Executives: Philippe Bordenave - Group

COO
Analysts: Lorraine Quoirez - UBS Delphine Lee - JPMorgan Jean-Francois Neuez - Goldman Sachs Jon Peace - Credit Suisse Flora Benhakoun - Deutsche Bank Maxence le Gouvello - Jefferies International Jean-Pierre Lambert - KBW Alex Koagne - Natixis Bruce Hamilton - Morgan Stanley Jacques-Henri Gaulard - Kepler Cheuvreux Geoff Dawes - Societe Generale Anke Reingen - RBC Capital Markets Stefan Stalmann - Autonomous Research Robin Down - HSBC Kiri Vijayarajah - Barclays Pierre Chedeville - CM-CIC Market

Solutions
Operator: Welcome to the BNP Paribas Third Quarter 2016 Results Presentation. [Operator Instructions]. And at this point, I'd now like to hand you over to Philippe Bordenave, Group Chief Operating Officer. Please go ahead, sir.

Philippe Bordenave: Thank you.

Good afternoon, ladies and gentlemen and welcome to BNP Paribas Group's third quarter results presentation. As you know, it's usually our Chief Financial Officer, Lars Machenil, who presents the quarterly results, but I am standing in for him this quarter as he has had a car accident and he's currently convalescing. He's fine and he will be back before year-end. Let's proceed now with the results presentation and then I'll be pleased to take your questions. Starting with our key messages on slide 3, in the third quarter we saw good revenue growth in our operating divisions which overall progressed by 4.8%, despite the low rate environment.

This confirmed yet again the effectiveness of the Group's integrated and diversified business model. Operating expenses of the operating divisions progressed at a lesser pace than revenues, implying a significant rise in the gross operating income of the operating divisions which was up 8.8% compared to last year. Cost of risk continued to decrease and stood at 43 basis points, with most businesses improving or remaining at a low level. This in turn resulted in a rise in the Group's net income which was just shy of €1.9 billion. Excluding exceptional items, net income group share progressed by 15%.

In Q3, we also had a significant improvement in the fully loaded Common Equity Tier 1 ratio which reached 11.4%, up 30 basis points. Thus in the third quarter, BNP Paribas delivered good results and confirmed its solid capital generation. Turning to the exceptional items of the quarter on slide 5, you can see that pre-tax stood at minus €455 million and that essentially related to restructuring costs of our acquisitions and CIB transformation costs. These latter, in particular, were higher this quarter, as we booked up front the cost of some departure plans in some countries. And in addition, there was as usual the accounting impact of the OCA, DVA which was roughly negative €200 million this quarter, while previous quarter -- the third quarter of 2015, it was slightly positive.

On slide 6, you can see the quarterly P&L, showing the Group's good overall performance. Net income, excluding one-offs, actually stood at close to €2.2 billion, progressing, as I mentioned earlier, by 15%. Turning to slide 7, you can see the aggregate picture for the nine months. As you can see, our cumulative net income improved by roughly 4% on last year and this in turn means that we delivered strong returns. The annualized return on equity even excluding one-offs which were positive for the period, stood at 9.8%, while the annualized return on tangible equity net of one-offs reached 11.7%.

When calculated according to the parameters set in our 2014-2016 plan, that is to say excluding one-offs and based on the 10% Common Equity Tier 1 ratio, the annualized return on equities stood at 10.7% in line with our target of 10% for the full year. Focusing on revenues of the operating divisions, you can see on slide 8 the breakdown of the performance for the three divisions. Domestic markets revenues were stable which is a good achievement, despite the low rate environment. IFS showed a 4% rise in revenues and CIB delivered strong revenue growth in the third quarter. So overall, a 4.8% increase for the total which shows the effectiveness of the business model.

The following slide, number 9, you can see the equivalent picture for the cost of the operating divisions which were 2.7% higher on the back of a rise in regulatory cost and the strengthening of compliance, coupled with business growth for some activities. As planned, cost savings accruing from our simple and efficient plan offset the natural costs' drift. Jumping to slide 10, you can actually see the updates on our Simple & Efficient plan which delivered an additional €150 million recurring cost savings in Q3, taking the total to €3.2 billion since launch. I remind you that from this year on, we no longer incur transformation costs related to Simple & Efficient. Moving to the cost of risk, you can turn to slide 11 and you will see that the Group's cost of risk declined significantly versus last year, to stand at 43 basis points.

If we take the different businesses in turn, corporate banking marked another quarter of low cost of risk. On slide 12, the cost of risk was still low in our French retail banking and even very low in Belgium, while it continued to decrease significantly at BNL, where it improved by 49 basis points year on year. We're particularly happy with the evolution in Italy, given what we see elsewhere in this country. It validates our strategic choice three years ago to de-risk BNL's balance sheet by repositioning on better corporate clients. In the other retail businesses, on slide 13, personal finance saw another decline, thanks to the low interest rate environment and also to the gradual shift towards products with a better risk profile, like car loans.

Europe-Med cost of risk was higher this quarter, on the back of higher cost of risk in Turkey, as expected. And lastly, BancWest cost of risk remained at a particularly low level. Moving now to slide 15, you can see the significant improvement of our Common Equity Tier 1 ratio that I mentioned, 30 basis points improvement, resulting from the combination of the results net of a 45% dividend payout for approximately 20 basis points and the contribution from the IPO of First Hawaiian Bank for roughly 5 bps, exactly 6 bps, by the way. Risk-weighted assets were stable, excluding the foreign exchange effect of the quarter which itself, I remind you, has an overall negligible effect on the ratio. Our Basel 3 leverage ratio was stable at 4% and our liquidity coverage ratio stood at a high 127%.

So to wrap up, this quarter saw a continued increased of our Common Equity Tier 1 ratio, illustrating our well known solid capital generation capacity. If you turn to slide 16, you can see that we have actually received the 2016 SREP pre-notification from the ECB, this slide is interesting, it's the first of its kind. BNP Paribas has been set a Common Equity Tier 1 ratio requirement for 2017 of 8% on a phased-in basis which includes a G-SIB buffer of 1%, a conservation buffer of 1.25% and a Pillar 2 requirement, P2R, of 1.25%. This doesn't include the Pillar 2 guidance which is not disclosed. It remains confidential between the bank and the supervisor.

Today with a phased-in Common Equity Tier 1 ratio of 11.6%, BNP Paribas is well above this minimum requirement. All this is on a phased-in basis and for 2017 this is what we've got as a pre-notification from the ECB. Now if we try to anticipate the equivalent of this announcement in a fully loaded basis, we can calculate that in case the P2R would stay stable, the fully loaded Common Equity Tier 1 ratio requirement for 2019 would stand at 10.25%, based on the gradual phasing-in of the conservation buffer to 2.5% and based on the G-SIB buffer hypothetic of 2%. Again, as I just said, it doesn't include the Pillar 2 guidance and this level of Common Equity Tier 1 ratio would thus be the binding constraint from 2019 for MDA restrictions. And for the sake of clarity, in 2017 the binding level for MDA restrictions is 8%.

