Svenska Handelsbanken AB (publ) (SHBA) Earnings Call Transcript & Summary

Good morning, and welcome to this conference call for the second quarter 2020. Together with me today are Carl Cederschiöld, CFO; Lars Höglund, Head of IR; and [ Annika Engler ], Head of Group Accounting. I will start the presentation on Slide #2. The second quarter has clearly been characterized by the ongoing pandemic. We have had a high activity in the bank to support our customers in different ways. We have also kept a high speed in our development of new tools to support both our customers and employees. And we have clearly seen results of these efforts. Swedish private customer ranks Handelsbanken highest of all banks in terms of how the pandemic has been handled, and customer satisfaction has further strengthened during the spring. We have also seen strong growth during the quarter. Both lending, deposits and mutual fund volumes have grown nicely. This is -- this, in my mind, is evidence of satisfied customers with a strong trust in the bank. During the quarter, 2 of our home markets entered into a really low rate environment, and that happened quickly. I will get back to that. But it has had a short-term impact on net interest income. We wanted to be well prepared to support our customers in the best way when the pandemic started. We secured early on an even stronger liquidity reserve. Given the turbulent start of the quarter, this came with an extra cost which is temporary in nature. The bank has also built further capital during the quarter. And at the same time, our credit quality is stronger than in a long time. Loan losses are at a very low level. I will get back to that. Looking at the second quarter result on Slide #5, we can see that it was stable. In Q1, we had some items directly related to the pandemic. It got an impact of over SEK 800 million, not least from the extra credit loss provisions we took at the outset of the pandemic. In Q2, revenues were down somewhat, which was expected given the turbulence in the markets and economies. Costs were stable, and loan losses were close to 0. Common equity tier 1 ratio strengthened by more than 1 percentage point. So all in all, a stable quarter even though some lines in the P&L moved quite a bit but yet in a way we could expect. On Slide #6, we took -- we take a look at net interest income. The strong volume growth added SEK 188 million in the quarter. We wanted to prepare for any size of growth, so we built further liquidity also for different scenarios of the crisis. That addition of liquidity came at a cost of SEK 236 million largely because of the volatile markets initially. For example, the bank opened up the European market for bond funding in early April. This extra cost will be much reduced in the third quarter and will be gone in the fourth quarter, all else equal. Norway and U.K. entered into a low rate environment very quickly. This had an immediate impact on deposit margins that fell by SEK 170 million. This change will not come back next quarter. Just like we have talked about during the quarter, one has to expect lending margins to come under pressure in the beginning of the crisis before new price levels have been established. The cost of that was around SEK 150 million. So out of the SEK 600 million drop in net interest income in the quarter, at least SEK 400 million is of a temporary nature. And again, given the big turbulence in the world during the spring, it is quite expected. Now turn to Slide #7, please. In most of our home markets, we have had a low and even negative rate for years. Now Norway and U.K. are also in a similar situation. Initially, as I mentioned, that comes with a cost. But over time and as you can see from the slide, there is a clear contribution from volume growth in the bank, not least in lending. Loan demand, of course, benefits from low rates for the customers. The latest quarter does look challenging, of course, on this slide, with a reduction. However, going forward, we have good reason to believe that volume growth will remain the big driver of net interest income rather than margins. Margins over time always move up and down. On Slide #8, we can see the total lending and deposit development in the bank. During March to May, the bank was the largest net lender to nonfinancial companies in Sweden with a share of over 25%. During the latest part of the quarter, we saw a declining need for new credits for the companies. But compared to 1 year ago, we have built a considerably larger corporate lending portfolio in all our home markets taken together. At the same time, corporate deposits have continued to grow. This underlines that the companies are strong, but they have also put a lot of trust in the stability of Handelsbanken. On mortgages, the stable growth has continued. We had expected a slowdown during the spring, but in most markets, it has been almost business as usual. We were largest on new mortgages lending in Sweden up until May but with a share below of our total market share. We have continued to launch new digital mortgage solutions for the customers and branches. These will both add value to the customers and make the work in branches more efficient to free up time for them. Household deposits have grown strongly during the quarter, and the earlier growth trend has strengthened further. And now to Slide #9, please. Fee income is down during the quarter, which was also expected. We had a strong start of the first quarter before the stock market crash and outflows in March. We see the impact here. During the second quarter, inflows have been strong. Stock markets were up, meaning that we are back on levels which, if they remain, bode well for this part of fee income going forward. Payments, roughly 15% of fee income during the quarter, have also been affected, as expected, by lower use of cards. The largest part of our card business is related to private cards, which have done better than corporate cards. We have also seen private cards developing in the right direction again during the later part of the quarter. We continued to develop our offering within the saving area, among others, and we can see that we have a nice potential still in our fee business after the natural drop we saw in the second quarter. On Slide #10, we take a closer look at the mutual fund business. Our strong position was further emphasized in the second quarter. In Sweden, we had 38% of net inflows in June. In spite of general outflows in March, all in all, we had higher net flows in all our home markets in the first half of 2020 than in 2019. Our unusually high advisory activity during the spring is an important explanation behind the good development as well as our continuous efforts within sustainable asset management. Now changing subject and moving to Slide #11. We continue on the path of change we started last year, making the bank more focused and with a less complex and more efficient structure. This means that we focus on our core customers and their needs. We develop our offering of the products they need and stop providing services that don't support the core customer business. It has essentially moved on according to plan during the pandemic even though some measures have taken a bit longer time due to closedowns. At the same time, our development, not least of new digital tools, has been done with a high tempo during the second quarter. We remain with our guidance of the total cost impact down SEK 1.5 billion, all else equal, by the end of 2020. Measures agreed and taken so far imply SEK 550 million in annual cost reductions. About half of that, SEK 70 million on a quarterly basis, can now be seen in our numbers. Our cost focus is strong and will remain so. We will get back to the positive position where Handelsbanken should be a truly cost-efficient bank, and this is definitely a step in that direction. And from here, I will hand over to you, Carl. Please continue.

