Swissquote Group Holding SA (SQN) Earnings Call Transcript & Summary

Ladies and gentlemen, good morning. Thank you for joining us to our press conference for our first half results 2026. We are here at our headquarter in Blanc, and I am with our CFO, Yvon Cardenas, and we will go through the presentation. I hope you had a chance of downloading our PowerPoint presentation. Otherwise, you can follow this here on the screen. We will go through the slides and then at the end, we'll have time for some questions. I will immediately start with the first slide. We had a good half year in terms of growth. The client assets are at the absolute record level. We're now close to CHF 100 billion of assets at EUR 96.3 billion. That's a plus of 19.8% compared to 1 year ago. And also in terms of net new money, we had -- it's almost one of the best half year we ever had, purely organic growth of assets at CHF 5.1 billion. So in total, we now have more than 1.2 billion and the revenues we generated with our clients amounted to CHF 364.2 million of net revenues. That's also a plus of 1.7% compared with the figures 1 year ago and a good pretax profit of 50.2%. So on the next slide, the net new money. I said we had sorry, we had CHF 5.1 billion of net new money in the first half. And if you see here on the left, that's almost one of our best results. I think it's the second best. It was only surpassed by H1 2025, where we had CHF 5.2 billion. So the growth story continues. very good results in terms of number of accounts. We added 64,000 new accounts in the first half 2026 for Swissquote and 24,000 new accounts for you. And you see here the final results at CHF 1.2 million. You also see the difference between classic trading accounts with Swissquote, where the average asset is at CHF 115,000 compared to you, where logically, the asset is much smaller. It's a neobank. So people have less assets on the account, but still CHF 9,500, that's probably if you compare this with our international peers, that's much higher than what you can find at other neobanks. So a nice growth of 6% from -- on number of clients from H2 2025 to H1 2026. In terms of net new money, this is famous CHF 5.1 billion. I was mentioning before. So we have our 3 hotspots where the main growth is coming from. So Switzerland is still our most important place where we are CHF 2.7 billion of net new money in the first half and Europe with CHF 1.7 billion. For Europe, we wanted to show you where actually the growth is coming from. So it's mainly Benelux, France and Germany, up to 74%. This is where we are concentrating our biggest sales and marketing efforts, and we had quite a success in those countries. Then Middle East, Asia, this is mainly our subsidiaries in Dubai with CHF 0.5 billion, CHF 587 million. That's also a very good number compared to previous to previous half year. And the rest of world is a little bit more modest, but it's more driven by some selective outflows organized by our company. We are a little bit more selective on our clients in terms of revenue potential than before. And from time to time, we do close a certain number of client relations. So CHF 5.1 billion, again, it's really a very good number. And you see the global dispatch here, we have 54% coming from Switzerland and 46% coming from our international operations. And also in terms of distribution, 54% is B2C. That's still our main driver and the rest is B2B or B2B2C is -- so a very good distribution in terms of net new money and also something that makes us really positive over the longer term in terms of our growth engine is thriving. And this is also the reason why we do not change our forecast for 2028. We still think that we will make CHF 0.5 billion of profit in 2028 and with about CHF 900 million of revenue. And this is purely based on asset growth forecast over the next year. With CHF 5.1 billion, we are in advance. Our average forecast that we use to forecast our future numbers is a growth of CHF 7 billion a year. And you see here that with CHF 5.1 billion, we are really ahead of our plans. Client assets here has now reached CHF 96.3 billion. Of course, it's the highest number ever. It has been a little bit helped with good market. And as usual, we have 15% of those CHF 96.3 billion that are deposited in cash. It's also the one element which is pushing us towards a category 3 bank in Switzerland. We are not far away. The one element that we are measured on to become a Category 3 bank is the total balance sheet, and we are at CHF 16.9 billion and the trigger is CHF 17 billion. So we are really very close to becoming a Category 3 bank. Here, the average assets has slowed down a little bit. This is not last due to lower volatility. So it's a weak it's a very short-term impact, and it will recover in the coming half year. The distribution of revenue shows a little bit what happened in H1. So we had good growth in fee and commission income. We had good growth on interest and stable revenue for eForEx and trading, but what really was missing in the first half was the revenue on crypto assets. There, we had -- we were -- in our budget, we estimated that in 2026, we will do CHF 85 million compared with what happens over the last 2 years, but we only did CHF 14 million in the first half, and this really is due to a very low volatility. If you follow a little bit the what's happening on cryptocurrencies and the biggest one of it, Bitcoin, it almost didn't change. It stayed around $63,000 to $65,000 for Bitcoin and very stable, very low volatility. And this, of course, is not favorable to trading with such a low volatility. So it's a little bit difficult to forecast the Bitcoin volatility over the next year. What we did for the second half 2026, we stayed a little bit conservative to say, okay, let's just double everything and more or less, and this is the new guidance for 2028. Now of course, if volatility would come back on our cryptocurrency, then this would completely change the figure. But Again, we rather stay conservative there, and we estimate that the second half will be like the first half. Next slide is the net revenue by customer profile. Here on the left side, you really see the domicile where we are making our money. It's still a Swiss business to 56%, but Europe is now picking up. It represents already 25% of our revenue. This is the place where we have a strong growth with our bank in Luxembourg. And it's also something where we do invest a lot. We invest in staff and people and systems in Luxembourg because we think we have a very good product that is appealing to European mass affluent market, and this is our strategy for the coming months and years is to continue pushing our brand and our products, and we know and we can see that we are successful there. Then number three is Middle East. That's mainly our operations in Dubai that represents 12% of our total revenue. Now by customer type, this is on the right side. And this also is a very stable figure. It's about 2/3 is the B2C business, that's still our main business dealing with direct clients and then 28% that's the institutional business or the B2B2C business. And this is a figure that is very stable and hasn't changed a lot over the last years. A look on the revenue by asset class. It's -- on the left side, you see that it's a well-diversified picture. Now the crypto assets here, as mentioned before, represents only 4% of our revenue. We estimated that it would represent about 10%. So it's not a crypto story anymore. So it can only be good news if suddenly the crypto revenue or the crypto stock would gain in volatility, then you would certainly