Then it follows that the anticipated level of total capital ratio requirement for 2019, again we see hypothetic that the P2R would remain stable and that the G-SIB buffer would remain also stable at 2%. It would lead to a total capital ratio requirement for 2019 of 13.75%, if we add the Tier 1 and the Tier 2 brackets of 1.5% and 2% respectively. Following this, we confirm both our fully loaded Common Equity Tier 1 ratio target of 12% and that for the total capital ratio above 15%. To complete this introductory part, please turn to slide 17, where you can see the continued growth of our net book value per share. At the end of September, our book value per share topped €73, while our tangible book value per share stood at €62.7.

It's a steady increase since 2008, illustrating the continuous value creation through the cycle. Moving now to the divisional results, starting with domestic markets on slide 19, domestic markets is our retail operations in the Eurozone. You can see that business activity has grown in Q3, the gradual improvement in loan demand in the Eurozone continued, as illustrated by the 3.1% loan growth of our domestic markets. And deposit gathering maintained a strong pace, with a 7.5% increase across all the networks. We're also continuing to evolve our digital offering, as exemplified by the recent announcement.

It's very interesting as well that we're bringing together our Wa! and Fivory of Credit Mutuel, in order to launch the common mobile multiservice payment solution next year that combines payments, loyalty programs and coupons. It will be developed in partnership with major players, like Carrefour, Auchan, Total and MasterCard, a major player in the retail world and payment world. As you can see, sizeable banking and distribution players are joining forces to propose an innovative solution to customers. Looking at the P&L, revenues stood at €3.9 billion, a fraction -- well, basically stable compared with last year, as the impact of the low interest rates was offset by a good performance of our Belgian retail and of our specialized businesses. Cost marked an increase, driven by the growing business lines, especially Arval.

But given the reduction of cost of risk, especially at BNL, pre-tax income increased significantly to €1 billion, up 9.1% on last year. It was very interesting to see that in such very low, persistently low interest rate environments, our pre-tax income of our domestic markets can increase by 9.1%. If we go a bit into more detail, French retail reflected the impact of the low interest rate environment, but we had enjoyed a good pickup in loan origination. BNL improved significantly its income contributions, thanks to the dramatic decrease of the cost of risk. Belgian retail confirmed good sales and marketing drive, delivering good income growth.

And the specialized business generated a strong rise in income on the back of a good overall business drive. Let's move now to slide 25, where you can see that our international financial services division delivered a good overall performance. The revenues on the P&L -- sorry, revenues progressed by 4.6% on a comparable basis and thanks to a lower cost of risk, pre-tax income improved to €1.4 billion, up 10.6% on a like-for-like basis. We'd now like to focus a bit more on each of the IFS businesses, starting with personal finance, on slide 27. PF continued to show sustained business activity in the quarter, with outstanding loans increasing by 9% on a comparable basis, driven by higher demand in the Eurozone and the effect of new partnerships.

Car loan business also continued to develop well. Revenues were up on a comparable basis, with the Belgian and Italian markets performing particularly well this quarter. Costs remain under control, while cost of risk marked a significant reduction. All this translated into a sizeable increase of pre-tax income which reached €411 million, up 13.5% at constant scope and exchange rates. Moving to our non-Eurozone retail activities, let's start with Europe-Med on slide 28.

Business activity progressed well in all regions, both in terms of loans and deposits. On the whole, Europe-Med contribution to Group results increased significantly to €165 million. I would especially point out here the increase in Turkey, increase of the result before tax in Turkey, where margins improvement more than offset the higher cost of risk I have already mentioned. Globally a very good quarter for our emerging markets retail banking activity. Moving to BancWest in the U.S.

on slide 29, the significant event of the quarter was the IPO of First Hawaiian Bank. We successfully placed 17.4% in the market at the beginning of August, generating roughly 5 basis points, exactly 6 basis points, as I said, to our Common Equity Tier 1 ratio in Q3. In the third quarter, BancWest business activity remains strong, with deposits increasing by 10% and loans by over 9%. We also continued to steadily increase the assets under management in our private banking which reached $11.4 billion, up 17% compared to last year. At constant scope and exchange rates, revenues were 4.4% higher on the back of strong volume growth which was, however, mitigated by a lower yield curve in the U.S.

as compared to last year. Costs were higher due to increased compliance costs, charges related to First Hawaiian's IPO and investments to strengthen the commercial setup. Given the very low cost of risk, BancWest pre-tax income stood at €210 million in Q3, a significant contribution, a bit lower than last year. Turning to our savings and insurance businesses on slide 30, you can see that our total assets under management broke through the €1 trillion mark for the first time this quarter. This good performance was driven by strong asset inflows in all our businesses, as well as a positive performance effect.

If we look at the insurance business on slide 31, Q3 revenues were 17% higher, reflecting the positive market evolution this quarter, compared to a decline in Q3 of last year, as well as the high level of capital gains realized. Turning to wealth and asset management, revenues showed a good overall resistance in an unfavorable context. To sum up, a strong income improvement this quarter in insurance, while our wealth and asset management marked good resistance in a still lackluster context for wealth management. Having completed the review of the retail banking and services businesses, let's turn our attention to slide 32 on corporate and institutional banking, where we're actively implementing the transformation plan in line with the plan timetable. In Q3 revenues topped €2.9 billion, marking double-digit growth.

It's a very good level for third quarter. Operating expenses increased at a slower pace than revenues by far, marking good overall cost control and benefiting from the effect of the cost saving measures. CIB's pre-tax income stood at €812 million, progressing by over 40%, 41%, compared to last year which is an excellent achievement. Still on CIB, if you could please turn to the next two slides, slides 33 and 34, to take a closer look at each of the three business lines. Starting with global markets which performed strongly in the third quarter, with revenues progressing by close to 20%.

As already seen with other banks, fixed income was the main driver of the quarter, the FICC business delivering very good performance in all business activities, as compared with the rather lackluster quarter last year in rates and ForEx. We confirmed our top position on all bond issues in euros and ranked number 9 again for international issues. Equity and prime services were lower compared to a very high base last year. When comparing results, it's important to bear in mind that in Q3 last year, by the way, our revenues, our CIB revenues, had shown 4% improvement, whereas our U.S. peers, for example, had shown an average reduction of 8%.

More specifically, U.S. peers had a poor FICC in the third quarter of 2015, minus 23% on average, while we were flat. And their equities were up by only 7% last year, compared to plus 20% for BNP Paribas. So the performance, the relative performance of this third quarter has to be assessed having in mind the basis of that comparison. Moving now to security services, this business increased its revenues by roughly 3% on the back of asset growth, both in terms of assets under custody and assets under management.

And finally, in corporate banking, revenue growth exceeded 9%. It was a very good quarter for corporate banking as well and was driven by higher volumes in all regions. We also had good contribution in that total from the fee income which increased 7.7% compared to last year. Transaction banking activities, such as cash management and trade finance, continued to develop well and marked good progress in the quarter. To wrap up, it's really an excellent performance for our CIB this quarter.

So to conclude, what I would like you to retain is the following. We delivered good revenue growth in our operating divisions, thanks to our integrated and diversified business model. The Group's cost of risk marked a significant decrease. We confirmed our solid capital generation and our Common Equity Tier 1 ratio stood at 11.4% at the end of September. And our return on equity stands in line with our 2014-2016 plan that's in the region of 10% which is quite good.

