Good afternoon to everyone of you. I'll try to make a very quick presentation, especially today because I imagine that some of you may have in mind some other topics and other worries than simply our results.

So let me start directly on Page 4 of the presentation where you have the main figures in terms of profitability for the group. I just want to make a few comments on this page.

First, keep in mind that from a macroeconomic point of view but also from a group point of view, the third quarter of this year was earmarked with a very significant rebound in the level of activity. And what we show on this page is that we have had very good operational performances which translate, for example, with a gross operating income which is up both for the group and for CASA.

Our net profit is very resilient despite the level of the cost of risk. I'll go back on this question later on.

And you can see that both for the group and for CASA, especially if you look at the underlying figures, the level of profit is very close or even a little bit above the one that we've posted last year for the same period. Solvency is progressing very rapidly and is now reaching record levels, both for CASA and for the group.

And last point on this page, the return on tangible equity for CASA at 10% annualized for the first 9 months of the year is, again, very resilient despite the very high level of capital that we have for CASA. If I go now to the next page, I told you that gross operating income were up both for CASA and for the group on the quarter.

This is the result of the top line, which is up in both cases, and cost line, which is down in both cases. So it means that we have significantly improved our cost-to-income ratios in the 2 perimeters.

In addition to that, we continue to have a very high quality of our loan books and the coverage of the NPLs, which continue to be very comfortable. We've continued this quarter to engage into operations leading to a simplification of the scope of the group with the sale of the residual stake that we had in Banque Saudi Fransi with the sale of a small Spanish bank that belongs mostly to one of the regional banks of Crédit Agricole, Bankoa.

And lastly, with the initiation of the sale process of CACF NL, our consumer credit operation in the Netherlands, which is now reclassified under IFRS 5. In addition to that, we can mention that the main element explaining the difference between the stated and the underlying figures in terms of profit for CASA and for the group are explained precisely by those last 2 elements, the reclassification of CACF NL and the sale of Bankoa.

If I go on the next page, Page 6, you can see an illustration of what I just explained, i.e., the very sharp rebound in the level of activity for the group with all the indicators that are shown on this page that clearly are rebounding very significantly, very dynamically after the lockdown period this spring. And so you see that in most cases, we are now not only catching up with the level pre-crisis but also above this level.

And this is especially important in terms of customer capture, where we continue to grow actually our customer bases in retail banking activities. If I go on the next page, just a few elements to highlight, the contribution of the group to this very specific period for our clients across the board.

We've been very active in granting to our customers state guaranteed loans, with a total amount of close to EUR 30 billion in France. And we have been also very active in granting our customers payment holidays with granted payment holidays to our customers in more than 500,000 occasions.

And at the end of September, 2/3 of these payment holidays had ended. And every time the normal repayment date has been passed, 97% of the payments have resumed without any difficulty showing that actually this payment holiday was very useful and practical tool in order for us to help our customers.

If I go on the following page, just an illustration of the reason why we've proven so robust and so able to navigate through this very special year without any major impact on our P&L. It's clearly the result of the strength of our business model, universal banking model, with a very solid customer base in retail banking activities, more than 50 million customers in our own retail network, plus the addition of a very significant series of strategic partnership, 24 as of today, which give us access to much -- many more customers in order to sell the products of our different specialized business lines.

And clearly, the combination of those 2 elements is providing solidity and resilience to the group very significantly. If I go on the next page, it's a summary of what I would call the robustness indicator of the group, which are summarized on this page.

A CET1 ratio of 17%; a very high-quality loan book with a level of NPL which is only at 2.5%, end of September; a very prudent provision approach which is leading to a coverage ratio of those NPL of 80%, above 80% at group level; and lastly, a liquidity position, which is very comfortable with liquidity reserves which are in excess of EUR 400 billion, end of September. Let me now dig a little bit inside the performances of Crédit Agricole S.A.

itself, starting with the revenues on Page 11. What you can see is that both on the quarter and on the first 9 months of the year, we've managed to post a positive evolution of the top line, plus 1.5% for the quarter and plus 2% for the first 9 months of the year.

It's a very good performance, which is clearly the result of the diversification of our businesses, and also the high proportion of our revenues being recurrent. And we've tried to calculate the proportion of our revenues which are linked to recurring activities or recurring contracts or recurring services that we sell to our customers.

And actually, we end up with the conclusion that more than 3/4 of our revenues are clearly recurring, i.e., providing resilience and solidity to our top line. If I go on the next page, the cost line, Page 12, what you can see is that on the quarter, we've managed to decrease a little bit the cost base by 1.2%.

And actually, it's been stable across the first 9 months of the year. But what is even more interesting is that in the business lines where we had to record a decrease in the top line, which happened in some of our business lines, we've managed, at the same time, to decrease the cost line in order to preserve as much as possible the profitability.

And it's another illustration of the efficiency of our method of monitoring the cost base. Each business line is responsible for its cost base, and it has to defend as much as possible its own cost-to-income ratio.

This is the most efficient way to monitor globally the cost base in a very decentralized group like our group. This is leading to a very impressive performance in terms of cost efficiency across the board because if I take a helicopter view on this issue of cost efficiency and if I look back to where we were in Q3 2015, I acknowledge that we've been able to decrease in 5 years the cost-to-income ratio of CASA by more than 11 percentage points, going down from 69.4% to 58.1%.

It's a very impressive and solid performance. Going now to the risks and the quality of our loan books.

As stated, at group level, we have 2.5% NPL ratio and more than 80% coverage ratio. We have also a very high quality of exposure at CASA level with an NPL ratio which is a little bit below 3.5% and with a coverage ratio which is close to 70%.

You may see that the coverage ratio on those 2 perimeters is a little bit down as compared to the end of June. It is almost totally explained by a technical factor which is the entry into force of a new definition of the default, which is, to put it in a nutshell, leading to a slight increase of the stage 3 for the nonperforming loan bucket without needing an increase in the same proportion of the provision because it's made of default under observation and not real default.

In terms of cost of risk this quarter, on Page 14, what you can see is that it is still significantly above the cost of risk that we posted in Q3 '19, but it has started to decrease as compared to Q2 '20. And it has decreased indeed quite significantly, minus 1/3 on CASA perimeter and almost a division by 2 on the group's perimeter.

We continue to have a significant proportion of the cost of risk, which is made of provision on the performing loans in our portfolio. So it's the provisions related to stage 1 and stage 2 assets.

And actually, in our outstandings of loan loss provisions, 1/4 of this provision for CASA and almost 30% for the group are made of stage 1 and stage 2 provisions. If I go now to Page 15, which is providing some details by business lines, what you can see is that actually, in all cases, we have the same type of phenomenon as the one I have described globally, i.e., a significant proportion of the provisioning which is made of stage 1 and stage 2 provisioning and a decrease of the level of additional provision as compared to Q2.

Maybe one element I can mention, which is very special to the consumer credit business, this new definition of the default has indeed triggered a significant move between bucket 2 and bucket 3. And so this is leading to a significant one-off where we have reduced significantly the bucket 2 provisioning and replaced it by the same amount of bucket 3.

