Good morning, everyone, welcome to BBVA third quarter 2020 results presentation. I'm Gloria Couceiro, Head of Investor Relations, here with me today is Onur Genç, Chief Executive Officer of the group, Jaime Sáenz de Tejada, BBVA Group CFO. As in previous quarters, Onur will begin with a presentation of group's results, then Jaime will review the business areas. We will move straight to the live Q&A session after that. Now, I'll turn it over to Onur to start with the presentation.
Thank you, Gloria. Good morning to everyone. Welcome, and thank you for joining our webcast. I really hope that you and your families and friends are all healthy and safe. I wish you the best in this complex environment. I go to immediately to the presentation, slide number three. Very positive in our view, financial results. On the left-hand side of the slide, you see that net attributable profit reaching EUR 1,141 million in the third quarter, an increase of 79.5% versus the second quarter, and the 4.1% growth in constant EUR versus the same quarter of last year. In our view, very good numbers, even when compared to normal times. This positive performance versus the previous quarter has been the result of three things. Resilient core revenue growth, I will talk to you about it in a second.
Cost control continues to be very robust, a better evolution on impairments. The second graph at the middle of the page, it shows the capital generation, very solid capital generation in the quarter. We have improved our position 30 basis points, leading to a fully loaded CET1 ratio of 11.52% as of September 2020. Already obviously surpassing our capital target. Last, on the right-hand side of the page, it is also important to note that we have maintained our tangible book value per share practically stable compared to the second quarter, despite strong market movements, very strong market movements, actually, in the third quarter. Let's jump into the page number four, slide number four. What are the top messages of the quarter? Five. First, our robust and resilient pre-provision profit growing at double digits.
We keep saying it in every single results presentation because that's what the numbers are telling us. Very robust and resilient pre-provision profit once again. Second, a very strong cost control and efficiency. Third, significant improvement in our risk indicators, which evolve within our upgraded expectations as we have shared with you a month ago. Fourth, once again, the strong capital generation. Fifth, our lead in digital gives us an edge in this environment. I will talk to you about this also in this presentation. Slide number five, looking at the summarized P&L. I would like to highlight the evolution of the net interest income, increasing 5.5% in constant EUR in the quarter, and the good evolution of the net trading income as well. Fees and commissions remain almost stable with a slight decrease of 0.08%.
All this, when combined, explains the increase in gross income of 5% in constant EUR. This evolution, together with the excellent performance of expenses, leading to an outstanding increase, in our view, in operating income, 13.5%. As I already explained, this quarter, we have seen a more normalized level of impairments versus the first half of the year. All combined, net attributable profit for the quarter is, as I said, EUR 1,141,000,000, 4.1% year-over-year increase in constant EUR. Slide number six, you see the evolution in the first nine months of the year versus the same period last year, so year to date. As you can see, we are registering a good evolution in net interest income, gross income, expenses, and operating income. All the core revenue drivers and also expenses and operating income behaving really well.
Our operating income, just pick one number, it has increased by 17.3% in constant EUR. In our view, a very strong reading, even in normal times. In terms of the nine-month results, when you look into the bottom line, the first quarter COVID-19 front-loaded provisions of EUR 1.4 billion. You would remember we did an extraordinary provisioning for COVID of EUR 1.4 billion in the first quarter, and another EUR 0.6 billion in the second quarter.
It leads to a negative impact, obviously, on the bottom line. Net attributable profit for the first nine months of the year is EUR 2,069,000,000, 36.6% decrease year-over-year. If we exclude, obviously, the non-cash, the BBVA USA goodwill impairment that we recorded in the first quarter. If we include that goodwill impairment, which was EUR 2.1 billion if you remember, the final reported net attributable profit is minus EUR 15,000,000.
Moving to slide number seven to shed more light into the revenues breakdown. Here you can see the quarterly evolution as well, not just the year-over-year, but quarterly evolution. First thing to note is, again, the 5.5% growth in net interest income year-over-year, and 4.9% growth in the quarter. This quarterly increase, it's supported by activity recovery in retail segments. The retail segments are coming back up slightly, and customer spread improvement in some countries. As you would see in some of the countries that we operate, we have improved, especially on the cost of funding in a dramatic way. Net fees and commissions has remained almost flat with a slight decrease of 0.8% year-over-year, negatively affected by the pandemic with the lower economic activity, affecting our credit card and payment system revenues, basically.
Also, there was a new regulation in fees capping certain commissions in Turkey that also had played a role. Nevertheless, the good thing to note on the fee curve is the fact that there is a very robust 12.9% quarter-over-quarter growth in fees. Again, mainly explained by the retail activity recovering in most of our geographies. Good performance on net trading income.
All in all, robust revenue growth, as I said, of 5% versus the third quarter of 2019, mainly driven by the good performance on NII, as I explained. Overall, this 5% growth, it was very important to post a growth in revenues in a year like this. We are very happy with the evolution of that figure. Slide number eight. There are two pages here that we are sharing with you to also maybe indicate the evolution in the coming quarters.
On slide number eight, regarding core revenue evolution and also the impact on some of our other metrics, let me show you a high-frequency indicator of card spending. These are BBVA figures. We monitor them very closely. They are the weekly credit and debit card consumption compared to the same week of last year in most of our markets, basically, you see it on the page. In March and April timeframe, as you would see, the slump was significant. For example, the weekly card spending in Spain, it was falling 60%. Mexico, it was falling 35%. Since then, what you see in the curves in the charts is that there has been a gradual yet consistent recovery everywhere.
We have to take these numbers in the context of uncertainty that we all live in, as we all know, the second wave of the virus is sweeping through some of our countries. We have to see the impact of the new measures taken by the authorities. These numbers, again, in all the markets, what you will see is we have caught up with the level of spending a year ago, as it stands now in all the geographies. As of now, from a pure economic point of view at least, the spending has come back to its normal levels, we have started to learn to live with the virus. Similarly, on page number nine, very quickly, in our own business, new loan production of the retail segment is also recovering.
Similar curve, again, bottoming out in April, May timeframe, and now coming back up in the last few months, and the latest readings is quite robust, as you would see in the page. Moving to slide number 10, expenses. Very good performance in expenses. If you compare the year-on-year evolution, expenses dropped by 3.8%. This is something that we have put a lot of attention this year, and we are getting the results, the fruits of that effort. We also maintain positive operating jaws, our core revenues, they are increasing by 3.4% in the nine-month period. While at the same time, the expenses are decreasing by 2.4%. The blended inflation rate in our footprint is 4.4%. Our costs behaving much better than the inflation. As a result, we maintain this nice jaws in our operating performance.
Finally, on the right-hand side of the page, as a result of all of this, you can see an outstanding efficiency ratio, 45.6% in the first nine months of the year, and it keeps improving. 438 basis points improvement versus 2019. Again, this is much better than our European peer group, as you can imagine. Moving to slide number 11, risk and provisions. Total loan loss provisions for the quarter was EUR 1,039 million, clearly lower than the previous quarters, as we already guided. We also have done a very strong upfront provisioning in the first and second quarters. In fact, the third quarter figure is very much aligned with the pre-COVID level of impairments. Thanks to a better-than-expected performance of the portfolio in all the countries, mainly deferrals in Mexico.
