Good afternoon, ladies and gentlemen, and welcome to the Springer Nature analyst call full year 2025 results. At this time, all participants have been placed on a listen-only mode. The floor will be open for questions following the presentation within the conference call. Let me now turn the floor over to your host, Tom Waldron.
Thank you, Anna. Good afternoon, everybody. Welcome to Springer Nature's full year 2025 results call. I'm Tom Waldron, head of investor relations. Today's presentation will have the following structure. Frank will start with a business update, followed by Alexandra with a review of our financial results. Frank will then come back to review our strategy before we move on to Q&A. Before handing over, let me briefly remind you, for revenues and adjusted operating profit, we present both reported figures based on actual currency rates and portfolio composition and underlying growth rates, which exclude currency and portfolio effects to ensure a like-for-like comparison. Our financial guidance for 2026 is based on constant currencies and the expected underlying performance of the business, excluding portfolio changes. With that, I will now hand over to Frank.
Thank you, Tom, and a warm welcome from my side. Let's start with a brief overview of the key highlights from our first full year as a public company. We've delivered very strong results, with revenue growing by 6% in underlying terms and AOP increasing by 9%. Our research segment continues to be the main growth driver, with our journals outperforming the market, especially in full open access. We've made great strategic progress over the year. We've consolidated our leadership position in open access with above market growth and continued to implement technology and embrace AI across the business. We've delivered a very strong cash flow performance and reduced leverage. With the good progress on 26 renewals and strong submission growth continuing, we entered 2026 with strong momentum and good visibility. We're guiding for 5%-6% underlying growth and further margin expansion in the year ahead.
Let's take a look at our segments. We operate in three segments, each with leading positions in their respective markets. By far, the largest is research, accounting for almost 80% of Springer Nature group revenue and almost 90% of group-adjusted operating profit. Around 2/3 of that revenue is contracted in agreements which typically have a three- to five-year duration. In research, we're the second-largest player in the market by some distance. We have the largest share of the top 50 journals by impact factor, and we're the world's largest publisher of academic books. In health, we lead our markets in Germany and the Netherlands and are growing the scope of our activities in international healthcare. Finally, in education, we have strong positions in key markets including Mexico, India, Brazil, and Southern Africa.
On the right-hand slide, there's a reminder of the broad distribution of our revenues across the globe. Before we get into the details of our full year performance, I'd like again to share some examples of research from across our journals portfolio in 2025. These demonstrate the value we create for our communities by making trusted knowledge accessible. I'll pick two from the six here. First, bit of a lighter example, an intriguing social research indicator study from our Springer portfolio. As a dog owner, I was very interested to see an estimate of the boost in life satisfaction from owning a cat or dog is actually equivalent to about GBP 70,000 a year. The authors say this is on par with the benefits of marriage. I don't want to be in a difficult spot here. Or regularly meeting with friends and family.
Second, on a more serious note, Google DeepMind's paper in Nature on AlphaProof. This paper presents a reinforcement learning-driven AI system for solving complex mathematical problems. AlphaProof reached silver medal performance when solving International Mathematical Olympiad problems, missing gold by just one point. This is a powerful example of the research ecosystem we enable. World-leading scientists, supported by high-quality, trusted platforms, uncovering new ideas and sharing discoveries that push science forward. The examples on this slide illustrate the crucial role that we play and the things we stand for. Trusted science, real-world impact, and sustainable growth. Let's now move on to our different segments, starting with research. Our research segment delivered more than 7% underlying revenue growth and nearly 10% AOP growth. We've seen continued strong development of our journals portfolio.
The 2025 renewal season was completed successfully at close to a 100% renewal rate. The global article market grew by about 8% in 2025, while our own article output increased by a little more than 12% as we continued to capture market share. The full OA market grew around 12%, and we've continued to perform well, growing around twice as fast as that. Submissions to our journals continue to grow very strongly at more than 30% across the portfolio. As you can see, we're gaining market share due to our investment, the quality of our portfolio, and better execution driven by technology. I will just give you a couple of examples. We launched 50 new journals, including two new Nature journals, with the remaining 48 in full open access.
We ended the year with 85 transformative agreements, 17 of which were renewed with 19 new agreements signed. Each time we sign a transformative agreement, we tend to see revenue accretion versus legacy Springer subscriptions. Our 85 TAs now cover over 4,000 institutions across the world. Europe is essentially fully covered by TAs, so the opportunities are in the rest of the world with a focus on the U.S. We currently have 40 active discussions around new agreements, with around half of those in the U.S. Finally, our technology investments are paying off. For example, our AI-based transfer recommendation tool saw a 40% increase in volume, adding more than 1 percentage point to our publication growth in 2025. While ARPI, our AI-based tool that checks editorial scope and quality, automated the screening of almost 500,000 articles.
