Banca Sistema S.p.A. (BST) Earnings Call Transcript & Summary

Good afternoon. This is the Chorus Call conference operator. Welcome, and thank you for joining the Banca Sistema First Half 2021 Results Conference Call. [Operator Instructions] At this time, I would like to turn the conference over to Mr. Gianluca Garbi, CEO of Banca Sistema. Please go ahead, sir.

Thank you very much, and good afternoon to everybody. I'm here with Carlo and Ilaria. And as usual, I make reference to the presentation that has been made available to all the participants. I'm starting with the result at glance in the Page 2 of the presentation. And the first thing is, let me say that we are very proud of the good commercial performance that we have also during this second quarter of the year where the factoring turnover was up 15% on a year-on-year basis. We are still seeing that tax receivable turnover did not increase, and without this component, the other component will have increased 23%. The CQ outstanding is up on a quarter-on-quarter basis as well as on a year-on-year basis. The pawn loans reached EUR 83 million in terms of outstanding, which is an increase of 4%. We have then continued to originate loans guaranteed by the government for client of our factoring activity and the stock has reached under EUR 34 million at the end of the second quarter. The net interest income end up to be at EUR 38.5 million with an increase of 15% on a year-on-year basis, which has been supported as well by the pawn as well by the broking activity, the SME guarantee and the lower interest expenses. The availability of extraordinary public fund led to some early repayment in some sector but that has not changed dramatically the situation on the performance of the public administration. We have registered a lower funding cost on year-on-year basis at 0.5%. The total income ended up to be at EUR 50.3 million with an increase of 9%. Recurring cost of risk is equal to 46 basis points, the EUR 4.1 million are the loan loss provision, which are up on a year-on-year basis due to a single position that we already have discussed on the first quarter, and that has had an impact of EUR 2.4 million and an additional EUR 1.4 million, which is only related to the time value that we have adjusted, taking in consideration a lengthening of the expected collection time of position under conservatorship of city. That means that we will expect to collect a bit later due to this lengthening of the position. At the same time, we will expect to have more late payment interest coming with the collection. The total operating cost has increased mainly to the consolidation of the pawn broking business, as you know, that after the purchase of Banca Intesa division on the pawn broking activity. The net income end up to be at EUR 8.5 million. And I would say that this is a substantially in line with the last year if we exclude the nonrecurrent items that affect balance sheet on the first half of this year, but also in a positive way on the first half of last year. Moving to the next slide. As you can see, the factoring outstanding trend shows a slight decrease on a year-on-year basis. And this is driven by 2 main effects. One is the less turnover on tax receivable. We continue to see that even though the economy is recovering, the invoiced amount did not generate yet tax credit to the size that we are used to see in the past. And then we have experienced some as I said before, faster collection from some cities in particular, which doesn't mean a change on the attitude of the public administration, but simply the government put forward additional resources, mainly to city that were close to distress. And even in the last support to the economy, they had EUR 660 million for cities that are in potential and possible future distress. From the last quarter, we have split the commercial performance into slides. So moving to the next slide, #4. As you can see, the CQ turnover was stable on a year-on-year basis. I think that here it's important to notice the better performance on a year-on-year basis on the direct component, which now represent 20% of the total. As you remember, our strategy was to invest more on this direct distribution of the CQ rather than acquiring portfolio. And the division is working to strengthen this -- the number of agents and intermediary network. We recently also have announced an increase -- a substantial increase of new agents that joined the Banca Sistema network for the CQ. And thanks to an organic growth of the ProntoPegno, where we registered a 4% growth in terms of outstanding. In the second quarter of this year, in line with the strategic plan, we have opened also new branches -- a new branch in Brescia. And we have reached also an agreement with Cassa di Risparmio di Asti to buy their portfolio, and we open and also a branch, of course, in Asti. Historically, the third Q shows a robust growth in terms of volume, but less auction because of the summer period. As we said in March, when we presented the strategic plan, we are glad of the diversification in the CQ and pawn loans because we have now registered some negative trend over the last 12 months since we have started to talk, and we see the consequence of the COVID. Now let me leave the floor to Ilaria for the next slide. Ilaria?