I will go more in-depth in the deferrals topic in the next page, but overall, the numbers are coming to be much better than our original expectations. As a result, our cost of risk improved significantly to 1.69 on a year-to-date basis, versus 2.04 in the previous quarter. Also on the page, you can see the NPL and the coverage ratios. NPL and coverage ratios, they remain almost stable versus last quarter at 3.8% and 85% respectively, which again shows that we don't see a deterioration in the underlying risk parameters at the moment. That said, we remain cautious and the credit risk anticipatory management, as we call it, continues to be one of our key priorities in this environment for any potential deterioration going forward.
All in all, good evolution in terms of asset quality indicators during the quarter. We maintain our recently improved 2020 cost of risk guidance to be in the range of 150 to 160 basis points at the end of the year. If you go to slide number 12, I would like to give you a bit more detail on the evolution of deferrals, which is underlying the numbers in the previous page. As you can see on the left-hand side of the page, total deferrals granted represent 8.9% of total loans. Out of them, 66% are related to retail and 34% to the wholesale portfolio. 70% of that total amount have already expired. We are seeing clearly the results of that expired amount. As I said, a very positive payment performance in our view, better than our original expectations.
As you see at the center of the slide, almost 80% of this deferred portfolio, which has expired, have already resumed payments. In that same chart, we gave a second deferral for 11% of the loans, of which more than half are related to high-quality mortgages, mostly in Spain. As you can see that at the top of the right-hand side. At the right bottom part of the page, a small graphic, but a very important one in my view. You see the resiliency of the delinquency buckets. With past dues on that in progress portfolio and so on, they should be flowing through the delinquency buckets. When you look into the delinquency buckets, the past dues, as a % of our portfolio, is completely aligned with a year ago. Completely aligned with a year ago. Again, good results overall.
We have been able to obtain these results. As we have put a lot of effort into this, we are applying an early and proactive collection management strategy, as I said, properly segmenting all the clients, tailoring the intensity of our collection efforts to every single segment. Again, good results so far on the risk side. Moving to slide number 13, great news in terms of capital generation. 30 basis points improvement in the quarter. You see in this page the breakdown. Following the waterfall, main impacts of the quarter, results generation, net of AT1 coupons contributes 29 basis points to the ratio. Second, negative market-related impact this quarter, detracting 11 basis points, mainly due to mark to market of our equity portfolio, together with some negative impact from some effects currency depreciation, 11 basis points.
The third, the last bucket, it explains +12 basis points, and that's mostly due to credit RWA. RWA impact is a positive one. Overall, 30 basis points in the quarter. Again, a very strong reading. This number, our fully loaded CET1 ratio, 11.52%, is 293 basis points above minimum requirements. We already surpassed our CET1 target range of 225-275 basis points. Our target range is already passed, we also obviously expect to continue to being above our target in the fourth quarter. There will be a positive impact from the software need treatment also of 19 basis points in the quarter. Things look good. On a phased-in basis, by the way, our CET1 amounts to 11.99%, this is the basis where distance to MDA is calculated, as you know. The distance to MDA as it stands now is 340 basis points.
At the bottom of the page, you should see the high-quality nature of our capital ratio. We remain among the leaders in terms of the leverage ratio, 6.4% on a fully loaded basis. AT1 and tier 2 buckets, they continue to be fully endowed on a fully loaded and on a phased-in basis. Last two pages on digital, page 14. As we have shown in the previous quarters, COVID-19 lockdowns have accelerated preexisting trends in digitalization and the adoption of new technologies. A good part of this increased digitalization is here to stay. Leveraging digital capabilities has proven essential. I would say differential in serving our clients. Let me put some figures to this.
Regarding the number of interactions with our customers, maybe only the middle part of the page, the interactions with our customers through our mobile app, it has gone up five times as compared to a year ago. Given our differential capabilities to go beyond servicing, which is very important, and execute sales through our remote channels, our digital sales have reached 64% in terms of units and 48% in terms of value.
On slide number 15, the last page of my section, we keep talking about digital all the time because we really believe in it, that there is something big is happening in our industry. Do we have a competitive advantage in digital? That is a critical question in our view. Let me give you some examples from Spain and Mexico that show that we do have differential capabilities in digital.
It is highlighted by an overproportional market share on areas that require strong digital capabilities. On the left-hand side, BBVA is the number one in banking app usage market share in Spain. 22.2% of the banking app users in Spain have used BBVA app. We are also number one in what we call do it yourself availability of functions. What you want to do and whether you can do that through the mobile app. It's 90%, this quality score, and it is very far from the 68% peers average. On the right-hand side of the page, in Mexico, we have a leading 39% market share in e-commerce acquiring, leveraging our digital and payments capabilities. These are just some examples, but they prove in our view that BBVA's advantage in digital capabilities and solutions is there.
We believe, this is one of the reasons, at least, that constitute the good performance of this quarter and the quarters coming along. I turn it over to Jaime for the business areas. Jaime?
Thank you very much, Onur, good morning. Let me begin with Spain. In Spain, BBVA Research, GDP growth expectations for 2020 remain unchanged at -11.5%, but 2021 forecast was adjusted to a 6% recovery, including now the initial positive impact from the European Recovery Fund. Loans have increased by almost 1% year-on-year, driven by the strong growth across commercial segments supported by state guarantees. Mainly as you remember in Q2, offsetting the leverage in the mortgage and public sector portfolios. In Q3, new lending flows in retail recover versus the previous quarter, especially in consumer loans. In the nine months to September, BBVA Spain delivered a strong pre-provision profit up over 16% versus last year, thanks to higher core revenues driven by the good performance of fees, thanks mainly to CIB asset management and banking services.
The better-than-expected performance year-to-date leads us to improve our guidance, and we now expect fees to increase slightly in 2020. It was also supported by higher net trading income, mainly due to higher ALCO portfolio sales and a remarkable improvement in operating expenses, down over 6% above expectations. We continue to expect costs to decrease by more than 5% by the end of the year.
The mid-teens operating income growth has been more than offset by higher impairments versus last year, mainly explained, as you know, by the significant front-loading of COVID-related provisions in the first half, but also by a base effect, as last year included provision releases from a mortgage portfolio sale. Looking at the quarter-on-quarter evolution, impairments continue to improve, aligned with our expectations to 80 basis points in the nine months to September, versus 100 basis points as of June.
By year-end, we expect cost of risk to improve even further to around 70 basis points. If our current macro forecast proves correct, 2021 cost of risk should be below 2020 levels. Let's now turn to the U.S. The faster than expected recovery has allowed us to revise the Sun Belt's GDP growth forecast upwards to minus 4% in 2020, and plus 3.4% in 2021.
Loans are growing at a very healthy rate of over 6% year-on-year, driven by commercial segments, which were supported by the drawdowns of credit lines in Q1, and the state guarantee loans under the Paycheck Protection Program in Q2. Lending growth slowed down in Q3, given the high amount of liquidity provided during the first half of the year. For 2020, we expect loans to grow in the mid-single digit range. Turning now to the P&L.
We had a very good performance of core revenues versus the previous quarter. On the one side, NII is up almost 5%, thanks to an excellent deposit cost management and demand deposit increasing. They represent now over 85% of total deposits in the U.S., while deposit costs improved by 18 basis points quarter-on-quarter. NII was also supported by lower wholesale funding costs and a higher contribution from the ALCO portfolios. We expect NII in 2020 to decrease by low single digits, consolidating the improving trend of the previous quarters. On the other side, we also enjoyed a sound fee growth, over 11% quarter-on-quarter, thanks to very strong fees in global markets and the pickup in activity after the Q2 lockdown. Regarding expenses, again, very good performance.
