Good morning, welcome to Unicaja Banco second quarter 2020 results. Let me start confirming that we have published the quarterly financial report and this presentation this morning before market opens in the CNMV website. Pablo González, our Chief Financial Officer, will present you the quarterly results, and afterwards, as usually, we will answer your questions. Pablo, whenever you want.
Thank you, Jaime. Good morning to everyone. These are the second quarterly results that we publish after the COVID-19 pandemic starting. We have now left behind the lockdown, which is a very positive news. This quarter, we have booked additional provisions for the future potential impact from the COVID-19. We haven't seen any asset quality deterioration so far, as I will explain later. You will see, since the pandemic started, we have been able to increase significantly our solvency and our coverage. We have further reduced our NPAs and still reporting positive net income. Before entering into the details, I would like to highlight, as you will see later, that Unicaja Banco financial position has been reinforced year to date. Moving now to the presentation, as you can see in slide two, we have three sections.
The first one includes the summary of the quarter, the second one, the results, and the third one, details on asset quality, liquidity, and solvency. If we move directly to slide four, I will start with the regular summary. Regarding the business, performing loans grew 2.7% in the first six months of the year. However, if we exclude the regular second quarter seasonal consumer loans, the growth was 1.1%. While new loans to individuals fell 38% compared with last year, new loans to corporates were quite stable. Customer funds showed a positive trend, growing 4.1% quarter-on-quarter, supported by a 6.5% growth in sight deposits and almost 3% growth in off-balance sheet funds.
From a P&L point of view, second quarter results reflect a negative impact in NII and fees related to COVID-19 lockdown, with net interest income falling 2.4% quarter-on-quarter, while fees decreased by 14.4% in that same period. This quarter, we have booked the EUR 47 million of gains from Caser, something that together with the strong trading gains mitigated the provision effort done in the quarter. It is probably worth noting that the gains from Caser, in our case, are not explained by the renegotiation of our distribution agreement or the disposal of part of our stake. In our case, we didn't sold nor renewed the agreement. In our case, it was a bit different because the distribution agreement included a clause saying that there was a change in control in the company.
If there was a change in control in the company, we had the right to finish the agreement. The change in control happened, so we reached an agreement with the insurance company, and in exchange of this agreement, there was a payment to resign or this, our right to finish the distribution agreement. Regarding expenses, total cost fell almost 6% quarter-on-quarter, or 4.3% when compared with first half of 2019. It is also worth noting that we have booked this quarter an additional EUR 78 million of COVID-19 related provisions, leaving the total provision in the first half of this year for COVID-19 at EUR 103 million. Excluding such provision, the cost of risk remained very stable and close to 2019 levels.
On asset quality, liquidity, and solvency, NPAs fell by 26.2% year-over-year, while coverage levels improved from 57% to 62% in the same period. Regarding liquidity, our position remains very comfortable with the loan-to-deposit ratio slightly below 70% and the liquidity coverage ratio at 346%. From a solvency point of view, our CET1, fully loaded, continued to grow, reaching, in June, 14.4%, which is 32 basis points above the previous quarter and almost 130 basis points higher than one year ago. A significant capital generation that maintains our solvency position among the highest of the sector. I will move now to slide six, where you can see the P&L details. Starting with the quarterly trends, both NII and fees reflected the negative impact from the lockdown this quarter. This quarter will be the bottom as we expect second half 2020 trends in core income to improve significantly.
The rest of the revenues in the quarter were supported by a strong trading income and the EUR 47 million of gains from Caser, including in the other income. Such gains more than compensated the contribution to the resolution fund of EUR 16 million accounted in that same line, leaving gross margin 8% above the previous quarter and 16% above the second quarter of 2019. In terms of costs, this quarter, we have reported a very positive trend, with total costs falling 5.6% quarter on quarter and 7.5% year on year. As we will explain later, such positive trend was focused in general expenses.