Thank you, Carina. Please go to Slide 12, where you can see our cost development, and you've seen this quite a few times before. It shows that the underlying cost growth continues to abate, now around 2.5% year-on-year. Looking quarter-by-quarter, you can see that costs are down somewhat both compared to the second quarter last year and the first quarter this year. The impact of the cost reduction that I just talked or that Carina talked about can be seen in the post other expenses. We have decided to step up the speed even further within the entire financial crime prevention area. This is a key focus for Handelsbanken, and we will continuously adapt to the changing world around us. It is of utmost importance to us that we do all we can to protect the bank from being used from criminal purposes. We will therefore spend more this year than we have previously said, SEK 1.5 billion rather than SEK 1.2 billion. In the first half of 2020, we spent around SEK 660 million, so it means some increase during the second half. Slide #13, please. As I -- or as Carina said in the beginning, we have secured a really strong balance sheet throughout the ongoing crisis. Our common equity tier 1 ratio improved by more than 1 percentage point to 18.71%. 1871 happens to be the year when Handelsbanken was founded, and that's why we used the 2 decimals this quarter. This means that we are 170 basis points above our target range. This will not render any particular action from us now. We think it's prudent and very much Handelsbanken to be extra well capitalized in a crisis. On top of that, we know that the capital requirements will move up going forward, as we have talked about before. Next slide is #14 and then moving into loan losses. Our loan losses in the quarter ended up just below SEK 100 million. This may come as a surprise to some, having such a low level in the middle of a crisis. So again, I want to start to remind you of the EBA stress test. It's now updated with the new EBA Transparency exercise published in June. Just like before, Handelsbanken has the lowest share of problem loans of all European banks. We are clearly below even the other Nordic banks. This is a very strong starting point. Most of the times when there are no loan losses in the systems, you will not see this difference. It will not be transparent. But it becomes more transparent in more challenging times. Moving on to another familiar slide, please go to Slide 15. This picture shows our long-term loan loss history. We are always humble about loan losses especially in a crisis, but it's clear that over many decades, Handelsbanken has distinguished ourselves in crisis by having a stronger credit quality than other banks. We are running our bank locally focusing on clients with a strong repayment capacity regardless of sector. On top of that, we have a very high share of secured or collateralized lending, typically in properties with low loan to values. All of this should help in a crisis like the current one. Slide #16, please. You need to look 13 years back in time to find a lower loan loss number than the one we have posted in the second quarter. We had SEK 11 million of losses in Stage 3 in the quarter, by far, the lowest level since IFRS 9 was introduced in 2019. This quarter, we have changed the macro scenarios quite a bit. In short, the 2020 scenario is now forecasting a GDP downturn of close to 5% in Sweden with unemployment close to 10%. In Europe, the scenario is even worse. As you can -- as you know, the macro figures are used to calculate the expected credit losses in our entire portfolio. This adds SEK 200 million to the loan loss provisions. In the first quarter, we made an overlay to specifically capture the impact of the pandemic, resulting in an extra provision of SEK 440 million. In the second quarter, that provision has decreased somewhat, which adds positively to the loan loss line. This is in spite of the fact that we have extended the universe of vulnerable sectors that we stress on the back of COVID-19. Finally, we have some mix improvements and other things that reduced the need for provisions. And all in all, we end up with SEK 97 million in the quarter. On Slide #17, you can see the scope of exposures that we have looked to include in the COVID-19 stress overlay, all in all, SEK 109 billion out of total exposures of close to SEK 3,000 billion. We have looked at the individual exposures in this group and excluded SEK 23 billion on an individual basis from a credit risk assessment. But still, the list of vulnerable sectors is SEK 27 billion larger than in Q1. We've also manually migrated all exposures in the vulnerable sectors with the risk class normal or worse into Stage 2 from Stage 1. Further, we are including all our non-property SME exposures into the stress as well as non-mortgage household exposures. On Slide #18, the scope of our total overlay is summarized. In terms of sectors added, it's primarily retail space properties that are now included in the stress. In total, just over 7% of our exposures are now included in the manual COVID-19-related overlay. But again, the change of macro scenarios used for the entire portfolio means that the IFRS 9 model is capturing most of the pandemic impact, which reduced the need for manual overlay compared to the first quarter. On the next slide, #19, we'd like to offer you some insight into our property management exposure. During the first quarter, we got a lot of investment questions around this, so we think it's prudent to be even further transparent to you here. About half is lending to residential property companies, and half is commercial. As you know, we do our property lending locally in every branch. We look for good customers with strong repayment capacity. That's the first line of defense. We never lend only on the basis of the collateral. The cash flow must be strong. But then, of course, we want the collateral as well, and then loan to values matter. As you can see, LTVs in our portfolio are generally very low. Residential properties in Finland have a high degree of government guarantees behind the collateral, hence, the higher LTV. In our report today, we also published tables where you can see more granular what types of collateral is behind our property lending in the different home markets. You can see that at Page 43. With average LTVs of around 50% and an extremely high share below 75% as you can imagine, the collateral we have from our strong customers are also resilient even to sharp property price declines. The same is true for our mortgage lending. Further back in the presentation, you will find very strong stress test results on our mortgages used to issue covered bonds. Slide #32 really provides a lot of detailed information about the manual overlay. We talked a lot about this one in Q1, and it's now updated with the new stress test. I just want to emphasize that even with a highly stressed PD value, provisions remain limited. One important reason for that is the low loss given defaults in our portfolio, as can clearly be seen here. Again, a higher share of collateralized lending is one important explanation. So finally, on Slide 20, to summarize. We continue on the path of change in the bank that we started last year to make the bank more focused and less complex. Our profit is stable, and cost development continues in the right direction. During the second quarter, we had a very high activity with our customers and also in our digital development. We have proven that we are ready to support our customers throughout this evolving crisis. This has been well appreciated, and our customer satisfaction has strengthened even further. Business volume had a strong growth. To be well prepared in this crisis, we took a cost to further build up the liquidity reserve. A couple of our home markets have joined the group of countries with a really low rate environment. Initially, that came with a cost, but over time, we believe it's supportive for the business. Our capital position is strong, and loan losses are close to 0 in the quarter. And we look forward to an exciting fall with a positive view. Our biggest hope is, of course, that the pandemic will ease its grip on the world, but we are prepared for all sorts of scenarios. And with that, I thank you for -- everyone for listening, and we open up for questions. Thank you.