Thanks, Gianluca, and good afternoon to everybody. As usual, from Slide 5 onwards, we cover the financials in more details. I'll first comment on the balance sheet. As shown in the table, total assets are down 6% versus year-end 2020. And in particular, loans at amortized cost is now EUR 2.7 billion and is stable versus year-end as a result of the following. Factoring assets are mainly -- are down mainly due to faster collections and the portfolio mix with lower average duration. CQ loans are up 3%, thanks to the good origination over the first half, and pawn loans are up 7%, thanks to the good business performance as well. The outstanding of the SME loans has significantly grown in the first half with EUR 134 million credit in the balance sheet at the end of June. Govies portfolio now amounts to EUR 639 million and is significantly down year-on-year with an average duration of 30.4 months, higher than in the past. On the liability side, due to bank's quarter-on-quarter increase is driven by interbanking while the ECB funding is stable at EUR 737 million. Due to customer quarter-on-quarter increase is mainly driven by the increase in current accounts that has more than compensated the decrease in Repos as a consequence of the govies portfolio reduction. Term deposits are substantially stable. Debt security quarter-on-quarter decrease is due to the redemption of a EUR 19 million product placement, the repayment of EUR 37.5 million Tier 2 bonds that was executed simultaneously to the issuance of an additional Tier 1 for the same amount. And finally, to a lesser extent, the reclassification of EUR 8 million additional Tier 1 under the item 140 of the liability side of the balance sheet. Shareholders' equity, therefore, includes since the second quarter, EUR 45.5 million additional Tier 1s, where the bulk is the AT1 newly issued. We now move on to the next page to discuss P&L. In line with the first quarter, the main factor driving the comparison between the first half 2020 -- and the first half 2021 and the first half 2020 is a different perimeter of the pawn loan business. We start analyzing interest income. First half interest income is up 4% due to higher contribution of pawn loans and SME loans. Factoring now represents 62% of total interest income, which is slightly less than historic contribution in the financial year 2020 and slightly lower in absolute terms year-on-year. Higher LPI contribution has compensated lower contribution from tax receivables. Going more into details of the LPI contribution, we registered the following. The contribution of LPI from legal action is equal to EUR 12 million compared to the EUR 9.1 million (sic) [ EUR 9.9 million ] in 2020, and LPI registered a strong performance in both accrual and extra collection components. Since the last presentation and more detailed information on LPI has been moved to the appendix so you can refer to that for a comparison over time. The contribution of the CQ business line has been slightly up quarter-on-quarter but down year-on-year due to a larger-than-expected impact of early repayments. Lower interest income in the quarter has been reflected on adjusted income margin reduction year-on-year. As already indicated, since the past quarter, we started to see a significant contribution of the pawn broking business also thanks to the acquisition. Indeed, the contribution to total interest income of this business line has been EUR 2.7 million versus EUR 0.5 million in the first half last year with a higher adjusted income margin versus the end of 2020. In line with Q1, the margin of the pawn broking has been confirmed above 15%, which is by far the highest among our 3 business lines. Total adjusted income margin was stable year-on-year as factoring and pawn loans have compensated the decrease in CQ margins. It's worth noting, however, that CQ margins are stable quarter-on-quarter, while factoring margins are slightly lower quarter-on-quarter. We now move on to total income on Slide 7. First half total income is up 9% year-on-year, driven by the increase in net interest income. Indeed, NII increased by 15% as a combination of a 4% increase in interest income and a 27% decrease in interest expenses. Total funding cost is 0.5% and is lower year-on-year in both the retail and wholesale component. Net commissions are down year-on-year due to lower factoring contribution that has not been fully compensated by higher commissions in the pawn loans. Other income includes EUR 1 million gain from the sale of factoring portfolio and EUR 2.8 million from the govies portfolio activity. The contribution of the govies portfolio was overall lower than last year as the contribution to the NII was significantly lower. In the pie chart at the bottom of the slide, we represent in line with the most recent earnings calls and the strategic plan