The strong revenues reported in the quarter and the mentioned cost-cutting effort left the pre-provision profit at EUR 134 million, 59% higher than in the second quarter of 2019, and almost 28% above the previous quarter, something that has enabled us to book EUR 110 million of impairments while reporting EUR 50 million of net income in the quarter. It is worth noting that the big bulk of impairments, EUR 94 million, were loan loss charges, of which EUR 78 million are COVID-19 provisions, and the remaining 16 are regular loan loss provisions. Let me highlight now the main trends of the half-year results. Net interest income was 5.3% below the previous year. We reiterate our expectation to finish 2020 at the same level as 2019, something that implies a significant improvement in the second half of this year.
Despite the low level of activity, mainly in the second quarter, and helped by some one-off fees reported in the first quarter, fee income was flat in the first half of 2020 compared with the first half of 2019. Here, the idea would be in our base scenario to finish 2020 very close to 2019 final year figure. Dividend income and equity method has been also affected by the pandemic, as you can see in the right side of the slide. Trading gains and other income have been strong so far, mitigating higher impairments. Gross margin grew 7% compared with the first half of 2019. Regarding costs, the trend is very positive, decreasing more than 4% year-on-year, leaving pre-provision profit at EUR 239 million, 24% above 2019. Finally, the previously mentioned provision effort left income at EUR 61 million for the first half of 2020.
If we move to customer funds in slide seven, you can see that total customer funds grew 2.5% year-to-date and 4.1% quarter-on-quarter. It is worth noting that after a few months decreasing, off-balance sheet funds showed a quarterly improvement of almost 3%. Trends of on-balance sheet funds were also positive, supported by a 7.1% year-to-date increase in side deposits. In slide eight, we show the regular credit and loss info. Gross loans grew 2.4% in the first six months of 2020, with public sector growing 9.6%, non-performing loans decreasing by 2.3%. Private sector growing 2.2%. On the right of the slide, you have the details on performing loans. As you can see, they grew 2.7% year-to-date.
Let me remind you that this trend includes around EUR 400 million of seasonal advances that we usually have every second quarter and that are included in the segment called consumer and others. In terms of segments, in June 2020, corporate performing loans grew 8.4% year-on-year, helped by the ICO loans, while mortgages decreased a bit less than 3% in that same period. In slide nine, we show the new loan production by segment. Overall, new production has been directly impacted by the COVID-19 lockdown, mainly in individual loans. Corporate new loans were stable compared with the first half of 2019, individuals' new loans fell by 38%. Total new loan production in the first half of 2020 was 25% below the previous year. It fell only 2% compared with the second half of 2019. In slide 10, we start the P&L review with net interest income.
As you can see in the top right, the quarterly decrease was explained by lower income from some products that are related to COVID-19 situation as delinquencies and set up fees that were not charged throughout the lockdown, negatively reflected in the second quarter of 2020 interest income. Such measures explain almost EUR five million of lower interest income in the quarter. On top of it, the rest of interest income added other negative EUR one million, mainly explained from the lower income from loans, which was partially compensated by higher contribution from our debt portfolio. On the other hand, interest expenses decreased a little bit, partially compensating the previously mentioned lower interest income in the quarter. As a result, the net interest income fell EUR three million quarter-on-quarter and bottomed at EUR 137 million.
A level from which we expect a significant improvement in the second half of this year, supported by the savings in interest expenses from the maturity of high yield deposits and an improvement from the impacts in NII related to COVID-19 lockdown. In slide 11, you have an update on our debt portfolio, where you can see that our overall balances grew during the quarter, following the increase in the liquidity position of the bank, explained by both the commercial gap and the higher ECB funding from TLTRO3. As usual, just to highlight that the big bulk of the exposure remains sovereign debt classified in amortized cost portfolio. The yield of the portfolio was 123 basis points in the second quarter of 2020, compared with 127 basis points on the previous quarter and 125 in the second quarter of 2019. Quite stable.
