a great sign, right? We can continue to work on the margin.

So margin expansion is an important part of it. Thomas mentioned about volatility, continue to work on that.

The line sizing started, we continue that process. The good news is you can reduce your line, charge more.

So the impact on the revenue isn't as much, right, continue to work on that. And the third issue that we really want to work on is growth, right?

So we're going to ask the underwriters to do 2 things, right? Renewal book, work on your margin, get ahead of loss trend.

Secondly, as prices go up, there are more and more opportunities that are attractive to us. 2 years ago, when the price was at x, we may not have been that interested.

Today, it goes up 25% last year, 25% this year. Also, it hits our number, and now we're interested.

So we're going to work on the margin, grow the business and deal with volatility.

Thank you very much, Scott. So you heard from Scott, no change in strategy, acceleration of the existing strategy in order to get to higher profitability.

The forecast for this year is €1.2 billion, and you've got the equation how to get there. People have asked me, well, what about the €1.4 billion?

Well, we are talking about exactly the same thing. €1.4 billion means with less volatility.

We've decided to buy additional reinsurance to reduce the risk exposure, which has a cost of €200 billion. So the €1.2 billion is the €1.4 billion minus €0.2 billion.

What have we achieved in 2019? Remember, the Ambition 2020 was focused on 2 pillars.

Focus and transform. On the focus side, we've clearly seen that we have gone back to growth.

When I started as a CEO, we were at 0% growth. We are now at 5% growth with the focus on the preferred segments, and the whole organization is fully aligned on this, which is really good.

You can also ask my colleagues here. When they talk about their business, we all talk the same language.

In addition to that, we've also achieved to increase our presence in China, and there might be a question later on about the Coronavirus. Does it put in question your engagement in China, it doesn't because in insurance, we don't look at the short term, we look at long term, we look at the demography, and this remains fully intact and fully attractive to us.

On the simplification side, we have clearly communicated the segmentation of our countries. 10 countries that make 80% of our profit, 6 countries with high potential that makes 16% of the profit and over 40 countries that make 4%.

We clearly communicated that we want to reduce our footprint and reduce the complexity of our countries. You have seen, we have delivered AXA Bank Belgium, AXA Life Europe, CEE with the 3 countries, Poland, Czech Republic and Slovakia.

On the balance sheet, we have changed the balance sheet from being 80% focused on financial risk to 80% technical risk, having a very resilient balance sheet, having a stronger solvency, having a solvency that doesn't rely anymore on equivalents, and certainly having a gearing that will come back to the place where we started. On the transform side, this shift that I mentioned earlier from Life & Savings or from financial risk to technical risk has been accompanied by significant transactions, the quotation and the IPO of XL -- of Equitable, which is done.

The Swiss Group Life transformation, which is announced, which is in the process of being done. And certainly, the XL acquisition and integration is on a good journey.

We are now completing the integration, and that means we can now move to a new phase, which is around acceleration of the profitability, acceleration of the simplicity of the Company. The one thing I'm personally very proud of is the customer experience because when you change a company, it is always difficult to find the right balance between change and making sure that you also progress on your existing business.

And there is a famous French saying, which is, "the sales need to continue while you rebuild your house," and I'm always reminded of this. And it's very nice to see that certainly, on the NPS front, so the customer experience, the customer satisfaction, we are at 75%, which means that 75% of our markets are at least at market average or above and certainly, the new brand positioning, the Know You Can campaign has helped us to also boost this both internally but also externally.

And then certainly, transform also means transforming into new business models, finding new business models. And it is also very good to see that we've tried new models that are starting to work.

So you might remember the vertical integration in Mexico, where we have built together with Keralty Clinics in Mexico City and other cities, where we are going and offering medical services to Mexican customers. We are in the process of doing the same in Egypt.

On the ING partnership, and Antimo is responsible for that. Across different countries in Europe, we've built a very innovative solution that is working really well and where we have really changed also the game when it comes to bancassurance in Europe.

And then on the climate side, we've been very innovative when it comes to parametric insurance, both in developed markets, but also in emerging markets. And this is really good to see that when you look at the number part, again, of Ambition 2020, you see that we are fully in line with what we said.

On the underlying earnings per share, we want to be in a corridor of 3% to 7%. We are on the average now of the 4 years at 5%.

There's 1 more year to go to the end of 2020. So we are on a very good journey.

When it comes to the free cash flow, we said between €28 billion and €32 billion. On a cumulative basis, we are at €25 billion by the end of 2019, again, on a good journey.

When it comes to the adjusted return on equity, the initial target was below the range that we have now. So now, we have already upgraded to 14% to 16%.

We are now at 16%, which I think is a very attractive return on equity. And on the solvency, I spoke about it earlier.

We want to be in the range of 170%to 220%. We are now at 198%.

And the 198% of today is a very different solvency to 4 years ago because we do not rely on equivalents anymore. So I'm very pleased with where we are.

And I will now hand over to Etienne, who will lead us through the details of the financials. Thank you.

Thank you, Thomas, and good afternoon to all of you. I'm very pleased to be here and to present this very solid set of numbers.

I'm going to start with the presentation of the earnings by geography of business line -- sorry, business units. The second presentation are the same earnings by line of business.

Just to have an existing view of the situation. I will then move to roll forward from underlying earnings to net income.

And lastly, we'll finish with the balance sheet items. Starting with the underlying earnings by geographies.

What you can see is that the growth is driven by France, Asia and international new markets, which represent quite a lot of different countries. Europe remains positive despite the restructuring of the group life business, so which is very sign of strong resilience.

XL is up versus last year, it's certainly below the expectations we had in '19, but represents a very strong upside potential for 2020. Moving to or starting with France, you see that the growth is solid, 4%, driven by Health and Life & Savings while P&C is relatively stable with some pluses -- strong plus in commercial line, offset by a negative on personal lines.

If we look at the net flows, which is -- it may be very interesting to understand the dynamic of France, where we have 2 good news. The first one is that the Protection & Health business is growing big time in terms of net inflows.

And on the savings side, you see that the growth comes from the unit-linked more than from the general account, which bodes very well for the future. And we think that AXA France is a very, very solid and giant, not only today, but for the future of this company.

When looking at the underlying earnings, the growth is driven this year by the P&C business. You see that the combined ratio is down minus 1.6 to 90 -- 11 -- 91%, which is a very strong performance, and it's driven not only by current year, but also by the prior year reserve development and by the expense ratios.

So a very solid business. The health is temporarily hit by an adverse claims experience in the group, group health business in 1 international contract.

It's very much related to the strong growth we have in this business, and we expect a reversion to the mean next year. Lastly, in Life & Savings, the resilience of our investment margin is remarkable.

It's certainly due to the fact that we have a capacity despite lower investment yield to manage a crediting rate in a very dynamic way. Europe, growing as well by 4%.

Here, with a very balanced growth among the different business -- lines of business, P&C plus 3%. U.K.

is a very strong contributor. We are #2 in the U.K.

market with an 8% market share. It's a combination of prices and new business.

Germany, which is a leading health insurer in its own market, the 5% market share, very strong price increases, leading the gross revenues. And lastly, Life & Savings growth 6%, strongly driven mainly by Italy, with strong multilines contract and you'll see with a very strong growth, not only in [geo] savings, but mostly also on unit linked.

So very strong momentum and very well balanced. Underlying earnings is up 1%.

As you can see, the P&C is extremely solid, with a combined ratio down 1.3 points to 93%, with an improvement across the board, across the countries. And due to both currency experience improvement and prior year reserve development improvement as well.

Moving to health. There, again, there is a one-off effect, nonrecurring of a very -- an exceptional experience in the U.K.