Thank you very much for your attention. I will now be pleased to take your questions.

Operator: [Operator Instructions]. Our first question is over the line of Lorraine Quoirez of UBS. Please go ahead.

Your line is open.

Lorraine Quoirez: I have three questions. The first one is on BNL, so basically I think you had the cost of risk guidance for 2017 of 100 basis points, but I'm not sure and I think the run rate was 80 bps. Is it something that you confirm or could we actually go below that 80 bps? In personal finance, a little bit the same question. I think the guidance was 180 bps, but I understand there is a structural change in mix.

Should we think about a lower run rate going forward? And finally, on First Hawaiian Bank, you IPO 17% so far and I think in the past you said you'd like to fully dispose of it by the end of 2017. Is that sort of timeline still valid please? Thank you.

Philippe Bordenave: On BNL, we confirmed that the run rates should be 80 basis points. Given the current level, I think that 100 basis points for next year is a reasonable assumption. So I confirm.

PF, yes, the business mix is changing because we're developing car loans, we're focusing our developments on car loans. Exactly the proportions of car loans in the new production is much higher than the proportion on the back book and hence the structure of the book is improving, because the cost attrition is the cost of risk on auto loans is lower than for other type of consumer loans. And so we consider that really the 180 basis points is on the safe side and yes, we can hope that -- well, it's still a bit new, so we lack some experience, but we could maybe reasonably expect that the run rate could be a little bit lower at PF especially as long as interest rates are going to remain very low. First Hawaiian, we had said that yes, we wanted to sell before the end of 2017. It's clear that given the pace at which we're improving the ratio on an organic basis and now that we have made the IPO, so we have evidenced the ability to sell.

And we're going now to optimize the sell down of the remaining stake. We're going to choose the best moments to pick the -- well, try and pick the best time to sell and hence we're no more in a defined timetable, I would say. We'll be opportunistic and if it goes somewhat beyond the end of 2017, it's not going to be an issue.

Operator: We now go to the line of Delphine Lee at JP Morgan. Please go ahead.

Your line is open.

Delphine Lee: Yes, thanks Philippe, for the presentation, three questions on my side. First of all, I just wanted to check on the capital improvement of 30 basis points this quarter. I can see from your slide 75 that there has been a lower deduction from -- on the regulatory deductions, they seem to be quite lower this quarter. CET1 does include 5 basis points from the IPO, but I'm just wondering if there's anything else that explains a little bit the move.

There seems to be 10 basis points which is not completely explained by results or by RWA reduction. Second question is on just the SREP, when we think about the new 8% requirement for next year, is there going to be some kind of impact in the way you think about your dividend policy in the long run? Or do you think it would be the main constraint is still your 12% CET1 target in the long run? And related to that, in terms of order regrowth, this quarter seems to be down slightly marginally, minus 0.5%. Has there been any active, proactive management of all the risk reduced debt beyond maybe some FX impact? I'm just trying to think about your long term guidance of maybe 1% or 2% and if that can be maybe a bit lower, so you can generate a bit more excess capital. And the last question is very easy, it's on the corporate center, if you could provide a bit of maybe guidance, because it's been quite strong for a few quarters now in terms of the capital gains. And if we strip out own debt or Visa, for the first nine months you're already at €670 million of revenue, so just trying to think what the new run rate could be, if there's any change.

Thank you.

Philippe Bordenave: Okay, on the corporate center we keep our guidance which is €60 million per quarter ex-one-offs, adjusted for one-offs. We think -- well, the corporate center indeed is relatively -- it has fared well, but it's not -- what maybe makes a difference between our corporate center and others is that we don't have any legacy part in the corporate center. So there is no part which would create regular losses or remaining losses or capital losses due to legacy assets. Coming back to your first question about the adjustments on equity, yes, it has gone down.

One of the reasons why it goes down is that some of these calculations of what has to be deducted from the equity are cash flow revaluations. And so the calculation can change from one quarter to another. And also it's also linked to the adjustments and the OCA DVA adjustments that are also higher this quarter, as you have seen and that has been passed in the P&L, instead of being in the deduction from the equity. About the dividend policy, we stick to our 45% up to the end of this medium term plan, meaning up to the end of this year, including the distribution of this year, for this year, happening next year. From next year on, we'll launch a new plan, the 2020 plan and we reassess our dividend policy.

Given where we stand, as we're already close to 12% now, we may adjust the dividend policy and increase somewhat the payout ratio. It will depend a lot on the regulation, the regulatory environment, depending on how it will evolve. We're facing a lot of uncertainty, as you know, at the moment, so it may be too early to elaborate on that. But we -- well, it's going to be part of the new plan. About the risk-weighted assets, in the risk-weighted assets the reason why the risk-weighted assets are down is foreign exchange effect on certain risk-weighted assets, mostly the sterling decline and also the Turkish lira to a lesser extent.

But this, as I said, on the full ratio basis, there is no impact because the numerator of the ratio is also impacted by the foreign exchange effect on the same proportion, so on the ratio there is no impact. So if you adjust for the ForEx, our risk-weighted assets were stable indeed this quarter and they were stable because on the credit side they were up due to the increase in the outstanding amounts of credit, as you have seen a little bit everywhere in limited proportions, but still up. But they were down on the capital market side because both counterparty risk was down and the VaR, the value at risk, was down as well, as you have seen. We have an extremely low value at risk this quarter. On the counterparty risk side it's more coming back to normal.

The level of counterparty risk was a little bit high at the end of June, because at that time we had a lot of volumes of derivatives business in the context of the Brexit which is now back to normal. So this is why the risk-weighted assets are stable, more on the credit side, less on the market side and that's it.

Operator: We're now over to the line of Jean-Francois Neuez of Goldman Sachs. Please go ahead, Jean-Francois, your line is open. Jean-

Francois Neuez: So a couple of questions here as well.

On the dividend, you made that comment that you may increase somewhat the payout going with your next business plan. The question is it still sounded a little bit in a sense shy, the comment that you made and now that you're close to the 12%, as you said, in a sense what is the reservation for not necessarily targeting to pay out maybe the entirety of what's left after risk-weighted asset growth? And the second thing I wanted to ask is on the retail side of the operation. Now we get a little bit of revenue growth, but what's very sticky is the cost growth here and I just wanted to try to understand the level that you've got when you look going forward for essentially the operating leverage in that division. Because otherwise, according to the last time you participated to the call, to do 10% with a non-tangible equity -- on equity rather, for the 12% capital ratio. Without that happening it looks tough to achieve.

And lastly, I just wanted to ask you whether you've made any plans already, now that we have the date for the Article 50 triggering in the UK, given that you have a sizeable operation in your investment bank there in Marylebone. Thank you.

Philippe Bordenave: Well, starting with your last question, we really don't expect a big change and we're very happy to be in Marylebone and to stay in Marylebone, a very nice building which we own, by the way. And we're well balanced. Well, we're lucky, being a continental bank, we have both significant teams -- market and CIB teams in London and in Paris and also in certain other cities.