Turning to books, the long-term trend is for digital books to drive growth as print gradually declines. However, in 2025, we've seen growth in print benefiting from a weak comparison to 2024. In addition, in the second half, we saw the pull-in from some orders into 2025 from 2026. Finally, within services, we have seen strong demand from corporate R&D clients for text and data mining solutions, but a bit more challenging environment in talent-related services in the U.S. Let's now turn to the developments in our other two segments, health and education. Starting with health. We saw good momentum in scientific affairs in our international healthcare business following the investments we made last year. Our Dutch events and books business also performed well, offsetting a lower advertising and events business in DACH, German-speaking markets.
AOP growth benefited from revenue growth, while Q4 was impacted by the comparison to the favorable product mix we saw in Q4 2024. Turning to education. After a solid performance in the first half, education was held back by delays in a new curriculum in one of our larger markets, South Africa, as we discussed at our nine-month results. Elsewhere, we had good curriculum sales in Argentina and India, and are looking forward to the launch of our new ELT content this year. We delivered 9% underlying growth in AOP, supported by continued progress in our operational excellence program in education called Elevate. With that, I'll hand over to Alexandra for financial update.
Thank you, Frank. I now walk you through our key financials for 2025 in more detail. It was a very strong performance. Reported revenue for the group reached EUR 1,926 million, with adjusted operating profit of EUR 544 million, which includes actual currency movements and a small impact from scope. We delivered strong underlying growth, with revenue increasing by 6% and adjusted operating profit rising by 9%. Free cash flow improved by EUR 79 million, reaching a total of EUR 298 million. Our financial leverage is down significantly year-over-year and is now in the lower half of our target range of 1.5x-2 x net debt to EBITDA.
Our proposed dividend of EUR 0.83 equates to a yield of more than 5% of our current share price, and we expect growth in line with our progressive dividend policy. The next slide shows how the outcome compares to our guidance. As you see here, we came in slightly above the midpoint of the guidance range. As a reminder, we raised our guidance in August of last year. The outcome here was well ahead of our original expectations for the year, driven by a very good performance in research. This next slide lays out the performance of our segments in detail. Group revenue increased by 6% on an underlying basis and 4% on a reported basis. The research segment was the primary growth driver with 7% underlying growth.
Reported growth of 6% was lower due to the strength of the euro against the US dollar and other currencies and a small portfolio impact. In education, underlying growth was 1%. As we told you last quarter, the business performed in line with our expectations in most markets, but we faced unexpected headwinds in South Africa with delays in a new curriculum. Reported growth of -6% is primarily impacted by hyperinflation effects in Argentina and the weakness of the Mexican peso and Indian rupee. Looking at adjusted operating profit. Group adjusted operating profit increased by 9% on an underlying and by 6% on a reported basis. For research, underlying growth was 10% and reported 8%. Currencies had a small negative impact due to the strength of the euro against the US dollar and yen, offset by two factors.
A slightly weaker pound, where we have more cost than revenue, and a positive portfolio impact from a divestment. In health, full year performance was driven by good results in international healthcare and the Netherlands, partly offset by a softer performance in the DACH region. In Q4, we faced a particularly tough year-on-year comparison in health, as Q4 2024 benefited from a very favorable product mix, as highlighted previously. The difference between underlying and reported results is attributable to the impact of a small divestment. In education, we saw the benefit of product mix and the efficiency measures that we began to implement in the years 2024 and 2025, with a strong underlying improvement in AOP despite the headwind in South Africa. Turning to the P&L. On top of the growth in revenue in AOP, we've also delivered very strong growth in adjusted net income and EPS.
This performance includes some positive elements that won't reoccur in 2026 and beyond. Our adjusted financial results benefited from lower debt balances following repayments, as well as from lower interest rates, which reduced our financing costs. In addition to that ongoing benefit of deleveraging, 2025 also benefited from a positive swing in other financial income and expense, which netted to a positive of EUR 47 million or around EUR 0.24 on EPS. During the year, we have worked to optimize our legal structure to reduce the volatility on the financial results in the future. There will still be some variability at period ends around the valuation of intracompany balances, but it should be significantly less than we have seen in 2024 and 2025.
Adjusted income taxes shown here exclude impacts related to purchase price allocation, but includes the one-off benefit from tax losses carryforward that we have previously explained. I'm pleased to report that our cash generation in 2025 was very strong. Operating cash flow increased by EUR 22 million year-on-year, driven by a higher operating profit and good working capital control. Investments declined slightly. We continued to invest in the business but benefited from a small year-on-year reduction in CapEx due to timing factors. Cash interest expense was significantly lower due to reduced gross debts and lower interest rates linked to our continued deleveraging. As a result, free cash flow rose by EUR 79 million to nearly EUR 298 million.
That strong cash performance allowed us to continue deleveraging, and we end the year slightly below the middle of our midterm target range, well ahead of where we thought we'd be at the end of 2025 at the time of the IPO. On the back of that strong performance, the management and supervisory boards are proposing a dividend of EUR 0.83. This is in line with our policy to pay a progressive dividend of approximately 50% of adjusted net income, adjusting for non-cash benefits that we saw in 2025 that I mentioned earlier. In addition to the regular dividends, we will propose to the AGM a five-year buyback authorization, providing another tool for optimizing capital return in the future. Next, I want to put the 2025 dividend and the buyback authorization into the context of our capital allocation framework.