It continues with a year-on-year decrease of almost minus 2.5% in Q3, a behavior we expect to be able to maintain till the end of the year. Impairments increased by 34% versus the previous quarter, mainly explained by a macro adjustment and the rating migration of some commercial customers. Having said this, cost of risk reached 169 basis points year- to- date, that of September, in line with expectation, and we expect the improvement to continue reaching around to 135 basis points by year-end.
According to our current macro prospects, 2021 cost of risk should be below 2020 levels. Let's now move to Mexico. We have fine-tuned our GDP growth estimates for 2020, improving to - 9.3% from 10% previously, thanks to the recovery in the manufacturing sector and the increased mobility after lockdowns. For 2021, we continue to expect GDP to grow by around 3.5%.
Loans increased by almost 6% year-on-year, mainly driven by corporates clients during Q1, and the good evolution of the mortgage portfolio, supported by good dynamics in the housing market. For 2020, we expect loan growth in the mid-single digit range. Moving to the P&L. Q3 shows a significant recovery versus the previous quarter, with net attributable profit increasing by 88% in constant EUR, quarter-over-quarter. Operating income increases 10.5% in constant EUR, driven by core revenue growth and cost controls. Regarding core revenues, NII grows by 11% quarter-on-quarter, as most deferrals in credit cards and SMEs have already expired in Q3, starting to accrue interest again. Let me remind you that we lost EUR 104 million because of the lack of NII accrual in Q2, and we have recovered roughly EUR 75 million in Q3.
Moreover, wholesale and retail funding costs decreased in the quarter, thanks to a very successful price management strategy. We continue to expect NII to be flat, slightly decreased in 2020. Fees improved 15%, one five, quarter-on-quarter, favored by the recovery in economic activity. As we have already anticipated, we expect core revenues in the second half of 2020 to be above the first half figure in constant euros. Costs remain under control, - 2.3% better than Q3 last year. For the whole year, we continue to expect cost to grow well below inflation. Finally, on impairments, they decreased over 40% versus the previous quarter due to lower macro-related provisions and the good payment performance of expired moratorias, while Q2 included some provisions related to COVID-19.
Cost of risk improves to 427 basis points from 495 as of June, and we expect cost of risk to be in the low 400s by year-end. Considering our current macro estimates, we think 2021 cost of risk would be below 2020 levels. Let's now focus on Turkey. Starting on the macro. We have improved our GDP growth expectations for 2020 back again to a flat growth, and maintain our projection for 2021 at +5.5%. Garanti BBVA delivered a significant growth in TL loans, almost 35% year-on-year, mainly explained by the commercial segment. Foreign currency loans decline 1.5% year-on-year in line with our proactive strategy to reduce FX exposure. For 2020, we expect TL loans to grow around 25% and the trend in FX loans to continue. Moving to the P&L.
Pre-provision profit in the nine months to September grew by 50% year-on-year in constant EUR, supported by the strong revenue generation and the focus on efficiency. NII was up significantly, over 30% versus last year, mainly explained by the excellent commercial dynamics and a significant improvement in TL customer spreads. In the third quarter, TL spreads started to decrease versus Q2 as a consequence of the interest rate hikes that we're seeing in the country. NII continued growing almost 8% quarter-on-quarter, thanks to lower wholesale funding costs and a higher contribution from the CPI linkers portfolio. For 2020, we expect NII to grow above 20%. We had actually, a very good performance of NTI thanks to FX results, gains from security sales, and a higher contribution from the global markets unit. Expenses grew 7% in the first nine months.
That's significantly better than the 12-month inflation, which is above 11%, improving the efficiency ratio to an historic level of 27.6%. This very strong pre-provision profit has enabled us to front-load provisions in these uncertain times. Given the effort done in the first half, impairments have improved significantly versus previous quarters in Q3, thanks to higher recoveries and the allocation of some of the front-loaded COVID-19 related provisions set aside in the first half, especially to FX loans to commercial clients in this quarter. As a result, year-to-date cost of risk has improved sharply, is down to 2% in the nine months to September. That's better than the 2.7% as of June and aligned with our expectations. For 2020, we expect cost of risk to be around 215 basis points. Actually, we expect a negative recalibration effect in Q4.
For 2021, based on our current macro prospects, we expect cost of risk to also be below 2020 levels. All in all, as Onur mentioned, excellent results that continue to prove Garanti BBVA's earnings resilience, with net attributable profit increasing by over 30% in current EUR in the first nine months of the year, and over 58% year-on-year in constant terms. Finally, South America. BBVA Research has improved its macro prospects for the region, not only for 2020, but also for 2021. That's mainly in the case of Peru. We now expect GDP to contract by -13% in Peru, and to recover by 10% in 2021 versus the 8% that we had on our previous estimate. Now some color on the main countries. In general, Q3 net attributable profit recovered significantly versus the previous quarter in all countries.
In Colombia, net attributable profit in Q3 increased 42% versus Q2 in constant EUR, explained by core revenue growth, strong NTI, and a reduction in impairments favored by lower NPL entries and high recoveries, reducing year-to-date cost of risk to 298 basis points and aligned with expectations. Peru also. Q3 net attributable profit recovers to EUR 45 million, growing significantly versus Q2 thanks to a very sound growth in total revenues, positive jaws, and also better provisions, thanks to a positive macro adjustment this quarter. Finally, Argentina, that also increases contribution this quarter, mainly due to provision release in the securities portfolio. Now back to Onur for some final remarks.
We are just one minute away from our allocated session for the presentation. Very quickly, I'm not going to go through the details of the last page. In conclusion, strong operating income, once again, core revenue growth, very strong cost control. Good numbers on those figures in our view. Significant improvement in risk indicators in the quarter as guided, aligned with our upgraded guidance. Solid capital generation, 30 basis points in one quarter. We will take that quarter any day. Differential best-in-class digital capabilities and solutions. That's the summary of what we have presented to you. Regarding 2021, there are a few things maybe we can guide you or we can help you understand better as we look into our own dynamics. First of all, our core revenue for 2021, our expectation is we'll continue to grow in constant euros.
Improving lending mix, retail new loan production recovering, and price management will be the key levers to achieve that. Cost control will continue to be a key management priority. I hope in every meeting we keep repeating ourselves, but it is a very important topic in our management discipline. We have proven it in the past, and it takes pain.
We have decreased our cost base 16% in the past four years. That 16% is 9% for competition. We are very much focused on this, and we will continue to be. Cost control, you will see also in 2021 to be a key management priority. Some of the learnings from COVID, some of the trends from COVID would obviously help. Finally, our cost of risk, we expect it to be below the 2020 levels, obviously. Good dynamics, but again, it's a very uncertain environment. We have to see how the health situation evolves, how the impact on the economy would be in the coming months. As it stands, we see some positive signals in terms of the underlying drivers of our business. With this, I conclude the presentation. Back to you, Gloria.
Thank you, Onur. We are now ready to move into the live Q&A session. First question please.
Ladies and Gentlemen if you wish to ask a question please press star one on your telephone keypad. Please be inform that in order to assure your audio quality we recommend that all questions are ask from [audio distortion]. The first question today comes from Ben Toms of RBC. Ben, please go ahead.
Hi, thank you for taking my questions. Two for me, please. It feels like on a daily basis we see headlines that Turkish lira has hit new lows. My question is quite big picture, really, but when management see the news about Turkish Central Bank decisions and currency weakness, how do you think about it? Do you bang your fist in frustration? Do you call up the treasury team and tell them to increase your hedging? Do you call the head of central bank and have a discussion? Do you sit back quite relaxed because you know you're well hedged and you have a good banking franchise in the country? I guess what I'm really trying to say is, does Turkey keep you awake at night? From an outsider, it's perpetually frustrating.