[Operator Instructions] Our first question comes from the line of Chris Hartley at Redburn.

The first one is on your reserving on that Slide 32 you talked about. I just want to ask specifically about the SME corporate line. So I think that you stressed the PD only by 10 basis points and where the stress seems to be larger than some of the other -- what you might talk about -- some of the other -- can you sort of tell us a bit more about what's in that book and why the stress is lower? And then second of all, just on the AML costs, please. Can you give us a bit more detail on why those expenses are increasing there? Is it something specifically found? And maybe how would you think about the level of those going forward?

Okay. Well, first of all, obviously, you see that we have stressed all the property lending you see above, defined as that we've moved in to the vulnerable sectors, and we define them and we're being transparent around them. So the other part of the SME business which are not included in vulnerable sectors are the ones we stress and make the line transparent on what the sums are. So in this sense, we're -- the limited or the ratio of the PD stress is obviously deemed to be lower on this part. And that's the reason why we stress it by just 10 basis on that line. And moving into the AML perspective then and financial crime prevention. We see 3 types of cost on that line. First of all, we're developing processes and systems for the KYC process. And that's something we invest in right now, and that's something which will prolong with us, that investment line, for a few years more. Second of all, we invest in systems and processes for transaction monitoring. I notice if the clients behave in a different fashion. That's also something we invest in and which will live with us for a few years as well. The third cost item on the AML or FCP line is rather to go through all the clients and create the correct legacy of customer information. And that's something which we have increased the pace on, and that's the reason we're increasing the pace now. We want to move fast forward there and have everything in check. If we -- we're not going to guide on 2021 cost yet, but it's obvious to us that IT investment costs will keep being elevated for a few more years. But gradually, the cost of updating the client information will abate with time.

Okay. And could I just maybe follow up on that? I thought you said regarding Oktogonen. So how -- to what degree is loan loss experience kind of incorporated into your thinking around Oktogonen? So say you have much better-than-expected loan losses but there is a bit of a slip on costs. Does -- is that still sort of -- is that good enough? Or sort of actually, the cost line, really what the focus is when it comes to thinking about how to -- or whether Oktogonen will get paid.

First of all, I think Oktogonen cost is obviously cost we like. So we want to move back as quick as possible and pay out Oktogonen. But obviously, it is a Board decision. And the conclusion earlier on has obviously been that we're not value generating enough. We believe that we will close this year with obviously ROE in a sense with the motivation of paying out the full Oktogonen. We will need to leave that question to the Board, and they will have to decide in the next springtime of if Oktogonen will be paid out or not. The reason why we haven't accrued an Oktogonen this quarter is the same as last one. We think it is not visible enough, the market circumstances, and we will wait. But if there's reason to -- we like to get back to pay Oktogonen, and it's a cost line we like in the bank. One of the few ones, I should say.

And our next question comes from the line of Robin Rane of Kepler Cheuvreux.

So how should we think about the credit losses from this point? So if we assume that the development of the economy is hypothetically exactly as you lined up in your base case, should we then expect that credit losses per quarter should go back to the level of where we saw past couple of years to 300 per quarter? That's my first question.