PPP layer, which is, as you all know, a reference number between the U.K. market in health.

Life & Savings, the transformation of the Group Life business and the amortization of VBI due to the low interest rates, explain the decrease. Without this impact would have been flat on the life business.

Showing, again, the resilience of our model. Asia and international is, of course, a source of growth, not only in top line but also in terms of bottom line.

Very well balanced in terms of top line between P&C health and Life & Savings, P&C, driven by Turkey, 18%, very spectacular growth. I must confess as well with very high pricing component, 2/3 of it where -- market where we are #4 player with a strong position.

Mexico, Thomas referred to the vertical integration which certainly makes our subsidiary -- our entity there, which is #2 in the market with an 18% market share, a dominant player, benefiting from a very strong new business momentum as well as tariff increases. And lastly, Hong Kong, with a solid growth last year, especially in the first half of the year, for reasons you may understand, the second half of the year was more of a slowdown, but the growth was so much so high in the first year that we finished on a positive note.

Underlying earnings, up 6%, driven by P&C and Health. The P&C, like in the other regions of AXA, is improving, minus 2.5 points, mostly this time due to the variation in prior year reserve development but with a strong growth, as we said.

Health is up 16%, mainly driven by Asia and notably, Japan. Life & Savings is down 4%.

And here, it's mainly due to our JV with ICBC, which has invested in the distribution networks of the bank and in an asset management services. XL, I would say, lastly, for this geographic presentation.

So as Thomas highlighted, the result has been up this year, but not at the level we expected. Why?

First, because we experienced an abnormal level of nat cat, €0.3 billion. So on the normalized level, we might have expected €0.8 billion.

Second, we had -- we observed, and we experienced a currency deviation from long tail lines in casualty, €0.2 billion. And lastly, elevated large losses all over the year, partly offset from -- by positive prior year's development.

So the question is what we do. And we have tackled issue with 2 components.

The first one is the pricing, very spectacular pricing, 10% -- sorry, revenues. But 8% price increases on the insurance side, with a very strong acceleration of the price increases quarter-by-quarter, as you can see here.

And to give you an idea, to give some flavor to this, this is an average, but actually the lines hit by social inflation and adverse experience have accelerated even more. In Q4, excess casualty in North America was up 29%.

At the moment, it's plus 43%, January. On professional, up 19% in Q4, it's up 22% in Q1.

There are not so many times in a cycle where you see such a high level of experience of price increases, and we intend to benefit from it. But of course, the impact will be partly in 2020, partly in '21.

The second aspect we wanted to tackle are the 3 problems I just described, with the experience of this year, which is first nat cat; second, large losses; third, casualty lines. And we have taken measures on these 3 issues over the recent months.

So we have not been waiting. We have addressed this in a very proactive way.

Starting with the nat cat. We have underwritten, less risk, minus 10% gross revenues in '19, there will be another minus 10% next -- in 2020.

And we have put in place a quota share, 15% quota share on the reinsurance nat cat, which we hadn't last year. Second, on large losses, we have set up an aggregate cover with an attachment point, which is not very far from the current level of experience, which means no further deviation is expected.

And lastly, on the casualty, we have not only increased massively the prices, we have also reduced our capacity. So line sizing in a market which is -- who's the capacity shrinking by 25%, which is actually both a risk today and an opportunity for tomorrow.

So the last point, which will tackle in the balance sheet is the reserving. Thomas said, he's comfortable with the level of reserve.

I can say that I feel very comfortable with the level of reserves because we are reviewing it regularly and very -- over the recent weeks. So based on the pricing dynamic, based on the measures we have taken to reduce the volatility and based on the adequate level of reserving, it makes the €1.2 billion underlying earnings very achievable.

And beyond, I mean, from 2021, we will get the full benefit of these measures, notably from the pricing. If we take a view by business line, I would say, in terms of messages that, first, of course, the P&C business is the main contributor but with a strong double-digit growth, which is very satisfying.

Second, Life & Savings business, down 1%, but in a context of very low interest rate environment, so very resilient. The third, Health is a steady growth business in line with the revenues.

And lastly, Asset Management, flattening result, what I call minus 3%, but with upside potential, and I'm going to get back on these 4 statements right now. Starting with P&C.

The increase is, of course, helped by the growth in revenue, the 5% growth in revenues, but also -- and mainly by the lower combined ratio. You can see here the numbers, minus 0.6 points on a reported basis.

Excluding XL, we would be at minus 1.1. And this minus 1.1 points is due to current year experience, minus 0.5, due to better business mix, a better pricing, a better frequency, partly offset by a more -- by higher severity.

On the prior year reserve development, minus 0.7, we have 3 main explanations. The first one is France.

Very strong profitability, very high level of reserves and the reserving ratio is increasing in France. Switzerland, same mechanism.

And lastly, in the U.K., Ogden effect with the change in this country, generating some surplus. So all of this allows you -- us to more than offset the decline in asset yield of 17 bps.

So a very -- a business which is in excellent shape. Life & Savings.

Life & Savings is resilient, as I told you, starting with something which might look odd, which is the investment margin. So further to the Swiss Group Life transformation, we had a slight decrease in reserves because we have transformed ourselves from an insurance provider to an insurance manager, I would say.

But the delta between the investment yield and the crediting rate is positive. Therefore, the 2 bps is at constant scope, so excluding any adverse or any exceptional items.

And I think this is a very encouraging number. Slightly lower technical margin due to non-repeat of the one-off from last year.

Higher fees and revenues, very much in line with the growth in revenues with the growth in volumes. So a very very expected number.

And lastly, we are -- there is a negative component, which is the -- which are the expenses, of course, expenses increased revenue with revenue, but we have 2 other elements, which are first the VBI due to lower interest rates, it's quite mechanical. And second, some more investment in notably in quality of service, to improve the quality of service.

And Thomas insisted on the importance of the Net Promoter Score and the quality of service with the customer first approach. And so this is absolutely something which we accepted.

On the health side, I told you, very steady growth, revenues up 6%, earnings up 5% with a balanced growth between the individual part and the group part, which we think is a very good balance, with a slightly higher combined ratio with a lower expense ratio due to the growth, but also with some less favorable group claims experience, and I -- we expect a reversion to the mean in 2020. Asset Management.

I call the performance of Asset Management sort of flattening. It's not for me a decrease economically speaking, because the assets under management are going up by 10%, mostly from the market effect and partly as well from a net inflow effect about €13 billion.

So the main explanation for the decrease is a non-repeat of performance fees of last year. So if we exclude that, put this aside, it's a relatively flat number.

What is interesting is the development of the alternatives. So the management of nonquoted, nonlisted assets.

This is an expertise, a few companies have. We have it at AXA Investment Managers with Real Asset, but also with structured finance.

It's growing big time, plus 13% to €137 billion, and it's high -- it has a high profitability. So if you make the math, in principle, everything being equal, you should expect an increase in earnings next -- this year, in 2020.

Let's move now to the roll forward from underlying earnings to net income. So of course, the number looks good with 75%.

But it's something which was expected because we had the last year, the goodwill impairment of Equitable, which is nonrecurring, of course. So what are the real elements to understand the difference between underlying earnings and net income?

First of all, the usual program of capital gain, mostly driven by real estate, France, Belgium. Second, change in fair value, we have 3 subcomponents there.

The first one is equities, equity hedges. We have equity hedges.

So when the equity markets are going up, the value of the hedges are going down. I want to reinsure -- reassure you the increase in the shareholders' equity is much higher than the decrease of the value other derivatives.

So economically speaking, no problem. Second, we have some ForEx hedges.

It's exactly the same. We have, with the weakening of the euro against the other currencies, a stronger balance sheet, stronger earnings.

