Welcome to RaySearch's presentation of the first quarter 2026. My name is Carolina Strömlid and I'm Head of Investor Relations.

As usual, our CEO and founder, Johan Löf, and our CFO, Nina Granberg, will walk you through the key highlights and financial results of the quarter. After the presentation, we will open the floor for questions.

Simply raise your hand in Teams to ask a question live. With that short introduction, I will now hand over to you, Johan.

Thank you Carolina, and welcome again everyone. I apologize for my voice, I have a cold as you can tell.

Before we dive into the results and highlights, let me start with a quick overview of RaySearch's business. RaySearch is pure software company and we focus on software for cancer treatment.

We have four different platforms: RayStation, RayCare, RayIntelligence, and RayCommand. What we see here is a schematic view of a comprehensive cancer center.

Although RaySearch has been focused on radiation therapy for the first 25 years, we are now about to expand into the other areas of cancer therapy: into chemotherapy, surgery, and [particle?] ablation for example.

Everything that goes on in a comprehensive cancer center will be supported by RaySearch software in the future. I think it's important to look at the long-term performance of RaySearch.

This diagram shows the revenue all the way back from 2008. You can see here that it's been a steady growth every year, except for the two pandemic year, 2020 and 2021.

In this graph, we also see the support revenues and how they are steadily growing year-over-year. I would also like to show another view of the same thing basically, where we also look for every year at the revenues for the individual quarters.

As you can see, for many of these years, there are fluctuations between quarters. We can look at 2017 for examples.

It fluctuates, but nevertheless there is a growth for the full 12 months. Another example can be 2022.

You see fluctuations, somewhat weaker Q2, but when you take all the quarters together, again, growth. 2024, the same thing, and 2025 even is as a similar situation.

This is not nothing unusual for RaySearch. There are of course underlying revenues from the support, about 40-plus%, and many smaller orders are the foundation for the revenue for every quarter.

Whether the quarter is really good or bad depends on a few bigger deals. It can be two, three, four different deals that determine whether the quarter is great or not.

They can fall either within a quarter or they slip to the next quarter or the quarter after that. This shows a very long track record for many, many years that even though there are fluctuations between quarters, the growth over the 12-month period happens anyway.

I just want to highlight that given that we had a softer quarter this quarter. All right, Q1 in brief: profitability remained robust despite the softening net sales in the first quarter.

Net sales declined by 12% to SEK 298 million. It was mainly due to stronger Swedish krone and a tough comparison with last year, which included a large order from Carbon Neutrons in China.

The organic growth was a negative 5%. As I mentioned, there were a number of deals that we thought would come in into this quarter, but that they didn't, simply.

It's a timing effect and they will come in in Q2 or Q3 at the latest. One example of an order around SEK 15 million that was stuck, it was ordered and the purchase order was signed, but was stuck in the Chinese customs.

That's something that was very unfortunate and not something we can control. That order is already in now as an example.

The recurring support revenue continued to provide stability, amounting to 123 million or 42% of total revenue. Operating profit reached SEK 68 million, corresponding to an operating margin of 23%, in line with last year.

Some business highlights: the demand remains steady during the first quarter with high customer activity across our markets. We secured several strategically important deals, including three new RayCare orders in key markets.

We received our first RayCare order in China together with an order for RayStation for proton therapy at facility at the Shanghai Proton and Heavy Ion Center. The solution will support an upright proton therapy system, which is a cost-effective pathway for clinics to introduce proton therapy.

The order is an important milestone for us in this large and growing market. In France, Institut Hartmann chose to replace its existing treatment planning and oncology information systems, Eclipse and ARIA, which are both from Varian, with RayStation and RayCare, enabling a unified workflow across a mixed treatment machine environment.

This was our first RayCare order in France. In Germany, several clinics selected RayStation to replace Pinnacle for treatment planning, and we continue to see business opportunities ahead of that system's planned end-of-life in 2027.

We're also proud that the first patients were treated with RayStation in Ukraine during the quarter. That's an important step in modernizing radiotherapy under very challenging circumstances in this country.

I will now hand over to Nina to go through the financial development in more detail.