If we move to slide 12, you have the fee income trends. Total fees fell 14% in the quarter owing to the impact from the lockdown, but also partially explained by some one-offs included in the first quarter of 2020, as we explained last quarter. However, if we compare net fee for the first half of 2020 with the first half of 2019, the trend was stable. As you can see in the right-hand side of the slide, for the second half of 2020, we expect fees to improve to a level somewhere between the first quarter and the second quarter of this year. Moving now to costs. As you can see in slide 13, operating expenses fell almost 7% compared with first half 2018, and more than 4% compared with the first half of 2019. As you probably know, cost-cutting is part of Unicaja Banco strategy.
We have a plan to continue to reduce our cost base. This quarter, improvement has to do with some measures driven by the COVID-19 situation that have enabled us to report lower general expenses in the quarter. In slide 14, we show impairment trends. As we have explained, we book in the second quarter of 2020, other EUR 78 million of provisions for COVID-19 potential future impacts, which added to the EUR 25 million booked in the first quarter of this year, explain the EUR 103 million of COVID-19 impairments that we show in the left-hand side of the slide. In the right of the slide, you can see that excluding this impairment, cost of risk represented only 17 basis point. When we include the COVID provisions, the cost of risk increases to 89 basis point, which is the high range of our guidance.
If we move to asset quality in slide 16, we have the details on the evolution of our non-performing loans that continue to fell one more quarter. In the second quarter 2020, non-performing loans balances fell 1% quarter-on-quarter, leaving the NPL ratio at 4.5%. As you can see in the slide, NPL trends remain similar to the previous quarter, with a small improvement, something that is positive. Gross entries remain very similar as the previous quarter at EUR 58 million, something that, together with the EUR 62 million of recoveries, mainly cash recoveries and EUR 10 million of write-offs, left total NPL balances decreasing EUR 10 million quarter-on-quarter. It is worth noting that this drop took place in a quarter where foreclosures were historically low owing to the lockdown, something that makes the NPL quarterly trend even more positive.
As you can see, the potential deterioration from COVID-19 was not reflected in second quarter 2020 NPL trends that continue to improve. In slide 17, we have included information related to the different mitigating measures introduced as the moratoria and the guarantees of the government for corporate loans. Starting with corporate loans guarantees, what is called the ICO lines. As at the end of July, we have approved 10,500 loans, representing a limit of EUR 750 million. This balance represents around 10% of total performing loans to corporates. However, this is not really comparable, because the 750 is the limit. While, for example, as at the end of June, the lines used were only EUR 426 million, so only 57% of the limit. Bear in mind that the EUR 750 million is the balance of the limit of loans formalized under the guarantee scheme.
The government guarantees for that limit represent close to EUR 600 million or EUR 330 million if we only consider the amount used, which is far less below the EUR 1.2 billion guarantees assigned to Unicaja Banco. In the case of the moratorias, as you all know, we have the legal and the sector moratorias. In the first one, we have approved EUR 8,600 million in mortgages and other EUR 3,030 million in consumer loans moratoria. Regarding the sector moratoria, the amount reached EUR 540 million in mortgages and EUR 25 million in consumer loans. It is worth noting that EUR 340 million of the mortgages and EUR 15 million of the consumer loans have requested both moratorias. After the three months in the so-called legal scheme, clients will benefit from other nine months in the sector scheme.
In relative terms, this means that EUR 800 million or 5.6% of our mortgages and other EUR 40 million or 1.3% of consumer loans are under moratoria schemes. In summary, total moratoria represented less than 5% of individuals performing loans, of which 95% are mortgages and only 5% unsecured loans. Something explained among others owing to our mix of loans, where the weight of mortgages is high, while pure consumer loans is very small. In slide 18, we have updated our credit risk exposure and NPL coverage details. Overall, NPL coverage continues to grow quarter after quarter. In the second quarter of 2020, NPL coverage was 61%, which compares with 52% one year ago and 54% at the end of last year. An extremely prudent coverage when considering that 87% of the non-performing loan balance are secured.