But of course, the level -- the overall level of cost of risk is the normal quarter-on-quarter level considering the context in which we are now. This is leading to the level of net income globally on Page 16.

What you can see is that actually, the decrease in the level of net income, both on the quarter and on the first 9 months of the year, especially restated from the contribution to the Single Resolution Fund, which, once again, has increased this year as compared to the previous year, we are minus 9% on the quarter and minus 8.8% on the first 9 months. I think it's a very impressive performance to be able to post such a resilience in terms of net profit when you see such an increase in the level of the cost of risk compared to the previous year.

Business line by business line, you see that the evolutions are differentiated. Maybe just let me mention 1 point on this quarter.

On the asset gathering business division, you see a decrease of EUR 42 million of the net profit. It is more than explained by an accounting issue, which is a simple fact that within the perimeter of Crédit Agricole Assurance, we have modified this quarter the way we account for the AT1 coupons issued by Crédit Agricole Assurance.

Up to now, they were accounted for against equities, and they are now accounted for directly in the P&L and the minority interest, but it doesn't change anything economically. If I zoom a little bit now in the different business lines, starting precisely with the asset gathering and insurance business division, on Page 18, globally, what you can see is that the level of activity has been quite dynamic, with especially a very high level of net new inflows for Amundi, close to EUR 35 billion, plus a positive market effect.

And as I said, the contribution of this business division to the net profit of CASA would have been stable outside this technical and accounting issue, which, again, doesn't change anything from an economic viewpoint. Insurance activities, to put it in a nutshell, it's been a very good quarter in terms of activity despite the fact that the total premium income has decreased, but actually, it has been decreasing because of life insurance activities.

And in life insurance activities, we've nevertheless managed to keep the positive level of net inflows. And again, we've managed to have a very significant proportion of the net inflows made of unit-linked products.

For the other insurance activities, the catch-up in terms of production of new policies has been very impressive. And so the premium income is up around 10%, both for P&C and for protection.

This is leading to a very solid financial performance because, again, if you look on Page 19 on the P&L, you'll see that the net income is up 1%, and this is only the minority interest which is leading to a decrease in the net income group share. If I go now on Amundi, I'll be very swift because you have seen Amundi's publication last week.

So very positive new inflows, revenues which continue to suffer from a management fees standpoint, but there is an almost complete catch-up, thanks to the cost line, and all in all, a profit which is up -- slightly up this quarter, the contribution to CASA's net profit. Retail banking, starting with LCL.

Another very good quarter at LCL, with all indicators oriented positively. It's been the case for the customer capture, for the evolution of loans outstandings, for the customer savings, for the equipment of our customers in additional services and products, and it's been also the case from a financial viewpoint with revenues up, thanks to a good combination of fees and net interest margins, costs down, cost of risk up.

But nevertheless, net income group share this quarter up 14%. Italy, Crédit Agricole Italia, a good level of activity and a good quarter in terms of catch-up of the production of new loans and new contracts with our customers; revenues which are almost stable; costs slightly down; cost of risk up significantly, but lower than the previous quarter; and a decent level of profitability, all in all.

The rest of the international retail banking activities suffered again this quarter, mostly because of the combination of a significant decrease in interest rates and flattening of the yield curve in the different countries where we operate, plus a ForEx effect, which is also negative when we translate into euro, the revenues generated locally. Nevertheless, we continue to monitor those activities with a very safe framework, and we continue to have a local self-financing of the loan book which is good and coverage of the NPLs, which is above 100%.

Specialized financial services. This is the division in which we have these very specific effects linked to the reclassification IFR -- under IFRS 5 of CACF NL.

But when it comes to consumer credit and when it comes to its activities, the production of new loans this quarter was very good, and actually, it was above the production of consumer loans in the third quarter of 2019. Revenues restated from this perimeter effect linked to CACF were actually only marginally down, minus 1.4%, and expenses are down more significantly, minus 2.5%, which is leading to a stable gross operating income and net income, which is only slightly down globally.

So a very good resilience and a very good catch-up after the period of lockdown in the second quarter. For the leasing and factoring activities, it was a more difficult quarter.

The production of new contracts was good, but the transformation of these new contracts into effective revenues is lagging a little bit, so we expect a better fourth quarter. Large customer division.

This is definitely the division in which the top line was the most dynamic, globally, close to 12% evolution of the top line with all the activity indicators positively oriented, both for the CIB and for the asset servicing activities. In the asset servicing activities, we continue to work on the integration of Santander Securities Services into our setup.

And we continue also to work on the integration of Kas Bank. We have had the latest authorization that will enable us to proceed to the mergers only a few days ago, actually.

So there's a lot of work ahead of us in order to generate additional cost savings. CIB, more specifically, on Page 26, what you can see is that we've posted an increase of around 10% of the top line, which is the combination of a very, very good dynamic on the capital market activities.

On the financing activities, apparently a decrease in the level of revenues, but actually restated from ForEx effects and from a base effect, which is linked to the fact that in Q3 '19, for the last time, we had booked in this division a dividend that we received from Banque Saudi Fransi. Actually, revenues are also up in the financing division of the CIB.

The costs are only modestly up in connection with the good level of activity. The cost of risk is definitely significantly higher than 1 year ago.

It's close to a multiplication by 5. But the biggest part of this provisioning is linked to performing assets.

And actually, this quarter, the stage 3 provision is representing only a very modest part of the global provisioning. And lastly, the level of RWA within CACIB has decreased quite significantly this quarter.

So this is also helping the profitability overall. Corporate Center, on Page 27.

We continue to follow the same path, which is a progressive reduction in the structural cost of the Corporate Center. And this quarter, we've benefited also from additional contribution, which is, all in all, leading to an exceptionally low level of cost for the Corporate Center.

Regional banks of Crédit Agricole, on Page 29. We are seeing more or less the same trends as the ones we see at LCL, i.e., a good level of customer capture and an increase of the customer base, a very dynamic evolution of the loan book, a very dynamic evolution and the production of new insurance policies on the basis of our customer base, leading all in all, to an increase in the top line, a decrease in the cost line, and due to these technical effects about the new definition of default, also a significant reduction in the cost of risk.

So all in all, the contribution of the regional banks to the net income of the group globally is up this quarter plus 12.5%. We can go now on Page 31 in order to take a look at the way we've managed to increase so significantly our solvency.

This is the combination of a reduction in the level of RWAs, which you see on the left-hand side of the page. The consumption of RWAs by the different business lines is down despite the dynamic of the production of new loans because actually, the new loans that we produced are home loans at LCL or CA Italia and where we've seen reduction is the portfolio of loans continuing at higher level of RWAs.

We benefited also from ForEx effect and from technical effect, which all in all led to this reduction in the level of RWAs. The other elements are more or less compensated.

So this is leading to this reduction in RWAs. When you combine that reduction with a good level of profitability after having reserved a dividend to be paid in 2021 with our permitting, plus some methodological effects, which this quarter are positive, this is leading to this very high level of solvency at CASA, 12.6%.

It's, of course, far above any regulatory requirement and also far above our target. For the group, it's more or less the same story.