But I would say for the capital markets it's mostly London and Paris. And they are well balanced and so we can say we can adjust to any scenario, almost any scenario of Brexit, soft or hard or whatever, without having to make dramatic moves like other banks that would be only in London or mostly, largely in London, would be obliged to really make a physical move which is always a waste of energy and time and money. We intend also to be in every country in Europe, we try to be close to the clients in every country in Europe. Now moving up to your questions about the dividend, we're happy to have that kind of question, it shows that we're rich. Clearly again it's going to be assessed in the framework of our new plan.

As I said, the reservation is mostly the uncertainty on the regulation and what exactly is done, how it will turn out, the CRD5, Basel 4 and so on, when all this will come to an end and how it will be fine-tuned. And so we're still in a heavy fog in front of all these discussions. On the operating leverage and the cost growth in retail side, while it's clear that the new plan will have a chapter of cost savings again, the cost saving is up, it's clear also that we'll have a chapter as far as retail is concerned like in other areas, by the way, that is going to be linked to the digital evolution. And digital evolution implies both some savings, because more things are done by the clients themselves and more things are -- more processes are completely automatized and industrialized and requires less manual intervention. But it entails also some cost because it's a lot of IT cost.

And so we have to -- well, all this has to be assessed in the plan, we're working on it and you will get the answer in due time when we release the plan. But we're clearly conscious that we have to work on the cost side and we do. Jean-

Francois Neuez: If there is any reservation on the payout it will be because of regulation and if that was to be settled, you wouldn't necessarily have any prejudice against hiking substantially the payout?

Philippe Bordenave: Yes, we can say that.

Operator: We're now over to Jon Peace at Credit Suisse. Please go ahead.

Your line is open.

Jon Peace: So I had two questions please. One was on French retail banking, just wanted to get your thoughts on when we might see an end to the negative year-over-year revenue trends. And then the second question was just on the cost of risk. I guess it's continuing to surprise on the upside.

How low do we think that might go and do you have any thoughts on IFRS 9, what that might do to provisioning and also just the impact on CET1? Thank you.

Philippe Bordenave: Well, the French retail is impacted by the low interest rate. So I think the answer to your -- the short answer to your question is, I think, to last until interest rates are increasing again somewhat. We try and, as you have seen at the level of our domestic markets, pull together, including Italy and Belgium and Luxembourg and the specialized businesses we manage to keep stable revenues. So that's the beauty of the diversified model.

We believe that we have more stability by being spread over several countries and within the Eurozone at the same time, we have the synergy of the asset liability management and the treasury. So our model ensures that domestic, as I said during the presentation, that revenues are stable and in such a quarter and net before tax even increasing by 9%. Of course, in the meantime, in all our countries, we're working at increasing the fees, because the low interest rate arguably is an answer as far as net interest income is concerned. But then you have the fees and on the fee side we can expect some improvement as soon as the stock exchange markets would improve. At the moment there is a strong reluctance of rich clients or even average clients, to invest in equity products or equity-related products which of course are the products where we have the most added value.

So it's clear that this attitude, kind of very shy attitude of our clients in front of the stock exchange is not helping to generate fee revenues. That may improve earlier than the low interest rate changes. On the cost of risk, I think I have already answered to Lorraine or Delphine at the beginning, it was Lorraine. Yes, we think that in Italy we're going to be go down fast at the moment and we should reach the 80 basis points within probably two years now or something like that. And in PF, we maybe -- we think that the new mix would help stay below 180 for quite some time.

And both are representing two-thirds of the total cost of risk of the Group. We don't see any early sign elsewhere of any deterioration, so we're confident on the trend. Again I think that at least partly that it's due -- well, it's helped by the low interest rate environment. Your last question was about the impact of, sorry?

Jon Peace: Yes, IFRS 9, thank you.

Philippe Bordenave: Yes, the impact of IFRS 9.

Well, we know these impacts will not be huge for us. Clearly we're working on it, but there are three chapters for IFRS 9. The impaired loans there is basically no change, so it shouldn't be a change. The watch list loans, I would say, are the loans that are problematic, what you call the group 2 or the stage 2. You know that we're already flagging those loans, we're already taking provisions on the portfolio basis.

So the way we have implemented IAS 39 was to consider that a deterioration of our internal rating for loan at a certain level was making that loan potentially impaired. So we had considered that there was indeed an incurred loss in that case. So we have created provision on the portfolio basis of those loans. So the only difference will be on the methodology which is more precise in IFRS 9. But the impact should not be huge and we have, of course, the full impact for the good loans, what they call the stage 1, where we'll have to provide for a one-year expected loss, even for the best loans, without any outlook for any loss.

This, of course, is going to be in addition to our provisions, but it should not -- at the same time we're talking about the best loans and we're talking about one-year expected loss, so it should not be huge. So we don't have yet all the -- we don't know yet how it will be integrated and interpreted by regulators and supervisors. So we're working on it, but we're not expecting huge amounts.

Philippe Bordenave: On top of this, we will argue that for such loans, for such change, I would say the quality -- the intrinsic quality of the balance sheet is not changed; it's even improved. The additional provisions are going to be -- yes, they are going to deteriorate our ratio but indeed they are going to improve the balance sheet symmetrically.

So we're going to argue that this should be offset by a symmetrical reduction of the Pillar 2 R logically. I don't know if we're going to be followed, but logically there is no reason why we should -- all banks should add to their balance sheet because of a mere accounting change which is translated into an additional protection within the balance sheet, instead of the protection on the prudential calculation. So one should be exchanged for the other.

Operator: We're now over to the line of Flora Benhakoun of Deutsche Bank. Please go ahead.

Your line is open.

Flora Benhakoun: I've got two questions please. The first question is on the capital requirements from the ECB. So obviously, it's clear from your slide pack that the MDA went down, the SREP we don't know anymore obviously. So my question is whether we should consider that the capital level that you need to reach to be able to distribute your dividends freely from the regulator is the MDA or is it still the SREP? And what happens if your capital, so Common Equity Tier 1 ratio is above the MDA, but below the SREP.

So what would the ECB do, you think, in that case? And the second question I have is regarding the recent deepening of the yield curve. I wanted to know whether you could make some quantitative or at least qualitative comments as to how you think this could impact BNP. Thank you.

Philippe Bordenave: Any deepening of the yield curve is positive for a bank that is transforming relatively short term deposits into longer term loans. So the answer is clearly that any deepening is favorable.

Up until now it's very, very small one, so we would like to see more. On your first question, clearly the constraints on the distribution -- MDA means maximum distributable amount and distributable amounts means distributed to the shareholders with the dividends. Also alternative coupon, the debate had been focused on the alternative coupons, the alternative Tier 1 coupons as being of this year, but what it entails in maximum distributable amount is restrictions in dividends, alternative Tier 1 coupons and even bonuses. So this is the regulation, so that's for your question the answer is here. The second part of your question is what would the ECB do and so they have answered, so I refer to the big document the ECB has issued on July 26, in order to explain their stress test exercise and the impact that it would have on the P2R and P2G.