We anticipate good long-term growth in cash flow, supported by revenue growth and cost control. Our financial results continues to be a favorable trend of lower expense as we delever and optimize our capital structure. On tax, we should be able to gradually reduce the tax rate we pay as we optimize our tax structure. With stable investment needs and strong working capital control, this will all translate into strong long-term growth in free cash flow. Our priorities for utilizing that cash flow are clear. We will invest in our business, maintain a strong balance sheet, look to value enhancing M&A, and pay a progressive dividend. Beyond that, if our strong cash generation leads to excess capital beyond the needs of the business, we will consider returning it to shareholders via buybacks. Finally, I come to our guidance for 2026.
As you will have seen in this morning's press release, we expect underlying growth in revenues of 5%-6%, with underlying improvement in AOP margin of around 30 basis points. We have good visibility in 2026, underpinned by robust submission growth at the end of 2025 and a renewal season that is tracking well. Since renewals commenced last September, more than 80% of contracts have been renewed, slightly ahead of the prior year. As you are aware, we will see a headwind to reported numbers from currency as shown on the slide, assuming rates as at year-end. Our strong performance in 2025 and our AOP guidance for 2026 means that we are on track to deliver our midterm target of 100 basis points of underlying margin improvement one year early.
We remain committed to further margin improvements beyond 2026. For 2026, we expect research to show a steady performance across the year, while growth in health and education will be weighted towards the second half. With that, I'll hand back to Frank for an overview of our strategic progress.
Thank you, Alexandra. Before we move to Q&A, let me take a step back and talk about our strategic direction and how we're positioning Springer Nature for sustainable long-term growth. As you know, we operate in an attractive and resilient industry. Over decades, growth in global GDP has translated into higher investment in research and development, which in turn has driven growth in researchers, which leads to more published articles. This positive trend has meant that the large players in academic publishing have delivered consistent growth over many decades. Following the transition from print to digital and the internet, we're currently going through two key developments. One, the shift towards open access publishing. Second, the adoption of artificial intelligence. Springer Nature has been embracing both for many years, leading in the transition to open access and deploying technology and AI.
The implementation of this strategy by our people is enabling us to outperform the market, maintain our reputation, and allowing us to grow sustainably and responsibly. In the next few slides, I'll explain how. Let's turn first briefly to open access. As you know, open access delivers clear value to the research community. Open access articles are downloaded more, cited more, and reach broader audiences. This helps researchers and their institutions build their reputations and the funders of the research to share outcomes. Importantly, it also aligns the revenue model of publishers with the value we create, publishing trusted knowledge. Springer Nature has been investing in OA for more than 25 years. We have a strong portfolio of open access titles, including Nature Communications, Scientific Reports, BMC, and the Discover portfolios.
In 2025, as I said earlier, we launched 48 new OA journals across those portfolios. We are the furthest advanced of all traditional publishers, with over 50% of our articles now published open access. We also ended the year with 85 transformative agreements covering over 4,000 institutes, giving us scale across regions and funders. Our full open access article growth continues to run well ahead of the market. Again, in 2025, we gained market share. This is exactly where we want to be, leading the transition responsibly with our strong portfolio and disciplined investment. Now let's talk about AI, representing the most significant technology shift of our time. There is no doubt that AI will accelerate research outcomes.
This slide carries a couple of quotes from leading figures in AI and research, and they all agree that AI will accelerate the discovery of new knowledge. The realization of this enormous impact of AI is spreading throughout the research community. A recent industry study showed that usage of AI tools has grown to about 84% of researchers, up from 57% a year before. In a recent Springer Nature survey of researchers, 50% said that they believe AI will increase their research output in the next 12 months. Researchers publish their research to be part of the trusted scientific record, to gain visibility within their community, and for essential validation. It's therefore no surprise that when researchers choose where to publish, they look for the qualities that reinforce that trust. Strong brands with high editorial standards, reputation, and impact.
In the same survey here, 98% of researchers indicated that they will continue to publish in journals. In short, researchers see publication in the journal as a key part of being a researcher. For editors and peer reviewers, being involved in the publishing process contributes to their community, and it increases their status in that community. These factors become more important in an AI world in which researchers need sources of trusted information verified by their community. As AI becomes embedded in the research process, trusted and verified knowledge is also a requirement for AI to function effectively. Again, that background, our extensive global networks, domain expertise, and trusted brands are critical and differentiate Springer Nature from its peers.
Our 2,000 publishing staff with deep domain expertise, most having an advanced degree, work with 200,000 editors, typically academic experts residing at leading institutions across the world. They work with more than 1 million selected peer reviewers and over 2 million authors to serve a community of more than 10 million researchers dedicated to creating a better world. More than 30 million monthly users interact with those communities across our platforms. At the center of those communities are Springer Nature's more than 3,000 journals, including more than half of the top 50 journals by impact factor. All of the top three most cited journals in the world, Nature Communications, and Scientific Reports, and the largest number of Web of Science index journals of any publisher.