Thank you, Johan. The quarter one sales numbers may give the picture that business is slowing down, but as you heard from Johan, that's not the case.

Order intake dropped 13% compared to last year. It was affected by currency and in relation to very strong comparison quarter with above 70% growth.

The activity from our customers and in our sales team remains on a high level. The book-to-bill ratio was 1.2 and order backlog end of March amounted to 1 billion 615 million, and 624 of those is expected to turn into net sales in the next 12 months.

Moving forward to net sales and EBIT: net sales was down 12% in the first quarter. More than half of the downturn is related to a weaker Swedish krona in 2026, although quarter one last year was the quarter when the value of US dollar, euro, and several currencies started to fall.

The organic growth was -5%, again, in relation to a strong first quarter 2025 and to RaySearch's normal fluctuation. As Johan mentioned, we had revenue that got stuck in customs in China and for other reasons delayed into later quarters.

License sales decreased with 24% and support sales with 1% year-on-year. Taking away currency effects from support numbers, the growth would have been 8%.

Despite the lower sales, EBIT was solid at 68 million and the margin amounted to 23%, which is the same level as last year. We have approximately 20% of our costs in US dollar and together with good cost control and slightly higher R&D capitalization rate, operating costs were stable.

In the 2026 quarter one, currency gains from revaluation of working capital gave us a positive 10 million SEK effect. The rolling 12 development graph over net sales and EBIT, as you see in this picture, is tipping down a bit in the 2026 first quarter.

As you can see, it also did in quarter two last year, just to turn upwards again in the next quarter, and then continue its pathway upwards. The rolling 12 EBIT margin amounts to 22%.

Moving on to the next slide and the revenue split: as Johan said, the support revenue was 42% of total revenue. That is a slightly bigger part than we usually have and that is related to the lower license sales in the quarter.

The lack of growth in support revenue was, as mentioned, mainly due to currency, and we also had some periodization effects from a couple of contracts that gave us a lower number in the first quarter. I want to highlight here that those periodization effects, it doesn't mean that we lose any revenue.

It just mean that revenue is sometimes moved between the quarters. The next slide and the cash flow: the underlying performance in cash flow was good in the first quarter and we had about 30 million inflow from contract assets.

However, this positive effect was offset by an active choice to pre-purchase inventory, something we did in order to mitigate effects from price increases in hardware. It is servers and computer that we normally buy upon order and transfer almost instantly to the customer.

Cash balance end of the quarter amounted to 439 million and cash flow remains an important focus area also going forward. With this, I'll hand over to Johan.

Online adaptive is a very hot topic in radiotherapy right now. Online adaptive in general allows you to modify the treatment while the patient is on the treatment couch and you take an image of the patient in treatment position, check the changes in anatomy, and then you change the plan to adapt for those changes.

A very strong feature here is that we have, with the combination of RayCare, deliver online adaptive on this machine. I'm going to show you, in a very simplified manner here, the workflow surrounding such an online adaptive treatment.

We have a patient here, as you see, to the right on the treatment couch and there is a TrueBeam. We also see the control room where an operator has screens for example, RayCare and RayStation.

If you zoom in on the RayCare screen here, we look at the treatment course for this particular patient whose name is Thomas Fischer. We switch to the treatment calendar and this particular clinic has three different machines: a proton machine, ProteusONE, Radixact from Accuray, and two different TrueBeams from Varian.

You see the various time slots here for the different machines. We zoom in on Thomas Fischer.

He's on an online adaptive protocol. The first thing we do is to ask the TrueBeam to acquire a cone-beam CT image, which is a three-dimensional image of the patient in the treatment position.

Via the existing interface between RayCare and TrueBeam, the VTI, we can transfer that image and the registration back to the RayCare. The operator opens the adaptive replanning module in RayStation and start that adaptive workflow.

The first step is to correct that image that we received from the machine, of course is not of the same quality as a diagnostic CT image. You see to the right is the image that is transferred from the machine, generated by the machine.

To the left, you see a corrected cone-beam CT image. We have used the diagnostic CT to extend the image outside of the field of view of the cone-beam CT imaging system, and we have also refined it using information from that diagnostic CT image.