We have a reduction in the level of RWAs for the same reason, so minus EUR 11 billion. We have a good level of reserves.

We provision, of course, the dividend to be paid in 2021. We have also some positive methodological effect.

And all in all, this is leading to a 90 bps increase in the solvency of the group in this quarter only and to a buffer of above 800 basis points above the Pillar 2 requirements. In terms of liquidity, I think I can go very rapidly because, again, we are in a very comfortable situation, thanks both to the monetary policy put in place by the ECB plus our very good capacity to access the collateral and to mobilize collateral whenever we need it.

It's also the same story for the market funding programs of the group, which are completed at 97% for CASA and above 80% for the group globally at the end of the third quarter. So all in all, and I am now on Page 36, I think that the conclusion can be quite simple.

A very solid operational and financial performance this quarter; a very robust group and a very robust CASA from a balance sheet viewpoint, again, reinforced by this good financial performance; and a group which is fully ready to continue to play its role in the management of this very specific crisis that we are navigating across. Thank you, and I will now take your questions if you have some.

This is Tarik El Mejjad. So I have 2 questions, please.

First one on the Italian M&A. So it's not a surprise.

So I mean, your shares have been underperforming the index by 15%, 17% since the headlines came. And -- I mean, I haven't heard you really denying it or -- and the markets either.

And -- so can you please just tell us if it's something you exclude doing? Or just maybe you can refresh what's your view in terms of consolidation in general, and in particular, in Italy?

And the second question is on the cost of risk. I mean you haven't -- you didn't give us, again, the guidance for the full year or expectations.

And I think in the press release, you mentioned that Q4 could be tougher for some corporate and retail. Do you still expect to add some more management overlay or stage 1 and 2 provisions as you did in Q3?

Or you think you are, at the moment, given the assumptions of second lockdown and so on, you are reserved enough, especially in the financing activities, actually, where we still see some higher level of stage 1 and 2 in Q3?

Thank you for those 2 good questions. Let me start with Italy, and let me start with what we do in Italy and what we do there since probably 30 years.

We are present in Italy not only in retail banking but also -- and I would even say -- but above all, with our specialized business lines. It's the case for the asset management, for the consumer credit, car financing, leasing, factoring, CIB, wealth management, name it.

All our specialized business lines are present in Italy. And in most cases, actually, they have managed -- I've forgotten to mention insurance, of course.

And all these business lines are present in Italy, not only on the basis of our own customer base in our retail banking activity but also thanks to a whole series of partnerships that enabled us to grow actually our business line beyond our footprint in retail. So it's the same business model than the one we had in France.

But simply, it is a little bit skewed, I would say. In France, we have the largest market share in retail banking activities.

And then we've built on the basis of this customer base. We've built all the specialized business lines, and we struggle in order to try and grow our market share in the different specialized business lines at the same size as the one of the retailer.

In Italy, we've decided that because our retail footprint was much smaller, we needed to find scope by those partnerships. And so globally, this is leading to a situation where 2/3 of the net profit that we've made in Italy last year, and it's the same case -- it's the same situation for the first 9 months of this year and it was also the case in '18, 2/3 of the profit comes from the business lines and 1/3 comes from the retail.

Does it mean that we don't want to grow in retail? Absolutely no.

We try as much as possible to grow organically, and all the performances that we post regularly show that we are indeed able to grow above the market, and more than the market, in the different product lines that are distributed by our retail banking branches. It's been also the case this quarter in home loans or whatever.

Does this mean that we are not ready to take opportunities to accelerate this organic growth? Taking advantage of, again, some opportunities, we've did it in the past, and we've did it, I think, quite successfully.

So always keeping in mind that we have a prudent approach on M&A transaction, we may take opportunities. But keep in mind that this is taking place in a global strategy, which is not a pure retail strategy and which is not a strategy on which we would assign the first role to M&A.

So this is the strategy that we follow in Italy. This is the strategy that we've been following in the last, again, 30 years, and especially, in the last 5 years.

And we intend to continue alongside the same lines prudently but decisively. Second question was about cost of risk and guidance.

So again, I don't like to provide guidance, especially in such an undetermined environment. It's -- the environment is really, really, really difficult to read across, especially from a sanitary viewpoint, and it has consequences on the economic situation.

So we haven't provided guidance on the cost of risk end of Q1 and end of Q2. We are not going to provide guidance end of Q3.

What I can tell you, nevertheless, is that we haven't changed the macroeconomic scenario on which we base our bucket 1 and bucket 2 provisioning in the third quarter. We are probably going to change this scenario for the end-of-year accounts.

Is this going to lead to a significant increase in the level of provisioning, i.e., bucket 1 and bucket 2 provisioning? Not necessarily because our bucket 1 and bucket 2 provisioning is made of 2 layers, 2 additional layers.

The first one is what we call the central forward-looking. So it's the effect of the macroeconomic scenario that is applied to all businesses across the group horizontally.

And then each business line may, if it deems it's necessary, to add up to this central forward-looking a local forward-looking, which is based on a more accurate view on certain portfolios or certain geographies. And this quarter, indeed, the biggest part of the stage 1 and stage 2 provisioning -- additional provisions that we've booked was the result within CACIB, within LCL, for example, of such local forward-looking.

And of course, if we modify a little bit the global macroeconomic scenario that is leading to the central forward-looking, this will integrate, I would say, all these local provisions that have been booked in the third quarter in certain business lines, again, like LCL or CACIB. So this is another reason why I don't expect a massive increase in stage 1 and stage 2 provisions.

So to summarize my answer, we will modify the scenario. It may lead to additional S1 and S2 provisions, but I don't expect a massive increase.

And as far as the S3 provisions are concerned, we will simply see what the situation will be at the end of December.

So if I understand, when you were setting your Q3 stage 1 and 2 provisions, you had in mind some kind of form of lockdown, and probably not national ones, or you would have to adjust for that, yes?

Yes. Actually, our scenario, which was established at the end of the first lockdown and end of the spring, beginning of the summer, proved in September to be a little bit more pessimistic than most of the official forecast.

So we found out that it was enough pessimistic to be used for the Q3 accounts. So we have now a few weeks' time to fine-tune and to try to assess as best as possible the consequences of the new situation we are facing.

But again, it's a moving situation. It's very, very difficult to precisely assess what is going to happen beginning of '21, end of '21, '22 and so on and so forth.

My first question is on the CET1 ratio. Could you just remind us of any headwinds and tailwinds, I'm thinking software, for example, as we go into the year-end?

And with your reference to the 60 basis points for 2019 dividend effectively in your ratio, I mean, how you think about the potential for paying something additional on top of your 2020 accrual? Or would you think it was prudent to hold a buffer for economic uncertainty or even for M&A?

And then the second one is just as a follow-up to Tarik's question, what would you consider as your normalized through-the-cycle cost of risk?

Okay. I'll start with your second question because actually, we haven't changed views regarding the "normalized cost of risk" across the cycle.