As a reminder, it is also very important to take into account that the big bulk of our loans are mortgages. As you can see in the bottom left of the slide, 71% of our loan exposures has mortgage collateral, and less than 20% of our total loans are corporate unsecured loans. In slide 19, you have an update on the foreclosed assets, of which balances have been more stable during last two quarters owing to the lockdown. However, we can say that the situation has improved by the end of the second quarter. Coverage ratio remains among the highest of the sector, disposals, despite representing fewer balances, continue to be formalized at very good prices, well above our net book value.
Actually, if you look the trend of released provision over book value on the top right hand of the slide, you can see that they reached 66.8% in June. Despite the lower volumes, prices are even better than the previous ones. In slide 20, you can see how overall NPAs have decreased significantly during the last year, representing in net terms, only 1.5% of total assets, with a coverage at 62%, well above the 57% at the end of last year, being one of the highest of the sector. Finally, I would only highlight that the lower NPAs balances, higher provisions, and the tangible book value increase continues to push down quarter after quarter our Texas ratio reaching a level close to 45% in June 2020.
In slide 21, we have the regular information of our liquidity position that, as you can see, remains very comfortable. Our loan to deposit ratio has moved below 70% in June, and the liquidity ratios with the LCR at 346% and the NSFR at 141% continue to be among the highest of the sector. Finally, we show in slide 22 the solvency position of the bank. Our regulatory CET1 reached 15.8% in June, and our total capital was at 17.3%, one of the highest of the sector and representing a EUR 1.2 billion buffer over our SREP requirement. In fully diluted terms, this is not considering the transitional benefits from IFRS 9 and from Basel III. Remember that we applied the loan calendar for the DTAs deduction.
The ratio improved 32 basis points in the quarter to 14.4%, of which around one-third is explained by the SME factor and the other third by the new rules for intangible, specifically software. The remaining, slightly more than 10 basis points, reflect the rest of the moving parts of the quarter, which are, among others, quarterly retained earnings, a small risk-weighted assets decrease. Bear in mind that the second quarter results always include a seasonal increase in risk-weighted assets from the seasonal advances that usually decrease in the following quarter. Also, let me remind you that our relative higher solvency is calculated in full under the standard approach. This means that it's considering very conservative assumptions, as you can see in our credit risk density in the chart at the bottom left.
Let me finish highlighting that in the first half of 2020, and mainly since the pandemic started, we have been able to further reduce the NPAs balances, increase significantly our coverage. We have improved by 40 basis points our solvency from the already relative higher ratios. We have also reinforced our liquidity. In summary, the financial position of the bank is extremely strong and, under our view, leave us in a very positive position to deal with the COVID-19 future expected recovery.
Thank you, Pablo. Let's move to the questions. We got lots of them. Let's go straightforward. The first one, Pablo, let's start with the NII trends, with the P&L. As usual, the first one is if we can update the dates of the high-yield deposits, the exact dates of the high-yield deposits maturing in the second part of 2020 and 2021.
Yes. We used to have EUR 1.2 billion of retail deposits at a very high cost of 4.3% that have started to mature at the end of this quarter. In June, EUR 80 million mature. In the next quarter, in this quarter, in the third quarter of 2020, there will be around EUR 500 million. In the fourth quarter of this year, another EUR 250 million. The remaining balance that are around EUR 400 million will mature in the first half of 2021. Considering the current cost of deposits, these maturities will enable the bank to reduce its annual interest expenses significantly, mainly from the third quarter of this year. This is one of the main tailwinds of the expected improvement that I mentioned of our net interest income in the second half of this year.
Pablo, the next one related to the ALM portfolio. What is the expected contribution from the debt portfolio going forward?