However, the value of the derivatives is going down. So all these elements are quite mechanical.

And lastly, we have interest rate swaps because the financial debt of AXA SA has been moved from variable to fixed. So when the interest rates go down, you have a negative value.

So these are the 3 components. Then we have, of course, the big part, which are the exceptional and discontinued operations, with the really -- and those elements are and where expected related to equitable AXA Life Belgium, AXA Life Europe.

Integration and restructuring costs, 75% of them being incurred at XL. Moving to the balance sheet.

So the shareholders' equity is going up, which is a good news, but also something which is quite mechanical because the main value driver is the change in unrealized capital gains. Fixed income assets are going up by €5.9 billion, motivated to the drop in interest rates of around 50 bps on average.

And equity is up €1.7 billion, motivated to the average 20% market increase in '19. Net income for the period and dividends are clear.

Change in pension benefits, driven by low interest rates, mostly in Europe. For your information, tangible net asset value is up by €10 billion to €36 billion.

Following the higher unrealized capital gains and lower that from Equitable Holding, which is something which is interesting, and I just wanted to flag it. Return on equity up to the 16%.

We think it's a very solid number, both on an absolute level and relative level. Solvency ratio, up 5 points 198%, which is, we think, a pretty good number in the low interest rate environment.

Banking, in line with our target range of 700 -- 107 -- sorry , of 170% to 220%. It's explained, of course, by the recurring operating return.

And we have, as expected, and very much in line with the key sensitivities we have always been publishing. The interest rates impact was a negative in '19, partly offset by the sell-down of our U.S.

activities. So as a whole, in the context, we think a solid number.

You look at the key sensitivities on the right-hand side, we should not forget them. But this is something which is manageable.

And we wanted to give you some indicative outlook for 2020. So I think most of the elements which are here listed are not new to you, operating retail, net of dividends, disposals, the conversion of AXA Group from a standard model to an internal model and the deleveraging, we intend to pursue to reduce our gaming.

However, I would like to highlight the seasonality. Everything will not be in AR.

So when you make your projections, don't forget from quarter-to-quarter, these elements will play, certainly, in '19, but not all at the same time and in a linear way over 2020. So here, no other signals, that's just the seasonality and the prospects to get some more Solvency II points this year.

We spoke about the adequate level of reserves, Thomas in his introduction. I get this message as well when speaking about the P&L of XL.

Here, we have 2 messages. The first one is that the level of prudence has increased at AXA, excluding XL.

While the second message, the level of excess reserves at XL remains sufficient. So if we go on the right, if we look at the left-hand side of the slide, you see that what we hear, what we compare is the IFRS reserves, which is accounting, and we compare this to the best estimate liabilities and discounted to make them comparable with the IFRS reserves, which are themselves, undiscounted.

In the previous presentations, we were comparing IFRS reserve, I think, with discounted best estimate liabilities, which made the -- which gave the impression that the number has higher. But here, this one is more comparable and more fair, I would say.

So you see that excluding XL in '18, we had €5.2 billion. Here with XL, we are €6.8 billion, with an increase excluding XL, which is relatively strong.

And despite all the prior year reserve development, you have been looking at. On the right-hand side, we want to explain to you how we manage the P-GAAP because I think this is a recurring question.

We started the year with €1.5 billion, which includes as well, AXA Corporate Solutions excess reserves. So all in, €1.5 billion.

There's excesses were unallocated. So what we have decided this year is to start allocating the risk we had identified in the opening balance sheet to certain lines of business, mostly on the long tail casualty lines.

And we have -- we are left with €0.8 billion at the end of this year. So the -- I would like to make something clear.

This P-GAAP is not a sort of buffer we booked just in case. It was a result of audit reviews, clearly identified a certain number of risks, including social inflation.

So therefore, it's absolutely normal to use part of it to -- when the risk appear and to allocate them. €0.8 billion is a number which is -- which has been reviewed in the recent weeks by the group.

So it's not only XL. You have always the entity making as a depreciation of reserve and the group review.

And it makes us comfortable with the level of reserves of XL at the end of '19. And I don't see the risk from now to have a deficit on this side, so relatively comfortable.

Cash at the holding level has been reinforced from €1.7 billion to €3 billion. Gross financial debt has been reduced by €5 billion.

€1.8 billion at the level of AXA SA, the rest being linked to the financial debt of Equitable Holdings, which has been deconsolidated. But it's very significant, and it explains the debt gearing.

So the debt gearing,we told you that we would get back to debt gearing level comparable to the pre -- to the levels before the acquisition. So our objective is to be between 25% and 26% this year, so 2020.

And everything being equal, if you reimburse between €2 billion and €3 billion level of financial debt, we are there. So it lies in -- it is in our hands, I would say, everything being equal.

So the drivers for the evolution of the cash, our first to cash remittance, €5.6 billion, which includes €0.5 billion of one-offs, including the Swiss -- the proceeds from the Group Life transformation in Switzerland. Dividend paid, which is normal holding costs, including the interest charges.

Then we have the M&A proceeds, I would say, pluses and minuses, but the Equitable sell down was a major event. The reimbursement of the external debt at the level of AXA SA, cash buffer we build, which is that there is, I would say, fungibility of cash within the group.

We used some of this fungibility to fund the XL transaction, and we have committed to come back to the pre-acquisition level. So we have reimbursed part of it this year.

We'll certainly reimburse the other part in 2020. And the other is a very, very different items, including margin calls.

This is the end of my presentation. I hand it over back to Thomas, and thank you for your attention.

Thank you very much, Etienne. And now we have the opportunity -- or you have the opportunity to go into Q&A, and we are first deal with the Q&A in the room.

And then later on, go to the digital Q&A. Let's start with Jon.

It's Jon Hocking for Morgan Stanley. I've got 3 questions, please.

Firstly, overall XL, I'm afraid. So firstly, looking at the casualty reserves.

So I think in the narrative, the Company results, you're talking about litigation financing, financial lines, et cetera? I wonder if you could give us a little bit more commentary in terms of what specific factors you take into account when you were setting the reserves at year-end?

That's the first question. Second question.

Just looking in the B slides, the B appendices, you have the B 17. If you look at the sort of volume changes versus pricing, it looks like the exposure actually went up the most in casualty and financial lines, which seems a bit surprising given those are the areas where the reserving issues emerge in the back end of the fourth quarter.

I wonder if you could talk a little bit about how comfortable you are with that mix of exposure movements over the course of the year. And then finally, just looking at the earnings bridge to get to the €1.2 billion.

I see there's nothing in there in the sense of missing another in terms of into rebuilding the reserve buffer at XL. So is the message effectively that you've got sufficient aggregate reserve buffer at a group level that we shouldn't be too concerned about, the sort of solo XL silo?

Very good. I suggest that Alban, you start with the first question.

And also the second question, and then Etienne, you'll go for the third question.

So the first question was on the quality of the reserves and what we factor in. So when we review the reserves, obviously, we look at different things, and to name only 1 or 2.

One is, obviously, the cost of litigations in the U.S. Only 2% of the claims that we have on the casualty go to court, 98% are settled, fortunately, obviously, before we go to court.

Nevertheless, the court settlements drive the trend for the overall reserves and the claims to be paid. So that's something that we look at and we look at it on a claims by claims basis.

Looking also, obviously, at the previous trends. The other thing that we look also, which is slightly different.

Those, sometimes, it's under the same word of social inflation. It's systemic risks such as [indiscernible], opioid, child abuse and so on.

And so what we do is we review our exposure over the last 10 to 15 years on what has been our underwriting on those lines in the past. And we reassess what could be the losses given what we hear, what we see in the market on those fronts.