From what we had assumed when we published the medium-term plan back in June 29, what we had said at this time is that -- was that the normalized cost of risk across the cycle at CASA was around 40 bps, and at the group level, globally, was around 25 bps. So we haven't really changed our mind.

We'll see if we converge towards these levels rapidly or maybe more slowly. But we have no reason to change our views on what should be the cost of risk across the cycle, considering the business mix that we have in the different perimeters of CASA on the one hand and on the group on the other hand.

When it comes to the CET1 ratio, what we are expecting for the end of this year is, as you've mentioned, the potential benefits from the decision on the software. It's very cumbersome to fully calculate this -- the impact, but it's going to be a matter of bps, not a matter of tens of bps.

So what I expect is that it could be up to 10 bps of additional benefits, both for CASA and for the group, not much more. And again, it's quite cumbersome to calculate because we have to go through each of our IT investments to see what is the remaining duration of amortization and so on and so forth.

In terms of headwinds, TRIM exercises continue, and we have a few billion of additional RWAs probably to account for, especially in the perimeter of CACIB. We don't know exactly how many billion, but would be -- would it be around EUR 5 billion?

I don't know. But it's going to be probably maybe even above EUR 5 billion.

We have still to analyze different TRIM reports and to discuss with the ECB and so on and so forth. And we have no precise indication on the calendar.

So headwinds and tailwinds are expected. But globally, nothing would change significantly considering the high level of solvency where we stand now.

When it comes now to this 60 bps dividend provision that we have kept, and actually, there was nothing else to do because as soon as we had the decision that we couldn't pay a dividend in 2020, we had nothing else to do than reintegrating the 60 bps of solvency into our ratio. Well, we are not going to hurry to take a view on what we are going to do with this amount.

I understand that the ECB intends to give some clarity on what is going to be possible to do in '21. So we'll wait for the ECB to talk and to express its views.

Definitely, we think that if there is any openness in the ECB's statements, we definitely belong to the category of bank that could benefit from this openness because of the high level of solvency that we have. And maybe one last point, if CASA consumes 60 bps by paying its dividend like it intended to do in '20 and was precluded to do, this consumes only around 15 bps at the level of the group globally.

So clearly, you see that the dividend question for us is not the solvency issue. Maybe just one last point.

When I talked about headwinds and tailwinds, I -- of course, it takes place within the global framework of 60 bps that we had explained regarding the implementation of Basel IV that we had stated initially when we published the medium-term plan. So it must be still inside this, I would say, rough envelope.

Very 2 quick questions for me. First of all, your AT1 costs are, obviously, down sharply for 9 months.

Is it okay to actually annualize this number as a proxy for the year and also for forthcoming years? And b, you were mentioning the TLTRO, I would say, reversible gain, if you can actually quantify that, that would be great.

AT1, well, what happened, as I explained, is that we had -- for accounting methodological reasons, we had to modify a little bit the treatment of part of our AT1, the one that were issued by the insurance companies. So we've taken back, I would say, EUR 43 million this quarter, which were directly accounted for against equity up to now and that we had to recycle through the P&L.

So again, it doesn't change anything from a net asset standpoint or from a earnings per share standpoint, but it changes a little bit the way the P&L looks. So this EUR 43 million is going to be increased in the fourth quarter by, I think, EUR 25 million, around EUR 25-additional million, which will follow exactly the same route.

So definitely going forward, around EUR 75 million of AT1 coupons from the AT1s issued by Crédit Agricole Assurance are going now to appear in the P&L. When it comes to the AT1s issued directly by CASA, we are on the same page, i.e., the cost -- the average yearly cost, which is not completely evenly broken across quarters, is around EUR 420 million, a little bit above EUR 400 million.

For the TLTRO?

And the TLTRO, excuse me, Jacques-Henri. The TLTRO, well, we've grown around EUR 125 billion of TLTRO III globally for the group.

Then we've allocated certain amounts of those TLTRO holding to the different business lines, each business line being entitled to a certain amount, considering its own contribution to the level of credit that -- the level of loans that are eligible to the TLTRO funding. So it means that a significant part was attributed to the regional banks, a part to LCL, to CACIB, to CACF, and even to FCA Bank, and of course, a part to CASA directly.

So this is for the amount of TLTRO. Then I will let you make your own calculation in terms of what it is generating in terms of revenues because we consider now that this TLTRO holding is part of our global funding, and each entity has a mix of liabilities, in which we have costly liabilities and cheap liabilities.

The TLTRO is clearly a cheap liability, especially for the first 12 months. But we have also some home purchase saving plans, which are very costly.

And so each business line has a combination of liabilities, which it manages with its own ALM tools. And so the benefits of the TLTRO is really spread between the different business lines, and each business line has decided whether it was going to modify or not the rest of its funding.

But globally, the group has taken EUR 125 billion of TLTRO III.

Phelbe Pace from SocGen. I have a follow-up question to the one related to Italy, actually, but even more generic to your approach to M&A, if I may.

So you are obviously in a very comfortable capital position right now, which allows you to embrace external growth, if you want to. But I guess despite the magnitude of your excess capital at the moment, it could still remain too limited if you wanted to opt for a larger, more structural deals.

So that is not something bolt-on. So 2 questions here, please.

First, would CASA have the appetite for larger M&A deals that couldn't be financed only with cash? And second, do you think that CASA has the capacity to conduct a larger acquisition in the near term?

Because obviously, we can imagine that integration could be significantly slowed down, and especially, more difficult to execute in the current context. So those are my questions.

Good question, again. First element of answer.

This is not the excess of capital that we have that decides our strategy or our M&A strategy. It's exactly the opposite way.

If we identify an operation which makes full sense from a strategic viewpoint that is accessible, that is coherent with what we want to do, and for which, all the operational aspects are under control. And you're perfectly right to mention the fact that it's probably more difficult to integrate some acquisition in the present period of time because it's difficult simply for people to gather and to meet.

Then if we are in such a situation, we identify the way we are going to finance the operation. But it's not the other way around where we say -- where we have a few billion of excess of capital, what can we do with it?

It's absolutely not the way we think about that. And are we ready to engage into operation that need -- that may need not only our excess of cash?

Well, in the past, we did an operation. Actually, it was through Amundi, but Amundi made a significant acquisition that required a capital increase, and actually, CASA provided guarantee for the success of this capital increase, but we've proven that it's perfectly okay for us if a large subsidiary like Amundi engages into an M&A transaction that is not financeable only with its cash position.

So again, what is important is what we do, not how we finance it. Of course, once we have decided what we want to do, we invest, put in place a financing that is coherent with what the group is, but we start with the operation itself, and of course, you've made reference to that.

The key element for us to identify and to tell if an operation is relevant, is, of course, to make sure that we have the capacity to integrate and that we have the capacity to generate the type of profitability that we require for such an operation, which, among other elements, is the capacity to reach a return on investment of at least 10% after 3 years, and mostly, by cost synergies or true synergies, I would say.

I have a bit of follow-up questions as well on M&A. One would be, in the past, I think you clearly said that you did not want to buy an entity and have to deal with issues around nonperforming loans.

I was wondering if this is still the case? And then I have another question, which is perhaps a little bit more generic.