You see that the left image is sharper than the right image. That's the first step.

The second step is to use AI, or in this case deep learning algorithm, for segmentation of the structures. That's done very rapidly in a few seconds.

That's completed. After that, the dose is calculated on today's image and or today's patient anatomy.

That's what you see to the left. To the right, you see the dose distribution that was the planned dose distribution, let's say the dose distribution that the doctor signed on.

The tumor here is the white contour that you see, that's the clinical target volume that we want to irradiate. There's a perfect match here between the high-dose region, which is red, and the shape of the tumor.

Then it drops off, the dose distribution drops off very quickly towards the healthy tissue surrounding. If you look at the left image, there has been some shifts in the geometry.

If we apply the same beams on today's patient geometry, you see it doesn't really hit the tumor. There is an underdosage in some part of the tumor up here, up at the top left part of the tumor, and there is an overdosage outside of the tumor in the lower right parts.

That is also illustrated by the dose-volume histograms. It should look like the dashed curve here for the tumor, but it is actually like the solid line, which signifies an underdosage of the tumor.

The other dose-volume histograms illustrate overdosage of rectum and bladder. We can also look at the clinical goals here for the plan that the radiation oncologists had approved and all green lights for all the clinical goals, but now this treatment would have some red lights here, we're not happy with that.

The next step is then to recompute the geometry of today. You can see on the left how the dose distribution is now shaped according to this new geometry, and it will now in a few seconds here wrap around and conform very well to that new geometry.

Yes, now it's done. What we can see here are a couple of things: now the dose distribution, the high-dose region, is perfectly matching the tumor shape again, and we get good protection of the rectum down here and also the bladder on top.

The dose-volume histograms for the tumor are overlapping, we have the same tumor coverage again as we had for the originally approved plan, and is all green lights. Now we're done with the adaptation.

I talked a lot now here. In reality, from when we get the first images that you saw in the beginning that were imported into RayCare, the procedure that I talked about after that: creating the synthetic CT, applying deep learning segmentation to get the structures, and then compute an adaptive plan, that whole procedure takes about 45 seconds.

It's much more rapid. I had to stop and explain what was going on, but in a realistic setting, this can be done in under a minute.

Because the patient is on the couch, we want this to be quick so that we can treat the patient as soon as possible. The next step here is that we say we're going to sign this adapted plan, and RayCare will transfer that adaptive plan via the VTI interface to the machine.

The machine will deliver this adapted plan to the patient. It's a dual-arc VMAT plan, it makes two rotations around the patient, one like that, and then the second arc is coming here.

That is how effective online adaptive can be on a TrueBeam. We expect the first real treatment to be done in this exact way in a couple of weeks in Belgium, so that will be a very important milestone for RayCare.

All right, to summarize: the first quarter was softer in terms of net sales, as we've seen, while profitability remained robust. Demand for integrated and automated workflows in cancer care remains strong and are supported by solid sales pipeline.

There is growing interest in online adaptive solutions, and as I said, we expect the first online adaptive treatments with the combination of RayStation, RayCare, and TrueBeam to happen within the next couple of weeks. I'm also confident that we will reach our operating margin target of at least 25% for the full year 2026.

To underline our long-term financial commitment, the board has decided on a new operating margin target, which is at least 30% for full year 2028. With innovative software, strong customer relationships, a global footprint, and a growing base of recurring revenues, RaySearch is well-positioned for continued profitable growth.

We will now open for questions and I will hand the word over to Carolina. Yes,

we will start the Q&A with live questions. Please raise your hand in Teams if you would like to ask a question.

The first question comes from Kristofer Liljeberg at DNB Carnegie.

Thank you. Good morning.

Follow up on what you said here about online adaptive: so when do you expect to have the FDA approval for RayCare online adaptive with TrueBeam available?

We expect to have that in the fall, around October timeframe this year.

Okay. Is that delayed from before or?

Perhaps.

It's a complex FDA situation with several different versions of both RayStation and RayCare that are sort of in the pipeline at the same time. There are dependencies between these versions and that's where we are now.

It will still be clear to US customers when the first treatments with online adaptive RayCare, RayStation, TrueBeam happens in Europe. They would be well aware of that.