The contribution from the debt portfolio in the second half of 2020 will be in line with the contribution in the first half. We have bought in advance the bonds that were expected to be purchased next year to replace the redemptions that we have. The purpose of these advanced purchases was to invest the liquidity coming from the TLTRO III that we have bring the maturity of the TLTRO to this year rather than waiting for next year. That was the original deadline. Also from the increase in the customer deposits that I mentioned in the presentation. All in all, the 2020 debt portfolio contribution will be above the 2019.
For the coming years, the contribution should decrease, especially compared with 2020, mainly due to the redemptions that I mentioned before, and because those redemptions and the new purchases have a lower yield than the bonds that we had, either that the bonds that mature or those that we sold to obtain capital gains. To offset this negative impact, in the future, we will have other tailwinds in the NII, like the better funding conditions of the TLTRO III, or the redemption of the EUR 1.4 billion of the high yield deposits that I mentioned.
Thank you, Pablo. Very related with this one, if we can please update our strategy regarding the TLTRO III funding, and its contribution to NII.
I already mentioned. Regarding the TLTRO III, we went in full for the whole amount that we were assigned in June, which is slightly above EUR 5 billion. We did it because the financial conditions between June 20 and June 21 are really attractive with a 50 basis points extra lower cost, but also because we feel confident in complying with the credit growth requirement between March 20 and March 21 to obtain the full benefit of the cost of this funding. Proceeds were used to replace the preexisting TLTRO, the TLTRO II. There were EUR 3.3 billion, as well as to fund any potential future commercial gap that may arise. However, as we expected, evolution of the business volumes are not clear as of today. We are investing the excess liquidity, as I mentioned, in fixed income, mainly in European sovereign bonds.
All in all, the larger funding and better conditions compared to TLTRO II will be a positive for net interest income in the coming quarters. If we comply with the credit growth requirements, as we expect to do so, the benefit versus the TLTRO II should be around EUR 25 million for the next 12 month, plus the margin that we may obtain with the investment of the EUR 1.7 billion of extra funding that we got from this new TLTRO. This will be very positive for the future net interest income.
Thank you, Pablo. The following one is if we can update our funding plans and issuance of subordinated liabilities.
Following this new TLTRO, and considering our very strong liquidity position, our funding plan for the next few years should not have any significant changes. We will continue to analyze long-term funding sources in the future to replace the TLTRO3, but it's still too early to quantify amounts of costs. In the short-term, no new funding issuance. Regarding the MREL, considering our current risk-weighted assets, the requirement of eligible liabilities for MREL remains below EUR 1 billion. This shortfall was expected to be covered initially before January 2022, but with the new MREL policy that lengthens the deadline up to January 2024, give us additional time and flexibility to comply with the requirement.
We still haven't decided the proportion to be covered between senior preferred and senior non-preferred. Considering the last liabilities trends, we could finally issue senior preferred in a larger amount of what we were expecting. These are good news in terms of cost, as expected volume requirements remain similar. We might issue a higher amount in the cheaper instrument, the senior preferred, rather than the senior non-preferred.
Thank you, Pablo. All in all, regarding NII, if we can update our guidance, the net interest income guidance.
As I said in the presentation, we reiterate, we maintain our previous guidance of reaching the same level of NII for 2020, the one that we have in 2019. Despite the small decrease in this quarter, we continue to expect a significant improvement in the second half of this year that will enable us to reach the 2019 level, at least in our base scenario. In order to do so, NII for the second half needs to be significantly above the first half, something that will be possible among others, owing to the maturity of the expensive deposit that I mentioned. There are other positive and negatives. On the negative side, new loans to individuals have decreased with the lockdowns, mainly consumer loans that have a relative higher yield.
However, this was partially compensated by the guarantees of the ICO lines for corporates and SMEs, but still having a net negative impact from the loan mix. On top of the expensive deposits, we also expect a relative higher contribution from the debt portfolio, and following the TLTRO strategy that I mentioned. All in all, with current information at this moment, our base scenario remains to maintain a stable NII for 2020 compared to 2019.
Thank you, Pablo. Can we provide an update regarding loan growth, please?