Secondly, on costs, in which your fees do you expect to be hardest to improve cost efficiencies in 2021? Thank you.
Thank you, Ben. On the first one, on the Turkish lira, obviously it's an important topic, we really actively try to manage it. Does that keep you awake? It's a concern, but it doesn't keep us awake because we have been managing it. If you go back to every single quarterly presentation that we have done in the past, I don't know, two years, Ben, the question on Turkey came, and the concerns on Turkey came from all of you. The thing that we said was the key concern area might be the FX because of the balance of payments of the country. We have to be watching the FX. Given that statement that we have been giving to you, we have been preparing for it.
You might have seen it in our numbers, given that vulnerability, we have reduced our FX lending book from $21 billion five years ago, 2015, to $12.6 billion in the last quarter in September 2020. We basically cut it by half. We have been reducing the dependency on the FX wholesale funding. If you look into FX wholesale funding in the same period, 30% down. If you look into our bond portfolio in Turkey, you would see that we do have a very sizable share of inflation linked, because FX over time obviously flows into inflation. You would see that our bond portfolio is mostly consisting of inflation-linked bonds, CPI linkers, as we call them. To cut a long story short, is that a concern? Of course.
Knowing that this is the vulnerability, we have been taking actions on all dimensions to make sure that we manage it through. Also the hedging you asked, obviously we do hedge our currencies, our volatile currencies. That is also one of the other levers that we use. The fact of the matter is that independent of the environment, we do have truly the best bank in the country, in our view. Again, proven by numbers, it is not a subjective opinion. If you look into the ROE or Garanti BBVA versus the ROE of the rest of the industry, pre-provision profit divided by assets, us versus the rest of the industry, you would see very good numbers.
If you have a good asset, if you have the leading bank in a country, and if you forecast the negative trends and prepare for them, you ride it through. That's where we are. Does it keep us awake? No, it doesn't keep us awake. Cost efficiency, I think the question is how are we planning to go further on this in the coming year? As we have done in the past, we will try to optimize.
We will try to use the trends of digitization and servicing our customers in a digital way. You have seen it in the presentation, the mobile app, transactions in the mobile app in one year, it has gone up five times. These transactions in the old days, in the very old days, would have been serviced in a much high-cost channel. We are reflecting that into our cost structure.
You might have seen it in the numbers, in the total staff size of BBVA, it has come down this year. All of that would be reflected in the coming year as well. We will continue to employ levers as we have done in the past. It's going to be one of our core management levers to create value. Jaime, do you want to add anything on this?
Well, maybe the hedging levels that we currently have in Turkey. We currently have hedged 50% of the excess capital that we have in Turkish lira, which leaves us a sensitivity to a 10% depreciation of the capital of around 3.5 basis points. Another data point maybe of interest, also our wholesale debt in US dollars in Turkey, has gone down over the last four years from $12.5 billion- $8.5 billion. We currently have an $11.5 billion excess cash in foreign currency. A very strong buffer to weather any circumstance.
Thank you. Thank you, Ben, for your questions. Next question please.
The next question comes from Andrea Filtri of Mediobanca. Andrea, please go ahead.
Hello, good morning. One question on the good detail you provided on moratorias. Can we consider the 20% portion of moratorias expiring, not resuming payment as a proxy for the potential parts where NPLs should come from in the future under a severe macro scenario evolution. Secondly, a willingly broad question to understand your thinking. It looks like to us, Europe is trying to pull together with the COVID-19 crisis. What are your expectations on dividend ban on an eventual bad bank project and on M&A, both domestic and cross-border in Europe? Thank you.
On the first one, in a stress scenario, should we expect more of the NPLs and the risk costs to be coming from the 20%? If that's the question, if I understand well. The answer is that most likely, that the intensity of NPLs in that portfolio would be higher, for sure. Meaning the percentage of that portfolio leading to NPL is higher than the 80%, probably. I wouldn't categorize that 20% group as the really bad sub-segment, because there are many sub-segments even within that 20%. Let me give you one example. As we have said here, the second deferrals, the 11%, some of that is basically incentivized by the governments in certain cases, by even us, because we want the customer to align the payment behavior with the requirements of the loan.
Most of that, for example, second deferrals, 51%, as you see it in the chart on page number 12, is mortgages and mostly in Spain. We don't expect a major problem on that portfolio at all. Some of that second deferrals is our wholesale clients, and we don't expect any problem in some of those sub-segments at all. Probably in that 20%, there might be a higher likelihood, but there are many sub-segments within that 20%, and I would say that we don't foresee a major problem. The other 8% within that 20% that you are mentioning, Andrea, is, as we said, in progress. In progress, meaning they haven't resumed full payments in most of the cases, but they expired, especially in July and especially in August timeframe. 60 days have already passed. We don't see those customers in the delinquency buckets.
That's why that little chart at the bottom of the page on the right-hand side is very important. We don't see that in progress bucket flowing into delinquency. They're paying. They're paying maybe one payment, or they are in the process of finalizing a restructuring solution. My point is that 20% don't take this as, oh my god, this is a very tainted area. There are very bright and good sub-segments within.
I would say that the likelihood of that 20% is probably higher than the rest, than the 80%. On the broader questions that you asked on Europe, the dividend ban, obviously, it's the decision of the supervisors. We are getting some signals that it will probably be lifted for the payments in 2021, but we don't know. It's their call, and they will do the meetings, I guess, at the end of the year.
We will see. My expectation is that it will be lifted. Let's see. It depends also on the evolution of the COVID crisis, on the health situation, and so on. We'll see. My expectation is that it will be lifted. My expectation is that if it's lifted, we are one of the ones. We have 340 basis points, distance to MDA. 340 basis points distance to MDA. When it's lifted, we are going to restart our usual shareholder remuneration program. On the bad bank, we'll see. Obviously, we are hearing the plans and the thoughts. At the moment, I don't think Europe needs one. At the moment, the numbers are not justifying that there will be a big need for a bad bank. This crisis is a bit different than the many other crises that we have seen. You should know this.
You are seeing it in the figures as well. This is a crisis where actually the balance sheet of most of the customers is actually strengthening. The deposits in the system is actually going up. It's a very heterogeneous crisis, so meaning some sectors are affected much more. Overall, for the economy, you don't see a depletion of savings and the deposits and so on. You see an increase in retail deposits, in commercial deposits, in CIB deposits. The governments are also helping the overall economy. I don't think a bad bank would be needed. Again, let's see. Let's see how it goes along. On M&A, I cannot comment on others. The only thing I would say is the cross-border, the M&A is done to create value, to create shareholder value.
I don't see a huge value creation in a cross-border one because the cost structures would not be overlapping too much except for head offices, I guess, because the distribution costs, which is still the key cost item for the European banking industry, cannot be optimized in a cross-border merger. I see less likelihood in cross-border mergers in the short term. For domestic M&As, again, it depends on shareholder value creation.
Thank you. Thank you, Andrea. Next question, please.
The next question comes from Adrian Cighi of Credit Suisse. Adrian, please go ahead.
Hi there. This is Adrian Cighi from Credit Suisse. Just a few follow-up questions, please. One on the cost and one on the capital return. You mentioned the cost control as a management priority. Is there any chance to be a little bit more specific as to where you see this landing next year? On the capital return, you mentioned just earlier that you are thinking or considering the restart of capital return. A number of your peers have been more forthcoming with their ambition for next year. How do you think about 2021 returns and maybe discuss your thoughts on share buybacks versus dividends? Thank you.