And -- I think at the beginning of this call, at least the Q&A session, you said that you run different types of models, obviously, through partnership where you distribute the products manufactured by the product factory, and also you own, effectively, distribution network. And so my question would be, do you think a business model, whereby you don't own at all distribution network, but you only work with partnership is a sustainable business model?

Okay. Thank you.

Well, what I have always said is that our business model was not to manage a portfolio of NPLs. I have always said that there are some investors which are dedicated to this business.

They do it very, very well. They generally generate high profits on that.

It's not our business. So I repeat the same thing.

In any kind of operation, what is interesting us is not the capacity of making money with the management of an NPL portfolio, and we stick to that clearly. When it comes to partnership and to the capacity of developing a business model that would rely only on distribution partnerships and not on the possession -- the direct possession of distribution networks, well, it's already significantly the case because actually, the universal banking model -- the full universal banking model, the perfect one, is only present in France and in Italy.

But when it comes to Spain, for example, you see that our insurance companies have engaged into partnerships in order to distribute their products without having the control of any distribution network. It's also the case for the consumer credit business, it's also the case for Amundi, so -- and even for CACEIS.

So Spain is a typical country in which we've managed through partnership to identify and operate different partnerships without having in this country a retail network and a direct customer base. So it's really something that is perfectly coherent with what we do.

A couple of questions from me as well. So one on cost delivery.

So as you mentioned in your presentation, that was very solid so far this year. And I was wondering, is there anything more that you can do or any new ways of doing business, especially, thanks to digital channels, that can lead to further cost savings going forward?

So that will be my first question. And then just a couple of quick ones.

So Corporate Center, in particular, is there any guidance on the revenue line because it's quite volatile. And then finally, you discussed already the impact of lockdown potentially on provisions.

But if I remember well, in the past quarters, you also had some impact on the insurance side because you decided to do some voluntary, let's say, payments even though you were not obliged because you don't cover pandemic risk. I was wondering if this is something that you're planning to do again now.

Well, let me start with the last question. We cannot rule out definitely things like that.

But we have no specific plan regarding such an operation for the time being. We've done a voluntary and very significant contribution to all our professional customers that were impacted by the lockdown in the first period of the year.

It was absolutely voluntary, not linked to our policies. And I think it was very well appreciated by the customers because of that.

But it's not a definitive, I would say, feature that is now embedded in our contracts. So nothing is planned for the time being.

Corporate Center, well, clearly, it is and it will remain rather volatile. What is important for the Corporate Center is the trend.

And globally, the trend continues to be positively oriented towards a reduction of the structural cost. But in addition to that, we have some more volatile elements that play a role in the Corporate Center.

We have some businesses which are outside the main business lines, and this year, they are performing much less than last year. It's specifically the case for the private equity business that we have kept in this Corporate Center.

And we have also, as you know, and this is also significantly impacting the revenue line -- we have also some accounting effects resulting from the elimination of intra-group debt issued by CASA and subscribed by Predica and Amundi, which are not accounted the same way by CASA on the one hand and by the investors on the other hand. So we have some elimination, which plays a role.

In most cases, there is a trend, which is, if I take this last example of accounting, elimination of intra-group debt, the trend is that, in average, it should converge towards euros, but it creates volatility. And what is important is really to acknowledge that all our efforts are leading to a progressive reduction in the cost of the Corporate Center.

Then as far as the cost line is concerned, I'll start by saying that we are already below the target that we had initially set for 2022. So of course, every time we have the capacity to do better than the target, we do it.

But please recognize that we have committed to a target of being below 60% in 2022 and that we are now significantly below this target. So we will continue to work on the cost base.

We will take advantage where possible of all the digitalization tools that are accessible. By the way, digitalization is not only and is not necessarily leading to a reduction in cost because we view digitalization as an additional tool, as an additional feature that we provide to our customers.

So it's not only and it's not principally ways of reducing the cost base. So we invest a lot.

We have pulled several IT resources between the different business lines in order, in the medium term, to generate additional cost savings. But this requires investments.

And the good news is that with the very low cost-to-income ratio that we've managed to post already, we have the capacity to invest.

So my first question was on -- just a simple question on the French retail. It's -- I see that the commissions were very strong in comparison also to BNP who reported yesterday, particularly driven by securities management, so the -- and also the insurance business.

And then the volumes of -- okay, so there could be some more transactionality for securities management. But the ones in -- with regards to insurance, which are up -- which were up 4% in the first quarter, 18% last quarter and 13% this quarter year-over-year, they don't correlate as well as they used to with the outstandings in life insurance that you disclosed for the same dividends.

And I just wondered what was the driver there? And what you expect -- and what you attribute the -- if you're on the increase in distribution margin, if I can call it this in that particular business line?

And secondly, with regards to LCL, so the net interest margin is holding up okay. And in fact, we see also from the aggregate data of Banque de France that front book margin seems to be increasing.

So I just wanted to understand whether there was still any drag to come from interest rates or whether the commercial margins, which we can see at Banque de France are a good indicator for the future of the margin in that division? And my second question is also on capital ratios, but rather more at the group level.

And obviously, the Caisses Regionales don't distribute much. The group capital ratio is now 17%.

And I hear you when you said that it's not because you have money that you all of a sudden start thinking of how to spend it. But nonetheless, that's a really, really high capital ratio.

And in a sense, it could argue that it's not employed very, very productively. And I just wanted to understand what is the outlook because it's building up really fast at group level.

For example, in the investment bank you might want to, I don't know, originate more and distribute them to revenue seekers such as the Caisses Regionales and things like that. So just to try to understand how the reallocation of that capital could work?

Well, let me start with the second question, which is, of course, difficult and interesting, but also, there's a simple way to answer it. Actually, it's structural with the regional banks that they continue to build up their level of capital because they don't have the same cost of equity as CASA, for example.

You know that the capital of the regional banks, so the prudential capital that is generating this level of solvency is mostly made of reserves that are accumulated and that do not belong to the mutual shareholders. The mutual shareholders, they own a mutual share.

So they are entitled to the repayment of the nominal amount of their mutual share. They are entitled to a remuneration of this nominal amount, but the reserves are accumulated without being allocable, I would say, to identify the owner.

So it means that we don't have this necessity that CASA has that whatever listed bank has, which is to remunerate all the capital -- prudential capital that is put at work. So it's a comfort, definitely.

It's very efficient in terms of generating, I would say, security and robustness within the group. It's also an element that plays a role in the global capital structure of the group because all the capital that we manage to build with CET1 at group level is what we don't need to cover with AT1 of Tier 2 debt or even TLAC debt when it comes to building the different capital ratios of the group.

So it's also -- paradoxically, it's a cheap way to build our capital ratios. So there's no need to identify investments that we could make in order to use this "excess of capital."

When it comes to LCL and to the fees that LCL has booked this quarter, you must keep in mind that insurance fees are not related only to life insurance activities. They are also related to non-life activities.