see this 4% growing again. But again, that's difficult to forecast, and let's see what the crypto market will do in the second half. Then the revenues by type of business, and we distinguish between transaction-based and non-transaction-based business. This is -- again, CHF 364.2 million revenue for the first half and about half of it, a little bit more is transaction-based, 52%. But the non-transaction based, which is the interest income, custody fees, the securities lending business this represents 48%. So a good development here, and this is also something we want to keep and even push a little bit the nontransaction-based revenue in the future. One of the initiatives we have that is working very well is the securities lending business that is now picking up. We have a very good solution, very attractive to our clients as we are sharing the revenue with our clients on a very good model. They get 50%, we get 50% and this is a very secure model that we have organized over the last 2 years. The headcount in -- as of 30 June 2026. Now we have in total 1,511 full-time equivalent. This is a distribution here on the left, you see it's mainly technology driven. So we are a tech bank, and we are proud to be a technology bank. That's our DNA. That's where we make our difference with all the banks in Switzerland and in Europe, 35%. And then sales -- the sales part represent 18% and 15% in our foreign offices. And then -- and this is a place where we have invested a lot in the past years and mainly to get ready for this Category 3 bank status that we now have or soon have. This is now 15% of our headcount is in risk and compliance. That's also quite normal for a bank -- for a bank of our size. Now if you look at the variation in headcount compared to the -- compared to December 2025, you see that the biggest part, 34 of the new headcounts we hired is in our growth initiatives, so mainly technology we hired 17 people. The growth of our foreign office, mainly in Europe, also pushed the headcount there at 15 and then a small growth on our new staff with 2 additional people. And that gives us a total of 1,482. And then the others, as mentioned before, this is growth due to our bigger status as being a bigger bank. And there, we hired 29 people. And in total, we have 1,1 -- so what's a little bit the forecast. We think it will slow down in the future. We have prefinanced the biggest part of our readiness to be a Category 3 bank. That should be okay now. And we also think that AI will bring a lot of additional efficiency. So we do not forecast to have the same rate of headcount growth over the next year until 2028. That's the strategy of our company. We think that we now have a good size to be able to manage the mission and the growth pattern of our bank. About the profitability, if we start from the CHF 96.3 billion of client assets, on average, generated 79 basis points of revenue and this is is CHF 364.2 million of revenue we generated and then 40 basis points on the CHF 96.3 billion is the pretax profit margin. And you see here on the right, the evolution of our profit. You see, first of all, that it did grow and it's now stabilized a little bit here. It's CHF 153.6 million net profit compared to H2 2025. But there in H2 2025, remember, we had this special one-off that was linked to the acquisition of the 50% of you from our joint venture partner, and that triggered this exceptional one-off in 2025. So we should rather compare our numbers with H1 2025. And then you can see here that we are stable at a high level, which I think is a good performance given the current economic situation, the uncertainty and also the special situation with cryptocurrency. For us, this is a good sign. So we are able to generate high profit even in difficult market conditions. But again, more important for us really is our growth engine. We want to grow the assets. We want to grow the net new money half year after half year. And of course, also, we want to get -- to have new good clients who are joining our system, and this has been achieved in the first half of 2026. Now about you, is our neobank that is now owned 100% by the group with 423,000 clients. We are the biggest neobank regulated in Switzerland. We still have a target of about 0.5 million clients by the end of 2026. So we will accelerate the growth a little bit. And for that, we have signed a contract with Young Boys. It's very important to us because it's -- football is a good brand carrier. And we think that goes very well with one of the most important and largest football club in Switzerland. And you see here a few pictures of our sponsorship agreement with Young Boys, and we're very excited, and we think this will be a great way of pushing the brand in Switzerland. we're not sure whether we have to because the growth is coming by itself. But I said in the second half of 2026, we want to accelerate the growth and go beyond the 6% growth we had here from H2 2025 to H2 2026. Very important for us also is that we're not only getting clients, but we have clients that are bringing their assets and they're also investing in securities. And out of the CHF 4 billion, you can see here on the left that about half of this 54% is in cash. That's a very different picture from the one I showed you before where the cash part represented only 15% overall for Swissquote. You, of course, it's a payment app, so the cash percentage is much higher, 54%. And it's mainly in Swiss francs. Of course, we are based in Switzerland, and most of our clients are Swiss, so up to 85%, sorry, in Swiss francs, the rest 12% in euro and 3% only in U.S. dollar. But on the right side, investments in securities represent 46%. So usually, people do not transfer assets -- their securities directly to you, they transfer the cash and then they invest in securities. And here, you see the very nice evolution over time. At CHF 4 billion, it's a very good number divided by the number of clients we have. This is this 9,500 average deposits per client, which if you compare this with M26, for example, you will see that this is much higher. So we have richer you clients than the rest of Europe. And this is not surprising even the possibility and what we can do with our account. The headcount on the left, speaking about you is that we have 64 people. They're now mainly located in Zurich, where we have our headquarter of you. You see that the split is between customer care, that's the biggest part, 60% and then products and marketing is 32%. All the operations and bank accounts are outsourced to Swisscote Bank. You is not the bank. It's a neobank, but the banking activity as such is performed by Swissce. Now the things that we are very proud of is our AI agent. We call it Yuhlia, and it's really working. It's actively used by our clients. And you see here on the right side, the number of conversations that are now taking place with Yuhlia. And it's a very good figure for us because, of course, whenever you do a conversation with Yuhlia, you may not call the call center because with Yuhlia, you can, of course, ask questions about the performance, about your portfolio, but you can also ask questions about general questions, customer support questions. And this, of course, is one call less in our center each time our clients ask the question on. It's the version 1. We will strongly develop our AI capacities there because, of course, we think this is the future, and we are just at the beginning of what AI can bring to such an ecosystem. Now we have a few slides on our balance sheet. And for that, I will give over to our CFO, who is sitting next to me, Yvon?