And they said that, I quote, if a bank will not meet its P2G, this will not result in automatic action of the supervisor and the courts, end of quote. It will result, quote, in fine-tuned measures of the supervisor based on the individual situation of the bank, end of quote. So one can assume that for them it's a kind of level that would -- well, I can tell you they are watching us very closely. And I think -- well, my personal interpretation is that if a bank would become too close to that level, there would be already discussions even before the level would be breached. And well, those discussions would result in probably decisions that could be of a different nature, depending on the reason why the level is breached.

So it's intended to be not automatic, to be not -- to be a level that is part of the supervisory dialogue between the bank and the supervisor and not something that is automatic and triggering publicly seen decision. I think it's going to be the essence of supervision of a savvy supervision. And the MDA level is something more brutal, so our MDA for us now, it's 10.25% at this stage. Of course, each year it can change, so it's also another tool in the hands of the supervisor. He can also move the MDA level, depending on the behavior of the bank and each year.

Operator: We're now over to the line of Maxence le Gouvello of Jefferies International. Please go ahead. Your line is open. Maxence

le Gouvello: Philippe, I will have two questions. The first one is regarding the fees in BNL.

You mentioned in your statement that the financial one was going up, but the commercial ones were down. Can you please explain to us why, because it's a bit counterintuitive when all the banks are trying to push up their fees on that side? The second question is regarding Belgium. You have very good top-line growth in terms of NII, good volume on mortgages, but you are always very shy about the level of renegotiation in Belgium when some of your peers locally are saying that the impact is not marginal at all. And the last element is regarding BancWest. I'm a bit surprised with your usual very cautious underwriting strategy that close to half of the growth of BancWest is focused on commercial real estate, especially at that moment in the cycle in the U.S.

Thank you.

Philippe Bordenave: Well, on the BNL fees, the banking fees are down somewhat this quarter, partly due to the fact that the repositioning on the corporates make that the new -- well, the former clients basis was generating more fees than the new one, because the bargaining power of the client was lower. So we're now developing a client space where margins are a bit thinner and fees are a little bit scarcer as well, because they are in demand a lot, those clients. The other reason is that it's less structural. It's just that this quarter we had a little bit less of mortgage origination on individuals and the mortgage origination is generating fees, one-off fees at the beginning of the mortgage and they were a little bit lower this quarter.

About Belgium, the renegotiations, it's a phenomenon that exists in Belgium as well, not only in France. And yes, it's true that -- well it goes with the low level of interest rate and it weighs on the revenues. We're not breaking down the revenues in different countries, but yes, renegotiations are there and they are going on the revenues and they make life difficult in Belgium as well as in France. In Belgium -- the difference between Belgium and France is not that much on the asset side, but more on the liability side, because in Belgium we were able to re-price down the cost of the carnet de depots, the deposit schemes, because those deposit schemes in Belgium are not regulated and that makes a difference with France. In France, the prime is that the rates for the customer are going down on the asset side.

But on the liability side, they are blocked by the regulations on the real estate saving scheme, the Plan Epargne Logement, on the one hand. And Livre A, the A savings scheme on the other hand. And together, the remuneration on the carnet depot in Belgium are now at 11 basis points to be compared with 75 for the equivalent in France. So this is why the situation up to now has been better in Belgium. Now, we're at the bottom of this repricing of the liability and so the evolution in the future could be a little bit more difficult for Belgium as well.

Maxence

le Gouvello: Sorry, on the French retail, as you give us just a bit more detail, the NII is down by 4% but roughly can you give us a rough idea of how much is driven by pressure on the asset side and how much is driven by your inability to move on the liability side?

Philippe Bordenave: Well, I think it's coming mostly from the deposit side, I would say. It's difficult -- after that, it's a question of ALM. It's the comparison between both which makes a difference. But you should not underestimate the liability side. Now on BancWest, on BancWest there is nothing special to say on the commercial real estate growth.

On individuals, the growth is 5% and corporate 10%. You know that the -- in the U.S. classification, when you -- each time you take a real estate collateral, it's called commercial real estate. So if you lend money to corporate in Europe and you take a land or a factory as collateral or something like that, it's going to be classified as a corporate loan with a collateral. In the U.S., going to be classified as commercial real estate.

Even if it's not really -- the client is not a real estate owner, he's not a company -- it can be a normal company of any industry.

Operator: We're now over to the line of Jean-Pierre Lambert of KBW. Please go ahead. Your line is open. Jean-

Pierre Lambert: I have three questions.

The first one is related to the revenues in the Eurozone retail domestic markets. They've been flat thanks to diversification. But when you look at the detail, it's a bit flat mainly thanks to Belgian NII. And so as you just pointed out, we're going to live in a different situation with the floor on what can be remunerated and how much you can adjust the pricing on savings deposits. So the outlook is probably not as favorable if Belgium cannot absorb, if you want, the negative trend.

And then that leads to the second question. Is there need for more radical cost initiatives going forward? You mentioned the plan, but if we take IT for example, you use IBM in three countries of operation, we've seen ING shifting its IT platform from Belgium to the Netherlands. Can you envisage something similar in your Group? The third question is related to Belgium. We've seen that bpost, where you have a 50% shareholding, is putting some more aggressive prices in mortgages, somewhere like 20 basis points below the market rate. And so I was wondering why you allow this to happen? Is it a segment of clientele you are not interested in? What is your point of view? Thank you.

Philippe Bordenave: Well, on bpost, it's a joint control, it's not an exclusive control. So it's not completely our decision. On the revenues in the Eurozone, I agree with you that we're going to be under pressure. We're moving all the levers, meaning as I said, also we're working on the non-interest income as well. Trying to improve the fee levels.

We also expect now the cost -- well, the opportunity loss on the revenue side that we incurred in Italy for the last three years, because of the re-de-risking of the corporate business, we expect that to come to an end now. And to the contrary, we expect all the commercial investments we have done with the new client base to start paying off in Italy. And after having said all that, I agree with you that we need more cost initiatives. So I take on your suggestion. We have also some -- well, some optimizations of our datacenters between Belgium and France already.

And with a complete setup in partnership with IBM, you're right. But beyond that, we're going to try and address all the potential cost initiatives. And this is really part of the plan we're preparing. And we have already started. In Belgium, as you are referring especially about Belgium -- I wonder why -- in Belgium, we have announced some further closure of branches.

The country is pretty much dense in terms of bank branches. And we think that we can organize our setup better. And recently, we have announced news of that kind. We're going to, for example, to close between 32 and 42 branches next year, as was announced recently by the Head of Belgium. And we're going also to outsource 20 more, 20 to 30 more, to independent agents, like it's possible in Belgium.

It's a business model that is relatively successful in Belgium. And so yes, we're acting. The only thing is that to be compared with other banks, including in Belgium -- I am not going to quote any name but you will understand -- we prefer to be relatively low key and to achieve these savings step by step and without making too much noise because we think it's more efficient than staying in bed for a certain period of time and then announcing a big change, a big cut which is necessarily creating -- unsettling the staff and the clients. We prefer to have a progressive evolution, progressive adaptation to the new way the clients use the branches, the progressive reduction of cost rather than announcing big steps. We feel -- well, it's less -- well, we prefer that way, it's our way.

But we're not inactive at all, believe me.

Operator: We're now over to the line of Alex Koagne of Natixis. Please go ahead. Your line is open.