These are improvements within existing modules. There are three modules in RayStation that you need to do the thing that I showed in the presentation.

You can purchase those already now, and then when the version is FDA approved, everyone who has those modules and they're all on support contracts, they will be upgraded to that version and then they will have that capability right away.

Do you see that US customers are happy to start acquiring RayCare before the approval for this purpose?

I think they can. Some customers already have this capability with RayCare and RayStation, they will have access to it, they can start to use it clinically once approved.

Okay. How do you view RayCare orders picking up here throughout the year?

Would you expect some improved momentum outside the US once you have done the first treatments

and then

the US later?

We already see better momentum than last year. I think we had four RayCare orders full year 2025 and we have three orders already Q1 2026.

It will start to ramp up now during 2026 and more in 2027, but you see already finally that starting to show some good momentum.

I agree with you. I think the concept of online adaptive and what you could do here with RayCare, of course, lowers the entry barriers.

Do you see potential for this online adaptive concept to accelerate RaySearch growth in the coming years?

I really think it has that potential because it is a very obvious thing that you want to do. It was very clear when I showed to you here what the benefit is.

We are extremely well positioned in this field. Historically, we've been very strong in online adaptive and sort of the whole company started around adaptive to begin with 25 years ago.

Yes, I think this can have overall a very positive impact on RaySearch and in combination also with all the new vendors that are coming into play with Hitachi's OAxRay and Leo Cancer Care's Grace and all these new machines, and they all want to enable online adaptive together with us.

Good. Then one question on the R&D capitalization that's higher this quarter versus previously, both in absolute terms, but also as percent of sales.

I think it was like 6% of sales Q1 and it has been more like 1% to 3% in recent quarters. How should we think about this coming quarters for the full year and what type of impact will capitalization have on the 25% margin target for this year?

Thank you.

To answer your first question, when you put it in relation to sales, you must also take into consideration that sales was a bit lower this month, that of course highers the rate. Besides that, we had more hours that was spent on the projects that we capitalize quarter one.

I don't see any special pattern in that. It can vary between the quarters.

Of course, we have increased the number of employees that work within R&D, so that means that we have higher expenses and with that a higher amount of capitalizations. I think that in general, I would say that we will keep the same.

Approximately we capitalize 70% of the expenses and I think that we will keep that rate also going forward with some fluctuations between the quarters.

For the full year, do you think it's fair to assume a similar, around 1% of sales as we have seen in both I think 20- yeah, the last 3 years actually?

Do you mean expenses or capitalization? I think it's hard to relate.

Capitalization to sales, because

the net effect of

capitalization,

that has been around 1% of sales recent years. Now it was 6% this quarter.

I know there's a seasonal effect, was higher in Q1 also last year,

but

I was just wondering because if you look at the full-year target, if you start capitalizing more, it's of course super easy for you to reach the 25% margin. I just want to make sure that's not what's happening here.

Thanks.

No, no, we will not reach the margin due to capitalization. As I said, I think we will continue capitalize about 70% of the expenses and the expenses will increase as we add employees into the R&D function, and as we had.

Thank you.

The depreciations

will also increase.

Since we do capitalize more because we have increased the R&D function, that will of course mean that the depreciations in the next coming year will increase as well.

Okay. Thank you.

Yeah.

Thank you, Kristofer. Our next question comes from Oscar Bergman at Redeye.

Oscar, please go ahead.

Thank you. The question I have first off is a bit speculative, I suspect maybe you won't be able to answer, but I have to ask.

The updated financial target of at least 30% EBIT margin by 2028, that's I think in line with analyst consensus and at least confirming my own, but I'm wondering what would be needed for you to go beyond 30%?

It's just more sales.

Would that be RayCare sales specifically or could it be done with just RayStation?

It would be RayStation, RayCare, and RayCommand.

Thank you. Then you mentioned some other order delays in Q1.

Just wondering if you can give some background to the reason for this or if we should just consider it business as usual?

It's business as usual in the sense that the customs situation in China was one very specific thing, but in the other cases, it's just that maybe the signing procedure wasn't completed on the correct day, that is within the quarter, but it was postponed into the next quarter and things like that. This is but I think we had an unusual What happened here?