You can see our brands, our content, and our deep domain knowledge makes Springer Nature a trusted and sustainable partner at the center of the global research communities based on human validation and powered by AI. Let's turn to our AI strategy in a bit more detail. Our AI strategy is focused on three strands. First, we are using AI to transform publishing, building a frictionless publishing process, serving authors, editors, and peer reviewers. Second, we embrace AI to disseminate knowledge, to enable researchers to use trusted knowledge at every point and every place in the research process. Third, most importantly, we use AI to maintain trust, to protect the integrity of the scientific records, our journal plans, and the communities we serve. Let's take a look at each strand in turn.
As you can see on this page, in publishing, we have implemented AI solutions across the process from submission to publication. These solutions have delivered meaningful impact, and they have been fundamental in allowing us to scale, reduce turnaround times, and most importantly, deliver high satisfaction levels. Authors, as I always say, they like to do research and not write or read. They can spend more time doing research, and editors and reviewers can make better and faster decisions. We've implemented nearly 60 AI assists supporting screening, editorial evaluation, retention, and research integrity, and are evaluating more than 50 AI assists this year. Snapp is the backbone for our publishing process and the basis for the implementation of AI tools. More than 50% of our journals and our community of researchers, editors, and peer reviewers are using Snapp.
We are driving further adoption based on the industry-leading advantages it offers. I'll highlight a couple of examples. Our Journal Finder helps thousands of authors find the right journal for their paper, driving more submissions. In 2025, more than 500,000 clicks to submit were made in the Journal Finder. RP, our editorial scope and quality checker, was used by nearly 500,000 manuscripts in 2025, helping editors rapidly build confidence that those papers are based on sound science. Our Peer Reviewer Finder helps editors quickly find the right reviewers for an article, something which has definitely helped us scale to meet current high levels of submission growth across the portfolio. This AI assistant generated over 400,000 recommendations last year. Our transfer recommendation tool streamlines the cascading of manuscripts across our portfolio.
This means that when a good paper is rejected by one of our journals based on scope, we can immediately recommend the right home for that article in another Springer Nature journal. As I mentioned earlier, in 2025, we saw a greater than 40% increase in transfer recommendations. Those additional articles retained in our ecosystem added more than one percentage point growth to our portfolio. At the other end of the process, ACDCX, our typesetting automation tool, reduced production time and cost by automating journal and book typesetting. ACDCX processed over 3 million pages in 2025 at a dramatically lower cost per page compared to the 2023 baseline. Turning to opportunities for AI in the dissemination of trusted knowledge. As AI changes how knowledge is found and consumed, researchers, institutions, and companies need to know where trusted information comes from.
We have been upgrading our discovery platforms to include AI-generated summaries, reading recommendations, and chat interfaces. Solutions powered by AI also help us to play in other parts of the research process. Nature Research Assistant is just one example that has cross-publisher trusted reference data and content at its core. Nature Research Assistant harnesses our domain expertise to allow researchers to query that content to generate verified outputs to help them to read and write more effectively. With more than 21,000 high quality users in our extended beta program, we have learned a great deal on how AI can accelerate processes around evaluation and discovery of research. Feedback has been very positive, with customer satisfaction rate over 80% outperforming general LLMs.
We also see an important market opportunity in enabling responsible reuse of our knowledge, not just by researchers assessing our full text articles, but by making our content available throughout the research process, taking provenance and author attribution into account. We recently launched ARC3, our content licensing proposition. ARC3 is our solution for AI-ready content, where we provide three key attributes, highly relevant content, validated metadata, and content enrichment. It's worth mentioning in this context that actually 2/3 of global R&D spend is in the corporate sector, and as such, this is a big opportunity. Lastly, and most importantly, the third element of our AI strategy, research integrity. With higher research volumes, research integrity becomes even more critical. Our reputation is our most valuable asset. We have invested significantly in both people and technology in this area across four pillars, prevention, resolution, deterrence, and continuous assurance.
We've grown our dedicated integrity teams and deployed a suite of AI checks across three dimensions. First, tools to detect image manipulation, fake references, and identify fabricated text in papers. Second, tools to flag bad actors, and third, tools to spot the patterns associated with paper mills. Papers are currently subjected to approximately 20 AI checks, and we have more checks planned for launch this year. Finally, as we're active contributors to the STM Integrity Hub, we're working with other players across the sector to raise the standards. Now, before we move to Q&A, let me briefly recap. As you have seen, we're extremely proud to have delivered a strong business performance in 2025. Research is the key driver of that momentum. We are a trusted partner in the constantly growing global research ecosystem.
We've built on that foundation with targeted investments to drive the shift to open access. We're harnessing technology and embracing AI to scale our capabilities, support researchers and institutions, and protect the integrity of the scientific record and tap into new revenue opportunities. In 2025, we've again grown ahead of the market, while customer satisfaction scores, journal impact factors, and other external measures of the quality of our research have risen. This all shows that we're growing sustainably and responsibly as we outperform the industry. This makes us confident and positions us well for 2026 and beyond. With that, I would like to hand over to Tom for Q&A.