That's something that we monitor systematically, and that has been the case for a long time because those systemic risks have always existed. And the third part that you should also keep in mind is that those casualty lines in the U.S.

are significantly reinsured through quota shares between 40% and 60% from one year to the other. So that, obviously, also reduces the impact of what I described earlier in terms of litigations of settlements.

The second question was on the growth in casualty and in growth in volumes in casualty and financial lines. So I think there, we need to be extremely clear that if you look at financial lines, you have a number of different aspects in that and different aspects in terms of geography, U.S.

versus Europe, and different aspects also in terms of customers, large versus mid-market. If you look at the financial line and notably, DNO, the issue was more in the U.S.

on large corporates. It is much safer, much sounder when you look at mid-market and European.

So we shifted the underwriting from the latter to the former. And you've seen despite the reduction in U.S.

launch, that overall, we have grown the volumes on selected segments.

Thank you, Alban. So to question, Etienne, on the €1.2 billion.

Yes, on the €1.2 billion, we have absolutely not reflected any release of provision of any kind. So the P-GAAP was set in place originally to face with the development of prior year claims up to the end of '18.

So this -- we consider that it will -- there will be -- the P-GAAP will neutralize the possible adverse elements from these years. When selling and this -- so this is neutralized.

When looking at 2020, and you'll see how the work forward is constructed, we just have a sort of economic approach on this, right? So we started with €0.5 billion, which embedded some good news, some bad news, but more bad news than good news.

And then from this, we look at the impact of our actions, the impact of the pricing. And the cost of the -- which is not the cause of earnings impact of the volatility measures, and this is how we come to €1.2 billion.

So in the €1.2 billion you don't have a positive from the reserves of a release, if there is some allocation within the P-GAAP, it will be only to compensate some new views on other risks, which have been identified during the P-GAAP and not yet allocated, which is a different mechanism. So we are not -- and I think this is a key message.

Why do we believe in this €1.2 billion? Because we have all the elements, which mechanically expect that in the normalized year, it will happen.

And as a CFO, I feel very important of this target.

Nick?

Nick Holmes at Soc Gen. A couple of questions again on XL.

I wondered if you could give us just a little bit more color on the decision-making behind the €600 million allocation, what were the key drivers? Was it environmental?

Was it changes in your own conservatism, if we can call it that? And also, with the €800 million, you sound extremely confident that this is a sufficient number.

Can you draw on any sort of illustrations relative to market, relative to your own perception of risk profile, why you do think that this is sufficient? And then secondly, a rather simpler question, I hope, is how confident are you that XL's internal model will be approved in, I think, you're saying, Q4, which is an acceleration?

Is it not? No, Alban, you're saying no.

We've always said end of 2020. So if it is in September, it's an acceleration.

Good. 2 questions.

I think, again, for Alban. On the reserves, how comfortable are we on the €600 million or the €800 million, and then the question around the internal model.

So on the €600 million. As Etienne said, when we bought the Company 18 months ago, we had identified a number of potential risk or potential issues, that led us to the booking of the P-GAAP reserve.

And now we're saying, okay, we have more certainty on the risks and the need for booking that -- for identifying those reserves. And that's related to exactly what I described a minute ago.

In terms of U.S. casualty, mainly where we had seen that there was a deviation because of social inflation.

And social inflation was -- did not appear in 2019 nor in 2018, we had identified that at the time of the acquisition. So it's just the realization of this.

Now the €800 million buffer. At this stage, as Etienne said, we are comfortable.

Now we know that it is a business where you have some uncertainty. You have potentially large cases, and that can go one way or the other.

But we think that the €600 million allocation today is sufficient, and that the €800 million is a good enough buffer against that potential volatility. Again, having done that analysis of our potential exposures on the path to casualty, social inflation and systemic risks.

And on the internal model, so we have discussed at length with the ACPR, the changes in the model and the fact that we want to include XL into our model. The formal application is in June.

But by then, we will finalize the discussion. And you know that you submit 6 months before the approval.

So we will submit in June in order to have XL internal model in December, as planned.

And, Nick, just to add to that, on the casualty piece. I mean, when you look at the whole market, if your prices are increasing by over 40%.

If your exposure is significantly reduced and if some competitors, which was the case even exiting the market, you have an issue of insurability. And so what you've seen always in the past that when there are these anomalies, at some point, it will come back to normality.

James, maybe you can just hand the mic, its easier to.

James Shuck from Citi. So 3 from my side.

I just want to go back to Slide A 30, if I can. So the margin over best estimate.

And I'm just initially thinking of the difference between the €5.2 billion and the €6 billion. Can you just explain to me what drove -- so excluding the XL part of it, what actually drove that increase in the margin over best estimate that you're showing there?

And kind of linked to that, if I look at that what's left in the P-GAAP margin in XL in relation to the reserves, it's 3% of XL reserves. And the total of the €6 billion is closer to €12 billion.

So kind of returning to Jon's question earlier, at some stage, will you look to increase the XL margin over best estimate more towards the group adequacy level because as the group adequacy level that's driving the expectation for the 1.5 to 2.5 points of PYD? That's my first question.

Secondly, Thomas, I mean, the €1.2 billion in 2020, that includes about €200 million to €300 million, probably, from legacy AXA Corporate Solutions. So if you look at the €12.4 billion acquisition price, this is post synergies you're doing this.

It looks like a kind of 7% or 8% ROI on the deal. I'd just like to get your insight into whether you think that's a sufficient number?

And then finally, just around the dividend. You -- I think you've been pretty clear that you would grow the dividend in line with underlying earnings.

You increased the payout ratio of target range last year from 40 to 50, to 50 to 60. We've got this year, which is depressed earnings.

So you can argue whether you've hit 50 or not. At some stage, you need to rebase that dividend up.

Otherwise, what's the point in increasing that target range? If we're not going to get that this year, then we're going to be next year, perhaps.

But just like some guidance over how to think about that payout ratio, please?

Very good. I suggest Alban, Etienne, on the first question around the margin of best estimate.

And then linked to the question, P-GAAP, 3% of XL reserves. Are you bringing it to the group adequacy level?

And then I will talk about the other 2 questions.

So on your first question, to the same extent, we review the XL reserves very regularly. We review the whole group reserves very recorded.

And we assess what is our best estimate liability for Solvency II purposes. It so happens that over the last year, the difference between the best estimate liabilities and the IFRS reserves has increased.

Why? Simply because we can have a change in the best estimate reserve, in best estimate liabilities that does not flow immediately through the P&L.

And it might flow afterwards. But then when you look at the PYD that -- the PYD positive that we release every year that comes fundamentally from that difference.

Do we want to bring XL to that percentage? No, not specifically because we want to manage this at group level, and XL is 1 division, among others.

The excess of a best estimate is not proportionately dispatched among the various entities of the group. For instance, Etienne had said that this year, we had positive project developments coming from Ogden.

So it can pop up in 1 entity and not 1 -- not another. And we're not planning to have a proportionate allocation in all our entities.

So on the 2 other questions. #1 is the question around the €1.2 billion in 2020, is it post synergies still an attractive deal.

Again, let's go back to the rationale of the deal at that time. The rationale was, how do we reduce our exposure and financial risk and how do we move towards technical risk.

Always looking at the question, how do we make sure that we are not too much exposed to interest rate risks, where we have no lever to act. So the first question would be, let's compare the exposure to equitable against the exposure to XL.

Fortunately, we have sold the totality of Equitable, but for those of you who have followed Equitable, you have certainly seen that in the last year, towards the end of the year, AXA Equitable has made a significant increase necessary in reserves when it comes to policyholder behavior. This was just below €1 billion.

This would have certainly hit us had we owned 100% of Equitable. For this amount, you can allow yourself quite a few natural catastrophes.