Thank you. Thank you, Adrian, for your questions. Let me address the first one on cost. I'm afraid we're not going to be more specific in the call today. I think that what has happened over 2020 and the years before, I think, more than speaks for itself. We are presenting an extremely good behavior in expenses in 2020. The acceleration of our transformation cost due to COVID has only improved the performance of our cost-to-income ratio. Both in Spain and the U.S., costs are down on a year-on-year basis. Cost-to-income overall, group level improved to 45.6%. That's a really significant improvement overall. As I said during my presentation, in a country like Turkey, cost-to-income is 27.6%, reaching an all-time low.
I think as Onur mentioned before, the focus on cost will remain, and it will continue to be one of our most important priorities.
Adrian, you also see it in the numbers in every single geography, despite very high inflation. In every single geography, we have delivered cost growth much lower than inflation this year. We have also, in the key geographies like Spain and USA, we are clearly in the negative territory. In the case of Spain, we are posting 6.5% decline in costs, and so on. I hope you do acknowledge the huge effort that we have been putting into this. When we put a huge effort and discipline, it typically works out. That's our intention for 2021. On the capital return, when the restrictions are lifted, our intention, given our share price, the level of our share price, obviously it might change and so on.
It seems like it would be high value added for our shareholders if we use a hybrid model of dividends and share buybacks. Our intention is to do both. We are getting messaging from also the regulator that the shareholder remuneration mechanisms, dividends or share buybacks, they are indifferent to the method. In that context, we do think, given our share price levels, a hybrid methodology would probably be the right way to go. We will see it when the time comes, obviously our decision-making authorities, our boards, and our general assemblies, they have a say on this, obviously. Okay, next one.
Thank you. Thank you, Adrian. Next question, please.
The next question comes from Mario Ropero of Fidentiis. Mario, please go ahead.
Hi, good morning. My first question is if you could explain the weak NII of Spain in the quarter and how much TLTRO NII did you accrue? The second question is a broader one. Onur was mentioning before that the digitalization of the bank may be conveying some competitive advantages. My question is, if in Spain you are perceiving that this competitive advantage is also giving you a competitive advantage in terms of loan pricing. Yeah, that would be it. Thank you.
In terms of loan pricing. Okay. On the weak NII in Spain, Mario, thanks for all the questions for both of them. On the NII, it's lower contribution from the commercial banking activity, the first reason. Meaning after this ICO loans, as you know, in Spain, there was this ICO government support program, and it was very active, especially in the second quarter. We have seen some decline in the loan balances. The volumes have come down 0.5% quarter-over-quarter. Customer spread has also come down by three basis points, as you have seen in the presentation. Mainly, again, because of the fact that the mix of our portfolio has changed towards more commercial wholesale than retail, and also because of Euribor. As you know, Euribor also affects our NII. Overall, all these FX have impacted our, what we call commercial banking related topics.
Second topic is securities portfolios. It has come down obviously lower given the rates. In the second quarter, you were focusing on the quarter-over-quarter evolution. There was this interest on arrears for a tax payment that we have received. There was this one-off. Going forward, the thing that I would say is what matters is the going forward and the fourth quarter. For NII in Spain, we maintain our guidance that NII still to increase slightly in the year. Why? As you have seen in terms of what is changing from the third quarter, first of all, the mix is partially started to change, as you have seen again in the presentation. Also consumer coming back up, credit cards slightly coming back up. It will be helped by that one.
The volume might be also better, and also the spread and the margin would be slightly better. We maintain our guidance for the year. On digital, is digital helping you in Spain? Is that the question that I got? There's a problem in the line a little bit. If that's the question, the answer is obvious, yes. You have seen it in the numbers again, as we are discussing about NII and Spain and so on. If you look into the highest margin product in the Spanish banking industry is consumer. In consumer, in the past three years, we have gained 2.8% market share. In this year, nine months of the year, we have gained 350,000 gross customers. 350,000 gross customers in Spain, in a mature market, in a stable market, in a COVID environment market like Spain, gaining that 350,000.
When you go into the underlying drivers of why it's happening, we are all acquiring them through digital. Acquiring the customer is the first step to monetize that relationship in the future. We do hope that in the future, we cross-sell to that customer base and we sell new products. Is digital helping us in Spain, in a mature market like Spain? The numbers are telling us that they are, and they are to a large extent. You were also asking about whether it's helping you on the margin of the loans in the retail and the SME space, which is most of the digital loans are being dispersed. Obviously. If you look into what you call one-click loan, it is a higher margin product than the same product that we have in the branch.
The channel also helps us optimize that spread in the loan book.
Thank you. Thank you, Mario. Next question, please.
The next question comes from Ignacio Ulargui of Exane BNP. Ignacio, please go ahead.
Hi. Thanks very much for the presentation and for taking the questions. I have just two questions. One is on NII performance in Mexico. We have seen a very strong growth quarter-on-quarter in local currency and in EUR terms. Just wanted to understand, whether we should expect additional growth quarter-on-quarter, in the coming quarters in the 4Q because of the end of the impact in NII from loan deferrals, whether the maturities of those that happened in September will be fully captured in the 4Q or that has been already seen in this quarter.
The second point is on the cost front, you have been sort of flagging for a long time the strengths of all this digital, but I was just wondering how you see that in Mexico and what could you pull to be more efficient in Mexico or in South America? We talk about the impact of digital in Spain, but I was wondering more on the EM economies, whether you would be able to capture it and if the digitalization is also an advantage there. Thanks.
Thank you, Ignacio. On the first one, the Mexican NII, yes, we do have some positive trends for the fourth quarter because in the third quarter there was still, as you said, the remnants of the deferrals that we have given and the net interest income that we have foregone for some of those products, like credit cards and SME loans, the revolving loans, they were registered in the NII account. Yes, one piece is the recovery from the deferrals. The second piece is in the presentation you are seeing it. We are focusing a lot lately, given the lower interest rates in most of the economies, but also in Mexico. We are focusing a lot on cost of funding and deposits. In the quarter, you see that in Mexico, the cost of funding has come down from 188 to 140. The trend is a clear one.
We would also get some benefit from that. Obviously, again, we have to also see what the volumes are going to be turning out to be. At the moment, it's very positive, but if the health situation deteriorates, it's only two more months, but our expectation is a positive one, but we have to also put a disclaimer. On the cost, and the digital, how it's helping in Mexico and so on. Obviously, the way that the digital helps different countries changes. In the case of Mexico, the situation is that we have increased our number of customers tremendously. Tremendously. Our target customers is now around closely 11 million, actually. It used to be three, four million less some years ago. We are serving 50% more customers with more or less the same branch network.
Actually, if you look into the sizable branches and so on, it's a lower, smaller branch network. How can you do the same business with 50% more customers with the same distribution network? As compared to before, the key change is digital. We are serving more of our customer base through digital means, which is a lower cost channel, and that helps us, again, to load more customers into the same cost base. That's the way that is changing. The dynamic of the country is very different from Spain. As a result, we use the digital in a way, again, on a customer basis, we still have a lower cost, but you don't see it at the broader costs because we are growing. We are growing tremendously, very crazy in Mexico. That's the differentiation. Do you want to add anything, Jaime?