As you saw in the presentation, the trends are quite different depending on the segment. While in corporates and SMEs, following the higher demand, the evolution is better. Individual loan trends, mainly consumer, are weaker. In our case, as I mentioned in the presentation, the unsecured lending to corporates, which is the segment that is currently growing in the sector, supported by the ICO loans, represents only 20% of our loan book. In mortgages, new production has decreased, not as much as in consumer lending. Net net, going forward, we expect weaker volumes than before the COVID crisis, we will probably not grow as much as the sector because of our conservative mix with less weight in unsecured corporate loans. There are also some potential good news. Our loan book is highly impacted by early amortization.
With the current situation and the moratorias, will probably reduce the previous speed of redemption and early amortization, partially compensating the lower new production in loans to individuals. It is too soon to have a clear view on the final volume trends, but it makes sense to expect prudent and conservative volumes ahead, at least in the next couple of quarters. That said, we expect to finish the year with performing loans quite stable compared to last year.
Thank you, Pablo. Related to this one, if we can clarify or provide under our view, what are the reasons to have relative lower balances also in moratoria, but mainly in the ICO warranted loans?
I cannot answer or provide details on what our peers are doing. What I can tell you is that the measures taken on the bank are positive for some clients and also for the bank. I think it's also for the sector as a whole. We continue to be very prudent with our credit risk approach, and the final balances of loans with moratoria and ICO lines are a consequence of the interest of our clients and the needs for such schemes. However, it looks like our clients do not need as much ICO lines as the sector, at least for the information that we have so far.
If we look at the information that we have from our risk and from outside our bank for each client, I can confirm you that with the latest available data, 95% of Unicaja self-employed customers with loans have not requested ICO lines, neither with us nor with another bank. In the case of SMEs, the percentage is 82%, and for large corporates, it's 70%. As you can see, because of whatever reason, Unicaja customers have not requested ICO loans, neither at Unicaja nor at another bank. We're confident on the financial situation of our customers. Another reason of our relative lower ICO lines is the mix of our loans, with more than 70% being mortgage-related and only 20% are unsecured loans to corporates. Probably another reason is the mix in terms of sector.
As we mentioned in the last quarter presentation, we have probably higher exposures to the sectors that were less affected by the COVID-19, they ask not so much ICO loans. Those sectors like construction and agriculture sector, these two sectors, in our case, represents, as we mentioned in the last quarter, around 30% of our total exposure. I can confirm that we have offered such loans to every single corporate and customer in the bank that could potentially benefit from it, because as you can imagine, commercial activity has been very intensive, but we only have formalized around 50% of the guarantees assigned to us. I think in the case of moratorias, we are more in line with the sector.
We have used the moratorias as an interesting tool for those clients that we believe could benefit from the terms, because we thought it was positive for them and also for the bank. As a result of this strategy, we had, at the end of July, EUR 800 million or around 5.6% of our mortgages, and only EUR 40 million or 1.3% of our consumer loans under moratoria scheme. In other words, only 3% of our total performing loans have requested moratoria.
Thank you, Pablo. Moving to fees, if we can please provide an update on our guidance.
I think, as we explained during the presentation, fee income has been negatively affected in this quarter by the lower activity as a result of the lockdown, also owing to the decision taken by the bank of not charging some fees to our customers in the special situation and in order to help them until the lockdown finished. Such a negative impact was concentrated in the second quarter of 2020. At the end of the quarter, the trend already started to improve, with activity levels growing and charging regular fees again. It is also worth noting that, as we explained last quarter, there were some one-offs in the first quarter 2020 fees. That also explain one part of the quarterly drop.
All in all, we expect to increase fees in the second half of 2020 to levels somewhere between the first quarter and the second quarter, something that, if it's finally confirmed, would enable us to finish the 2020 final number in fees very close to the 2019 number. That is at least our base case scenario. This guidance is considering, obviously, an upward review of our fees from the third quarter onward, something that, together with the recovery in the transactional business, would probably help to change the second quarter trend.