Second point, I wish AXA Corporate Solutions would have made €300 million. Unfortunately, they made only between €150 million and €170 million, and would have been hit exactly in the same way on the topics that we've been discussing.

So no separate treatment. When we come to the question around the €1.4 billion or the €1.2 billion, at the end of the day, nothing has changed.

The €1.4 billion has been designed in AXA XL, or XL at the time, plus corporate solutions plus the cost synergies. We are well on track when it comes to the integration, so the cost synergies are flowing through.

We still have not achieved everything, but they are flowing through. We are working on the profitability of the portfolio, which you've seen, we have taken very clear measures in up pricing the portfolio, again, 10% increase of a premium, 10% increase on net earned premium of €10 billion is certainly a significant upside that will come through the earnings.

And then thirdly, we have taken a decision which was after the combination of the acquisition of reducing volatility. So when we talk about €1.2 billion at reduced volatility over €1.4 billion, we're talking about exactly the same thing.

So for me, nothing has changed. The strategy is exactly the same.

And again, coming back to my first point, have we taken the right decision of shifting our risk profile from financial to technical risk, definitely in this environment. Your last question on the dividend.

When we speak about the dividend and when we look at how do we determine the dividend, we always said we want to have an adequate and attractive growth in the dividend. So what we'll do, we'll look at where do we need to position the dividend, how should we be increasing the dividend and the payout ratio is a resulting figure of it and not a determining figure.

And so that's why you see, this, as a Board, we have been very comfortable to have a 7% increase in the dividend, which is more, as you stated yourself, than the 5% underlying earnings per share, but we took account of the fact that we had more than normal natural catastrophes. And by the way, this was the same for the full year of 2018 that came to the 7% dividend.

And that's how we will continue to look at it. The first question is, how do we get to an attractive dividend growth and the resulting figure is the payout ratio.

And I guess, what you're saying, between 50% and 60%, we are still at the lower end, and there is a big room for flexibility.

Could I just clarify 1 point there? Just, I mean, one way of hitting that is to shrink earnings, I suppose, let's hope you don't go that way.

But if you want to go --

I wasn't thinking about it this way.

But if you want to -- I mean, if you're going to raise the target range, I mean, are you categorically ruling out a rebasing of the dividend to the midpoint of the range at some stage and then grow in line with underlying earnings?

The reason why we changed the range at that time was exactly the change in the risk profile and the taking account of the fact that we have more volatility through natural catastrophes. You've been following AXA for a long time.

The volatility of Equitable, which was also very high, was always absorbed to the DAC. So you didn't really see the volatility because it was inherent in the deferred acquisition cost.

Here, we don't have that mechanism. What is volatile here is volatile at the bottom.

And so we said, in order to take account of this and to make sure that we have the flexibility of breathing, we need to move from the initial payout ratio of 45% to 55%, to 50% to 60%, but always in the spirit, #1 criteria is we need to have and determine what is attractive dividend growth and the payout ratio is the resulting figure. Let's move -- sorry, I can't see who it is because the light is very difficult.

It's William Hawkins from KBW. Picking up on one of the points you've just made to James and you've made a lot about the improvement in the volatility in XL.

I'm surprised that you have been clear about the earnings hit of your derisking but you haven't been clear about the capital benefits of your derisking. I mean, if you bought more reinsurance protection, your volatility has gone down, there should be some capital release from that?

So if you're able to say to us, look, we've freed up €2 billion of capital from this, for example, that would usefully offset the downside to earnings. So I'm surprised you haven't been clear about the capital freed up, could you be clear about the capital freed up, please?

And then secondly, with regards to the cash buffer rebuild, Etienne, that you showed on Slide 31. You showed that €1.9 billion of cash buffer rebuild.

Can you just remind us where we are with regards to the stock on that number? My original memory, correct me if I'm wrong, was that when you did the XL deal, you talked about cash at hand of €3.5 billion being spent.

If we just subtract the €1.9 billion from that, are we saying that you've got another €1.6 billion to go? Or is the number bigger or smaller, Alban?

Those were your 2 questions. So thanks, Will.

First one is for Alban. Is the capital release from the further protection.

Second one for Etienne, where is the stock on the cash.

So I'm afraid, I have to be a bit technical for a second, so bear with me. The -- when you reduce the -- the way our internal model is built, there is a benefit of diversification.

And when you look at the investment risk, there is little diversification from one country to another, because when you have a crisis, that goes everywhere. So when you hedge your financial risks, you do have, immediately, a benefit on your capital.

Going to your question on hedging or reducing our insurance risks. Because there is a lot of diversification, because when you had a windstorm in Japan, you don't have, at the same time necessarily, a casualty issue in the U.S., reducing hedging part of the book doesn't lead to a significant amount of capital reduction because of the benefit of diversification from what is done locally to what is needed at group level.

You see? So if I were to do a 50% push of my whole book in P&C, yes, you would see it.

But reducing the cat nat from €500 million at 20 years to €400 million, you would hardly see it in the capital required.

Farooq, I think, yes. Sorry.

So the second question was --

Sorry. sorry, sorry, sorry.

Yes, the stock. I'm sorry.

Too fast.

Yes. So around €2 billion.

I'm Farooq Hanif from Crédit Suisse. Firstly, on the walk to get to the €1.2 billion.

When I look at the €600 million uplift, I think to myself, you talked about €400 million pretax synergies. You had €200 million of XL's large losses last year.

Pricing drive the top line, but also margin expansion, presumably. So what's the negative factor in that €600 million that you haven't really talked about, is question 1.

Question 2, I personally am quite surprised that you are reducing risk in property. I agree that you've got a reinsurance business, and it's volatile, and you probably want to reduce exposure there.

But given where -- given the strong increase in property pricing, also the fact that it's got a short-tail business. So yes, you reduce volatility, but you don't get it sort of biting you, if you'll excuse my French here.

Then at the backside, 5 years later, it was what we're seeing in social inflation. So why tilting more towards long tail business, which pricing might be going up, and it's going up for a reason?

Just wanted to understand the philosophy. And then lastly, on the cash holding, I mean, you started to show this number now?

And I was kind of wondering what significant it has? So next year, obviously, you'll get this -- it sounds like another €2 billion hit from buffer rebuild but over time, we'll see this as build given your payout ratio, given your earnings.

So I'm just kind of wondering what level do you -- would you like to have at the holding, what -- why are you showing this number, in essence?

So I suggest, Etienne, you take the first question, why €0.6 billion were the negative factors. Then the question on reducing property are quickly answered.

And then Etienne, you'll go back to the holding question. When it comes to property and when it comes to taking risk, it's always the question of what you'll get for the risk.

You are absolutely right that prices are increasing. And what you see as a phenomenon, and Scott will certainly agree with that, that prices in primary insurance are increasing far more than prices in reinsurance.

So you have roughly a factor of 1 to 2. So we have in primary insurance to 10% to 12%.

In reinsurance, 5% to 6%, if you look at the renewal. So yes, prices are increasing, but you have to ask yourself, are they increasing enough so you can make a decent return.

And what you've seen, certainly in the property area is that there is a lot of alternative capital streaming into this market because they have exactly the same logic that we have applied with our own portfolio, how can I go into risks that are not correlated with financial risks. And if I was a pension fund not being subject to Solvency II, not maybe being subject to the same economic rationale that an AXA is, not having a 16% return on equity, they will come in with a very different view, and they will write that risk for, I don't know, 5% return on equity.

This is not what we can afford. So you have to see how do you balance your exposure that you, yes, get the right price and get the premium increase, but also that you get the right return.