I don't think you can make that kind of judgment of our portfolio. Many of the questions around the credit losses in the last quarter and most likely this quarter as well is between understanding the way we do our credit decisions vis-à-vis the majority of the banks. And many of the banks do top-down ratios. And yes, I think the logic you're posing could hold true for a bank like that. If we go back to normal times, a top-down approach to credit lending should -- might go back to the normal ratios. We do bottom-up. We live -- we have local branches in the local community offering lending to the clients, which we think are good and that needs it. So we don't like credit losses at all. We like to approach 0 on that line. What the outcome will be is going to be fairly idiosyncratic, I think, in the future. So I don't think you can actually guide on the credit. We think we -- or at least history tells us we have a quite competitive credit loss line, but the future will have to tell.

Lars Höglund here. I'd like to add one point to Carl's point there. I mean if you look back even back to '08, '09, the loan losses we had at that time, you could clearly see that these were focused on a very small number of exposures, so idiosyncratic cases again. We had another case last year in Sweden. So -- but the portfolio, as such, has not really moved from the changing macro environment. So guiding on loan losses is something we don't do for that reason.

Thanks for that, Lars.

Okay. Sorry to dwell on the loan losses, but would this sort of bottom-up decentralized approach make you more -- do you think it would make you more reluctant to take upfront provisioning than peer banks, for example?

Could be. I don't know. Or if you work top-down, you might have less transparency on the individual exposures. What we've done is, first of all, we do IFRS 9 modeling with the new macro updates, then we do the COVID-19 overlay. And then we go through exposure by exposure, look at the individual levels and make a decision and see if that's a reasonable amount. If you do lending top-down into sectors with a diverse set and with no specific knowledge to the individual exposures, you might have more challenges with having a view on the individual credit. So I guess it's up to you to decide on...

But I think -- I mean, just to clarify also, I think the fact that all our credits are being monitored from our branches locally, of course, means that they are really close to the customer and can really take quick action if something starts to deteriorate. So I think it's rather the opposite. I mean from a bottom-up point of view, I would say we are quicker in taking action, including making provisions if needed.

Okay. And then on lending margins looking ahead, so we've seen various interbanking rates going down again this -- to the end of this quarter. So how do you see -- do you see easing on lending margins from this point, I mean, excluding the more temporary effect from funding, et cetera, but more looking to the competitive picture on, I guess, both household and corporate line?

Well, if we go back to Slide 8, obviously, there are multiple things happening beneath the surface, so I think we'll have to dig into it. Obviously, lending margins on the corporate side has been poor this quarter, and that's due to a lot of the volume is obviously RCFs and guarantees. And obviously, we -- and they are issued at prenegotiated levels. So this quarter has a poor credit margin on that part. That would most likely adjust going forward, obviously. Then you are going to mortgage lending, and we see obviously quite a fierce competitive pressure there. On the quarter, we see fairly little adjustment, a slight deterioration, but obviously, we see the new lending at lower levels than book average. In most of the crisis, you normally see credit margins picking up. I think we're much more humble this time because most of the banks will be in a good situation with strong balance sheet, and they want to be in the competition to win all the lending. So I think the competitive climate will be more fierce this time. Then perhaps it's worth to mention as well that looking into the interest rate net margins on this quarter, obviously, we -- if you look about the deposit side, obviously, that Norway and U.K. entering into more or less a 0 rate climate. That has -- we have noticed periods on roughly SEK 80 million on these parts. So SEK 80 million should adjust, and they should more have adjusted now or very soon into the third quarter. So there are multiple issues playing in this margin picture. But generally, we believe that will be -- there will be pressure still on margins, but we have a lot of short-term effects in this quarter which will abate.

Okay. And then just a quick follow-up on Oktogonen. You said that the final decision is normally taken during the spring. Given the uncertainties that we see now, should we expect that you don't make basically any allocations in the coming quarters? But if you take -- if you do an allocation, you'll do that in -- like in Q4?

I don't think you should expect anything on that line. If we think the picture is transparent and we think it's easy to judge what the consequences will be, we will take the cost then as soon as possible.

And our next question comes from the line of Magnus Andersson of ABG.

Just a very quick follow-up on the Oktogonen since it's quite important, just to make it crystal clear. The reason why you did not make a provision this quarter was, as I read you, the continued uncertainty and not related to the fact that your costs, all are equal -- all else equal, will be SEK 300 million higher than we thought earlier because of this AML -- new AML guidance. Is that correct?

Agree.

Yes. Okay. Good. Then on NII, I mean, you clearly state that the SEK 236 million will come back -- be back in Q4. And then I guess up to the SEK 400 million, you have the SEK 80 million you mentioned on lower deposit rates. And I guess that up to the SEK 400 million, the rest was due to lower -- to improved lending margins in home markets where you have the minus SEK 151 million. I was just, for modeling reason, wondering, you also write about one-off costs of the insurance guarantee of SEK 57 million. Is that just a Q2 thing? Or has the level increased there for -- so that we should put that into Q3 and Q4 as well?