Thank you, Pablo. Moving now to one of the topics of the quarter, because we got plenty of questions regarding the costs evolution, the expenses evolution. If we can update the cost-cutting plans, expenses update, additional cost-cutting measures, and the expected trends.
Okay. As you all probably know, during the last couple of years, we have booked a significant amount of restructuring cost in provisions that will enable us to crystallize additional cost cuttings until 2022. On top of such plan, following the COVID-19 impacts, we have taken some additional measures to save as much costs as possible, explaining the quarterly decrease in general expenses. Following the pandemic, the situation has changed significantly, and we have taken these measures to compensate the lower income with lower cost. All in all, cost-cutting remains part of our strategy. The idea is to keep crystallizing additional savings going forward and become a much more efficient bank than we are so far.
Thank you, Pablo. Moving to loan loss charges and the cost of risk, if we can update our guidance, please.
As we have discussed in the past, our conservative loan mix and higher coverage leave us in a relatively good position to absorb any potential deterioration on asset quality. That said, as you can see in the trends reported today, we haven't seen such deterioration so far, and the NPL balances continue to decrease. For the second quarter in a row, we have booked more provisions, a bit more than EUR 100 million in the first half of this year. These provisions have been booked mainly for a potential future deterioration of our credit risk exposure following the worse than initial expected macroeconomic trends. In order to do so, we have taken advantage of some of the trading and the cash gains, confirming our strong result generation capacity one more time.
We have increased our NPLs coverage by almost 10 percentage points in the last 12 months. From 52% in the first half of 2019 to 44% at the end of last year, to the current 61%. These coverage levels, together with our loan mix and the level of collateralization of our book, leaves us in a relative good position. It remains difficult to quantify the final cost of risk, as you can imagine. As we guided last quarter to a level of cost of risk between 50 and 90 basis point for this year, we were in the low part of the first quarter, and we have moved now to the higher range this quarter. At this point, we still cannot be much more specific on how much provisions will be needed to absorb COVID-19 crisis.
We expect to keep recurring cost of risk at the current levels, so very low levels. On top of this, we will continue to analyze the situation if additional provisions are required. All in all, we expect to finish the year with a cost of risk in the previous range guided between 50 and 90. Obviously, it will depend on the evolution.
Thank you, Pablo. The next one is on NPL trends. The NPL balances decreased in the quarter. When we do expect NPL balances to start to increase going forward?
Overall NPLs continue, as you could see, to fall this quarter. The expected deterioration has not taken place yet. There are a lot of measures in place, as you all know, like the ICO, the moratorias, the so-called ERTEs, which is the temporary leave from the companies supported by the government. That will probably explain part of the lag effect, but that will also enable some clients and customers to avoid future problems paying their loans. Some businesses will be affected more than others, obviously, even within the same sector. In all cases, the potential deterioration has not been reflected in our exposure so far. The provisions that we have already booked, as you can imagine, are mainly stage 1 provision. Year to date, we haven't seen an increase in stage 2 or stage 3 exposure.
Actually, stage 2 is stable, and stage 3 has decreased, as you know, from EUR 1.4 billion to EUR 1.3 billion. The big bulk of the provisions booked so far for COVID-19 are provision made after considering higher expected loss from worse macroeconomic trends, but not owing to a specific sector, loans, or customers. On top of this, we have the watch list loans balances, which is one early warning indicator that we use in the bank, were in June below the level of December 2019. We expect to have more color and more visibility in the second half of 2020. Until then, we are already working to identify potential future NPLs. For example, we have included in every single commercial agenda and contacted every single customer that has been affected by unemployment benefits or salaries under these temporary leave schemes, ERTES.