Yes. So a few words about the balance sheet. So the balance sheet has continued to grow. In the first half, we can see that there was a 5% increase in the total balance sheet. So we could increase the cash deposits of customers in all currencies, which is positive, in particular, as interest rate expectations have changed as well during the period of time. I will highlight 2 particular categories, loans that are mainly Lombard loans. We have a plus 12% increase in Lombard loans. So we see that the appetite of customers to invest this year and a 12% growth in the 6 months, I think it's a positive development. We have continued to increase the investment securities portfolio. So we use part of the balance sheet to lock duration. We could find nice opportunities during H1 as interest rate expectations have been changing from time to time. So I think we've been good in selecting the right timing and the right time of opportunities. As Marc mentioned, interest income is over initial guidance and we will most likely have a higher interest income in 2026 compared to 2025, which is as well a positive development. Last comment on the liabilities. You see the structured product business that we have. So we as well issue our own structured product business, our own structured products. And you see that there is a 21% increase. So it's growing. We could see before that structured products represent something like 5% of net revenues. These are popular products in Switzerland and while the volume of issuances is growing positively. Total assets, we are at CHF 16.9 billion at the end of June. So very close to the limit to be classified as a Category 3 bank. You can imagine that the 17 have been most likely crossed in July when you look at the growth of the balance sheet in H1. So most likely, as I mentioned in the press release, we should be classified in H2 as a Category 3 bank. If we move to the next slide, this is the usual slide that we propose on every conference. Again, volumes are higher than initially expected. Interest rates have developed better than initially expected. Swiss franc remained flat. We think that Swiss franc will remain flat in H2. Interest rates in USD have not decreased. That was what the initial guidance expected. They remain relatively stable, and we had a hike on euro interest rates. So the positive aspect on the interest income is that interest income is likely to be higher in 2026 compared to 2025. And we will enter 2027 most likely with higher rates than we entered 2026, which was something that we were not expecting. At the same time, the margin lending portfolio and the investment securities have continued to grow. So we could as well secure interest income for the future.