I would say that, if you see the numbers by country, penetration rates of the digital in many Latin American countries are even higher than in the rest of the footprint. In Colombia, 73% penetration rate as of September. In Argentina, 71%. As it's been the case over the last few years, we've been able to increase in a very significant fashion the number of clients in the whole region without necessarily increasing the physical infrastructure that we have in place. Servicing through digital has become extremely efficient. Also it's quite impressive how we've been able to improve our digital sales in the region, both in terms of products and in terms of PRV. Much higher in both instances than in the North. Even more relevant, I would say, in Latin America.
Thank you, Ignacio. Next question, please.
The next question comes from Stefan Nedialkov of Citigroup. Stefan, please go ahead.
Yeah. Hi, guys. Can you hear me?
Yes, Stefan. Yes, very well. Please go ahead.
Hi. Okay. Thank you, Onur. Two questions on my side. Back to the moratorias. When I look at the Stage 2 ratio of the moratorias, it's around 24%. Kind of close to the ratio of loans that are being restructured or not paying right now. That 24%, can you give us a sensitivity of the cost of risk for 1% going into Stage 3? If that's too detailed for you, can you just give us some color in terms of how the 24% Stage 2 ratio informs your cost of risk thinking for next year? How much of that 24% do you think ends up in Stage 3, and how much of that is already incorporated in your guidance for next year? The second question is on dividend. How much are you guys accruing at right now? I think as of 2Q, you were doing 49%.
Is that still the case? Thank you.
Stefan, on the first question, we don't have a guidance for next year. The only guidance is it's going to be better, which is a very high-level guidance at the moment because we have to see, and we have to come to the end of the year. The crux of your question, I think, goes down to whether, what we call the transformation ratios from one stage, from one delinquency bucket to the next bucket, is changing in the current environment. Again, that's the crux of the question, if I understand well. On that one, what I can tell you is the signals that we are seeing is that the transformation ratios are not changing. Meaning within that Stage 2, what % goes to Stage 3, in our view, by different products and by different portfolios, is not changing. That's a very good signal.
We are registering the buckets and the transformation ratios remaining stable. On the dividend, we are not accruing any dividend in our capital this year, you don't see any number. As you also know, though, we have been one of the very few banks who paid dividends in 2020. If you remember, given the fact that we have done our annual general assembly on March 13, we legally had to pay. After that date, the ECB came out with the recommendation, but we had to pay because it was legally binding on us to pay. We are one of the very few banks who have paid dividends in 2020. Given the suggestion and the guidance from the ECB, this year, we are not accruing for dividends.
Depending on what comes out next year, how the situation changes for the supervisor, then we can reconsider, we can look into it in 2020. Our expectation is we will start payments in 2021, but we are not accruing anything for this year's results in our current results. On the first question, Jaime, do you want to add anything?
Well, I think in page 44 of the annex, we are providing what I think is a very interesting data point to try to weight the size of this Stage 2 level. The Stage 2 currently represent of this defer portfolio, represents 1.3% of the total loans. As deferrals mature, this concept of deferrals in practice will almost be disappearing by the end of this year. And the real focus that we need to have is, what's the size of this Stage 2 versus the overall portfolio? I would definitely take a look at page 44 to really weight the size of the problem going forward.
On that same page, on 44, Stefan, I would highlight the bottom of the page as well. If you look into our coverage ratios, it is very clear on where we stand versus the rest on that same page of 44.
Thank you. Thank you, Stefan. Next question, please.
The next question comes from Carlos Cobo of Société Générale. Carlos, please go ahead.
Hello. Good morning, and thank you very much for the presentation. Just to clarify the dividend comments that you make. As you are not accruing, that means that you are not likely to pay dividends against 2020 results. Is that what you just said? Just to understand you properly. The first question would be NII outlook in Spain and Mexico. Next year we have the TLTRO arrival and TLTRO III step down in Q3 of 2021. Could you please quantify both things, assuming a 25% drop in your TLTRO arrival? What's the combined impact in NII next year? In Mexico, same thing, guidance for NII. You seems to be presenting a slightly more conservative, now flat to slightly down NII in 2020 instead of flat, which is what I had in the previous presentation.
Could you please explain the drivers, and what is the gross contribution from the securities portfolio in Mexico? That seems to be coming down substantially on the back of lower inflation and lower rates. How much is the gross contribution, and what do you expect for the next year? Lastly, a question about the cost of risk, and this is probably a question that you've discussed plenty of times with investors, but it's a bit counterintuitive. You are not the only bank doing it. When you presented the guidance in Q1, everything was based on a more benign economic forecast and also a shorter lockdown. Now we're having longer pandemic, longer mobility restrictions, and weaker GDP prospects, and you and other banks are guiding to lower cost of risk.
It's understandable that the public support has been very important. I don't think anybody was planning to have mobility restriction for more than six months or so. How do you really see that not having an impact in cost of risk next year? Thank you very much.
Very good questions, Carlos. I'll take one and three. Why don't you take, Jaime, the second one? On the dividend topic, just to make it very clear. Our intention is to start the payments in 2021, obviously, because the restriction is that we cannot pay for in 2020. That payment and the basis for that payment, what it would be, depends on the new guidance that we would be receiving in January, also the evolution of the numbers and the situation. The basis is not defined yet, is not finalized yet. What I was trying to say is, our expectation is to start the payments in 2021. Given the guidance and given that expectation, we are not accruing, as also suggested by the supervisor, we are not accruing in our capital figures.
The second one, do you want to take it, Jaime, and then third, then I'll come back.
Yes, sure. Carlos, I'll give you the sensitivities that we have to rate movements in the different geographies. The only one that has really changed slightly in Q3 is the euro balance sheet sensitivity. It's now - 8% in - 100 basis points decreasing rates. That's the 12 months impact of a parallel move. The reason why it has increased by 2% mainly has to do with the higher volume of sensitive exposures, meaning mainly central bank balances and the high-quality liquid asset portfolio that we're building in the ALCO, due to the high liquidity generated by the balance sheet in these previous quarters. Also by the maturity of part of the hedgings that we have in the mortgage portfolio. In the remainder of the footprint, things remain quite stable versus previous quarters.
100 basis point decrease hits us by -4% in NII in the U.S., -1.5%, more or less in Mexico, 1.7% down in Turkey. In the different geographies, things remain quite stable. On the contribution of the ALCO portfolio to the NII in Mexico, well, actually, it has improved in Q3 versus Q2. As you can see, we've increased the size. Actually, we've been increasing the size of the ALCO portfolio quarter-over-quarter since the end of last year in Mexico, in contained levels. Betting on the rate decreases that they actually took place. This quarter, the size of the ALCO portfolio increased by 11% in Mexico, and the contribution to NII was EUR 30 million. That is EUR 28 million more than what it contributed in Q2.
On the third question, which is a very good question again, Carlos, I would say a few things on this one. The first one is, because you are saying that as compared to what you were expecting back in March, April, it's actually coming out to be worse, or it is turning out to be longer and so on. Well, that's not what we see at the moment in the numbers. First of all, on the forecasts. If you look into the forecasts of GDP growth, even for this year, for a clear number of geographies, we have upgraded positively our GDP forecast. Not only us, the whole market. Our expectation for Peru at the end of first quarter, was - 15% GDP contraction, and it's upgraded to - 13. That - 2, is it important? It is an upgrade.
In the case of Mexico, it was -10% at the time. Now it's -9.3% for 2020. These are forecasts. What I care about is also some real data. That's why in this presentation, we have put that page number eight. That's a number that I track very closely. Every week this comes up, and I see what is happening in the economy in terms of spending. At the time, in the first quarter, we were hoping that a V-shaped recovery would happen in the economies. When you look into page number eight, what you see, again, in Spain, in the April timeframe, the spending was -60% down. 60% down. Now it is back to where it should be.