Thank you, Frank. We'll now move to Q&A. As a reminder, we ask each analyst to limit themselves to two questions initially. If you do have additional questions, we'll be happy to come back to you at the end, as we've done on previous calls. With that, I'll hand back to Anna.
Thank you very much. Dear ladies and gentlemen, if you are dialed into the conference call and have a question for our speakers, please press nine and then the star key to enter the queue. I repeat, the combination to state a question is nine, star. If you wish to cancel your question again, please press three and then the star key. For now, we are looking forward to your questions. Please press nine, star. There's a couple of questions incoming. The first question is from George Webb, Morgan Stanley. Over to you.
Yeah. Hi, Frank and Alexandra. I've got two questions to kick off with, please. Firstly, you continue to see very strong journal article submissions growth. You mentioned more than 30%. Feels like submissions growth has been running at very high levels for a while now. Could you talk a little bit about what you think has been driving that consistently strong pace of growth and the extent to which you think research submissions have already been benefiting from researchers using AI, or whether you think that's still ahead over the coming years? And then secondly, you mentioned also on the topic of AI, the different ways you're deploying it across the publishing process. You mentioned 50 AI assists under evaluation compared to the 60 that are already implemented and scaling.
It feels to me there's still a lot of margin and productivity gain to potentially play for, but then that's kind of set against the expectations for 2026 of 30 basis points of underlying margin improvement versus nearly 80 basis points in 2025. Do you have any view on the magnitude of potential productivity gains that future AI assists could offer for operating leverage compared to what you've already unlocked? Thank you.
Yeah. Thank you, George. I will take the first question, and Alexandra will come back on the second one. Basically, I think a valid question. I mean, I think in the past, we've always said that this industry was gonna grow at about 3.5%-4%. If you look at what happened over the last two years, at least from an article perspective, it has definitely done better. Now we know that there's kind of, you know, let's say strong underlying fundamental drivers of article growth, which are mentioned in the presentation around GDP, research and development, number of researchers, and driving number of articles. I think there are indeed two other factors that are driving article growth. I think one that we shouldn't forget is actually the regional mix.
I mean, I think historically, the U.S. and Europe were the primary drivers of article growth. Over the past couple of years, we have, of course, seen a huge surge in output from China, and now we're actually seeing other parts of the world also contributing and becoming part of the global research ecosystem. You know, India has shown tremendous growth over the past couple of years. We see a country like Turkey picking up. Latin America is playing a more important role. I think, yes, from my perspective, there's two factors that are driving increased article growth. That is one, the geographic mix, and next, we do already see the impact of AI and AI driving article growth.
I think you saw that also by the research that I mentioned, that close to 100% of people actually feel that, you know, they will see an increase in article output. I think against that background, we are still confident given, you know, the quality of our portfolio and the targeted investments that we're making, that we should be able to outperform that market.
Yeah.
With that, maybe Alexandra?
I think the second one. George, coming back to your question asked earlier about our margin expansion. As you said, we have been really pleased about the underlying margin expansion that we have delivered in 2024 and now also in 2025 with all this 80 basis points. What you can see currently, and that's the way how we have given this morning the guidance, where we expect a revenue increase 5%-6% and a margin expansion underlying of 30 basis points. We feel confident about this, seeing the submissions that we've received and also where we are currently with our renewal process.
In terms of the tools that we are using, we continuously expand the scope of the implementation, but you have to be aware first that's a process that we are going through, and we are also constantly adding new ones. Considering all of that together, I think it's early in the year. Currently, what we foresee for 2026 is a 30 basis points margin expansion.
That's very helpful. Thank you.
Thank you very much. The next question is from Steve Liechti, Deutsche Bank. Sorry if I mispronounced your name. Liechti, I believe.
Yeah. Hi there. Steve Liechti, for the record. I'll do two. Thank you very much. One is on contract renewals. Can you just flesh out a bit experience in the U.S.? I think that there's fewer renewals in the U.S. this year. Just correct me on that. Then secondly, just are you seeing any price pressure on transformative deals given the AI efficiencies that you're and platform efficiencies that you're instigating internally? That's the first question. Then the second one, can you just flesh out about ARC3, your licensing solution to corporates? I don't know how big that is in terms of revenue right now, but as you say, I think you alluded to, it could become a lot bigger.
Just any detail you can give us there in terms of potential for the long-term growth. Thanks.
Thank you. I'll take both questions, Steve. On the first one, if we look at the renewals, as Alexandra already mentioned, you know, we're actually slightly ahead of where we were last year. Maybe just to put things in perspective, as we mentioned in the presentation, 60% of our global research revenue is actually contracted, so that's typically contracts with a duration of three to five years. In the U.S., it's actually a touch higher. It's about 70%. Now remember last year we got a question, you know, following what's happening at the NIH, how does that impact you? We indeed said that in 2024, we actually had a relatively high share of active renewals in the U.S. for 2025.