Okay. Thank you, last slide.

Yes. So on the equity side, I would say no major change. Capital ratio remains solid and relatively stable. You can see that the equity has not grown much. This is mainly explained by 2 items, while the dividend payment that we paid during H1 in accordance with our dividend policy. And as well, you can see that we slightly increased the portfolio of treasury shares that we have. We've invested something like CHF 55 million to CHF 60 million in treasury shares in H1. We are now above 3% of the share capital. So overall, the equity remained solid, stable and no particular change in the capital ratio.

Okay. Thank you. I'm coming back to comment a little bit our investment in AI. So as we mentioned many times, we are -- we think that AI will change the way we do banking. We also think that we are better prepared than other banks because we have developed sovereign infrastructure at Swissquote. We have invested in technology. We have invested in systems. We have invested in people. I should give a number over the last 18 months, we have probably invested about CHF 30 million in our AI capacity with staff and CapEx. So what are we doing with those investments? We have the 4 pillars that you can see here on this slide. The first one, which is kind of obvious, this is the client interaction automatization. So I already mentioned urea before that, but we also have chatbot developments for Swissquote. So when you want to go in contact with our agents, you have the possibility to address the issue you may have with our chatbot. So it's an AI chatbot. It's very different from the previous chatbots you could see. It's working well. We have above 70% of the feedback do rate the conversation they have with the chatbot as good. So meaning that the issue is solved. And we are, of course, working hard to develop the Phase 2 of our chatbot. This is due to be delivered in December 2026. Then internally, AI-assisted software development. So it first shows that we have a potential of 25% to 30% reduction in development cycle time. So that is -- it's very important for us. We are pushing our engineers to use AI as a tool. It's not automatic. You have to first have to organize your network, you have to organize the software. You have to to acquire a license and then you have to train your people to use AI capacity, but we really think this is the future. We also think that we will reach full AI maturity in our development staff by 2027, 2028. By then, we think that AI has the capacity of doubling or tripling the productivity. This means that we will be able to bring new softwares and solution to work faster to the market. It's not free of charge. It's really a dedicated investment in people, technology, infrastructure, and we think that we are very well prepared and organized for this AI change that is affecting -- that will affect, as you know, many industries in many different sectors. Then one very important element for us is the -- we call it payment intelligence. It's related to the monitoring of everything that is happening on our accounts. So it's -- we now have 100% AI coverage. So each time you do a payment or you do a transaction, we have the capacity of monitoring these transactions using AI tools. So we have developed many tools to do this, and there's one specific development we call it Diego internally. We think that the AI agent will be a breakthrough software solution to secure the transactions. You are aware that -- we are in a market that is challenged by fraudulent activities, account takeovers, new accounts and banks like us do need to invest in technology to secure completely this part of their business. And they also have tools that are very advanced and using frontier technologies to secure our banking activity. And then last but not least, for the products. So these are the product-facing tools. If you are a client of Swissquote, you have certainly noticed a lot of changes. We have developed wets that do analyze AI, do analyze stocks with AI capacities, and we have developed many, many different tools -- and this is there, we expect a lot of new things to come and tools that are very advanced for our clients. Now revised guidance 2026. So again, we haven't been very creative to have a full guidance for 2026. We basically took what happened in H1 and we doubled it, and this is the results for the full year. Now the latest guidance, CHF 730 million and CHF 365 million pretax profit margin for 2026, a little bit on the conservative side, I have to say, especially given the lackluster trading activity in crypto, but at least this is something we know that we can achieve in 2026. And so we'd rather stay there a little bit conservative. So I have mentioned it before. We haven't changed anything on our guidance 2028. We still think that we can do CHF 0.5 billion pretax profit in 2028. We are reassured by our growth pattern that is unchanged in the first half of 2026. So we're quite confident that we can achieve revenues of CHF 900 million with a pretax of CHF 500 million in 2028. A look on the margin on assets here, not much to comment here on this slide. You've certainly seen it and also are classic in our presentation in 2026. Maybe, do you want to say something on that? No, yes, it's more to basically help to understand how we see the distribution of the net revenues. And now with the revised guidance. Obviously, I think if we will compare as I mentioned, I'm not sure whether you could hear me, is that the revised guidance for 2026 is based on doubling the H1 2026 figures. So we haven't been very creative there. We just have taken our numbers and we multiply it by 2. It's a little bit on the conservative side, especially if you think that, for example, cryptocurrency could revive a little bit in the second half, but let's stay conservative. We know that we can achieve CHF 730 million of revenue in 2026 and CHF 365 million of pretax. That would be about the figures we achieved in 2025. if we take away the one-off on the profit side we had in 2025. And here, Yvon, maybe comments on marginal assets.