And actually, the portfolio of non-line -- non-life contracts at LCL is building up rapidly because LCL has significantly accelerated in -- the equipment of its customers in non-life insurance policies. And so this is generating an accrued amount of fees going forward with potentially a certain volatility which is linked to the level of risks within this portfolio, the sinistrality.

But nevertheless, this is generating an increased number -- an increased amount of fees as well as the fees generated by the life insurance policy. And then net interest margin.

Well, everybody is focused only on home loans, which is not the only component of loans generating margin for LCL. But it's true that for home loans, we have seen this quarter, a narrowing -- quite significant narrowing of the spread between the front book and the back book.

And roughly, it is now around 30 bps when it has been historically closer to 100 bps, then around 50 to 60 bps. So 30 bps means that if things going -- if things continue to evolve like they've been evolving in the last period of time, we may envisage a convergence of the yield of the loan of the front book and the back book probably around '22, which would be a good thing because this would alleviate the very heavyweight that is penalizing the net interest margin of retail activities.

A couple of questions for me. One is on the volume outlook for France.

If I look at the latest ECB lending survey, it seems that there is a bit of a slowdown of expectation from a very high base in your lending, even excluding the state guaranteed program is still very solid. So if you can give us some indication on outlook in the various bits of the loan book?

And then a couple of quick questions on Italy. One is if you can give us the detail of the moratoria loans and how they are evolving there?

And the second one is just to come back to the M&A, but very quickly. If I understood well, you said that you would only consider, and that makes sense, if it arises, and for a retail franchise, it could only be interesting if include also the ability of cross-selling for your asset management among the -- or the insurance?

Am I understanding it correctly, what you mentioned?

Thank you, Azzurra. Volumes in the French market, yes, it's been performing well across the board.

In loans to businesses, actually, we had a very sharp slowdown in the second quarter. So what we've seen in the third quarter has been a catch-up after a situation where almost all our activity was made of state guaranteed loans.

So there's a catch-up effect. But there's also a certain dynamic.

And actually, when you analyze precisely the GDP growth in the third quarter, you see that actually investment from businesses contributed significantly to this GDP growth in the third quarter. So there has been a catch-up not only in lending to businesses but also in investment from businesses in the third quarter in France.

So what are the prospects? It's very difficult to tell actually because, again, we are in a very uncertain and moving situation from a sanitary standpoint.

And this is leading to some impact on the rest of the activities. And this is definitely creating some wait-and-see attitude, especially when it comes to investing.

But to summarize, yes, there has been a very good dynamic in loan evolution in France, and this has partially translated into also a very significant increase in the volume of money that is put on side deposits, both by households and by businesses because at the same time, the spending of these liquidity lines has not been as fast as the granting of those liquidity lines. In Italy, we've granted around EUR 1.8 billion of state guaranteed loans, which were more or less the same as in France, not exactly, but more or less.

And we've also provided around EUR 800 million of payment holidays of moratoria. Those have been extended up to the end of this year or maybe January '21.

So for the time being, we have -- we haven't seen -- we have not the -- we don't have the capacity to acknowledge how it's going to perform simply because we continue to be in the payment holiday period. Then M&A, I'm not going to repeat what I already said.

Our business model is made of cross-selling and is made of making different business lines working together. That's the only thing I want to say on that.

I've got 2 questions. The first one, again, with regard to the AT1, I'm sorry, but I got a little bit confused.

So you say now, we should expect more than EUR 400 million per year, which is what I had in mind before Q3. But is that now EUR 400 million per year or more below the line as well as another EUR 200 million per year in insurance?

Or is it the sum of the 2 that is EUR 400 million?

No. No.

Yes, Guillaume, on this question, it's going to be, to be precise, EUR 405 million for the full year of '20 below the line, i.e., directly against equity and around EUR 75 million above the line, i.e., through P&L. And so the pace going forward for what is going to be accounted for on the minority interest line is going to remain around EUR 75 million a year unless there is a new issuance or a repayment by Crédit Agricole Assurance.

So this is going to be the pace going forward, EUR 75 million above the line. And as far as the AT1 issued by CASA are concerned, so the EUR 405 million this year, may be a little bit higher next year, but I was giving an indication of EUR 420 million, simply because actually, we have increased the outstandings by issuing an additional AT1, which was in September, if I remember correctly -- October…

Yes, EUR 750 million.

Yes, exactly, the EUR 750 million one. So this is adding up a small amount of coupon, and EUR 420 million would be with the full year effect of this new AT1.

But I don't want to provide any indication of any kind of what we could do going forward with all these outstanding AT1s.

Sure. Okay.

And the second question, again, I'm sorry, because these are big numbers, in the Corporate Center, I don't really understand the comment you made on TLTRO. So is it fair if I say I take Q3 revenues, and I consider that within that, you have accrued 50 bps extra on the EUR 125 billion -- or the CASA portion of the EUR 125 billion, and that this will materialize for 4 quarters and then stop?

Yes. The question being, and I don't want to provide details on that, the exact proportion of what has been kept by CASA versus what has been lend by CASA to the different business lines, but the reasoning is this one.

Okay, fine. The last question is about the international retail outside of Italy.

Obviously, they suffered from lower rates and lower FX. Is there more headwind to come from lower rates or the current level is the new run rate?

And the follow-up question on the international retail ex Italy is that now the ROE of those businesses despite a good cost of risk -- or despite a higher cost of risk is now approaching 0. So the question arise now about would it not make sense eventually to get rid of those businesses in order to be a bit even purer in terms of business model?

Well, first, the run rate is now -- unless there is additional moves in the interest rates locally, we are now at the run rate because -- so the minimum run rate because the latest rate cuts took place in the second quarter. So definitely, this level is the new normal before the effect of all the efforts made locally to increase the top line by improving the revenue mix, by generating additional commissions and so on and so forth.

So of course, we are not satisfied with this level of activity and this level of profitability. Then strategically, what you can see and maybe what you can acknowledge is that every time we think that within the scope of businesses that we manage, there is a business that doesn't make sense, that doesn't generate the minimum level of profitability that we want to post and that is not strategic, we don't hesitate to dispose off it.

So for the time being, we've sold in several transactions, our 31% in the capital of Banque Saudi Fransi, we've sold Bankoa, we've sold the Bulgarian activities. We are in the process of selling the Romanian activities.

We are in the process of selling CACF Netherlands. So we don't hesitate when we deem it necessary to fine-tune our perimeter in order to be more coherent and more strategically aligned.

Yes. The first question I wanted to ask you is regarding the Switch unwinding, just to see if you would potentially consider unwinding also the other 50% that you have not committed to, if this is a possibility for the future?

Then going to Crédit Agricole Consumer Finance, where the revenue margin is rebounding this quarter, and I see from the slides that you mentioned a positive change in the product mix, but just if you could elaborate on how the product mix is changing there? And the very last question is regarding the wealth management business, where the costs were significantly better this quarter.

Just if you could elaborate on what was the driver and how sustainable that is?

Yes. Let me start with the Switch.

First, we still have 2 years to go. But during those 2 years, we must fulfill our commitment to unwind 50% of the Switch mechanism.

So there's another 15% of the total to go, which, before it can take place, needs that we see the clawback mechanism work for the triggering of the end of June. So before any move, we must wait for the clawback mechanism to work.