presentation, the breakdown of the total income contribution of the 3 business divisions. The factoring division continues to represent the lion's share, although its relative weight has decreased with respect to year-end and also last quarter in favor of pawn loans, in line with what we envisage in the strategic plan. The CQ contribution to total income is pretty much unchanged with respect to past quarters. Now we discuss costs on Page 8. We do not have much comments with respect to what we already highlighted in the first quarter. Indeed, total operating costs are up year-on-year mainly due to the consolidation of the pawn loan business unit. Personnel expenses are up year-on-year due to higher number of FTEs following the acquisition and as well administrative expenses. In addition to the pawn loan cost increase, administrative expenses are higher in the first half as we've registered higher servicing costs and advisory expenses. D&A is also up due to the rental cost of pawn loan branches. We now move on to the next slide on funding. On funding, there isn't any relevant change to present in terms of our funding strategy since previous quarters. Cost of funding, as we said, is equal to 0.5%, which is lower year-on-year and stable quarter-on-quarter. The 1 thing that is worth commenting on the funding is that the wholesale component has further decreased in percentage terms since last quarter, reaching now a 36% relative weight. The most recent decrease is due to the combination of different dynamics occurred in the second quarter, which are, in particular, the maturity of the EUR 19 million product placement, the replacement of the Tier 2 bonds of EUR 37.5 million with additional Tier 1 for the same amount, the reclassification of EUR 8 million of loans as I mentioned before, and finally, the reduction in repos as a consequence of a reduction in the govies portfolio. ECB funding, as we mentioned, is stable quarter-on-quarter. Retail funding increased quarter-on-quarter, both in relative terms and those in absolute amount, thanks to the increase in corporate current accounts. Now we turn to Slide 10 to discuss asset quality. As you can see from the top left table, gross nonperforming exposure is down quarter-on-quarter due to the past due decrease related to factoring. As you will remember, we've had an increase in past due in Q1, driven by an introduction from the 1st of January of the new definition of default. As foreseen during that call, in the second quarter, we managed to reduce part of the increase registered in the first quarter. Within the gross NPEs, we've changed the classification of some exposure from 1 bucket to another. In particular, exposures versus municipality in conservatorship has now been all classified as bad loans following the recommendation of Bank of Italy, while in the past, most of them were classified as unlikely to pay. As a result of that, there has been an increase in bad loans and an equivalent decrease in UTPs. The new classification was not prompted by a different assessment of the probability of recovery neither by deterioration in credit worthiness. Which means that the business appetite versus exposure towards municipalities may enter the conservatorship process has not been affected by this change. Loan loss provisions in the first half are equal to EUR 7.8 million and are up year-on-year compared to the EUR 5.1 million last year. The higher-than-usual provisions are a consequence of the amount already registered in Q1 for EUR 4.1 million of which EUR 2.4 million were one-off provisions already discussed and also mentioned -- just mentioned by Gianluca and of new provisions for EUR 3.7 million set aside in the second quarter. Of that amount, EUR 1.4 million can now -- can be considered nonrecurring as they are related to a fine-tuning of our estimated collection times for positions versus municipalities in conservatorship. These one-off provisions would be, however, compensated as Gianluca mentioned, by higher collection of LPIs which are currently on balance sheet. As a result of that, first half cost of credit risk considering the above-mentioned provisions as nonrecurring stands at 46 basis point. In line with the last results update, we'll continue to represent in details our exposure to municipalities in conservatorship that, as we've just said, are now are all classified as bad loans. As you can see in the bottom left table, such exposure amounts now to EUR 130 million in total and has slightly gone up in the last 6 months. In the bar to the right, we represent the amount of LPI related to those exposures which is now EUR 60 million that as stated many times now are off balance sheet is not eligible to be considered within the accrual perimeter according to our accounting policy. I now hand the floor back to Gianluca for his final remarks.