Yes. So here, you have a bit of details to how compute the net revenue distribution that we have now included in the revised guidance 2026. So here, you have a comparison between the new guidance and the previous years. I think what I could comment is the changes between the initial guidance and the revised guidance. So you see that on interest income side of things, the situation is better than initially expected. We were expecting a decline in interest income in 2026 compared to 2025. Now we expect the interest income to be higher in 2026. So there is a positive development on interest income. On the crypto assets, as many times commented during the conference, the environment was weaker than expected. So there is a change compared to the initial guidance. We think there could be a recovery somewhere in 2026. But we, for the time being, do not rely much on it. And any recovery will probably not be early enough to compensate the delay we have on the crypto asset income. One last comment is about client assets. So having in mind the target of CHF 7 billion per year that we have, it means that with the current level of client assets, we could be very close to CHF 100 billion in 2026, which is significantly ahead of what we expected. So on client assets, the positive aspect is we might be significantly ahead of plan at the end of 2026.

Okay. Thank you, Yvon. And I will just comment a few slides in the appendix. The first -- this is something we have a very precise look. We want to know if the clients we hire, the new clients joining the system, if they are equally as profitable as the old clients. And you can see this here. So among the 797,000 Swiss accounts, we had a growth of 7% in the first half and also 7% on the new accounts. Now this 7% transform into revenues. And you see here that with 7%, we made 4% of revenues. Why not 7%? Well, this is a normal distribution. It's -- of course, the clients we hired on the 30th of June didn't have time to contribute to the figures. So normally, you should expect these numbers divided by 2, so 3.5%. So you can see that in terms of revenue, it's a good number. So we have normally active new clients, so a very stable figure there. About the market share, we try to measure our market share in Switzerland. This is a stable figure. So first of all, you can see that the addressable market is growing. So this is according to many studies we have seen. So the addressable market for us in trillion is growing over time, and it has reached now CHF 2 trillion, and that's only for Switzerland, of course. And the good news is that our market share is growing in a growing market. So we now reach 3% overall market share in Switzerland. So it also means that, first of all, we still have a high growth potential here in our home market in Switzerland, and we are growing our market share, which really is basically good news. Now a few last information about our road map, development road map in H1 2026. First of all, we're very proud of having achieved a MiCA license that was very important to us in -- as we have an active crypto business. You know that now if you don't have a license, MiCA license in Europe, you're not able to provide cryptocurrency trading as of 1st of July. And writing time, we have received our MiCA license. We have also developed the trading in silver. It has been very attractive lately. And of course, as mentioned before, we have developed a lot of new enhanced security features on our -- for our bank accounts. Okay. Now our last look, and then we go to the Q&A session. There will be many occasions to meet us. We are present at the UBS Best of Switzerland Conference. And then also in September 2023, we have the Bank of America Annual Financial CEO Conference. We will be at a Swiss Equity Conference. And then in March 18, we'll have already the presentation of our full year results for 2026. So thank you for joining us here this morning. I guess we have now time for a few questions.

[Operator Instructions]. The first question comes from Tam Haley from UBS.

My 2 questions then, please. Firstly, the 2028 targets, I think you've been very clear about those. I mean that is very strong growth from your implied 2026 guidance to be 17%, I think, every year in pretax profit over the next 2 years. Could you confirm for us what your assumption is for crypto as a percentage of the 2028 target? And then my second question in terms of the net new money, very strong, GBP 5.1 billion. I think you've consistently beaten that EUR 7 billion target now for 3 years, I think. Was there anything unusual about the H1? Was there something in the Middle East or the Dubai flows that were elevated? I know there's usually an H1, H2 seasonality, but just trying to understand how we should think about this sort of level of flow going forward beyond 2026.

Okay. I take the second question and give the first one to Yvon. So for the second question, I think you -- it's true that the Middle East helped a little bit and -- but not massively. Actually, you can see here the uptick in the net new money was at CHF 0.5 billion, CHF 587.3 million to be precise, and that's higher than the previous half year. And this really is due to the situation in the Middle East where people or expats where they try to put the money offshore, and this helped a little bit. But of course, compared to the CHF 587 million to CHF 5.1 billion, so this is only 10%. So really the growth is coming from our main jurisdictions where we are, which is Switzerland and Europe.