Then the first priority will be to take a look at the last 15%. The other 50%, 5-0, is definitely a possibility.

We are not committed to any type of timetable, but it's definitely a possibility. Business mix in the consumer credit operations.

Well, there's different moves. There is a tendency which has been triggered by regulation leading to a permanent decrease in the proportion of revolving loans and a permanent increase in the proportion of amortizing loans.

So this is producing a series of different effects. The first effect is that it reduces a little bit the net margin that we generate.

But going forward, it improves the cost of risk by definition. There's a second move that is creating more or less the same effect, which is the fact that we have more and more loans that are granted on the basis of dedicated investments like UTA, like some equipments for the house.

So again, this is generally a little bit less rewarding in terms of margin, but better in terms of risk profile that we will acknowledge going forward. So it's the combination.

And of course, we try and adjust permanently our offer in order, first, to suit to the needs of our customers; and second, to be as efficient as possible from a revenue, cost of risk combination. And we have, I would say, accompanied the pressure that we had felt on the margin in the last maybe 2 years by significant efforts to reduce the cost base in order to adjust our scope of business.

But I think that on this consumer credit issues, maybe I can give you rendezvous because we intend to give a workshop, a dedicated workshop on consumer credit activities, like the ones we did a few years ago on insurance, and last year, on CIB, and it may take place in next December, probably on the 8th of December. So we have the occasion to provide much more clarity and details on the evolution of consumer credit activities.

Okay. And on the wealth management cost, please?

The wealth management, excuse me, I've forgotten this one. Well, on the wealth management cost, it's true that since 1.5 years, we are working very, very hard in trying to improve the profitability of the wealth management activities.

And actually, we've managed to decrease and to engage into a tendency to decrease the cost base of the business. In the first 9 months, last year, it was -- the cost base was close to EUR 560 million.

So this year, it's about EUR 520 million. So it's a significant decrease.

And we continue our efforts in order to adapt to a context where definitely the revenues are a little bit harder to generate.

Just a quick clarification on the NIM in CACF. Did you also use the TLTRO III there?

Or did you benefit basically from lower funding cost?

A little bit, and differently, in the different activities. But CACF in France, it has the status of a bank, so it's entitled to access to the TLTRO.

It's also the case for FCA Bank. It's not the case for Agos, for example.

So it's not been the case everywhere.

Just a couple of follow-ups for me. First of all, on Italy, would you mind just giving us a little bit of an update on your thinking around Agos, I think, your discussions about the JV.

So just if you could give us a bit of color on what's going on there? And more generally, on partnerships.

I mean just trying to think about in Italy, where do you think you still need to grow? Or where you think you have room to grow in terms of the revenues of the business lines in Italy?

And then second question is more on capital and the usage of the excess capital. I think you mentioned Switch 2, but just trying to think, would you consider buybacks to redistribute some of the excess?

Or would you ever consider waiving a little bit the payout ratio from the current 50% level? If you could share your thoughts around this?

Okay. Agos, so the situation is very clear.

We own 61%, BPM owns 39%. We have an agreement through which they distribute Agos products in their network.

And there is a specific provision in the agreement that is giving them the right. But of course, the present environment is not necessarily the best one to do so.

It gives them the right to trigger an IPO of Agos. So of course, nobody really wants today to trigger such an IPO in the present context.

So discussions are around the idea of what we do with this provision? Do we extend its validity?

What is the context in which we could contemplate extending the validity? And so on and so forth.

So it's rather technical, and it doesn't change anything on the fact that the distribution agreement is up and running and is covering now all the branches of the BPM, which was not initially the case. Partnership in Italy or elsewhere.

Well, it's a permanent search for us to what we think when we assess the strength of the group. What we think is that in several business lines, asset management, consumer credit, insurance, life and non-life, and potentially, also leasing, factoring, custody, we have -- and we are ranking among the top players in Europe.

We have skills. We have very good products.

We have a very recognized expertise. And every time we have the capacity of finding an access to a new customer base, if it's compatible with the other partnerships that we have, is a good opportunity.

So we are not saying, well, in Italy, we need 5 additional market share points -- percentage points in that business or this business, all the business points, and it's -- actually, it's not stirred by Crédit Agricole S.A. itself, it's mostly driven by the capacity of the heads of each business line to identify new partners with whom it can work.

So really it's a matter of opportunity, but be sure that we are permanently trying to generate new partnerships. And in the last quarter alone, and this is on Slide 8 of the slide show, we've been granted a license in China to build a new joint venture between Amundi and Bank of China, in which we are going to be the majority shareholder.

We've concluded a partnership with Europ Assistance in France. Europ Assistance, which is in the scope of Generali.

And we've been buying the remaining 25% stake in the capital of GNB Seguros that we hadn't before with the consolidation of the distribution agreement with Banco Novo (sic) [ Novo Banco ]. So clearly, we are working permanently on trying to expand those partnerships.

Capital. Well, we are not going to talk about this question of "excess of capital" or undistributed dividend before we have some clarity going -- coming from the ECB.

So I think it's no use to think of all the theoretical possibilities. We'll see when we have the new framework, what we can do, what is relevant, what is coherent with our DNA, and we'll provide a clarity on that.

But clearly, we have the capacity to pay a dividend, and CASA in the long run continues to position itself as being a dividend player.

One -- first question, on Page 41, you mentioned in the note that the sensitivity of the ECL if you weight your unfavorable scenario by more than 10 points or something like that, whatever. But I was wondering if you proceed like U.K.

banks, for instance, which prefer to indicate what would be the deterioration of provisioning based on the central scenario weighted at 100% in case, for instance, of a degradation of 100 basis points to the GDP. I wanted to know if you had made this type of exercise, which, in my view, is more pertinent to feel exactly what is the risk of a degradation of the economy that this type of exercise has made?

No. No.

This idea that you can summarize all the banks' parameters into a big spreadsheet and that you can try to assess what is going on and what is going to happen if you move a number in that cell or this cell is, I think, a little bit of a fantasy, excuse me to say so. Because…

I do agree. I do agree with that, but I am talking about the methodology because you do the same by saying…

Yes, we have to do the same.

You exactly do the same. But I was talking about the methodology of taking one scenario weighted at 100%…

Rather than…

Rather than 3 scenarios that I manipulate, I would say.

Yes. Well, at the end of the day, when you weight the different scenarios, you end up with, I would say, a synthetic scenario which you can also, I would say, reverse engineer.

You can say that I want to have that endgame scenario and I adjust the weight of the 3 base scenarios in order to generate the scenario I want to end up with. So we can do those calculations.

I don't know if it's going to improve significantly the image that we are going to provide to the market. But we are going to see what we can do in terms of going this route.

No. I was just doing this remark because I participated in a debate with Ernst & Young recently.

And we have noticed that this exercise made by U.K. banks led to an increase by more than 60% of ECL in this type of exercise, while I see that for French banks, we only see a small increase in provision in case of this exercise.

That was just my opinion.