Thank you, Ilaria. So let me turn to the last Slide 11 relating to the regulatory capital. The core Tier 1 ratio and total capital ratio, of course, are well above the minimum requirement. Tier 1 ratio benefit from the replacement of the Tier 2 bond with an additional Tier 1 bonds for the equivalent amount. Let me only conclude that we expect higher results for the second half of the year, driven by further growth. And let me also say that in the month of June, we have closed an inspection that we have received from Bank of Italy. So I will leave the floor to the audience for questions.

[Operator Instructions] The first question is from Christian Carrese of Intermonte.

First one, I will start from the last statement that you did the Bank of Italy inspection in June. Just to understand if you have booked in the quarter already some additional provision to take into account maybe some indication by Bank of Italy? Or should we expect some additional provision in the second part of the year? The second question is on dividends. Maybe I missed in the press release, but I wanted just to be sure that the dividend payout for 2019-2020, the DPS [ EUR 0.20 ] would be paid by the end of the year or there is any change on that? The third question on the margins in particular CQS. There was Lexitor cost adjusted decision, then there has been the sustaining base decree. If you can elaborate a little bit on these 2 items, if there will be any impact in terms of fees going forward for the past loans already issued. And finally, on cost of risk, we saw some extra provision in the first half. We saw better loans going up, but the creditworthiness is not deteriorating. I was wondering if you can give us an update on cost of risk for 2021, what you are seeing and what are your expectation for the full year?

Okay. Thank you, Christian. So yes, about the Bank of Italy inspection ended up in June. And we have followed the informal recommendation of the inspector. So we do not expect any adjustment on the second half of the year coming from the request from the regulator has also been mentioned, and these are answer also to the last question on the cost of risk. On one hand, 1 of the request was the reclassification of the city in distress from bad loans -- from unlikely to pay to bad loans, which is -- for us is the same RWAs, so it's under 50% RWA. Is it more as a formality as far as we can see because at the end, it's still city under distress. And so the creditworthiness of this counterparty do not change. So we expect to collect 100% of the capital plus the LPI associated to that. One of the other change that we have made was related to the time value. We adjusted the time value on the city under distress, which is a one-off adjustment. And this one-off adjustment on one hand create some additional provision because of the time value. On the other hand, if we'll end up to be true, that it will take more time to collect, we will have more, late payment interest at the rate of 8%. So on the bottom line of the P&L, we will make more money rather than less money. As been mentioned before by Ilaria, if you look at the perimeters of the city under distress, we have today under EUR 30 million that are in this category, and there are EUR 60 million of late payment interest that has been already accrued, but there's not any single penny of the EUR 60 million the P&L. So this means that there are all money that will be collected when the payment will be made. And this under EUR 30 million are accruing an 8% implicit LPI. One of the things that to add to that extent, which I think that is very relevant, not for the size, but for understanding this concept of LPI. In the second quarter of this year, we have done the first ever transaction on selling only LPI related to city under distress. So basically, we have found an investor that was keen to buy part of this late payment interest that were connected only to city under distress. This means, in other words, that the EUR 60 million that doesn't appear in our P&L, that doesn't appear anywhere in our balance sheet have a market. And today, we know also the fact that we can eventually even sell this type of LPI. Clearly, we don't have any interest -- necessary interest to do it because we are very happy have an 8% yield associated to city. So we do not expect at the end to have a cost of risk that will increase in the second half of the year, but we stay within the 50 basis point that is in line with our plan. In terms of dividend, we didn't make any further announcement. So -- means that the general assembly of Banca Sistema have already approved the payment of the 2019 and 2020 dividend. As soon as the regulator will have eventually lift their restriction, this had happened last week. So unless there will be any request because under the statement from Bank of Italy is written that Bank of Italy will have a dialogue with the banks on an individual basis in light of the process of -- the annual process of SREP, which haven't started yet for us. If nothing change from October, we will be able to distribute the dividend without going through the general assembly because the general assembly have already approved it. In terms of the third question about the CQ margin, the effect of Lexitor. On one hand, I believe the situation of Lexitor has been cleared up, thanks to the [indiscernible]. So now it's clear that the fee connected to the upfront do not have to be returned to client. Let me say that for us, it was not big deal in a sense that Banca Sistema started to originate directly only 2 years ago. So we don't have much background or backlog of this type of request. In the portfolio that we have purchased from other seller, from other intermediary, the responsibility will have eventually end up to be in the end of the seller because clearly, we did not collect upfront fees, but were originally the originator that has collected that. But now I think that is good for them that they don't have to return this money and good for us in a certain way because clearly being the owner of the credit today, we have also been involved with many complaints from retail that we have to manage. So it was from an operational standpoint was quite time consuming. Going forward, I think that the market has moved more into a yield that include all the upfront cost. So our view is that we continue to maintain the split between the yield, the interest and the fees. Even though taking into consideration the prepayment, we will make an impairment on the amount of the upfront fees that based on our experience will have to return to individuals. So that means that this will be neutral for us because if as usually happen, a large percentage of those individual will renegotiate the CQ loan and we have to return the upfront fees. We have already put aside those fees for them to be returned. So I do not expect that will create anything, certainly enough from the past because we were not there in the past, but not also for the future. In general, I think that the -- we do expect a slight increase on the yield because even though that the change of law have resolved the problem of the past. On the other hand, that confirm the fact those fees have to be -- the upfront fees has to be returned to individuals going forward. And this means that there's an additional cost that has to be taken into consideration. And we do expect that the market will reply a little bit, not much, but maybe a few basis points just to take that into consideration.