No, it's a good question you're posting, Magnus, and I will have to be a bit more clear on this one. First of all, the temporary effects to the NII we see are FX of SEK 90 million. They are the government guarantee, as you're alluding to, which is SEK 60 million, and that's a one-off, just hit in the Q2 results this year. Then we have the notice period of SEK 80 million estimated. And then we have the liquidity reserves of roughly SEK 240 million. This accumulates to SEK 465 million, above SEK 400 million, which we rounded it off to, but SEK 465 million. So 3/4 of this NII drop, we believe, are temporary.

And also just related to that, when I look at your Note 17, we can see that the cash and balances with central banks, U.S. dollars went up to -- I mean it's usually been between SEK 50 billion and SEK 100 billion. It went up to SEK 350 billion or more than that in Q1, and it remains at SEK 300 billion. Is that the number that will come down during the second half or -- because I guess that's part of the problem?

Yes. I think it's -- I don't know if you're correct. It will most likely come down because the stress levels in the market has come down and then especially financial companies put less cash in our deposits. But they obviously -- and we don't make money on them, we place that cash on Fed. So that balance will come down. But it's -- but the NII, and that will obviously hit perhaps the average margins, but the temporary effect is rather to the answer I gave earlier on. And please, Lars?

Yes. Just to add, I mean, part of that increased balance with Fed was also financed not only through overnight deposits but clearly through the issuance that we did in early April, so 3-month paper typically. So -- and that obviously came with an elevated cost early in the quarter and, at the same time, receiving basically no return from the Fed. So you have a double whammy, you can say, lower return from the Fed but also temporarily higher funding cost for that position.

Okay. Good. And then just can you confirm that you haven't taken any formal government support during the quarter, i.e., from the TLTRO or the SEK 500 billion Riksbank facility?

Yes.

Okay. Do you think that will be a relevant argument or an important factor in the potential dividend discussions during the autumn?

I don't make any guidance on that or do any judgment on it. We will have to wait, and then we'll leave that to the Board and the AGM to decide.

And our next question comes from the line of Sofie Peterzens of JPMorgan.

Yes. Here is Sofie from JPMorgan. I was looking at your Stage 2 exposures in the report, and they went up quite materially. But then when I look at the provision done, they only went up around SEK 200 million versus increase in the Stage 2 exposures of almost 60 -- sorry, SEK 25 billion. Could you just explain why Stage 2 exposures went up but why provisions didn't really go up?

Sofie, it's Lars here. So basically, the reason -- the primary reason why you saw the increase in volume in Stage 2 was the fact that we decided to migrate the COVID exposures into Stage 2, the ones with a risk class normal or worse. We migrated them into Stage 2. So that obviously lifted the volume in that stage. If you look on -- well, either in the note in the report or on Slide 33, you can then clearly see also the impact of provisioning of that move, which was SEK 37 million. So -- but the main reason for the increase in volumes in Stage 2 was the move we decided to do.

But -- sorry, I'm just a little bit confused how you're going to have over SEK 20 billion increase in Stage 2 exposures, and you just take -- and you have macro declining by 5% versus your expectations in the first quarter, but you need the SEK 37 million increase in provisions. It just sounds very, very low.

The SEK 37 million increase is related to this move only. So -- but again, I think if you look on Slide 32, you get part of the answer to that question because we have also added the loss given default numbers for COVID-exposed sectors. And as Carl and Carina talked about earlier, we have a lot of secured lending. But certainly also in the other sectors that are exposed, you can see we have fairly low loss given defaults. So even if you make harsher assumptions on PDs, the provisioning in the model or in the overlay doesn't -- isn't that much impacted.

But the -- going back to probably of default, how do you arrive at some of these? I mean hotels, for example, 1.6% probability of default under stress when we know that most hotels not just in Europe but in Sweden are under immense pressure and loss given default of 21% for hotels.

Yes. Yes.

Are they based on historic numbers? Or how do you arrive at the probability of default?

Yes, so I think we have to leave time for other participants soon, but we can discuss this further in the afternoon, Sofie. But just to take that question straight away, I mean hotels, in our case, means a lot of hotel properties, as we talked about in Q1. Hence, the low loss given default that we have collateral in the hotel properties. But then, yes, I mean, looking at the PD values, they are based on our historical experience of PD values in stress in the different sectors. We can dwell more into this, if you want, in the afternoon bilaterally.

Okay. And just a final question. On Slide 34, you say that core equity Tier 1 improved by 50 basis points due to other. Could you just give details on what this other is, that healthier core equity Tier 1?

No. I mean it's -- when you look at the capital development, you do have a lot of different small impacts related to calibrations or models and other things. And over time, they typically sum up to 0. But this quarter, they happened to be a relatively large positive, but it's a sum of a lot of different things.

And our next question comes from the line of Johan Ekblom of UBS.

Sorry to dwell on the asset quality question, but I'm just trying to understand, with the macro assumptions you've done, the loan loss ratio year-to-date is not only below peers, but it's actually below your own historical average. So should we read that the book today is materially safer than it's ever been in the past? In that, I think you've annualized 5 basis points of loan losses year-to-date. And I think going back to 2000, the average is 6 basis points or so. And I guess your macro assumptions don't seem to be materially less severe than the peer group. And yet of the banks that have reported so far, we're looking at somewhere between 3 and 6x the average loan loss level over the same period. So how can the outcome be so different? Is this just the effect of IFRS 9 that models between banks differ so much? Or is there anything else we're missing here?