If you asked me back in March, April, would we be back in the month of June at the end of June and then stay there despite the fact that the health numbers are not doing so well? I would have probably said probably not, but we are seeing it in the numbers. In terms of spending, we are seeing some good signals in the real data rather than the forecasts, if you don't believe in the forecasts. That's one clear thing to put on the table. The second thing that we will put on the table is the fact that this is a different nature crisis. The governments are kicking in a major way to help the broader economy, and we fully support that, by the way, because this is a crisis that is affecting even the good companies, even good establishments, in a negative way.
In other crisis, governments do not typically kick in because they don't want to save the unproductive companies. This is a crisis where everyone is affected and the fabric is being pressured. The governments are kicking in to help the good companies. That is being reflected again in the balance sheets. The deposits of the system in every single segment, in every single country, is going up. Retail households, they have more money in their pockets as compared to before. The crisis is such that it's a heterogeneous one. We are not saying that it's a big crisis, but affecting very differently. The broader economy seems to be okay, but certain sectors, certain households, are affected much more.
Given these dynamics that we have seen in the past few months, and so on, in the figures, that's why we are saying that cost of risk is turning out to be better than what we expected. Again, on cost of risk, these portfolios are out of deferrals since August, and 80% paying their installments or payment dues. Then a good number is in process and asking. These are very good numbers, Carlos. Very good numbers. It goes back to the nature of the crisis, and I think the level of support that the governments are giving to the economies.
Thank you. Thank you, Carlos. Next question, please.
The next question comes from Sofie Peterzens of JP Morgan. Sofie, please go ahead.
Yeah. Hi, here is Sofie from JP Morgan. My first question would be around capital. If you could just give an update on where we stand with Paraguay, the insurance TRIM, when we should expect these headwinds and tailwinds to materialize, and how we should think about capital as well. I recognize that you're not going to say or give details around how much dividend you potentially would pay, but given that you haven't accrued for any dividends, how should we think about the capital headwinds potentially from there as well, assuming the ECB allows you to pay dividends? My second question would be on the cost saves. You have had very strong cost management this year. I recognize that you also mentioned that you will continue to focus on costs next year. Costs are down almost 4% year-over-year.
How much of these costs will basically reverse once COVID is behind us? Are there some costs that basically come back, such as travel, entertainment and so on? Thank you.
Do you want to take the capital one, Jaime?
Yeah, sure. Okay. On capital, we’ve been generating very large levels of capital quarter-over-quarter. I think that will continue going forward. Remember that what we’ve said at the end of Q2 on the impact of things that we had pending were, first of all, on Paraguay, we expect a positive capital impact of around six basis points. It will be materializing between the end of the year or the beginning of the next. On the safe side, let’s put Q1 next year. The joint venture with Allianz for the non-life products, we expect an additional positive impact in capital of roughly seven basis points, also expected for the beginning of next year.
What we also need is to update, as we did in the Barclays conference, we updated the positive impact from the intangibles that we currently now expect a positive impact of 19 basis points that will materialize in Q4. Then we were expecting 15 basis points of regulatory and supervisory negative impacts in 2020, coming both from TRIM, mainly on the low default portfolios and from adjustments to PDs. We don't know when that will materialize. That 15 basis points negative impact, maybe at the end of this year, maybe in 2021. That will be the pro forma ratio that I think you should calculate.
On the cost savings, on the second question, Sofie. Some of the cost savings of this year is there to stay. Some is obviously transitionary. One component is variable compensation. In variable compensation, one-third of the savings that we have done was due to variable compensation. It's completely driven also by the results. If it goes back in terms of costs for next year, it means that it's a great year. We will take that year in any case. Some of that is going to be linked. If the results don't turn out to be as usual, variable compensation will continue to be at a bit of a lower level. Some other pieces of that cost reduction is on, again, the staff size. As you might see in the detailed numbers that we are publishing, we have reduced our number of staff by 2,800.
Most of that is there to stay because we are using the trends in such a way that we optimize our cost base, especially on distribution and also in head offices in central services. It's a mix. Today, we are not providing any guidance for costs. We will do that in the next quarterly call. What we are saying is it will continue to be one of our bright spots in the coming year.
Thank you. Thank you, Sofie. Next question, please.
The next question comes from Daragh Quinn of KBW. Daragh, please go ahead.
Hello. Hi. Thanks for taking my questions. One question just on the capital generation and the kind of disconnect between the CET1 generation and the slight decline in NAV per share. Even if we exclude the risk-weighted asset component, organic capital generation and markets were still around 18 basis points, but that doesn't seem to have flowed into NAV per share.
Just if you could comment on that and on the outlook for that metric for 2021. Just to revisit the economic outlook, you sounded relatively upbeat versus your previous forecast and highlighted a number of countries where you've upgraded estimates. Listening to the ECB yesterday, they clearly think the outlook has deteriorated, and the outlook for Q4 is potentially for weaker economic activity versus Q3. Just trying to get a sense of, do you disagree with that outlook? Do you have a more positive outlook? Just how your outlook compares to the weakening data and outlook we're seeing from the central bank. Thanks.
Okay, I'll take the first question. On the tangible book, on the evolution of the tangible book and tangible book value per share. The main reason why the very strong P&L performance of the quarter doesn't flow to the tangible book mainly has to do with the FX differences that were generated during Q3. Mainly the depreciation of the Turkish lira. It was 16% in the quarter alone. Also the depreciation of the U.S. dollar, that also hit us by over EUR 400 million in the quarter. The Mexican peso behaved quite decently. Those, Turkey and the U.S., were definitely the most important countries. Then the second aspect is the evolution of the held to collect and sell portfolios. Particularly on the equity front, particularly Telefónica. Telefónica itself alone drained around 350 million euros in the quarter alone.
The forecast of the economic forecasts, Daragh, the fact of the matter is, I don't think anyone knows. It's a forecast, and it's a very uncertain environment. The only thing that we can comment on is what we currently see in the numbers, and that's why we put it in the documentation on what we are seeing in terms of spending in different countries.
We have upgraded. It's not us only, by the way, it's many other agencies and investment banks and so on. Most of us have upgraded some of the growth forecasts. In our case, we have upgraded the U.S., Mexico, Peru, as compared to our second quarter forecasts. There is some positivity as it stands. I do think that there is a big amount of uncertainty, even our forecasts, even someone else's forecasts. There is a lot of uncertainty.
We don't know how the situation would definitely evolve. In Europe, there are these new constraints. I don't see the impact of those things yet. We have to see. We have to see the impact. Again, the only thing I will tell you is, again, if you go to page number eight, if you go to spendings, the health situation was much worse in the last month in Spain, in terms of new cases, as compared to even April timeframe. Despite the fact that the health situation was much worse, despite the fact that there was some sort of confinement, even in these days in the country, the spending index to 100, it's 100 versus 40 in April.
These are real numbers that are kind of leading us into this positive mood a bit. We are not seeing at all the reflection of the overall situation on cost of risk, as we mentioned. There is a lot of uncertainty. We have to see how the situation evolves.
Thank you. Thank you, Daragh. Next question, please.
The next question comes from Jernej Omahen of Goldman Sachs. Jernej, please go ahead.
Good morning from my side as well. Can you hear me well?
Yes, Jernej . Yes, very well.