Yes. So to comment on 2028 and perhaps to add on what Marc has just mentioned, Eli, I think that well, we have, let's say, unfortunate crypto volatility in H1. On the other side, what is interesting is to connect this trend with the level of net new money. We near record levels. The record was last year in H1 last year when we had a strong momentum on crypto assets. And when I think what H1 has demonstrated is while there is a diversification in the business model, while we could not more than compensate, we could compensate this decline in crypto assets. But there is a low correlation between the capability of the group to acquire new customers and crypto. And I think this is something that was a bit challenged back on time is how much Swissquote is relying on crypto momentum to acquire new customers. And now we see that we could most likely 3 years in a row, overachieve net new money targets, whatever crypto environment. And from my CFO perspective, this is something extremely positive as a learning lesson from H1 is, well, even with a very weak environment of cryptos, we're very close to record levels of client acquisition. Now to come back to your question on 2028, well, we remain confident and we think that the assumptions are intact. mainly because of what I've just been mentioning. If I remember the assumption that we have published back in early 2025 for the guidance, supporting the guidance 2028, the first key assumption was net new money, CHF 7 billion per year of net new money. We will most likely overdeliver 3 years in a row. So we overdelivered in 2024, in 2025 and most likely in 2026. And I think the growth will have been to have a slowdown in client acquisition. So this one, I think we're very comfortable. And most likely, we will overdeliver on the level of client assets. We cannot predict market impact, but we could easily be at least CHF 10 billion above the initial level of current assets forecasted for 2028. And this will help to compensate if we have a bit of pressure on the revenue margin on assets. As well on the mix of revenues, the mix of revenues was 60% transaction-based, 40% non-transaction based. We see that we are now already at this level. It's even a better mix than we initially targeted for 2028. This is certainly supported by interest rates and interest rates are basically staying much longer high than we initially expected. So we'll provide probably more information with the full year results 2026, but our confidence comes to the fact that the critical assumptions for 2028 -- they are intact and we most likely are overdelivering. But for sure, in 2026, we have a short-term volatility, but this we knew it. I mean we never expected the journey to 2028 to be linear. We knew we could face more adverse scenarios. We could face short-term volatility. But what we have tried to do is to protect the most critical underlying assumptions. On the level of crypto assets for 2028, initially, we had in mind they could represent 10% of the EUR 950 million. This is something that we'll update with the full year results. I think that while we could challenge the ability of Swissquote to reach this 10% with crypto assets. I think it's still feasible, easily feasible. On the other side, the share of interest income could probably higher than initially expected. So I think one could compensate each other. And we see at the same time that we have new revenue streams growing, structured products, securities lending. So we have initiatives that should as well help and should be significant in 2028 compared to 2026.

The next question comes from Christoph Blieffert from BNP Paribas.

The first one is on net interest income. Your NII guidance basically implies a decline in the second half versus H1. If you could share the underlying assumptions with us, this would be helpful. And the second question is on crypto. You have been highlighting a CHF 5 million loss on crypto inventories in the half year report. And if you could help us better understanding your market-making activities and the remaining value at risk, this would be helpful.

Yes, Christoph, thanks for the the very key questions that you're raising. So the net interest income forecast is built by the CFO. And historically, I think the net interest forecast is relatively conservative. So I think the positive aspect is we expect interest income in 2026 to be higher than in 2025, and this was not what we initially expected. H1 was extremely positive. Balance sheet was growing. Interest rate expectations were volatile and changing, so we could capture as well good opportunities. In H2, interest rates could still continue to grow. There could be a hike in euros. In USD, we have not bet on these increases. So we have assumed interest rates to remain flat in H2, and we have been slightly more conservative in the growth of the balance sheet. At the same time, we know we have a few expirations in investment securities in H2, and we have been as well a bit conservative in our capability to renew these expirations at the existing rate. So I think we have a good level of interest income for H2. This is much more than initially expected. Then, yes, for the time being, it's slightly below H1, but it's still an excellent number compared to what we had in mind 6 months ago. The other question about the crypto assets, I think it's important to highlight it. We have reached more or less CHF 14 million of crypto assets income in H1 2026, but the level of brokerage was higher because we have the above CHF 5 million negative mark-to-market. This is sort of a one-off that is impacting the numbers. CHF 5 million is not significant when we look at overall net revenues, but it's significant when you do CHF 14 million of crypto asset income. Where does it come from? You know we have this S crypto exchange. So in an exchange, you basically have sellers and buyers. You need liquidity. This liquidity is provided by market makers. But what we do is we as well bridge our exchange with other exchanges. Could it be Coinbase? -- could it be other venues, other exchanges that are known in the market. So to ensure we always have the best bid-ask spread across all these venues, we basically have our own internal market-making technology that is here to bridge the liquidity from our exchange to another exchange. And there is no way to bridge liquidity between 2 order books without basically bearing a certain risk. So the initial bid and half that we have on an exchange is provided by the sweet call crypto inventory. And when you offer 52 crypto platform, even if you will buy $100,000, $200,000 for each crypto asset, basically, it creates an inventory of around $5 million to $10 million. And this inventory is fair valued every time we close the books. So last year, it was slightly positive. I think it was CHF 2 million positive, but we did CHF 85 million. It's not significant. In H1, prices have decreased by 40%, 50%. And I'm afraid then we had to recognize this negative mark-to-market that we're not forecasting to happen anymore in H2. And this is important when understanding the numbers we have forecasted for H2. The total value of this crypto inventory at the end of June is CHF 8.9 million. So this will be technically the maximum additional downside that we could face. Should the situation recover, let's say, should prices recover, we will probably have more volume activity and as well recover part of this unrealized mark-to-market. I hope it's clear enough. Christoph. I know it's a bit technical.