Well, it's a really good question. Thanks for that one. Yes, I think that this actually reflects a better asset quality over time. And I do think that most of all, we have improved the asset quality if you look back long term now. And as well, we have improved it even further with strength -- with moving the bank into the more focused banks. So yes, we do believe actually that we have better asset quality. And we obviously have lower absolute numbers now with a higher portfolio. But then again, I do think that we're obviously not used to making manual overlays when we do credit provisionings. And now with Brexit, we've done it, and now we've done it with COVID-19 as well. And that obviously puts a cushion on our IFRS 9 modeling, which should -- if history is a good guidance to the future, which should guide on expected credit losses quite nicely. And obviously, as we said before, we have low correlation between macro statistics and defaults. So we believe we have better asset quality now, yes. We believe that the more top-down approach you have, obviously, you're more likely perhaps to know less about your exposures and then move into even further provisioning. But that's for you to guide on. We're fairly -- we're good with the provisionings we've done now. We're -- we know the exposures on the individual level, and we're fine with these results.

And then just a final thing. On dividend decisions or thoughts, how do you think about countercyclical? I mean you mentioned that they might very well come back at some point. When you think about your capital planning, what kind of -- I mean, I guess the target, that includes no increase, but presumably it's part of your thought process when you think about capital planning in a slightly longer-term perspective.

Yes. We obviously want to be well capitalized for the future. And obviously, if we add back the countercyclical buffers, that would take the SREP up from 14 to ish-16 or 15.9-ish. If you then -- obviously, we got the SME support factor a bit ahead, and that's adding 0.5 as well. So adding these ones back, we are at roughly 16.5, and then we're mid in our target range. So I think that might give you some guidance to the way we think.

And then when you talked about increases in capital requirements going forward, apart from CRE in Sweden and Norway and Basel IV, is there anything else you have in mind?

No, I just think that there's a lot of uncertainty going forward. So we will have to wait for the autumn a bit and see what's going to happen with the crisis and with dividends and et cetera, and then we can plan even further.

And our next question comes from the line of Andreas Hakansson of Danske Bank.

Yes. We've gone from most areas so just a few follow-up. Quick one just on Oktogonen. Could you just confirm, is it so that if you decide that there's not going to be a dividend for 2019, that you can't pay any money to Oktogonen?

It is so, yes.

Okay. Then next point, sorry, coming back to the asset quality situation. Just so I get the numbers right, how big portion of the increase to SEK 25.5 billion of Stage 2 increase, how much of that was just the move in the COVID exposures?

We can come back with the number, but it was a very big part of it.

But still, your coverage ratio is now below 1%, which is significantly below, I think, all peers. You don't feel that you want to be on a conservative side and you rather see potential write-backs in the future? I mean I agree with some of the other analysts that it looks quite aggressive, don't you think?

Well, let me start and then, Lars, you can add. I mean being a bank which makes lending bottom-up, doing it locally with a good knowledge of the client, and then we provision in the way we do. I mean it's natural for us to try to guess the line as good as we can. And it might be -- and I'm not going to judge the way other banks does this, but we obviously want to have the correct credit provisioning every quarter, and that's what the ambition had been in this quarter as well.

So basically, if there would be another bank that says that to take more overlay in general provisions now, which they expect them to meet specific provisions in the future, you would rather not see the same way? And then shouldn't we really then expect that if there's an increase in actual losses, that you're going to see higher loan loss provisions in the future since you haven't taken it upfront?

But we obviously had added the COVID-19 overlay. So that is...

Sure. But it's not a huge overlay, I would say.

Well, how long is it...

We might not come to a conclusion. It's just quite striking numbers.

No, but -- and we agree with that one. Obviously, it is a low number compared -- on absolute levels compared to other banks. We're fine with this one. We think it reflects our credit quality.

Just also just to add...

And can I ask you one -- yes.

Yes, again, I mean, I think a big difference -- without judging any other banks at all, but a big difference between our portfolio and some other banks is simply the type of sectors we're exposed to and the share of portfolio exposed to certain obvious problem sectors. And also, that's very well reflected in our Stage 3 provisioning versus some of our peers.

Sure. No, I'm asking about Stage 2. But can I ask, since you have this decentralized model, and I know that each branch basis get its own P&L, and they've been measured on that and compare it every quarter between everyone. Do you think that there is an interest in this -- on the broker launch -- branch basis to keep provisions down as long as they can, hoping things are going to be fine? And if you would take more overlay provisions, is that allocated down to the branches or is that kept on a central level?

It is allocated down to the branch levels, and it is supervised by the group credit department. So they're obviously in charge of the credit process and the way we do the provisioning. And then it's definitely correct that it's done on a local level and all the credit losses are divided out there.

But I mean on the contrary to what you might be suggesting, Andreas, I mean the local branch has no incentive to try to hide things or delay things. On the contrary, they will start working with the customer, I would claim, earlier than the more centralized bank would to avoid problems. And I would argue that's one of the key reasons why our loan losses have been lower for a long, long time, that we take early action with the customer to simply avoid a loan loss from happening.