Alex Koagne: So from my side, I have a few questions.

The first one is on the CIB. Could you please elaborate a little bit on market share gain, both in America and in Europe? I think that you put more -- this was one of your targets. Can you just update us on how are you performing so far? The second one is on margin expansion in Turkey. Can you just elaborate on how you see margin going forward? Is there any increase that we can expect for the coming quarter? The third one is still on the cost initiatives in the retail banking. I'm just trying to understand what could be the size.

So the question is, if you look to your core countries, domestic country, in which country do you consider that you have more a revenue issue than the cost issue? Because the cost to income is not always a relevant indicator. So I'm just trying to see in which country you think that you have more a revenue issue than a cost issue? Thank you.

Philippe Bordenave: About CIB, we're gaining market share. I think just -- well, we're following, there are consultants and so on following the total revenue pool, monitoring the revenue pool for the full CIB market by market, segment by segment and providing us with the evolution of our market share. So the only way in CIB is the share of the revenue pool.

Because otherwise, you have the rankings but it's only for the new issues and only for part of -- a small part of the total business. Indeed, as far as derivatives over-the-counter trading, you have no statistics. So the only way is to track the total revenue and the revenue pool. While we see that we're gaining market share, both in Europe and in the U.S. and that we're maintaining our market share in Asia in a growing market, we -- well, the reason for that is not -- well, it is also that we're very good.

But it also I think has to be counted. It's so that a certain number of competitors are withdrawing from the business or from certain segments or certain geographies. You have seen the -- well, one very good example, very well-known is what happened to RBS when they were obliged to drop most of their CIB business everywhere -- almost everywhere, except maybe in the UK and Ireland. And while we're proud to announce this quarter that on the cash management business -- because for the cash management, they were keen not to stop the business on the continent without providing their clients with an alternative solution. And so they launched an RFP to have -- well, to have another banking group to be able to refer their clients to the follower.

And so we won that RFP at the time. It was one year ago or two years ago. And then we went -- or they introduced us to their clients. Of course, the clients themselves, they organized RFP because they didn't want to rely only on one solution. But we were part of this.

We were at least in the panel that was systematically -- we're in the panel for their clients and we won in several cases. And we're counting that we're now ensuring the cash management for 215 new groups, corporate groups. So the number of companies much bigger. So it means that clearly this is creating a gain in market share, obviously, for cash management. And it's also indirectly producing closer relationship with 215 groups.

Because when you manage the cash, the money of a group, of course you are very close to it. It's a kind of intimacy and you have the opportunity to propose other services starting with the ForEx. Because it's about international cash management, but also other service. And hence we -- well, for us it was an opportunity to increase our core market share in several products. But beyond the RBS case, you have a lot of other cases.

You know that the Swiss banks have considerably reduced their FICC business. It's another example. So we have -- well, I am not going to multiply you. You know the story yourself. So this is what -- I guess we're not the only one either, to increase our market share.

But clearly, we're one of the winners in this big change in the market with the new regulatory pressure on the CIB business, new regulatory requirement, obliging a certain number of banks just to stop and to withdraw. About Turkey, I said that -- and would like to stress -- that Turkey is of course going through a difficult period following the failed coup. And it has entailed a slowing down in the business environment in Turkey, less growth. And it has entailed some increase in the cost of risk for corporate especially. But at the same time, the margins have widened, because the country risk perceived is higher.

And so overall, we -- our TAB, the Turk Ekonomi Bankasi, our subsidiary in Turkey, is doing fine, really and increasing its results, the net results. In this environment, they're very cautious on the risk side, very active on the cost side as well and benefiting from the increasing margin. On the cost initiative on retail banking and the cost/income ratio, I would say that what's happening in the retail is more than -- well, I think it's more than a reduction of the revenues due to the low interest rates. Of course, this is important with -- this is clearly important. But it would be a mistake to address the issue only with that angle.

I think we have to address -- infamously, we have to address the digitization of the retail and to understand that the clients are changing the way they use their bank. And so we -- and this, here again, there is -- what is at stake is both revenues and cost. Because we have to be careful not to keep branches that would be progressively empty because clients would not come any more because it's not the way they want to work with us anymore. And we also have to be careful not to lose some segments of the business to fintechs or new entrants that would have a favorable environment starting from scratch and favorable regulation as well in certain cases. So we have to work really simultaneously on the operating efficiency, but also on what we call the client -- the customer journey.

The client experience or the customer journey. Meaning that we have to adapt to what our customers want. And so we have started redesigning our processes in order to both make the customer experience seamless and easier. And also, behind that, to have the end-to-end process much cheaper because it will be completely automatized. So this is a complex transformation which has to be rolled out on all our domestic markets.

And to that respect, I would say the differences in cost/income ratio are less -- I would say less relevant than the threats coming from the evolution of the business. But the good thing is that we have the clients, the clients are accustomed to work with us, they tend to be loyal. If we're able to provide them with the right service, they are going to stay with us, because we're reinsuring compared with new entrants. And so we can work on the real scale and the real life. We can change the bank in real life.

And to that respect, what we're building, for example in terms of electronic wallet with Credit Mutuel, it makes sense to work together because it provides us with a better critical scale at the beginning. And also, we're sharing the cost. Again, all that entails significant IT cost. And so it makes a lot of sense. Because if we can -- I think due to all our relationships with the clients, with the retailers and so on, we can really -- we're better positioned than fintech to build a real-size, real-life business which meets, really, the needs of a lot of clients.

Operator: We're now over to the line of Bruce Hamilton of Morgan Stanley. Please go ahead. Your line is open.

Bruce Hamilton: Two questions from me. Firstly, just on the wealth management division.

I'm just trying to understand; the revenues are quite weak both QonQ and year over year, despite asset growth and flows being reasonably robust. So I just wanted to try and understand what's going on there. And then secondly, just going back to the regulatory fog. We're approaching the final proposals at Basel, there still seems to be quite a big gap between the comments of no significant capital increase and the numbers done bottom up. So would your base case be that this ends up being pushed down to a European level and there's lots of recalibration before it passes into law and therefore, we don't really get much clarify? Or what odds would you put on Europe just simply refusing to adopt? Just interested in your latest thoughts on that, thanks.

Philippe Bordenave: Well, on the wealth management, the new cash of today is making the revenues of tomorrow. When you get the new cash, initially very often it's just deposits. So initially it doesn't make money, even it can cost, given current rate. But then you are going to discuss with the client and progressively it's going to be invested in more value-added products. Initially, it's invested in very simple products.

And so the new cash of this year is probably a good sign for the revenues of next year. But it's not completely happening exactly at the same time. On the regulatory front, I'm not going to comment, sorry, because we're -- well, it's too confusing and too uncertain. And unfortunately, I am not in charge. So we will wait until they make up their mind.

Operator: We're now over to the line of Jacques-Henri Gaulard of Kepler Cheuvreux. Please go ahead. Your line is open. Jacques-

Henri Gaulard: A couple of questions. Point one, sorry to come back on that SREP thing but I would like to widen the debate a little bit beyond the dividend right.