And I would add that, of course, the -- for the inventory to go down to 0 would mean that the entire crypto market would go down to 0, which is not our baseline as such.

The next question comes from Daniel Regli from Kantonalbank.

I have basically 2 follow-up questions on previous questions from other analysts. One is on net new money. Obviously, last year, we also had a very strong H1 and then the second half year was, let's say, more "normal". Should we expect kind of a similar seasonality this year? Or was there anything which would lead you to assume that H1 could continue to be much stronger than H2? Or is basically the jury is out for H2, and it could well be that we see another CHF 5 billion of net new money in H2? And then my second question is following up on the kind of CHF 500 million guidance for 2028 and the assumptions behind. Obviously, I think you originally had assumed a 90 bps margin on assets for this CHF 500 million. And can you just tell me, do you kind of still commit to this 90 bps longer term? Or has this kind of changed? Or have your assumptions changed in this regard?

Okay. Thank you, Daniel. Very good question. So I'll take the first one, and Yvon will take the second one. So about the net new money, it's true that we have a little bit of seasonality, but it's not systematic. In H2 2023 was less good than H1 2023. And then the situation completely reversed in 2024, where the second half was the strongest. In H1, it went in the other direction. So it's a little bit difficult to forecast. It -- sometimes it's a little bit market dependent. But I think really the growth and especially the growth coming from Europe is pushing net new money. So we are anticipating good net new monies for 2026. Now whether it will be just double, it's difficult to say. Our average forecast is that we will be able to reach half of what we are expecting for the year, the CHF 7 billion. So that will be at least CHF 3.5 billion. So CHF 3.5 billion and CHF 5.1 billion, that would be at CHF 8.6 billion, that would be also a super good number, but that's probably the lowest number we can achieve. potential good news there in the second half.

On your question for 2028, Daniel, so currently, we stick to the 90 basis points. Why? Because we have and we will roll out more and more products and services that are not necessarily related to the level of client assets. And you will be helpful in this regard. So 90 basis points remains the underlying assumption. I think when we look at mid-2026, I understand it could look challenging, but we have growing products, and we have as well more products and services that are not necessarily correlated with client assets. That being said, I think if somebody would be skeptical about the capabilities of Cisco to reach 90 basis points, I think the level of client assets and the level of net new we have is likely to compensate in such a scenario. Basically, net revenues will be a function of client assets and revenue margin. So you can over-deliver on client assets and slightly underdeliver on revenue margin, you may reach exactly the same number of net revenues.

The next question comes from René Locher from ODDO BHF.

So a few questions on -- or 2 questions on costs or expenses. So the first one, Slide 22 marks, I was wondering have you already expensed the CHF 30 million investment in AI? That's my question. Because actually an interview, I guess, was in. mentioned that you're going to spend CHF 30 million in AI, but over the period 2026 to 2028. So again, my question is already expensed and now we get the benefit. That's my first question. And the second one also on expenses, some push actually on operating leverage. Now if we dig a little bit deeper into operating expense, I can see depreciation is up quite heavily, and I have seen that was depreciation of proprietary software. So what was the reason here? And also marketing expense was up 14%. Just wondering, yes, was it the contract with the Young Boys.

Do you want to take the second one? I'll take the first one.