Something happened here with the [unclear] with the presentation. Sometimes we are unlucky and several of these orders miss the end of the quarter.

Mm.

It's not the first time this happens, but we are very vulnerable. It looks much worse than it actually is.

If you look historically, it has happened many, many times, but then you zoom out 12 months and then it's extremely robust in terms of growth.

I agree. I'm not sure if you mentioned the Pinnacle conversion in the report or the conference call, maybe you did, maybe I just tuned out, but just wondering if you can define maybe the percentage of sales that came from converting clinics and if we should expect that the conversion window should be closed sometime during Q3 or very early in Q4 at the latest?

The Pinnacle, it was 22%, right, in Q1?

Of license sales.

Of license sales. Of all the software license sales, it was 22%.

I think it will continue throughout the whole year, the Pinnacle replacements, because there are many deals still ongoing. We've been very successful in Germany converting Pinnacle, but there are still deals left out there, and still in the US, some.

I think it will continue all the way into the end of the year, actually.

Interesting. I just have two more quick questions, if that's okay.

I think in the Q3 report, you initiated the six months trial period campaign with the certain customers and you opened up the additional modalities. Just any sort of status update on this would be appreciated.

We only have the two pilot regions still, which is Benelux and UK. It has been very well received in both regions.

I think in Benelux, they are more active on this and they spent the time because it takes time for them to evaluate as well. In the UK, they're happy about the offer, but they have less time

from

their everyday work to spend time on this. Of course they should, because it's a good investment on their side.

We have learned also that it takes some effort from our side to train on this new functionalities that we open up, so that they can get a good experience when they try it, we have to train them in using these functionalities. I think it was a good approach actually to do it in a limited fashion to start with so that we learn.

It was more complex than we had anticipated, but so far so good, I would say.

Will you now go back to sort of the sketching table and do some tweaks or will you expand the campaign? [Swedish background speech]

What happened there?

[Swedish background speech]

Basically so. I would say yes.

[Swedish background speech] I think someone else is on the line.

Just

the final question, Johan. You have a very strong market share in proton therapy.

I'm just curious about what the other players are doing here to increase in their market share or are they maybe not focusing on this area for some reason?

[Swedish background speech] I think the others are simply not very focused on this. We don't see much competition in in this area.

It's one, yeah, I know it's only Elekta and Varian that are involved here. Varian left proton therapy, in a sense, when they well, when they cancelled their machine production, the ProBeam.

I think they may and then sort of the Eclipse support of protons also faded away. I think they are maybe returning, I'm not sure.

I think Elekta won a proton order somewhere for for treatment planning recently. We we still have almost 100%, I would say, of the new sales for the treatment planning is for RaySearch, from RaySearch.

It's still a great area for us to a great segment for us to to be active in, and we we focus really hard on that and we work with all all the different proton vendors and there are many that that we support. It's a lot of activity in this field all over the world, but mainly in in Asia.

China, they have a plan to build 100 particle centers, combination of well, I think 70% protons only and 30% including carbon ions. They have a very high ambition, and Korea, Japan, these countries, Taiwan, they are building particle centers, a lot of particle centers, which is very good for us.

Great.

Thank you very much, och krya på dig. Tack.

Thank you.

Thank you, Oscar. Next, we have Mattias Vadsten from SEB.

Mattias, please go ahead.

Hello, can you hear me?

Yes.

Great. I have a question on the France order that you talked about.

If you could elaborate a little bit on the key reason for them to replace the Varian software suite. Also, I wonder if Hartmann's Halcyon system will be affected by this change of software?

Also, when do you expect integration with Halcyon with RayCare to be possible? Sort of seamless integration.

Yes. This is a let's say medium-sized center.

They have, I believe, two CyberKnifes from Accuray. They have one TrueBeam from Varian and three Halcyons from Varian.

Six machines in total. The total order was I believe SEK 35 million.

There will be a phased approach with I think we already recognized, I think, 11 million of that in Q1, and then the rest is coming later. They will go live with RayCare combined with CyberKnife and TrueBeam beginning of next year, and then the three Halcyons beginning of 2028.