You have now this, I would say, capital requirement of 10.25% at MD level which is [indiscernible] last year was 11.5% comparable, including what would have been the P2G this year. And I have problems to believe that that P2G could be as high as 1.25%. So the question to you, really beyond the dividend is, do you really have to manage the bank at 12% Common Equity Tier 1 ratio, frankly? Even in agreement with the regulator? That would be the first question. And then just a couple on the divisions. At personal finance level, the consumer credit had a very, very good -- very good trends, with loans up 9%.

I'm a bit surprised by the fact that the revenues are up just 0.9%. And I wanted to know -- because intuitively, that would be a division less sensitive to interest rates than others, but maybe I'm missing something. And lastly, we had a bit of a hike on the cost of risk in Turkey. And I wanted to know how we should proxy the cost of risk in Turkey going forward? Should we stick to 189 basis points or should we go back to something a little bit more -- a little bit lower? Thank you very much.

Philippe Bordenave: Jacques-Henri, good questions as always.

In the P2G and the 12%, what we think is that we're keeping our current targets. We had set those targets last year, 12%, 15% for the total capital. We have decided to keep them as they are, because -- the reason being, again, the uncertainty on the regulatory front. So we think that everything is so uncertain about what's going to be asked from us in the years to come that it's not the right time to change anything. So this is basically why we have decided to stay there.

If, one day, things are settling down and if one day there is a stable landscape, if one day they stop changing every year then we'll revisit and maybe we'll be able to fine tune that figure. About PF, the reason is -- well, there are two reasons. Obviously one reason -- there is one good and one less-good reason. The less-good reason is that there is some pressure on margins, like everywhere. So the growth in revenues is not strictly proportional to the growth in volumes.

It's a pity but it's normal life. And the other reason, it's a better reason, it's that, as I said, we're refocusing on more car loans and margins on car loans are lower than on usual consumer loans. And the cost of risk is also lower. So it's another business model, somewhat. But we consider that it's better to pay a little bit, if I may say so, on the revenue side and then to be able to lower the structural cost of risk.

And then on Turkey, well, Turkey is evolving a lot. It's also in a region which is currently under stress. So it's difficult to anticipate what will happen. I can't tell you. We're --really it's difficult, it's a little bit day-after-day evolution and we adapt locally to the reality, the local reality.

But it's difficult to know how things are going to evolve in the future.

Operator: We're now over to the line of Geoff Dawes of Societe Generale. Please go ahead. Your line is open.

Geoff Dawes: I've got two questions.

Should be quite quick, slightly related. The first is, if I look at your immediately-available assets, you had a very big jump in the third quarter to pretty much the highest level I can remember seeing, at least for a few years. Is there any particular reason for that? Is there anything going on behind reinforcing that pool of liquid assets? The second question is, if I step back and take a top-down look at some of the money market fund data coming out of the U.S., it seems there's been a very, very big fall in money market fund U.S. dollar lending to French institutions. Again, do you have any comment on that, is that something that you're partly involved in, are there any specific reasons for that? Thank you very much.

Philippe Bordenave: Indeed, your questions are interrelated. It's especially because we had -- well, we had anticipated that the change of rules applicable to the money market fund in the U.S. starting in October would create a potential shortage of short term funding, that we had anticipated that and we had decided to push a little bit more our deposit raising elsewhere and in order to be a little bit on the safe side, by the end of September. And this is why, in practice, the change turned out to be relatively smooth. So we have seen a decline in the deposits of money market funds, especially a reduction in the maturity of those deposits to a shorter maturity.

But in proportions that were less dramatic than we expected. And at the end of September, we were a little bit over funded with an equity coverage ratio at 127% as you have seen which is obviously too much, because the threshold is 100%. So we're probably going to come to a lower level both in liquidity coverage ratio and in liquidity reserve by the end of the year.

Geoff Dawes: Costs associated with this or costs in the future that would be associated with the change in funding?

Philippe Bordenave: The costs have been incurred already. The additional cost to that has been incurred mostly in the quarter.

There will be some costs in October as well and then probably we'll come back to a more normal level by year end.

Operator: We're now over to the line of Anke Reingen of Royal Bank of Canada. Please go ahead. Your line is open.

Anke Reingen: I just had three questions.

The first is on your -- in your ROE calculation, you show €350 million annual cost for remuneration of sub debt. And I just wondered how we should think about this in the next years, especially with more AT1 issuance? And then on the SREP, I'm sorry to follow up, but obviously the 10.25% [indiscernible] do you think it's somewhat flexible depending on what will be coming out in Basel IV for 2017? Or is this really -- you say it's still subject to confirmation but how much flexibility is there? And then lastly, I was just wondering your progress on the risk-weighted assets reduction in the CIB division? If I understand correctly, it was €6 billion out of a €20 billion at the first-half stage. And I just wonder where you are now? Thank you very much.

Philippe Bordenave: On the sub debt, yes, the more we issue alternative Tier 1 the more costly it is if that's of use. As everything is public, you can make the calculation yourself.

We're not issuing very often but we already did that twice over the last 12 months or something like that. So yes, it will cost a little bit more with time. The cost of security, I would say, for the senior bondholders. On the SREP, I can't answer your question really. It's -- the Basel IV is more about inflating the risk-weighted assets than changing the level of the ratio.

But it would be logical indeed if the risk-weighted assets are inflated that the requirements go down in percentage. But I don't know at all what are the intentions of the SSM. And as you know, we're in a complete uncertainty and year after year, we have to adapt to new rules and new ways of doing. And on the risk-weighted assets of CIB, the reduction was €7 billion. We had a reduction of €7 billion.

Operator: We're now over to the line of Stefan Stalmann of Autonomous Research. Please go ahead. Your line is open.

Stefan Stalmann: I have a couple of questions around the French retail market. And in particular, the mortgage market.

It seems that you're now growing at least in line with the market again, after having lost market share in mortgages for a couple of years. And if I understand you correctly, you're always a bit hesitant to engage with these very long-dated mortgages and the potential interest rate risk that results from that. But that may have now changed. Could you talk a bit about how you weigh these two things? Market share gains and losses on the one-hand side and the interest rate risk of these mortgage products on the other side? And second and maybe related, there was an interesting article the other day, I think it was in Les Echos, discussing how the French banks were basically preparing to raise account fees again next year in France. Could you maybe provide a bit of color on whether that is actually the case and what could be expected there? Thank you very much.

Philippe Bordenave: On your second question, I don't have any comment to make at this stage. I am sorry. On the first question, yes, we thought that there was a limit to the market share we could lose on the French market, because we're not that big. Usually, it's the market leader that has to put in play part of its market share in order to try and go a little bit against the trend. We cannot do that ourselves with our 8% or 7% market share.

So yes, we tried to be more reasonable than the competition. It didn't work, obviously. It didn't make the competition more reasonable. And so at some point, we decided to stop in order to keep a critical size. We can't be too small in -- nobody can be too small in a market.

So at this stage, we have stabilized our market share. And it entails some rebound in the production. So in terms of in short term -- well, if you look over one year, we have stabilized the mortgage loans volumes. But on a longer -- if you compared with second quarter, we have indeed increased. So we went through the trough and now we're rebounding in terms of total outstanding amount in order to catch up with -- as you said, with the market.