And at the moment, in particular, in reinsurance, we are, in some areas, and notably in Europe, not at the right levels of return. Therefore, the exposure reduction, you've seen that the 10% exposure reduction that we mentioned is linked to the reinsurance nat cat area, and that's the rationale.

Etienne, question number 1 and 3.

So we start -- the question will refer to €0.6 billion. The €0.6 billion reflects mainly the pricing effect, net of the possible claims inflation.

That's it. So when you start from the €0.5 billion, of course, you have €0.2 billion deviation in large losses.

We have not said, for the large losses, that there would be a reversal to the means because this works for the natural catastrophes, we have models. For large losses, it's not the same mechanics.

So we have -- we are not averaging the large losses. They are in the €0.5 billion, and we consider that this day.

I don't say this is what will happen. I just said this is how we build our forecast, right, in a very structured way.

So it's mostly the pricing at an inflation. There are, of course, some cost synergies.

But as Thomas was saying, they are well advanced. So if we speak about big numbers, I really think this is the pricing, net of further claims inflation.

Knowing that what I said for large losses, it's so true for the social inflation impact, it's in the €0.5 billion, and we don't say, it comes back to 0 immediately and we'll rebase everything. So €0.6 billion encapsulates all these different items.

The second question, why do you show that? I must confess that we'll get always a lot of questions about the cash variation in the holding.

And therefore, I thought you would be pleased to see that and that you would not ask us why do you show us what we want you to show us. So thank you very much for your question, but I was not prepared for this one.

So seriously speaking, we are, on the long run, we are happy with some cash buffer at the -- level. Just in case there would be a very stress test for, let's say, 1 year of financial catastrophes or whatever.

So between, let's say, €1 billion and €3 billion, we're happy. The intention is not to build like Microsoft, hundred billions of euros in cash and not returning it to the shareholders.

So the message here is really bringing clarity on what are the flows so that you can better understand. There is no hidden message.

And we will not put as a headline, the level of cash in the holding in the front page, never. It's just to give you a hint of where we are and how we manage the cash.

Just a quick comment, rather the question. It seems to me that you've lost the cost synergies although or it's the general revenue and cost synergies because they weren't quite there in 2019.

We didn't really talk about them in 2020, you are going to come, but now that seems to be gone. So yes.

No cost synergies starting in '18 continued in '19. I must confess that in terms of cost, we had some slightly positive element, nonrecurring for next year.

But the synergies will continue to play in 2020. But if you look at -- if you compare that to the impact of pricing, it's much lower.

But don't forget that it was already in the '18 and '19 numbers. So when thinking about '20, you cannot take the €250 million total announced because part of them were already in the previous numbers.

Let's just move, sorry, I can't see. Because --

Ashik Musaddi from JPMorgan. Sorry to go back on that €600 million number again.

I mean, if that is net of price, net of the claims inflation, I mean, was that already expected when you gave the €1.4 billion number because pricing, more or less, happened last year. And €1.4 billion number was given before that, if I'm not wrong.

I mean, I've just started following AXA more closely for the last 6 months. So any thoughts on that?

I'm just a bit confused on that one. So that's the first one.

Secondly, you mentioned that some of the pricing improvement will follow in 2021 as well and more volumes as the 10% volume growth will follow in 2021. So how should we think about that?

What sort of numbers are we talking about? Is it an incremental number?

Or is it still within the range of around €1.2 billion? And lastly, just 1 simple question, maybe a bit naive question, but you mentioned that the volatility, it would most likely go down?

I mean, what are we talking about here in terms of volatility. I mean, let's say, we have another 2019 and 2020, what sort of XL number would be?

What sort of profit XL would print?

So Etienne, it's best if you take those 3 questions. I think for the third question, we should maybe have a look at one of the slides, B 39, or whatever it was.

So 59 -- I can't remember.

52.

52. Sorry.

Exactly. Etienne?

So I must say that back to '18 when the target of €1.4 billion was set, a lot of things were different. So we expect certainly a lower level of large losses and we expected social inflation, but not by the magnitude and the speed of increase we are facing now.

So -- and of course, on the -- at the same time, we didn't expect such a high level of prices. So I think it's pretty clear.

And therefore, a lot of things have changed. And if you look at -- at the end of the day, what Thomas was saying, what's the difference between €1.4 billion and €1.2 billion.

It's just the same with less volatility and less risks. So this is another way to look at it.

So yes, certainly, we don't expect so much volatility on the market -- on this market. And to make the link of your second question, which is what happens in '21, '22.

Of course, the €1.2 billion is our forecast as of now for this year. We don't set a target for '21, for '22.

What you can just mechanically expect is that the pricing increase as we see today, we continue to produce some positive -- incremental positive results, but we don't want to quantify at all, the -- and it's -- we don't expect a minus for '21. It's -- we expect mechanically a plus, but we don't want to quantify it today.

Let's have a look at the last question, which is, is volatility actually going down.

So the horizontal line is the best estimate in terms of cat charge in our forecast today. So if you were to draw the line more on the left-hand side, we would come with 1 out of 2 year with everything at €800 million.

So when you look at what happened in '19, you see that our experience was much above €1 billion, it was between -- so 1 out of 5, or 1 out of 10 years. We were not far away from €1.3 billion, something like this, net of insurance.

So the supplementary measures we have taken is -- are now bring us to a result, which is the right -- the dark blue bar, which shows a decrease in terms of volatility of around 25%. And second, what's interesting is that you see that there is an impact short-term for the 1 in 5, but it's flattening as well up to 1 out of 50.

So it's a very -- it's -- and this is where it's most effective, especially if you look at the euro numbers. So when looking back at '18 and '19, we would have saved in '19, $100 million.

In '18, $200 million. So there is a real impact at the XL level.

Let's move over to this side. Far right.

Michael [indiscernible] from Berenberg. I had a question.

So you mentioned that -- so XL earnings will go up. It sounded as if life earnings will go up because they were a bit depressed this year.

It sounds if asset management also, you're very confident if I had add all these, I'd get to €1 billion more, which would mean that your earnings would be up like 13%, 14% this year. And I just wondered, is there anything I'm missing either on the plus or maybe on the negative side here?

The other thing is on the debt. So you have a slide where you use the word, limit, limit down, I hope, of cutting the leverage to the bottom end of the target range, 25%.

And I just wondered what will nudge you in that -- at what stage will you decide or we'll do the maximum or we'll do the minimum, and what will influence that decision? And the last question is, is that France which seems to be the star of the show today, is that how sustainable is that €1.7 billion?

What can we expect this year?

Etienne, 3 questions for you.

Thank you. Maybe you missed the contribution of AXA Equitable Holdings, if I understand your question well.

So you will have positives from -- so asset management, even if it goes up, it will not be very significant in absolute numbers. But the contribution of AXA Equitable will disappear fully next year.

So I think this is the, maybe, the missing element in your projection. Second, on the level of debt, I would say the -- it's pretty simple.

The -- I told you, the main drivers reduction of debt, but you must have available debt, if you see what I mean. So that, which -- where we have the possibility to reimburse in advance.

So if you make the math, you will understand what kind of number we are talking about. And then it's purely a question, and this is an information we cannot disclose, what are the debt, specific debt, we are going to reimburse.

And your third question, it was for a Jacques.

AXA France is €1.7 billion sustainable. I think that was the question.

Thank you very much for the question. Concerning AXA France, we have a very, very solid engine that is growing each year at the same level.

So it's not new that we are performing. So the question is are we going to perform in the long term, definitely, yes.

If you look at, for example, the P&C business. The P&C business, we are a 90 combined ratio now, but if you look at what are the level of the improvement is definitely a very strong level.

It's level concerning the current year reduction of claims, and that is due to our underwriting, which is each day improving, and it's also due to the reduction of the -- the decrease of the expense ratio, and that is just the fruit of the work we are always doing in order to be better. And if you look now at the life business.