Gianluca, very clear. Just a clarification on distressed conservatorship. We saw the loan loss provisions booked in first half. Looking at the moving parts at top line level, do you expect some rebound in the second half? So looking at last year, we expect total revenues to be more or less in line with last year, up or down? And if you see that there is a lack of revenues, are you prepared to sell some of the LPIs you've got in your book?

As I mentioned, we have a nice treasure of LPIs in our book, both in city that are now under distress as well as in city that are under distress. So we are able to use on an opportunistic way in order to make sure that at the year-end, we are able to generate revenue that we would like generate. So I do not rule out that we may sell some of the LPI just to be in line with our target.

The next question is from Manuela Meroni of Intesa Sanpaolo.

The first 1 is on the factoring. The outstanding declined by 3% quarter-on-quarter. Tax receivable are not growing. So I'm wondering if you can share with us the outlook that you have on factoring business, both in terms of volumes and margins going forward? And if you expect to see some increase in tax receivable within the end of the year. The second question is on cost. Cost declined a bit in the second quarter compared with the first quarter. So you can please comment on that. And if you -- if we can take the second quarter cost as a guidance for the next 2 quarters. The third question is on the cost of risk. Just first of all, I would like to know if the reclassification of municipalities in conservatorship had an impact on loan loss provisions in this quarter. And if we can assume that the cost of risk in the second half 2021 may be in line with the first half, excluding the nonrecurring items close to 46 basis point. Then on the new definition of default. I'm wondering if the process of -- concerning the new definition of default has been concluded or we may expect a further impact in the next quarters? And finally, on your last comment concerning the second part of 2021, you said that you expect an improvement in profitability. I'm wondering if you -- this improvement can be significant, so if the second part of the year can be significantly above the first part? And if you can please elaborate a little bit more on what are the levers to achieve this improvement?