Actually, if you look at. Typically, we do, given the duration of the contracts, three to five years, we do about 1/3 of our renewals every year. Now, if I look at the U.S., let's say 2025 or 2026 was a little bit lower, given that we had a lot for 2024, for 2025. If I look at 2026 for 2027, it's probably gonna be about 40%, so a little bit more than 1/3. Actually, 2027 to 2028 will be higher, and that's about 50%. Actually, from 2026 to 2027, it's not significantly higher than normal. Actually, the majority is more sitting from 2027 to 2028. Sorry, that was complicated dates, but I hope you could follow it. Then the other question was around TAs and price pressure.
Well, typically, if you look at the TAs, of course, there's always price pressure in our market. I mean, I've been in this industry for 30 years, and I've never experienced anything else. I think it just goes with this industry, and that's essentially the role of librarians to basically look at how can they, you know, optimize spend at their institute. I think what is different with TAs is that the attention tends to shift more towards the volumes that we publish as opposed to the price per article. Because at the end, volume growth is probably more important in the growth of TAs than, let's say, price growth. That's actually something we prefer as well. That's why typically we see that, you know, all our TAs are renewed. Secondly, we typically see that TAs are performing better than subscription revenue.
Subscription revenues in Springer portfolio was typically 1%-2%, because that's library budget growth, and TAs tend to grow 1%-2% better than that. That I think was the question. Sorry, I got carried away. The other question was around ARC3. Well, ARC3 is essentially our licensing proposition. It's. And as I mentioned, we do see a big opportunity in the corporate market, keeping in mind that actually 2/3 of global R&D spend is sitting with corporates. If you look at our revenues, and I think we're not that different from other research publishers, it is less than a fifth of our revenues at Springer Nature in research. There is a big opportunity.
ARC3 is essentially a proposition in which we sell or license our content for AI purposes. It's on the basis of an MCP server, which is great because that way we ensure that we can actually track usage and attribution. Essentially, we do it on the basis of an inference- RAG model. We don't, let's say, sell our content for training purposes as such, but we actually like to use our content to refine the answer set and make sure that the answer set is more precise. If you look at the revenue impact of ARC3, I think it's still early days. I mean, if I look at Springer Nature, most of our revenue growth today is pretty much driven by the journals portfolio. I think long term, this is, for us, definitely an interesting opportunity.
Can I just do one follow-up? To be clear, is that ARC3, is that books or is it journals and books?
It's journals and books. It's everything. Basically, the way we license our content for AI, typically, we actually make slices of our content because, you know, especially in the corporate market, they're typically interested in a specific part, whether it's maybe the material science part or the chemistry part or maybe the life sciences part. That's basically what we do with ARC3. We actually slice and dice the content. The nice thing is that what it does is it actually makes our content available throughout the process. Instead of just researchers reading articles, we can actually now also effectively make use of the supplementary material that we have together with the articles, like the research data.
Great. Thank you so much.
Thank you so much also from my side, Mr. Liechti, and sorry again for the mispronunciation. All right, moving on. The next question is from Nick Dempsey, Barclays. Please, over to you. Mr. Dempsey, can you hear us?
Oh, can you hear me?
Yes, we can. Hello.
You're now right in the middle of the 1.5x-2 x net debt to EBITDA target you set out at the IPO. You mentioned you're gonna stick with the kinda current dividend pattern. Do you have a pipeline of acquisitions which will prevent you going below 1.5 x net debt to EBITDA? Or you mentioned buybacks on the call, but do buybacks seem like the most sensible thing to do when you've got a very low free float and that's hurting your liquidity? Just some more thoughts about what you're gonna do with your cash as you've now reached your gearing target. Second question, just on, in the fourth quarter, you mentioned that books was helping organic growth to research that was pulling forward from 2026.
Can you help us out by quantifying that a bit? Roughly, what would Q4 research organic re-growth have been if you haven't had that books effect?
Okay. Thank you very much, Nick. I suggest, Alexandra, you wanna take the first one, and then we'll.
Yeah.
We'll talk about the books one. Yeah?
Absolutely. I'll take the first one. It's more about our capital allocation. Nick, you're right. Starting with the latter one, yes, we are authorized, or we asked for the authorization of the share buybacks. Definitely it will also depend on trading and liquidity to use that. From our perspective, and that's what I laid out in the presentation today, the best use for our cash flow is really to fund our organic growth, and that would always come at the first place. We invest in our strong balance sheet. Yes, we have this target range of 1.5x-2 x, and still have some benefits for us, to further deleverage and have this strong leverage ratio for refinancing purpose and also discussion about interest rates.
We continuously look at M&A. On one hand side, that can be accretive to our growth, but also has leverage through technology parts, but it has to be the right M&A that fits to our portfolio. Yes, as I laid out, we will remain committed in the term to progressive dividends.
Thank you, Alexandra. Yeah, maybe Nick, to come back to your second question about the growth and the pull forward in books. Yes, that definitely has an impact, but it had an impact on our print book line. If you look at the impact on the overall research growth in the last quarter, it was actually not really material. I think the primary reason why we saw higher growth in Q4 2025 was actually that the comparable in Q4 2024 was relatively low. Remember, exactly a year ago, we got the question the other way around. People then said, "Hey, your growth in the last quarter was a little bit lower.