The life business, we are at a tipping point now, and we are jumping in a new era where we have a wonderful, wonderful potential of cost. And this potential rely on 4 pillars that you have to have in mind.

First is a reform of the pension in France. We -- and this reform is done, it's not the reform that is discussing now at the assembly, I am speaking about the reform done that went into effect in October with the backload.

This has created a tremendous potential of development of our volumes, but also offer profitability. We -- as you know, just as a reminder, we have now an attractive product, you can take -- you can go upon the payment with a lump sum, you got -- you have portability.

You have the remaining -- the keeping, let's say, of the favorable fiscal treatment. And on this market, the pension market in France, we are a leader with roughly 18% market share.

And if you look at the expectation of the government on this market, the government expects to go from -- in 3 years -- by 2022 to go from €220 billion asset under management to €300 billion asset under management. It's a 40% growth in 2 years.

And if you look at the revenues that, that involve, it's a triple of the revenue each year to reach this target, and if you look at the first month on this new market that is opening. And you look at it, we -- on the first 3 months, we are -- we have improved, we have increased our revenue is 60%.

And if you look at our market share, it's even more than the last market share. We used to have, our market share is 22%.

And if you look at the end, and if you know that this market as a 60% plus unit-linked rate compared with the market we have in saving with, let's say, a €27 million rate at the bucket, you can calculate with math, what will give for the leader in France has a very strong track record on this market and in taking the opportunity. And you know that we have been the first one in France to deliver this product, a few days after the law took into effect.

Second pillar, Eurocroissance is now €200 million of euro, it has increased by 40% compared with last year. And these Eurocroissance, I think that in the shift of the behavior that our client must have in her head due to the change of the environment of low interest, these Eurocroissance will be a very strong answer, a very strong response to what they need because it will permit them to have a guarantee at the end of the contract, at choice, but also to have a potential of a good performance.

The third pillar is innovation. Innovation on Health.

We have a large bull bar in front of us because you know that we have always had a very good UL rate in France, in AXA France compared with our peers, and I can tell you that this year, with 39% UL rate, we have even grow the gap between us and the market. The gap is now 11-point, and we are -- we can still grow it through innovation.

Innovation, we did it, we structured fund, that was in the early '20s. But we did also in 2012, with real estate, we are doing -- we will do it with private equity in the few months, with thanks [rampant] that have softened, let's say, the whole concerning private equity in that.

And the last but not the least, is will concern our contract, the term of our contract. We are able to adapt the term of our contract to the low environment we face -- the low rate environment we face.

We are -- we can modify this term in order to adapt the guarantee to what we can serve. And also, the last one, the underwriting policy.

You know that. All the French insurer are taking now more stringent policy underwriting to raise the unit-linked rate.

So with thanks to our track record, thanks to our expertise, thanks to the strong links we have with AXA Investment Managers in France. We are very, very confident in the capacity we have to grow the underlying earnings in the coming years.

Thank you, Jacques. Michael, you must be convinced now.

So. Let's just move forward to you.

Colm Kelly, UBS. Most of my questions have been asked and answered, but I have a few remaining.

In AXA XL, to what extent have the changes announced today impacted on the diversification benefit do you expect to achieve with the merger of the internal models. So the risk exposure here is measured by best estimate of liability has increased, but also there's an accelerated mix shift that is different to what was expected at outset from the transaction.

So to what extent will that impact on the capital synergies as dictated by the diversification credit? And then secondly, related to that, you have the guidance for capital synergies of 5 to 10 percentage points.

But I suppose, what's driving that confidence ahead of discussions with the ACPR around what that number will end up at? I mean, it's difficult to predict what the outcome of those discussions between regulatory authorities and companies will be and the diversification credit is always a key lever of change in those discussions.

Then lastly, on the operating free cash flow, which typically uses a good measure of dividend progression and capacity. The operating free cash flow was down 5% this year.

But hasn't it either exhibited growth over since 2015, it was pretty much the same number. So I mean, how important is this in setting the dividend.

I mean, it's -- the cumulative free cash flow is one of the main targets within the Ambition 2020 plan, it's not exhibiting any growth at the moment. So I suppose, how is that supporting confidence to grow the dividend at 7% this year?

And when I look at the next plan, how does that give me confidence in the progressive dividend strategy within that plan, if the cash flows themselves are not exhibiting growth?

So before Alban is answering the first 2 questions, let us remind ourselves that we're here to discuss the full year results 2019 and not the plan 2021 to something. So when we are ready to discuss the plan.

We should go back to the question around cash flow. I don't know if Etienne wants to make statement afterwards concerning Ambition 2020, but let's be clear, we're not talking about the plan today.

Alban, on the question about the diversification benefit, business mix shift. And then the question on the capital synergies, what is our confidence?

So the -- your question on the change in business mix and its impact on the model. It's a bit the same question as the one earlier on the release of capital or the absence thereof.

It's not significant enough, that change, to impact the amount of capital that we need to hold at group level. So don't expect to change coming from that shift.

And on your second question, we started the discussions with the ACPR. We have explained to them how we plan to integrate XL in our internal model.

Now those discussions are still being held. We have not come to a conclusion yet.

And I would say, nothing leads me to believe that the 5 to 10 points is -- wouldn't be achieved. But I don't want to put pressure on the ACPR by saying this, and so that will come in the next month.

Good. Let's move on.

Let's go here. Sorry, I can't.

Yes, the cash one. I think I just checked with Etienne.

We'll come back to that when we talk about the plan because for the existing period, we have seen €28 billion to €32 billion, we are at €25 billion. There should be no reason to believe that we are not achieving the range that is targeted for Ambition 2020 and what everything else that concerns afterwards, let's deal with it when we talk about the afterwards.

Thomas Fossard from HSBC. Two questions left on XL, sorry to come back to this one.

The first question would be regarding the claims inflation you're seeing coming into your book? Actually, could you help us to better understand how this compare to the 8% price increase in commercial lines?

Or even better, how the claims inflation is currently versus the 14% price increase that you have recorded in Q4 stand-alone that will help us to better understand what is potential underwriting margin improvement in 2020 or 2021 that you're talking about? The second question would be related to Slide B 17, where, in fact, just splitting XL by different business lines.

Sorry to insist again, but I think that from this slide, it's still difficult for us to understand what is XL exposure to the most tricky line of the U.S. casualty market at the present time.

What do you have in terms of business on lead umbrella, on the excess liability, which is this part of the book where you're ensuring Fortune 500 in the U.S.? I think that this type of granularity we're still missing.

And also on the premium in the reinsurance side, where you've got order of €3.5 billion. What is the volatility of the business which is behind?

Good. Thank you very much.

Etienne, Alban, do you want to talk about the first question. We'll take -- sure, Alban.

Let's not go in so much.

So on the claims inflation, I think the main message that we want to give is, we do believe that we are ahead of the trend, i.e. that the price increases are greater than the claims inflation that we're seeing.

That's a very strong message. But then when you look at the claims inflation and the price increases, what we see, the mix -- the price increase that you show -- that we showed is an average.

On excess casualty in the U.S., as Thomas said at the beginning, we see a 40% price increase in January, 40%, 5 -- 4-0. And so depending on the layer you're at and the business you're in.

You would not have the same price increase because you wouldn't see the same claims inflation. Obviously, when you are in the umbrella layers, you do see a significant claims inflation.

But that's also where the price increases are greatest, and that's also a place where we have significantly reduced the layers there. Typically, on that one, we used to provide a €25 million capacity, we now provide a €10 million capacity.