So I'll start perhaps. So on the marketing, well, first of all, I have to say that the level of Qantas acquisition is very positive. And this is the most important for me as the CFO. Then on the marketing spend, we now consolidate You. And when you look at the numbers of H in the tables in appendix, we show basically the incremental contribution of you. You see that you in its level of maturity is spending a significant amount of marketing. The incremental contribution of you, when you compare it with the incremental revenues, you're close to 50%. So yuhlia is growing -- I think they grew accounts by 20% last year. So they're growing fast, and this needs to be supported by marketing. You need to establish a brand in Switzerland. So this probably explains a bit the fact that -- I understand you are surprised by the level of marketing expenses. On depreciation, what you have to keep in mind, and I think we've mentioned it in the communicated press as apparently you read the French, Rene, we say that during the acquisition of Yuh, you have this accounting specificity of purchase price accounting, we have recognized intangible assets. We have recognized goodwill, but as well intangible assets that we have to depreciate. And the goodwill is not depreciated, but we have recognized certain assets that have to, and they will temporarily increase the depreciation cost of Swissquote. So this is coming from the acquisition of Yuh. Have a look and very happy to guide you more into details later on after the call if needed. And now I give over to Marc.

Yes. Thank you, Rene. So when we speak about AI expenses, there are 3 segments to consider. The first one is staff, many people. You hire the specialists that are able to deal with AI, mainly building the infrastructure, building the data infrastructure and also building the gateways to the various AI and large language models. Then you have CapEx, you need to build your infrastructure. So even though a large part actually is happening in the cloud, but we -- if you want really to build a sovereign data warehouse and the AI system, you also need to buy hardware. So we did acquire H20 systems and chips in -- mainly in 2025. And then the last figures, these are the consumptions. So whenever you go outside of your internal systems, you consume tokens and these tokens are expensive. They're going up and up. And by the way, this is the biggest threat for the industry in general, not only banking, but the industry, you're getting addicted to these systems that are sold at a discount price for now. But sooner or later, the price of tokens will go up. and these companies will want to make a business out of that. This is a huge difference with an Internet infrastructure. So Internet was a common good, whereas the AI infrastructure is mainly in hand of private companies, mainly U.S. company based. So this is -- it shows even more -- it's even more important that you can build your own internal AI infrastructure because when you consume when you have large language models that are built in your premises, then the tokens are almost free of charge. So on one side, staff, CapEx and tokens. So staff, we -- I think we hired now most of the people. So this -- and -- and to give you a few numbers, overall, in 2025, we invested CHF 20 million in those 3 segments I mentioned before. In 2026, we will invest CHF 15 million, and then it will go down to CHF 10 million and 10 million for '27 and '28. '27 and '28, the expenses will mainly be the cost of tokens. This is the part where we are consuming tokens of those frontier models, even though we try to throttle it. But when we say we will -- we are installing or using the autopilot or advanced systems to double or triple the efficiency of our development team, it comes with the cost and the cost is the consumption of token. So the number you read of overall CHF 60 million from '26 to '28, this is correct. But it's probably more in a little bit of CapEx and lots of token consumptions rather than stuff. But if you -- in your assumptions, if you take CHF 15 million, so that would be CHF 5 million more of what we have already expensed in 2026. And then as of '27 and '28, you can add CHF 10 million in our cost for AI.

The next question comes from Miriam Killian from Deutsche Bank.

I have a question for Yvon. As you're approaching the EUR 17 billion threshold for FINRA Category III banks, could you maybe quantify the potential incremental capital requirements and cost burden associated with that? That would be helpful.

Yes. So thanks for the question, Miriam. I think it's very close to questions we received on the Q&A chat. So -- we most likely are today above the CHF 17 billion. When you look at the growth of the balance sheet in H1, basically, you can assume that in July, we have most likely crossed this threshold. So we'll be notified soon, I think, by FINMA that we are now officially a Category 3 bank. However, my understanding is that we have been treated as a Category 3 bank for a certain time already. And we have increased headcount in control functions in the last month. We did so as well in H1. So obviously, when you are a Category 3 bank, you have more regulatory scrutiny. The regulator is looking more closely at you. I think this is already the case for a couple of months. It was the case in 2025, and I think it is the case in 2026. So the main change is the minimum capital ratio that will increase. We already commented about it in the full year results, but the Category 3 bank has at least a minimum capital ratio of 12%. And as well, there is increased regulatory supervision. Should you go to the website of FIMA, you can, for example, see that category 3 banks are basically subject to on-site visits of FIMA. They do -- generally, they rely on external auditors, but they will -- as a Category II bank, they will as well do their audits themselves. And in 2026, for example, we have scheduled 3 on-site visits from FINMA, which is generally between 2 and 3. So I think today, the numbers of 2025 and 2026, they already reflect the cost of doing business as a Category 3 bank. This will be my CFO estimate.