Okay. Let me take some of the questions, maybe most of them. I will leave only the second one to Ilaria about the cost. Factoring outstanding tax receivable, yes, there is a decline of factoring outstanding even though there was significant growth on the turnover. The outstanding has been reduced because of tax receivable as well as because of some speed up of payment from public administration. And that was the cause of the reduction of the outstanding. Now we already have seen in -- even on the first months, so we are already at the end of July, some recovery on the tax receivables. So corporates are starting to probably increase the amount of invoice -- increase the invoice, then there will be tax credit to be sold. And tax credit has a duration that is longer than the commercial factoring. So in other words that means that the outstanding will go up if we buy assets with a longer duration. It's also fair to say that in the first part of the year and also at the end of last year because of the uncertainty on the new definition of default, we have been -- we stay a bit cautious on assets where we know that the duration and the payment was too long, just to avoid to have too much risk of been obliged to apply until 50% RWA. Now let answer to your third question, whether the new definition of default has been concluded. Let me say that it seems that it's a very complicated and very legal driven interpretation, counter interpretation and so on. As I said, we have received an inspection from Bank of Italy. Clearly, we had the possibility to exchange views also with the regulator. And so what we have applied now has been applied under the direct on-site supervision of the regulator. Whether this means that anything is concluded there? Let me say, I hope so. I hope that we have reached a good point of interpretation, whether there are still a point outstanding that could be subject to interpretation in favor or not in favor because I would say that it goes in both direction, and let me add more in the direction that could be in our favor because everything that was not our favor, probably has been already embedded in while the regulator were on site. So we think that probably we can have a situation on the following months that remain pretty much stable in terms of interpretation of the new definition of default. The cost of risk, I confirm what I also said before, we expect that in the second half of the year, we remain in -- we remain with the cost of risk of below the 50 basis point. And reclassification of the loan loss provision did not have an impact. What did have an impact was the time value that has been applied, but not because -- not due to the reclassification. For us, the reclassification do not have any impact. We were -- the credit were the same city under distress. The RWA was the same. So the only thing that has been changed was the expectation of collection that has been lengthened, and that means that creates more time value. As I said, if this will be true at the end because we will know only when we collect means that we will have more like an interest when we collect. So if it's true that it will take more time. It's also true that we will make more money. But unfortunately, this second component of LPI cannot be accrued in the balance sheet. In terms of profitability, I don't think that it's different from the past in a sense that the second part of the year has been always more profitable in terms of the P&L than the first part of the year. When I look at the result, taking away the one-off component as Ilaria said before, actually, the second Q was slightly better than the second Q of last year if you take out of this one-off component. As also before Christian Carrese said -- ask whether we could expect to eventually sell more receivable in order to achieve the target. I think that I hope that I was clear when I said that we have a treasure where we sit on. This is an amount of LPI that overall more than under EUR 50 million. And we can use them if we see that we may want to have a more revenue in order to be in line with our budgeted target. I will leave then to Ilaria to comment on the cost decline and whether there is a new level of cost is confirmed also in the second part of the year. Ilaria?

Yes, sure. Yes, the main difference in terms of cost between Q1 and Q2 was the contribution to the Single Resolution Fund. The contribution in the first quarter was higher, which was the ordinary component was higher than the contribution, the extraordinary component that affected the P&L in the second quarter. So really, the cost base related to Q2 can be excluding the contribution to the Single Resolution Fund can be a good estimate for the cost trend up to year-end. Let's remember that in the third quarter of the year, we would expect the contribution to the deposit scheme, which is in the region of EUR 2 million.

Just as a follow-up, usually in the third quarter of the year, you update your model for the calculation of LPIs. I'm wondering if you expect some positive -- some positive income coming from this process also in 2021.

I think that last year -- then I will leave to Ilaria maybe to answer. But last year, the adjustment was very minimal on the third quarter. And I think that this year, we have to do all the calculation, but if there will be any adjustment, I don't think that we could be more than what was last year, probably even less. So it could be marginal. But maybe Ilaria if you can further comment.

No. That's absolutely correct. We are not expecting a significant impact, neither positive nor negative. Let's remember that the adjustment both relates to the accrual percentage as well as the time value. As Gianluca said, already last year the impact was not relevant, not significant, mainly due to the fact that the accrual percentage have now reached sort of steady state compared to the past where we were actually making year-by-year progressive changes. So the accrual rate is more or less -- has more or less reached the level whereby no further changes to be expected in the future?

The next question is from Luigi Tramontana, Banca Akros.