Does that mean that your expectation for the year are coming down? Yeah, in both cases, I think it's fair to say that we are an annual business and our guidance is based upon annual growth rates, and we don't tend to look at quarter-over-quarter growth.
Okay. Thank you.
Thank you very much. Moving on. The next question is from James Tate, Goldman Sachs. Please, Mr. Tate, over to you.
Hi, Frank and Alexandra. Yes, James from Goldman Sachs. Got a couple questions, please. I think, firstly, you've guided to a midpoint of 5.5% growth for the group this year. Could you just help give us a flavor of the mix between the divisions? Do you expect health and education to be year-over-year broadly in line with market growth of 3%, and then I think this implies research growth of at least 6%. Is that the right way to think about it? Given research grew 7% in 2025, what are some of the moving parts that are driving this deceleration in growth? Is it mainly low share gains in open access or something else to think about? I guess secondly, we know that NIH are looking at ways to manage its publication costs.
I guess, where are we in that process? Are we still in the consultation phase? Any update on this and how you're thinking about it would be helpful. Thank you.
Yeah.
Yeah.
Thank you very much, James. Maybe I shall start with the second question, and then,
Okay. No problem.
Alexandra will come back on the guidance one. On the NIH, I think, you know, as we shared early on in the presentation, we are a globally diverse business. I think also in the past, we mentioned that the U.S. accounts for roughly a quarter of our global revenues and about 12% of articles, and half of those articles, about, let's say, 6%, comes from our U.S. government funding, and about 2/3 of those are funded by the NIH, so then you come to about 4% of our total articles being funded by the NIH. If you look at the situation of the NIH, I think on the positive side, of course, we've seen that the budget for the NIH for this year has been relatively the same as last year.
I think it's kind of a little bit higher than last year, so that's a positive development. Secondly, they also announced that they were gonna look at, let's say, the publication cost. They basically identified five different options. They asked the industry for feedback, which all happened in the last quarter of last year, and they initially said that they would come back with a decision before the end of the year. Now, they haven't done so. Then we heard that they would come back to that before the end of February, which they also haven't done, that as well, which we actually feel is a positive 'cause it means that the feedback that we and other, let's say, stakeholders in the industry have provided is actually taken seriously. So that's a positive.
I think at this stage, it's too early to say which option they're gonna take. We'll have to see. I think the only thing is that we should keep in mind that, let's say, the lead time between submission and publication is on average across the portfolio about 200 days. It gives you a bit of a sense of what that could mean for this year. I think on top of that, also something to keep in mind is that actually if you look at our average APC across the portfolio, it's just a little bit more than EUR 2,000 per article. Only, let's say, less than 5% of the journals that we have have actually an APC above $5,000. I think that just puts it a little bit more in perspective.
Okay. I'll take the first question about our revenue guidance. Yeah, we had set the guidance between 5% and 6%, and our expectation is that all segments will contribute to this strong growth. It's early in the year, but when we look at our research business, we have the continued strong growth in our submissions and also the good progress, as you alluded before, in the renewals. That gives us really confidence about the growth trajectory in research. With regards to health and education, our current assumptions are broadly in line with their markets, so we would expect, I would say, a kind of normal year for both segments.
Great. Thank you.
Thank you very much, Mr. Tate. The next question is from Conor O'Shea, Kepler Cheuvreux. Your line is now open. Please go ahead.
Yes. Thank you. Thank you for taking my questions. So my two questions, just to come back on some of the granularity in the implied guidance in the research business. If you just in the more, say, volatile revenue lines, books and services, at this stage for 2026, are you assuming again, you know, overall modest growth in books and what are you assuming for services? That would be helpful. Second question is on education. I understand there's a lot of moving parts there, including currency and so on, but would you at this stage expect another year-on-year decline in margins in 2026? Thank you.
Yeah. Thank you very much, Conor. I'll take the research question, and then Alexandra will come back on education. On research, yes. I mean, if you look at the way we built up our budgets, of course, it is at the product level, and we assume certain growth rates. In this case, we would expect our books to grow, you know, driven by digital. You know, as I mentioned in the presentation as well, we expect print to continue to decline over the long term. We expect actually our services part of the business to show good growth, very much driven by solutions like ARC3, but also our TDM. Overall, if you look at our research business, I think we feel confident about our research business. You know, it's strong submission growth.
We've gone through a strong renewal cycle so far. Basically that comes back to what Alexandra earlier said about our guidance. You know, we feel that we're on a good trajectory. At the same time, it is early in the year. I think that's something that we should not forget. That also explains why if you look at our guidance range of 5%-6% and the margin improvement, that's where we, at this stage of the year, feel kind of safe. On the education one, maybe Alexandra, you want to take that one?
I think the education side. Just kind of when I look at education in 2025, because we have seen a strong momentum in the curriculum business, a little bit softer as we have been in the last year of ELT. We have seen the shift in with regard to South Africa. With respect to underlying margin expansion, I really have to say we have been pretty pleased to see the progress that also education had on the Elevate program. When I now look into 2026, I expect this will be most probably a normal year for education. Also I see the additional potential for underlying margin improvements. As we also have said, this is a strong business.