And our next question comes from the line of Nick Davey at Exane BNP Paribas.

So 3 questions, please. The first one, on SME lending rates, just wondered if you have seen any repricing going on in the SME segment in recent weeks. The second question would be about this funding position in the U.S. and the low-risk nature of your bank has come up a few times. I think you used Handelsbanken as an adjective earlier to mean low risk. So my question would be, given that this funding position in the U.S. is introducing more volatility into your P&L than the loan loss side, is this something that you're structurally comfortable with? Or do you think, over time, perhaps diversifying the funding sources makes a better sense? And then the third question is, thanks for providing in the report some details around repayment holidays given -- I think the number is now about 50,000 clients across the different geographies in sum. So I just don't have in mind a number of how many clients you have in these geographies. So could you just touch on these 50,000 repayment holidays in any country? Is that a more meaningful portion of the book? And I just wondered if it sort of squares away with the 2 bps of loan losses at the group level. So just any more detail would be helpful.

Let me address the 2 last questions, and then I'll leave the first one to Lars. But then again, going back to the financing strategy we have, obviously, what we do is we are a risk prudent bank. We believe in having really good liquidity reserves. We believe in matching the lending with the financing structure. So obviously, it's not that it is a U.S. question. During April and May, to some extent, the funding costs were elevated at perhaps roughly 1 percentage point higher than pre-corona levels wherever you were in the world. We fund ourselves in dollar to some extent all the time. We haven't changed that one, and we don't believe we need to change that either. But the similar numbers would have happened whatever currency we would have funded ourselves in. And obviously, since then, levels have dropped down and now are comparable to pre-corona levels. So the extra funding cost of SEK 236 million will abate over the next and slightly into Q4, and that will drop off. The third question was, yes, the repayment and the amortization holidays. We are 33,000 in Sweden out of these figures, which are then obviously -- and in Sweden, we can't see any reason at all that that's because clients are strained, rather it is just when you leave the offer on the table to skip the amortization and to do other things, 33,000 of our clients have chosen to do so. In the other home markets, the range goes between 400 up to 1,500, I think, so a very, very minor part of the client segments, all in all.

Yes, even lower than that in the Netherlands, around 30 customers, but still a very low share of the portfolio. Yes. Then on SME lending repricing, I think in general, Nick, you can say that, I mean, initially in the quarter, the corporate lending pricing was a bit all over the place. And I think it's gradually sort of stabilizing and, of course, finding new levels, reflecting the new market conditions, so to speak. But I wouldn't describe it as any significant repricing but more adjusting to new -- the new normal.

Our next question comes from the line of Antonio Reale of Morgan Stanley.

It's Antonio here from Morgan Stanley. I want to ask you about asset quality. I think that we've discussed enough. Two quick, one of which is a follow up on NII. It was a bit surprising, to be honest, to see such a big drop in the quarter. I understand the drivers, and you mentioned that some of this may reverse later in the year. So my follow-up is really, leaving aside the government fees and the increase in liquidity reserves, which you've already discussed, out of the other 2 drivers, how much do you expect to recover? I'm thinking about the lending margins, which I think you quantified in the SEK 151 million and in the rate cut in Norway and the U.K., which I think was SEK 170 million. So how much of those 2 would you expect to be able to reverse later in the year? And my second question is actually linked to some extent, is more focused on the U.K. Could you just speak to what your expectations are in terms of net interest margins and loan loss provisions in the division going forward, please?

It's Lars here. Starting with the first one, out of the SEK 170 million for lend -- sorry, for deposit margins in Norway and U.K., as Carl said, around SEK 80 million of that is related to the notice period. So that should be clearly temporary. Then in terms of lending margins, we don't guide for that. We don't slice out how much of the SEK 151 million was temporary. All we say is that it was an unusually big impact on the lending margins in the quarter, which is quite natural given the big turbulence we had in the beginning of the quarter. So you will have to make your own assessment where that will go from here. It's not representing a structural change in margins in that sense. It is quite a bit temporary. But again, we don't guide on -- we don't know really how that will play out going forward. Then on the U.K. NIM, again, I mean, we don't give any guidance as such. But clearly, we have seen for a long time in the U.K. a big pressure on mortgage margins, clearly. Now in this quarter, the mortgage growth was basically 0. So let's see what happens going forward. But we definitely come from a couple of years with quite some pressure on mortgages in the U.K.

And perhaps if I can add, Lars, to the deposit margins, obviously, when rates drop down to 0, margins might be lower -- adjust to a lower level. But as Carina was pointing out at Slide #7, if we look back in the history of our other home markets, which entered a 0 rate policy, we've actually -- our experience is that we had increasing volumes of the lending business, and volumes are the only decisive factor actually in the NIM figure going -- or in the net interest income figures going forward. So margins tend to come and go, but volumes tend to be decisive of the net interest income. So future will have to show what the outcome will be, but our history says so.

Our next question comes from the line of Riccardo Rovere of Mediobanca.