Just to make sure that the situation, and to analyze with the customer if they need any solution to put in place. With the target analysis, identifying and managing these cases in detail. We are also closely monitoring all the legal moratorias that have not requested the sector moratoria, just to identify if some cases, this would be a positive solution. So far, most of the people that finished the legal moratoria that didn't engage in the sector moratoria keep maintaining the payments. All in all, so far, we haven't seen delays or a clear deterioration in those legal moratorias that have already finished. The first ones, as you can imagine, because we start in March, the first one is finished in June. We are not seeing signs of a specific deterioration in those portfolios. As you can see, we're doing our job.
We are reinforcing the provision levels in order to generate extra buffer and monitoring very closely the situation on a client-by-client basis. In this sense, Unicaja Banco, under my view, is in a relative good position with one of the highest coverage of the sector and a very prudent loan mix.
Thank you, Pablo. This one is specific on the DTAs. Do we expect the COVID-19 to reduce or to limit somehow the future realization of deferred tax assets, Pablo?
No, we don't expect it. As at June, 56% of our total DTAs were guaranteed by the government, and the remaining 44% are deducted from our solvency ratios. Bear in mind that in exchange of this government guarantee, we pay a bit more than EUR 50 million every year. That said, under our expectation and considering the potential impact from COVID-19, we continue to expect to generate enough results in the future to realize our DTAs.
Thank you, Pablo. We are running out of time, so moving to solvency. We have several questions. The first one is if we can clarify the regulatory impacts, considering the core equity tier one fully loaded. What will be the impacts from IFRS 9 transitional arrangements and the sovereign prudential filter and the details, Pablo?
Okay. This quarter, there are a lot of regulatory changes in solvency, so let me summarize the main ones. From a regulatory point of view, as I explained before, our CET1 fully loaded reported today includes around 10 basis points of the Quick Fix, the SME factor, and other 10 basis points from the new rules for intangibles, specifically for software investment. In our case, the sovereign prudential filter is not material, so we didn't take that into account. Finally, the IFRS 9 transitional arrangement is something that only applies for the regulatory ratio, not to the fully loaded. In our case, this IFRS 9 represents around 10 basis points, so it's not really a big impact.
Very related with this one, Pablo, we got lots of questions asking us to clarify it. If it's the core equity one fully loaded considering the IFRS 9 transitional benefits is already answered.
I think that's already covered.
Moving to IRB models, if we can update the situation, Pablo.
Despite we have been working internally with the internal models for some time now, we still haven't received the formal approval to apply them in terms of solvency. We continue to use the standardized approach in full, as you all know, meaning that we consider very conservative risk-weighted assets densities. We have continued working with the supervisor in the process throughout the lockdown. As you all know, we prefer not to consider the benefit in our solvency plans, and at least until we have more clarity or more visibility from the supervisor.
Thank you, Pablo. An update on dividends and the accrual of dividend in 2020?
As you all know, the ECB confirmed last week the extension of its recommendation of not paying dividends and buying back stocks until January 2021. In the announcement, the ECB said that they will review the recommendation in the fourth quarter of 2020, and that once the uncertainty requiring this temporary and exceptional recommendation subsides, banks with sustainable capital position may consider receiving these payments. You all know that our aim is to pay cash dividends as soon as possible, and we understand that considering our results, solvency, and financial position, Unicaja Banco will be among those banks receiving dividends and buybacks as soon as possible. In the meantime, let me confirm you that we are considering in our solvency ratios only part of the 2020 results.
Out of the EUR 61 million of net income reported in the first half of this year, in solvency terms, we are only considering EUR 27 million, so the rest have been accrued dividends.
Thank you, Pablo. One final question regarding our views on the sector consolidation, please.
As we have said in the past several times, we remain open to analyze all the options that could be positive and interesting to our shareholders.
Thank you, Pablo. We will now finish. Thank you all. Thank you for listening to our webcast. If you need further info, do not hesitate to the IR department. We'll be delighted in attending as many requests as we receive. Finally, for those taking a brief summer break during next days, have fun, disconnect, and keep safe. Thank you very much.
Thank you.