The interoperability work, or the interoperability project between RayCare and Halcyon is ongoing with Varian, as we're working together on that. There is no definite timeline yet, I would prefer not to communicate that now, but for sure before they need to start beginning of 2028.

[Swedish background speech] Why did they switch? I think there were several reasons, but they're very happy with the level of automation in both RayCare and RayStation, with scripting and other means and streamlined workflows, and especially since they have non-Varian machines as well, the two CyberKnifes, I think that helps.

I think they believe in the power of this bundle, RayStation, RayCare, and also the vision going forward for this software. It's a very important order for us and we're happy that we got it.

It's fair to say they would never switch if it wasn't for the latest news flow around RayCare, TrueBeam?

No, I think this is a direct consequence of the interoperability that's an important factor, of course, that's an enabler. That's a requirement, but that would not in itself be enough for them to switch.

It really opens up the door when we have the interoperability in place for TrueBeam and we also have a plan for Halcyon.

Good. Then my next question is regarding growth.

It's been a lot of discussion around Q1, but just so sort of what makes you confident to go back to solid growth rates already in Q2 here?

As I've said over and over again that you fool yourself if you look at a single quarter for RaySearch. There are a few deals that determine the quality of the quarter and they can end up on either side of the quarter.

That's what it is. We have such a good track record, if you believe history at all, then look at what has happened since all the way even further back.

The reason why we picked 2008 was that we got the first revenues from RayStation in 2009. We could go even further back and we still have this very steady growth and fluctuations in quarters.

That's with the current license model, that's what we have to live with. Either people understand that or they don't understand that.

I don't think it's so difficult to understand, but it seems to be very difficult. If we had a subscription model, then this would be much smoother.

We don't, because our customers don't want that right now. Most of them don't want that.

We adapt and either you look at RaySearch on a 12-month or 24-month window, then everything is fine, or you focus on the quarter and then you have to be nervous all the time.

Thanks. Last one from me would be, do you have any worry that growth will fade off in 2027 specifically due to the vacuum created from Pinnacle contribution fading off for you?

Absolutely not.

Thank you very much.

Thank you.

Thank you, Mattias.

Someone is vacuum cleaning.

We will now take a question from Magnus Bärnarp at Bärnarp. Magnus, please go ahead.

Yes, hello Johan. My question was partly answered, but I would like to know about the EBIT margin expected to reach 30% by 2028.

If you could give me the details for how such a margin improvement will be achieved? Could I assume that the margin improvement will be gradual or will it mirror the ramp-up of RayCare?

No, I cannot say anything else than that we will have at least 30% EBIT margin in 2028, full year 2028. It will be gradual increase of margin.

I don't see how I can give you more details except how that will be achieved.

Okay. Secondly, I would also like to ask because you have this change in the board since Günther Mårder are going to a trial, and I just wonder if he will not be convicted, will he be welcomed back to the board?

That's a good question. I haven't thought about that at this stage.

I think it's, as I've said before, it's very unfortunate what has happened. I really like Günther Mårder as a person, of course this is awful.

It's not convicted yet, we have to keep that in mind. I think that's all I have to say regarding that at the moment.

That was all my questions. Thank you very much.

Thank you.

Thank you. We have two written questions also.

The first one is, what is the status of the Ortega order? When will the RayCare delivery take place and in which quarter will it be invoiced?

In quarter one we delivered RayStation to two Ortega sites, and RayCare deliveries to those two sites is planned for quarter three and quarter four.

Yeah. The next question is-

We can add to that that those are the first two Ortega sites,

and then

the remaining seven will be delivered during 2027 and 2028 is the current plan.

Yes. During which quarter will the order from Shanghai Proton and Heavy Ion Center, SPHIC, in Shanghai be invoiced?

I don't have that information. Do you, Johan?

The revenue for RayStation happened in Q1 and the revenue for RayCare hasn't happened yet.

Okay.

That will happen later this year.

That concludes the Q&A session. Thank you for joining us today.

Should you have any follow-up questions, please don't hesitate to reach out to us. We look forward to connecting with you again on August 13 for our Q2 results.

Have a great day. Thank you.

Thank you.

Thank you.