Oh, wonderful. Okay. I have three questions, please. I'm sorry to continue with the dividend question. I just have one clarification here. You said that you expect the ban to be lifted, but that you're not yet sure what the basis of payment would be. I'm just wondering, there isn't that many options for what the basis of payment could be. It's either the 2020 profits or you start distributing 2021 via an interim dividend. Is that what you were trying to say? That you're not sure that 2020 will be distributed, that you might just switch on from profits as they start accruing in 2021? That's question number one. Question number two is on this slide eight, which you referred to multiple times, and I also think is an excellent slide.
Now, the last data point here is the 18th of October, and you mentioned you get this data on a weekly basis. And given that things really started deteriorating sharply on the public health side of the equation post the 18th of October, I was wondering if these charts still look the same way. If the latest data points that you have continued to be resilient. My last question is, you asked an interesting question before, on M&A domestic versus cross-border. The response was logical, which is M&A is there to create value. What I want to ask you is this question in reverse. I think M&A is there to create value, but the option of disposing of a geography, of rationalizing, if that creates value, that should probably be considered as well.
I was wondering, when you look at your geographic portfolio and you look at the opportunity set as you currently see it, whether shrinking rather than expanding geographic portfolio is something that you sometimes consider. Thank you very much.
Very good questions, all three of them. Very quickly on the first one, the dividend. I guess I'm not that clear because this is the fourth time I'm getting the question. The basis of whether it's going to be 2020 results or whether it's going to be 2021 results on which we would be paying in 2021 is not clear. Yes, that's what I'm saying. We don't know yet. Our expectation and our intention is to start the payments. Depending on what the guidance would be, depending on how the situation evolves. It's going very positive at the moment, we don't know. We will see the accumulation of profits and capital in the March timeframe and so on, we would decide. It can be a different basis depending on what the situation is.
The expectation, you asked me about the expectation, and on that one I wanted to be very clear that the payments, our expectation is that the payments would start in 2021. On which basis, we don't know. The second one, the spending in Spain. I have the two figures, actually. First of all, the 25th of October figure, it's 96% as compared to the index of 100. If you call me today later, I will give you the last week's number. The fundamentals are still the same. It does that fluctuation from time to time, but I would assume that today also would be a good figure. On the third one, M&A is shrinking an option also in terms of value creation. Absolutely. We have done it. I mean, we have sold Chile a few years ago. Chile was an important franchise for BBVA.
We sold our pension businesses in every single Latin American country some years ago. We are selling Paraguay. We sold Puerto Rico. The key thing is what is the basis on which you take these decisions? The basis is, again, shareholder value creation. It sounds like too fluffy and too high level, but that's how we operate really as management. We look into shareholder value creation. If we don't deliver above our cost of equity in any country, if we don't have any conviction to deliver above our cost of equity in any country, and if we think there is a better natural owner for that asset who is willing to share the additional value creation with us, we are very much open to portfolio optimization on both sides. It can be a plus or negative, but it has to be value creation for the shareholder.
Thank you. Next question, please.
The next question comes from Benjie Creelan-Sandford of Jefferies. Benjie, please go ahead.
Yes, good morning. I just have two questions, please. The first one was on the U.S. and asset quality. The NPL ratio obviously jumped quite sharply quarter-on-quarter, and looking at slide 40, Stage 2 balances were also up. I'm just wondering whether you could give some more details about the drivers of that, and also what you think the appropriate level of Stage 3 coverage in the U.S. would be, because at the moment it looks a little bit light versus the group average. The second question was just on Mexico. I mean, the fee income trends were very strong quarter-on-quarter. Can you give any more detail around the drivers of that and how you see that going forward?
Also just more broadly in Mexico, whether you think the current environment is providing you with a potential competitive advantage versus peers, and do you see the prospect of taking more market share again in the current environment in Mexico?
Do you want to take the first one?
Yeah, I'll take the first question. You're right. The NPL ratio increased in the quarter in the U.S. by 79 basis points. Two main reasons. The first one is the decrease in the denominator of the ratio. We had a significant de-leveraging in the U.S. in the third quarter. Lending draws are down by 9.3%, that affected significantly the ratio. On top of this, we did have some additional NPL entries in especially the wholesale portfolios, mainly related to the sectors that we think can be most affected by the crisis, mainly oil and gas and commercial real estate. On the current coverage ratio that we currently have is 95%. It's not easy to set an specific coverage ratio for Stage 3s, especially in a country like the U.S. where it is very much affected by individual names, mainly large corporates.
We need to do an individual analysis on its counterparty to really set the right coverage ratio. What I can assure you is that at every point in time, we will have what we think is the right coverage ratio. We've proven in the past that many times we sell NPL loans in the U.S., we tend to generate a profit. So coverage levels historically have always been strong.
Mexico question, the fee income is progressing very nicely in the third quarter, as you said, because the fee income, 60% of that in Mexico is two things. Close to 40%, slightly higher than actually 40%, is payment systems, credit cards, and then another 15%-20% is on asset management. In both categories, they are coming back up very strongly. I did mention, it is tied back to the economic activity. Economic activity and spending goes up, you immediately register more fee income. Are we expecting that trend to continue? Yes, we do, the economic activity, hopefully, is going to be a strong one also going forward. You were asking me in the previous question, the spending index for Spain in the last week. That same number for the last week is 105, 104.5 in Mexico.
It is hanging up in there, and that probably would feed into the fee income. Are we getting the benefit of the fact that the crisis is happening in Mexico? For sure. I keep saying it, and I'm not saying it as a subjective opinion. It is proven by facts. Our banks, in most of the geographies that we are present, is either the best bank or one of the best banks in the country, proven by the return on equity that we post as a bank over time versus the competition in that same country. In that context, our bank in Mexico, in my view, is clearly one of the best banks, if not the best bank. I think it's the best bank, but I don't want to be too subjective on my own babies. It is a great bank.
In these type of situations, flight to quality happens, and we are seeing it. In every single product, we are gaining market share.
Thank you, Benjie. I'm afraid we are running out of time because we have the press conference afterwards. Probably we have time for the very last question. Next question, please.
Yes, the last question comes from Ignacio Cerezo of UBS. Ignacio, please go ahead.
Hello, good morning. A couple of quick things from me. Back on the dividend question, if I may. If you can give us any color about the possibility of ECB imposing a cap on the dividend payout as an approach to separate banks that can pay dividend or not, or do you think it's just going to be related to the amount of money you make and the capital you have? The second question is on Turkey. If you can give us the sensitivity of the capital of the local unit, not the group basis, in terms of the FX moves from here, and how sensitive is Garanti local capital base to the lira depreciating additionally from here? Thank you.
Thank you, Ignacio. On the first one, the possibility of a cap, we don't know. We haven't had that detailed interactions, so we don't know. The only thing I would say is, in our case, I would look into our case. We have 340 basis points, again, distance to MDA. More importantly, we distribute our dividend policy is a very consistent one, as you all know. We distribute 35%-40% of our profits typically in a year. It has been the case since the time that we established this policy, which is again, a long time. That 35%-40%, when I compare that to most other banks in Europe, it sits very well. I would not expect anything. Let's see what comes out. Our policy is a prudent one, so I wouldn't expect negative consequences. On the second one, Jaime.
Yeah, the local CAR ratio, capital and equity ratio sensitivity to a 10% depreciation of the Turkish lira versus the US dollar is 48 basis points. 48.
Perfecto. Gloria.
Okay, we need to end it here. Thank you very much for participating in this call. Of course, let me remind you that the whole IR team will remain at your disposal for any questions you may have.
Thank you to everyone for joining in, and stay safe, stay healthy. Bye-bye.
Thank you. Keep safe.