And I think we have no choice, we have no choice. And we're going to manage the interest-rate risk it entails. The only thing where we still maybe have a differentiation with our competitors is that we try and focus more our production of new loans on the relatively not-too-long maturities. I mean up to 20 year but not up to 30, in order to -- well, at least in terms of maturity, to have some kind of reasonable average maturity in our books because production of 20-year loans makes an average after -- it makes 10 years on average book in [indiscernible] and it's manageable. When you go to 30 or more, it's more difficult.

So this is what we're doing. And yes, you have noticed it and I think we had no choice but to change our stance.

Operator: We're now over to the line of Robin Down of HSBC. Please go ahead. Your line is open.

Robin Down: Sorry, I waited an hour and a half to ask the same question that Stefan just asked about the French mortgage market strategy. But maybe I could follow up quickly on the capital side. We're due to get the GSIB numbers from the FSB in the next three weeks or so. I was just wondering if you've harbored any hopes of dropping down a bucket on that? You look like you were quite close to the border using last year's numbers, if you like. And then can I just make a quick observation.

And this is something I have made to IR in the past. But the way that you're doing your calculation of the TNAV and the ROTE on slides 72 and 73, they do feel quite inconsistent to me in that one is using a re-evaluated book and one isn't. So just maybe be an observation that it would be nice to have numbers in a more consistent basis going forwards. Thanks.

Philippe Bordenave: Yes, well on the GSIB, I don't have any clue at this stage.

We have to wait until the final decision. On the calculation of the return on equity, we always calculate our return on equity, be it return on equity or the return on tangible equity, on the basis of non-re-evaluated balance sheet because we believe, as you know, we believe that the revaluation of bonds, of available for sale bonds, is not real. At the end, you are going to be reimbursed only for the face value and what's happening in between is anecdotal. It's not a trading book. So we consider that the non-re-evaluated equity is more representative of our real equity that the re-evaluated one.

So we have always calculated the ROE and ROTE the same way. Now for the TNAV, we give both. We give both because we want you to know both. Arguably, for a shareholder, being -- well, potentially selling and buying every day, the re-evaluated figure makes sense. For us, we manage the bank for the long run, the re-evaluated figure doesn't make sense for us as managers.

So this is why the return on equity, we consider, is -- well, we think it's -- it's always the way we do it. So it's not inconsistent, it's just that -- it's very consistent for the return on equity, return on tangible. And for the TNAV, we give both because we want you to know both.

Operator: We're now over to the line of Kiri Vijayarajah of Barclays. Please go ahead.

Your line is open.

Kiri Vijayarajah: Yes, good afternoon, Philippe, a couple of questions if I may. Firstly, on the corporate center, you had a big jump in the headcount there last year. And I wondered how that's developed this year? How's that buildout now largely done? And when you think about your strategic review, you're clearly looking at a lot of your costs. Will you be -- is the corporate center headcount there also part of the remit? Secondly, just a quick follow up on the First Hawaiian.

You said you'll take longer than 2017 to fully sell down. But in terms of just helping us with the modeling, do you expect to still deconsolidate the RWAs by the end of 2017? Or is that going to take even longer than that as well? Thank you.

Philippe Bordenave: We don't know. It will depend on the market. As I said, we'll take time.

We're not going to waste time either. We're going to make -- to sell down, as I would say -- we're going to find the right balance between as soon as possible; we don't see any -- we don't have any interest in keeping the residual stake for a long time. But at the same time, we don't want to sell at a bad moment, because we're not in a hurry. So we're going to balance those two things and try and find good opportunities, good windows in the market as soon as they will occur. That being said, because we're one shareholder and we're selling in the market in a very spread way, we're going to keep a majority and a controlling majority until we have less than 25% stake.

So 25%, well we have to sell down to below 25% before deconsolidating. So it will take some time. Your first question on the corporate center headcount, I have not understood, sorry. What do you mean by headcount on the corporate center?

Kiri Vijayarajah: Yes, just looking at your divisional disclosure last year, during the course of 2015, there was a big jump in the headcount there. And I wondered how it's progressed this year? Because clearly you don't disclose it intra-year.

And also, then linking that to the strategic review and cost plans that you're already currently working on?

Philippe Bordenave: Okay, I didn't know the answer. So thank god I have a good team here. The answer is that for the headcount, the control functions; compliance, audit, the control functions are not -- because they are integrated now, they are not spread between the operational divisions. And so they are allocated to the corporate center as a conventional way but their cost is distributed to the businesses. So in order to symbolize their independence, the people are not affected in our reports to any business line.

And so they are kept in a corporate center figure. But in reality, of course, their cost is allocated to the happy beneficiaries of their services. So this is the explanation of the big jump you are seeing. Because we have staffed those teams much more. And in the new plan, we're probably going to continue.

Because the trend is certainly not to reduce the staff of the control function.

Operator: Our final question for today is from the line of Pierre Chedeville, CM-CIC Market Solutions.

Pierre Chedeville: Two remaining questions. First question, I'd like to come back on partnerships and I would like to better understand how this will work? For instance, Matmut, are you going to create a joint venture? Are you going to pay a bill for managing the plant or building the factory process? And you mentioned that with Credit Mutuel, you will share costs. But from a legal point of view, how will it work? Who is going to pay, according to what types of industry expertise, etc.? It's not very clear how it will work from financial point of view and from a legal point of view.

So could you give us a little bit color on that? According to [indiscernible] I imagine that you are contemplating probably other partnerships in the future. So how do you see the way partnerships are going to work? And the last question is, do you have a setup size regarding your SME fund that you announced this quarter in coordination with the domestic market business? Thank you very much.

Philippe Bordenave: On the last question, to start with, because it's easier, it's shorter, it was €500 million, the target. And we have reached the target, it's fully subscribed. So it attracted very -- a range of top-tier institutional investors very quickly.

And it's now closed already. On the partnership, what we do is that we usually -- well, almost every time, we create a joint venture, so a common entity, with the partner. And we -- this entity is billing its services to -- and entering into contracts with us for the services that are provided to us. So it's a way -- it will work both with Credit Mutuel and, on the insurance side, we want to -- we think that on the insurance side we have some upside that is not yet fully tapped at BNP Paribas. Our penetration of our clients in terms of providing casualty insurance for their home, for their car and so on, is not very high.

We see that other competitors have higher success in that respect. And so we're looking for a way to revitalize this business. And so we found that it was an opportunity to -- we had an opportunity to open a request for proposal to several insurers and so we're in the process of selecting a partner. It's not yet official but you have mentioned a name. I'm not going to confirm nor to make any denial.

But yes, we're working. And once the things are going to be finalized, we're going to sign and to create a joint venture with the selected partner for this type of casualty insurance product. And then we'll have a distribution agreement between this joint venture and our network that is going to distribute the product of this joint venture. So this is the way it's going to work. Okay, that's all? It seems that there is no other question.

Philippe Bordenave: Okay, then thank you very much. And well, you will see us for the full-year results and this time with Lars. Bye, bye.

Operator: This now concludes today's BNP Paribas third quarter 2016 results presentation. Thank you all for participating and you may now disconnect your lines.