On my side, first question is on the evolution of factoring margins. This decline in income generation in factoring, what is it due? Is it related to a different mix between NHS,, municipalities, tax receivables, football receivables, et cetera? Or is there any pricing pressure in the market? And so what do we have to expect going on? And the second is, again, on the cost of risk and provisions. Before the reclassification of the municipalities from UTP to bad loans, the coverage of the bad loans was over 50% and of the UTP was 15%. So what I understood is that this reclassification won't have any impact on the coverage of those positions. Is it right? And if so, why do we have to expect a cost of risk of over 40 bps in the full year, so still high also in the second half? Is it due to this impairment on CQ fees you were speaking about that you are going to take in the second half?

I will pick up the first question, and then I will leave Ilaria to maybe further comment on the cost of risk. If you go to the Page 6 of the presentation, to be hones, I don't see the pressure on margin because when you look at the adjusted income margin, we take into consideration the fees connected to the factoring, and the first half of 2020, we had 5.1%; on the first half of 2021 is 5.2%. What has changed was the outstanding, so the amount of profit, not the yield. And this is justified, as I said, with 2 main driver the outstanding. The tax receivable because they have a longer duration, and that means if we buy assets with shorter duration, clearly, we need to pick up outstanding in order to be able to generate the same in absolute term not in real-term amount of revenue. And the second was a speed up of some payment due to extra money that the government has put forward to some of the municipality and that allowed them speed up some payment. Without this means that everything is resolved and there's no problem. Absolutely, this is not the case, but they try to use the money as much as they can in order to give money to the real economy to the supplier. And to a certain extent, we are 1 of the suppliers because we buy from suppliers. And as you know, the public administration is obliged by law to pay in the credit in chronological term, so they cannot pay somebody that have invoiced them after the invoice that we have bought, but that's going chronological order. So the attitude to try to draw to put some additional money to the economy, that certainly has created some speed up of payment, which end up to reduce the outstanding. But if you look at this adjusted income margin, I don't see the yield pressure on the factoring product. Maybe, Ilaria, if you want to comment on cost of risk, even though let me only say 1 thing, that it is correct that the coverage is not going to change. So the sum of the 2 will not change. But maybe, Ilaria, if you want to comment more.

Yes. Sure. Yes, let's start from what the coverage rate was in Q1 before the change we made to the classification of the cities in conservatorship. The coverage ratio of the bad loans in Q1 was around 52%, 53% while the coverage ratio for the UTPs was around 15%, which means that the position versus cities in conservatorship had a lower coverage than clearly position that were already classified in bad loans. Now we moved these positions from UTPs to bad loans, but we didn't change at all the methodology -- our methodology in order to calculate the provisions related to those positions apart from the adjustment to the right time value, which Gianluca mentioned and which I mentioned as well, which was not related to the movement of these exposures from 1 bucket to another, but were simply related to a fine-tuning of our model -- underlying model suggested by the Bank of Italy. The result of that is that the amount of bad loans clearly increased by an equivalent amount of a decrease in UTPs. The positions, the exposures that were moved to bad loans didn't really have a changed coverage. And as a result of that, the coverage ratio of the bad loans, the average coverage ratio decreased because the result of the combination of a fewer position with a higher coverage rate with be -- much positions with a average coverage rate was such that average coverage rate for the category went down to around 27%. For the same rationale, the coverage ratio of the acquisitions that remained in UTP moved up again simply because of the maths, and it moved from around 15% to around 29%. Now going forward, provided that we are happy and satisfied with these levels of coverage, we are not expected to make any additional change at least given the current background to the coverage ratio. And as a result of that, the cost of risk in the second part of the year will readjust towards levels which are more in line with what we have seen in the past, which means around or slightly below 40 basis points.

[Operator Instructions] Mr. Garbi, there are no more questions registered at this time.

Okay. Thank you very much. Thank you also for the question. Let me wish to all of you some nice weekend, but also some happy holidays for the one that will take some day off, and we will meet the next conference call for the third quarter results. Thank you very much.

Ladies and gentlemen, thank you for joining. The conference is now over. You may disconnect your telephones. Thank you.