So that's the kind of underwriting measures that we have taken on umbrella, but the same on others. But obviously, not all of them are affected the same way when it comes to inflation, and therefore, our price increases need to be selective, but significant.

Thank you, Alban. Scott, maybe it's an opportunity for you to talk a little bit about XL and by business line.

And I think it would also be important to shed a bit of light on the social inflation is not a general problem everywhere, it's very focused on some parts, like rail, like commercial auto. Could you talk a little bit about that so we get it more into perspective.

Yes. And in the casualty number, you see that is a basket, as Alban says, that's a basket of goods.

There's motor in there. Our work comp book, we have United States is in there.

And so it's a geographically spread book plus different products. It's not just excess casualty in the United States.

But yes, so social inflation, there's lots of conversation about that. It affects, as Alban says, it affects the excess casualty, particularly on the larger risks more prevalently than the smaller middle market accounts, right?

Then there's just the jury award you see in the United States and the tendency for the plaintiff's bar to ask for more money from bigger companies. You can see it in the settlements and you can see it in the court cases.

So it is not a universal number across the entire portfolio that says everything goes up. But it is still all going up, it just goes up at different levels.

Let's move on.

It's Kamran Hossain from RBC. On results being going, I guess, we haven't hit Coronavirus yet.

So let me just ask about that.

It took a long time.

In terms of, I guess, being prepared for -- I think we can probably ignore the claims in the environment and -- due to Coronavirus. But if you look at SARS post that happening back in the early 2000s, there's a huge spike up in kind of health sales following that.

How prepared are you to kind of take advantage of that? That's the first question.

The next Q, sadly, are both on XL. Coming back to...

Don't apologize for it.

Coming back to the €0.6 billion. To what extent are you already recognizing what you hope on margin improvements on long tail price improvements there?

So we've seen claims inflation spike, to what extent do you think that doesn't happen again? And the second question is, that slide on cat exposures is fantastic.

But I would expect that your cat, your kind of average nat cat number should probably come down because on the 1 in 5-year basis, it comes down quite a lot. So is the 4% nat cat number for AXA XL still the right one going forward?

Good. Thank you very much.

I'll do the Coronavirus one. And then Etienne, if you could do the other ones on the €0.6 billion, and the margin improvement, and the question around does the cat exposure come down.

So on the Coronavirus. Look, the short-term is certainly that we are being affected by it, both operationally, so just to understand, people could not work anymore in the Company.

They have to work from home, which causes always operational difficulties. We have been dealing with it, both in Hong Kong and in China, you see that because people don't leave their home anymore, there is also a negative impact on sales.

Luckily, we've been investing significantly on our digital capabilities. So for example, in Hong Kong, we've launched a service, still under the leadership of Etienne at that time, called Emma, where we have most services being digitally present and the digital interaction numbers have skyrocketed.

So that is actually very good to see that if you can't move physically anymore, you can still move on a digital basis. You are absolutely right that a reminder function of this crisis, means that people are thinking more about health.

And when I think about the upside, that is exactly that. We are focusing on expanding our Health Business with now the fully-owned AXA Tianping.

You've probably seen that we have launched 3 products recently that are exactly focusing on the health area, always in combination of, I've got a health insurance offer plus. I've got a physical offer of health care delivery.

In this case, with Tencent-trusted doctors. So yes, we see a great opportunity once the dust is settled, and we're getting back to normality, and that is exactly where we want to focus on.

Etienne, on the other 2 questions.

So €0.6 billion, the famous now €0.6 billion XL. What I can tell you, I can reiterate.

In the €0.5 billion starting point, we have minus €0.2 billion of casualty deviation. We didn't touch this because we think there is a lag between the pricing, some lag between the pricing and the effect on the recognition of the claims inflation.

So the €0.6 billion reflects a level of price increase, which is substantial, and which is supposed to grow more or less in line with the further potential inflation. So this is an estimate that you have a lot of assumptions behind that, as you can imagine.

And I think the opportunity, just to come back to a question which was related to the Page B 17 on the casualty line, and saying that around 20% refers to the lines particularly hit by the social inflation. So if I come back to the question #3, which is the cat exposure.

I'd say that the main benefit of what we are doing is to reduce the volatility. This being said, it's true that the average load of cat nat has come back from the sort of 4% plus to a 4% minus, which is mechanical, mathematically.

However, in our forecast, we are very near to still to the 4%. The main benefit is -- comes from the volatility.

So last question. And Andrew Crean, you've got the mic.

So I can't prevent you from asking that question.Go ahead.

3 questions. One -- only 1 on XL.

Oh, that's a record.

So firstly, as you're going back, and this is on XL. So going back to James Shuck's question.

I think the €0.8 billion excess of the best estimate is a 2% margin, which looks a bit scarily low. Is it right to apply it against all the reserves of XL rather than just the casualty lines?

I noticed that the P-GAAP reserve of €1 billion was applied just to the casualty line, in which case, you get a margin of 5%, which is slightly more comforting. And then secondly, coming back to Colm's question, could you address the question related to 2019 as to why the cash flow went down 5% in 2019?

So we're not looking forward too far. And then thirdly, your reinvestment rate on new business, I think it was 2% in the first half, 1.7% in the second -- in the full year.

So it was about 1.4% in the second half, and it's probably now down at about 1%. What is the outlook for investment margins.

Do you intend to keep cash in crediting rates to maintain your margin? Or do you look to the policy holders getting rather bad deal?

Thank you, Andrew, for those 3 questions. I suggest, Etienne, you take the first 2.

So the question around, do we have to relate the reserves to casualty only or the total? Second question as well, why has the cash flow gone down in 2019?

And then, Alban, you take the question on the reinvestment rate, and also maybe linked to the question of the crediting rate. And if you need some help from Jacques, because that is very much linked to France, he can happily do that.

Etienne.

So the P-GAAP identified a certain number of risks at the end of 2018, most of them being related to casualty. But of course, they are basically unallocated.

The 0.6% this year are focused on the liability lines on the casualty lines because these are the ones where we will observe a deviation. So the €0.8 billion is left open.

I expect that the big bulk of it will be there to cover casualty lines because we don't see negative or much less negative deviations on the other lines. Second, fundamentally, if we want to increase our earnings and dividend at 5%, let's take that as a sort of average.

Of course, the cash upstream, the operational -- operating cash upstream, will have to grow -- to follow this path going forward. It's purely logical.

So then from year-to-year, you'll have some variations related to specific country situations, which are, I would say, nonrecurring. So generally, we share the view that this cash flow streams, more than the operating free cash flow in themselves, which is a proxy, have to increase by the same level of pattern than the dividend and the earnings per share.

Alban, on the reinvestment rate.

So on the reinvestment rate. You're right in saying that in 2019, the investment rate was around 1.7% on average.

We expect the same in 2020. And the way we achieve that is by benefiting from something that Etienne mentioned earlier, which is the expertise of AXA IM in alternatives.

So we are progressively increasing the share of alternatives in our balance sheet. We're at 20%, we want to move to 25%.

That's a higher yield without more capital, and therefore, it's good for everyone. In terms of margin on the Life side.

So we've done projections with the whole group and notably, with AXA France, with the current level of interest rates and reinvestment rate, similar to the one I've just described. And over the next 10 years, we believe that we can both keep margin similar or close to what we have today and still give a very decent return to our policyholders, notably when you compare with the negative interest rates that you would have today on the market.

Good. Thank you, Alban.

We, unfortunately, have to come to an end. I know there's still plenty of questions.

We will still be here. So if there is any urgent questions, we can deal with them bilaterally.

I want to thank all of you for being here, for your numerous questions. Thanks to my colleagues for having supported us here.

And I wish you a great afternoon, and hope to see you soon. Thank you.