Sorry to get back to asset quality one second. One -- a couple, if I may. First of all, I don't understand, when I look at Slide 17, if I understand it correctly, your total credit exposure is SEK 3 trillion. Take out SEK 700 billion sovereigns and institutions, you are left with SEK 2.3 trillion. And if I understand it correctly, the PD stress has been conducted, based on Slide 32, on 220 -- kind of SEK 220 billion. Sorry, if I understand it correctly, it's like saying that we use only kind of 10% of the credit exposure is vulnerable to COVID-19, and the rest is not. And I don't understand. First of all, if I understand it correctly. And first, if I understand -- second, if I understand it correctly, why should it be the case, and 90% of your book should be immune to everything that is happening? This is the first question. The second question I have is, when I look at, again, Slide 32, has LGD been stressed in your COVID overlay? Or it's just the PD that has been stressed? And the final question I have, sorry to get back, but when I look at Slide 33, the SEK 21 million, so everyone has downgraded GDP expectations, starting with the IMF, the European Commission, the ECB, everyone. Now just to add in more challenging numbers into your models should give you a negative number because the quality of the book cannot change in 3 months. The book, your large book, cannot be changed in only 3 months. So I don't understand why, from a statistical standpoint, plugging, I would imagine -- or even in banking numbers, and then time will tell whether those internal models will be right or wrong. But I don't understand how can it be possible when the GDP is downgraded globally, that you get a positive number. Just logically, I don't understand it. Sorry for this, but just to understand because the feeling I have is that the way you calculate your credit losses is closer to incurred losses. Even that you see no defaults out there, which is probably the reality, you charge less because there is -- because in the real world, nothing is happening or not much is happening. Just -- sorry for this, but just to understand the way you're thinking.

Sorry, Riccardo, we're running out of time, so let me try to answer your questions. So first of all, 90% of the portfolio, as I say, not stressed, well, I disagree because we have changed the macro scenarios quite a lot as we disclosed. And that is obviously impacting expected credit losses in the entire portfolio. And the reason we do this is obviously COVID-19. And then your question on whether the LGD is stressed, no, it's the PD which is stressed. And this is a 1-year stress we do. So LGDs remain the same. And then I'm not really sure I got everything right on your last question, but you were sort of questioning the fact that we had SEK 154 million positive from the impact from change in the portfolio. And yes, I mean, the portfolio has changed since Q1. Our portfolio keeps changing and, I would say, improving every quarter. And you can see that on migrations, we take out volumes that have a higher risk, and we take in new volumes with a lower risk. And on top of that, this quarter, you also had shorter maturity on some of the exposures. So that gives you that number.

No, no, the question was mostly on how can I get to SEK 21 million when everyone has downgraded GDP estimates globally.

Yes, sorry, let me take that quickly. We have a slightly lower overlay this quarter, SEK 21 million, as you say, simply because we have updated the macro scenarios. So the entire portfolio is now capturing better the impact of the COVID pandemic than it did in Q1. So that's why on the margin, the overlay has decreased somewhat. And Riccardo, more than happy to discuss this more bilaterally, but I think we have one more question on the list. Let's try to take that quickly as well.

Yes, yes, yes, makes sense.

The next question and the final question comes from the line of Martin Leitgeb of Goldman Sachs.

Just a couple of follow-up questions, please. And the first one, you mentioned competition and the expectation of competition to remain intense. Could you just differentiate between mortgages and corporates, how you see competition evolving in those 2 key product areas? And then secondly, in terms of capital and capital headwinds, just looking at your strong capital print this quarter up 110 basis points, you continue to provision for a 40% payout ratio in terms of the 2020 dividend. Could you just remind us what the kind of the capital headwinds you're expecting either this year or over the medium term for the group? And finally, in terms of dividend and potential dividend resumption, I was just wondering, is this essentially mainly a question of having more certainty on the outlook and then to recommence with similar dividend mechanics as we had them before, meaning annual dividend? So do you also see scope that the dividend approach might change in the future, if uncertainty were to prevail, maybe going to quarterly, et cetera?

Let me start on the capital side, and then I'll leave over to you, Lars. Well, firstly, as you say, yes, we're 470 basis points above our target. We like to -- as we said the last quarter as well, we like to plan our capital situation based on a normal situation. Right now, we have an un-normal situation with the countercyclical buffers being removed. But we plan for them to get back sooner or later. We know that the CRE floors will add -- in Norway will add 0.3. We have a lot of uncertainty with dividends. We obviously have an uncertain market level. So we don't see this as anything which will change our strategy right now. What we pointed out with the 40% accrual in Q1 was that, that was -- should just be seen -- that shouldn't be seen as a long-term guidance on the dividend level we make. It should just be seen as we -- the history and the way you do accrual should have implied 75%...

Yes.

75% accrual ratio going forward. And that we thought was not a fair assumption being a bank with a growth ambition. So that's the reason for accruing 40%. And we -- it's so far too early to tell what we will do in the future.

And then final comment perhaps on competition. I mean as Carl said, we can clearly see that banks are in good shape. And we also know that we -- and we have proven again, I think, that we have very good customers. And I think all banks want to bank with good customers. So I think -- I mean we can expect competition to remain tough both within the mortgage side and on the corporate side for the good customers.

Okay. Thank you all for this, taking your time to listen to us. All the IR people are happy to get back to the questions you want to. And otherwise, we will meet some of you during the coming days on the road show. Thanks all for listening.