This can leverage technology, and we also have a very strong cost focus. When there are no surprises in their respective markets, definitely we see there also for education and in line with market development.
Okay. When you say underlying margin improvement, that excludes their currency translation.
Yes. Underlying would be excluding exchange currency.
Okay. Great. Thank you.
Thank you very much. The next question is from Konrad Zomer, ABN AMRO-ODDO BHF. Please over to you.
Hi, good afternoon. Thanks for taking my two questions. The first one on your research division. You reported an acceleration in Q4, both in terms of revenues and in terms of margins. Can you maybe give us a bit more insight as to where that acceleration came from? My second question is on your guidance for 2026. I remember last year, your guidance was deliberately a little bit cautious in March, and as a result, you raised it both in Q1 and in Q2. Obviously you were then on track as your first year as a listed company. Now you're into your second full year. Has the process of coming up with guidance changed, or are you still deliberately cautious at this point of the year? Thank you.
Should I take the first one?
Yeah.
Konrad, let me take the first question about research and the fourth quarter. You're correct. We have been very pleased with the 8.6% growth we've seen in research in Q4, but that's something that we had expected. Maybe you can remember, at the same time last year, we have seen a weaker performance in 2024 Q4 about research because we had missed some of the one-timers that we have seen in 2023. Bottom line, what you can also see in research, there's usually in the fourth quarter a bit more of transactional revenue that is causing some of those wins.
In particular, this revenue has also been driving a better margin performance in the fourth quarter that also has some topics around some smaller aspects that has been positive for us as part of the closing process. This is the kind of justification of the margin development.
Yeah. Which in a way is a good segue if you look at, you know, where we are in terms of guidance and how we look at our guidance. I think, you know, in principle, we would expect a continuation of our performance. Last year we achieved 6%. Now it is early in the year, and as Alexandra said, there are parts in our business that are a little bit more transactional. We have to see how pipelines will build up for those product lines. I think education is a good example. I mean, actually, if you look at education outside Southern Africa, we would have actually delivered market growth in line with market growth. I would say it's. I don't think it's being cautious.
I think it's just being realistic, you know, being early in the year, knowing that, let's say, a significant part of our revenues, we have good visibility. Being, let's say, article growth with the lead time and submission growth, looking at our renewals, but of course, also parts of the business that are more transactional, and we have to see how those will develop throughout the year.
Okay. Thank you.
Thank you very much. At the moment, there are no further questions in the queue. With that, we are closing the Q&A session. There is a follow-up from Steve Liechti. Please over to you.
Thanks. I just wanted to hear you say my name again. Thank you very much.
I tried my best.
No, that's it really. I got two really quick ones. One, just on the Nature Research Assistant, can you just remind us whether you're gonna be able to charge people for that, or are you gonna bundle it into anything or on how that works? Sorry, maybe I should remember that. Just anything you can flesh out there. Secondly, you mentioned on the tax rate, which is high against international peers, and you said you might be able to bring that down, I think you said in the midterm. Just in terms of what is a realistic expectation for your sort of midterm tax charge and anything that's not too technical in terms of how you could achieve it? Thanks.
Yeah. Maybe Steve, quickly on the NRA one, Nature Research Assistant. It's important to recognize that actually NRA is, it brings together things that we're also doing in other parts of the business. If I look at some of the stuff that we're doing on Snapp, actually, that is also part of NRA. In that sense, it's not just a standalone product, but it's part of our total AI initiatives. Now, as we have said last year, we said, you know, also, if you look at the adoption of new products in our market, those tend to go relatively slow. It's interesting that the adoption of technology goes extremely fast. The adoption of new products actually goes a little bit slower, which I think is driven by, let's say, you know, reputation, et cetera, things like that.
For NRA, we don't expect it to have a significant revenue impact, you know, this year or next year. I think if we look at pricing in the long run, I would expect it to either be part of our institutional agreements, if we look at transformative agreements or subscription agreements. I think we also see an opportunity to maybe price it as a standalone consumer type of model.
All right. I'll take the tax rate. If we still see the current tax rate for the midterm between 32% and 34%, I know at the time of the IPO, we have guided 32%-35%. Where we stand, I would say it's rather 34%. The two topics that we will address after the midterm, reduce this tax rate are primarily about interest deduction limitations that we currently face in Germany and U.S. When we implement those projects that we have currently in mind, we will be again in a position to deduct interest in those countries, and that will have then an impact of around 400 basis points for us.
Great. Thank you.
Thank you very much. So with that, I'm closing the Q&A session for today. Thank you very much for participating, dear ladies and gentlemen. I'm handing the floor back over to the host.
Thank you very much, everybody for participating. I hope you've heard that we're actually really proud of what we achieved in 2025, our first year as a listed company. I think we have delivered great results, but I think equally important, we've really made good progress on our strategy. I think you've also seen that we have quite a lot of confidence for 2026. With that, I want to thank you for your time and attention and wish you a lovely day, including you, Steve.