But it's difficult to assess only the risk by looking at the provision flows on a single quarter or even on the full year. The starting point is the outstanding of provisions that you already have booked in your balance sheet, and as far as the group is concerned, we have EUR 20 billion of provisions in our balance sheet, which already covers our NPL by 80% at group level, 70% at CASA level.

It's above the average of our peers in the European space.

Okay. Okay.

Another question, Page 11, you explained that one of your strengths is the fact that net interest income only represents 37% of your net banking income and that is due to the saving businesses -- part of the savings businesses in your business model. But it is also due to a monetary policy and also due to the consequences.

So my question would rather be, what would be, if you had a dream, the best contribution of net interest income in your business model? You see what I mean?

Well -- yes, yes, I see what you mean. I think that the situation we present for the first 9 months of 2020 is not that different from the one we had for the full year of 2019.

So it's quite stable actually. What is true is that if you take a helicopter view and if you look back on the last maybe 10 years, you will see a sharp decrease in the proportion of net interest income in the total of our revenues for 2 reasons, because we've grown significantly the non-interest margin generating activities, and because also on the retail banking activities, the net interest margin has been very much compressed going forward.

But when we presented the medium-term plan 1.5 years ago, we already stated that net interest income was representing around 1/3 of our revenues and was supposed to reduce by 2 percentage points across the medium-term plan. So it means that this is a feature that is already identified and actually really used in our business model since not only the lockdown period.

Okay. And my last question relates to Europ Assistance.

In your view, the change of the partnership was in the idea of improve the cost center or to improve revenue center? Or the…

Well, it's not the change of partner, it's a change of business model because when we were with the previous partner, we were a simple customer having a long-term contract and the assistor was providing us the services of an assistor. We are now in a completely different situation where we are a partner in the capital of the French operation of this assistor.

So it means that, of course, we have provided our business to this joint venture, but we have a joint venture with Europ Assistance. So it's a different strategy.

And actually, it illustrates exactly what we've been doing in the non-life insurance activities since we started probably 20 years ago. We start by learning in being the customer of another insurer, more experienced than we can be or a reinsurer, if necessary.

And then after a certain progression on the yield curve, we try as much as possible to be autonomous and to be able to operate ourselves the business or to be -- to engage into a joint venture for a partnership. So this is…

It's a Chinese strategy.

Maybe Chinese. I don't know.

They won't use it with Amundi.

We are going to be the majority shareholder of the new operation that we are going to build with Bank of China in China.

The first is -- sorry, just a follow-up question on the M&A. Could you please remind us of the financial criteria for deals in terms of EPS accretion and return on investments and over what time frame?

And then secondly, just coming back to the cost control, I mean, clearly, there's been, I mean, very good performance in terms of operating leverage, and I mean, being continually positively surprised. And as you say, you are already 2022 target.

Can you maybe just talk about what has been -- or where you have done better than you expected when you put up the plan? And I'm not wanting to send negatives here, but is there a risk that year 2020 benefited from less travel and entertainment, and then we could see a bump up maybe going into next year when things return to somewhat normal?

Thank you. Well, in M&A, we've got different criteria, but the most important one is that we want any M&A operation to generate a 10% return on investment after 3 years and to generate that return on a very secured manner, i.e., either by purely cost synergies or revenue synergies which are credible enough because you know that generally in M&A transactions, we can be very ambitious on revenue synergies, but this is generally more difficult to reach than the cost synergies which are more accessible.

So this is clearly the main criteria that we would look at in any transaction of this kind. In terms of cost control, I think that what we've done well and that what we will continue to do the same way is, again, to decentralize the responsibility of the evolution of the cost base across the business lines because who knows better what is the relevant cost base at LCL than the CEO of LCL.

He knows exactly where he's heading towards -- where he's heading to in terms of revenues. He knows exactly what he needs in terms of a branch number evolution, in terms of IT developments.

And so he fine-tunes far better than what CASA could do the level of the cost base. So we -- this is exactly the way we want to continue to do things, i.e., to decentralize the responsibility but also regularly, to assess the evolution and to make sure that the improvement action plans are identified and are put in place.

It's true that probably in the last 6 months we benefited from a reduction in certain expenses, especially travel and entertainment expenses because of the COVID crisis. We have had also, at the same time, some additional costs linked to the COVID crisis, which are not negligible.

So all in all, the net economy is probably real, but it's not so important. And it is probably here to stay at least in part because I think that going forward, we will not have the same intensity in terms of travel and entertainment as the one we used to have before.

So clearly, we might see a rebound in certain costs after the end of all this sanitary crisis, but we are not going to go back to the level where we were before.

Yes. So overall, good performance on CIB overall, but specifically in structured finance, just wondered how has the deal pipeline looked as you went into October?

Because seasonally, 4Q is usually a little bit quieter in terms of kind of the CIB financing revenues. And I'm wondering if I need to allow for a bit of a steeper fall off into fourth quarter this year?

And then more broadly, if I look across all of your loan books, I wonder if you told your loan officers to tighten lending criteria at all as we head into this kind of second lockdown? I know one of your peers was stressing yesterday how the second lockdown is going to be very different to the first.

So just kind of -- it would be helpful to hear kind of your views on that topic, please?

Yes. Well, for the CIB revenues going forward and the pipeline, the pipeline is correct, i.e., we've been significantly improving in the last years our positioning towards our customers, which means that every time one of our clients wants to engage into an operation -- a financing operation, bond issuance, whatever, we are more and more often one of the possible banks to be chosen to lead this operation.

So from this viewpoint, the pipeline is quite satisfactory. Then the moment when our clients are going to trigger the operation will depend on the circumstances, will depend on their need, will depend on the evolution of the sanitary crisis, will depend probably on the way the American presidential election ends up and so on and so forth.

So we can -- we depend very much on clients' decisions regarding the timetable of their operations. But every time one of our clients wants to engage into an operation, the chances that we are chosen to lead the operation have increased in the last period of time.

Then did we strengthen our credit criteria? We cannot say so.

What is for sure is that the present lockdown period is not the same than the previous one. That is for sure.

In the first period -- in the first lockdown period in France, it was very simple. You had a few activities that were forbidden, like the restaurants, the cafés, the bars and so on and some retail shops.

You had a few activities that were compulsory, and actually banks belong to the essential activity category and had to keep their branches open. And for the rest of the activity, which represents the biggest part of the French economy, the motto was you work only if it's possible remotely.

So this was definitely leading to a very sharp decrease in the level of activity. Now the motto is completely different is, you work, and every time it's possible to work remotely, you do it remotely.

But the rest -- for the rest of the time, we know that all the businesses, all the factories, all the offices have put in place some sanitary measures, some sanitary protocols that make it possible to work on-site safely. So this is why I believe that the effects of this second lockdown are going to be completely different on the global output and on the GDP evolution.

And actually, it's already been estimated by some economists that say that it's going to be probably at least halved in terms of impact as compared to what it was for the first lockdown.

Okay. Well, thank you.

Thanks to everyone of you. And I don't know if we are going to meet in December like was initially planned, but we're going to talk.

Have a nice afternoon. Bye-bye.