Ladies and gentlemen, good morning and welcome to today's press call and telephone conference at Siemens AG. To begin, we would like to inform you that this call is being recorded and webcast. After the presentation, there will be time for questions. If you would like to ask a question, please press the star key followed by the one key on your phone to register to ask a question. With that, I'll turn the floor over to today's host, Simon Krause, Head of Media Relations. Over to you, Mr. Krause.
Good morning and welcome to today's telephone conference for the fourth quarter and the results for the 2026 fiscal year. Like in the previous quarters, I'm pleased to welcome our CEO, Roland Busch, and our CFO, Veronika Bienert. Welcome. Now, before we begin, a couple of notes as always.
This morning, we released our results for the fourth quarter and the preliminary results for fiscal 2026. The presentation, the speeches of our managing board members, and all other materials are available on our website at siemens.com/press. You'll also find a recording of this conference call there. Now let me give you a brief overview of the agenda. Following the presentation, Roland and Veronika will be available to answer your questions, and our press call will end at 9:00 A.M. And as always, I'd like to briefly draw your attention to the safe harbor statement, which you'll find at the beginning of our presentation. With that, I'll turn the floor over to Roland.
[Non-English content]. Thank you so much. Good morning, everyone, and welcome to today's telephone conference. Fiscal 2026 was another extraordinarily successful year for Siemens. We delivered outstanding order growth, supported by high structural demand in our key markets. Revenue growth also accelerated further and reached 8%. Earnings per share before purchase price allocation accounting, or EPS pre PPA, grew even faster, increasing by 14%. This growth reflects our strong operational performance. Free cash flow, the ultimate measure of our performance, was up 12% to a record EUR 12.1 billion.
And once again, we reached our guidance. We also made progress in our strategic key performance indicators, or KPIs. Our digital business continued to show strong growth. Revenue was up 16% to EUR 10.9 billion. As a result, this business already accounts for close to 20% of Siemens' total revenue, and this is without Siemens Healthineers.
We're fully on track to double our digital business revenue by fiscal 2030. A key catalyst for this rapid growth is the continued strength of the software business at Digital Industries, or DI. Revenue now exceeds EUR 7 billion. Annual recurring revenue, or ARR, was up 12% year-over-year. Very importantly, artificial intelligence is driving growth at DI's software business and beyond even more strongly. In fiscal 2026, around 5% of the revenue generated at DI's software business was already attributable to AI or industrial AI, but the opportunities created by industrial AI go well beyond software. For us, industrial AI is the defining growth opportunity, and Siemens is ideally positioned to win in this sector. Why Siemens? Because we have a clear head start.
For years, our strategy has been to combine the real and the digital worlds, and industrial AI is making this combination more valuable than ever before. Industrial AI requires more than powerful models. It requires real-world industrial data. It requires deep domain know-how. How do engineering, physics, and complex systems interact? This is precisely where Siemens is unrivaled. We have a strong, unmatched portfolio in industrial AI. Our software enables customers to build the most comprehensive physics-based digital twin of their products and their production processes, all the way down to the simulation of molecules and atoms.
We know exactly how things are built in the real world. We give AI the context it needs to act in the industrial world, and we do this in business-critical environments. Our customers trust us here. Our AI agents are already transforming the engineering workflows from design to the shop floor.
We have a global footprint covering the industry, infrastructure, and Mobility markets. Our Eigen Engineering Agent, the first of its kind, demonstrates how fast we can be. Since its launch in April, we have expanded the agent across the entire engineering value chain. We have broadened its reach across industries such as semiconductors and consumer packaged goods. And most recently, we expanded it into automobile manufacturing. Customer adoption is very strong because what matters in industrial AI is not only what technology can do.
What matters is the value our customers create with it. Take our customer Andritz, for example, a global supplier in the machine building sector to several key industries. The company has around 30,000 employees at more than 250 locations in 80 countries. Andritz is already using many products from our automation and software portfolios.
At a company of its size, engineering is performed at a scale of several million hours every year, followed by several million more hours for installation and commissioning. Improvements enabled by generative AI have the potential here to drive massive gains in productivity, cost efficiency, and overall competitiveness. That is why the metals unit at Andritz is using our Eigen Engineering Agent. The AI agent has the potential at Andritz Metals to fundamentally change how engineering is done, because today, the information comes to the engineers.
They do not have to search for it, and that information is grounded in the company's own real-world industrial data, its standards, and engineering know-how. Because today, engineers can generate and refine software for automation interactively. They do not have to write it manually. Every result is checked against the relevant specifications, also by people, so that the engineers can trust the AI.
Instead of navigating complex documentation, engineers can ask questions and receive immediate answers. All this is based on a deep understanding of specifications, drawings of plants, and references from other factories with decades of engineering experience. At a customer this big, the impact is significant. Thanks to our Eigen Engineering Agent, Andritz Metals can now perform the same engineering tasks twice as fast. And because the customer's individual engineering rules are captured and reused, the capacity for these processes grows without adding headcount. For us, that is scaling at its best.
Machine building is a key sector for other growth verticals like semiconductors, aerospace, defense, and robotics. For us, this is another opportunity to scale. We use digital twins, data, networked automation, and industrial AI in order to further increase the value we deliver to our customers and in order to grow with them.
In short, we scale industrial AI by combining the real and the digital world. Andritz, our customer reference, illustrates the power of our ONE Tech Company program. We are accelerating our profitable growth with five growth levers, all supported by our highly synergistic portfolio of hardware, software, and services, and by stringent investments. We are turning our strength in digitalization and AI into growth in attractive verticals and regions, and our strategic growth levers are reinforcing each other. Lever one, grow AI. AI in operations has now moved from piloting to our customers to scaling.
The semiconductor sector is rapidly expanding its capacities. Customer demand for digitally enabled manufacturing and AI-driven productivity in facilities and operations is enormous. GlobalFoundries is one example. Our partnership for this company's new AI-enabled control center in Singapore proves why customers trust Siemens in critical operations.
We combined GlobalFoundries' semiconductor manufacturing expertise with our strengths in industrial digitalization. The AI-enabled control center will, upon full deployment, integrate more than 1 million critical data points to support capabilities such as AI-driven predictive maintenance. The customer detects potential issues earlier, reduces unplanned downtime, and can respond faster. Lever two, grow digital. This week, we signed an agreement to acquire Lauterbach, a leading supplier of debugging and tracing solutions for embedded software and system-on-chip architectures. We want to combine Lauterbach's industry-leading debugging technology with our hardware-supported verification solutions and thus further expand our competitive advantage.
This combination will enable customers to develop and test their embedded software earlier, even before the real chip exists. The planned acquisition is another key step in the growth of our digital business. The acquisition of Altair and Dotmatics is also accelerating our growth in this area.
It is enabling us to accelerate our transformation into an AI-powered technology company. Altair and Dotmatics are both developing very successfully. Their integration is fully on track. At Altair, we have met our ambitious revenue targets for fiscal 2026 and are realizing the first revenue synergies. The combined organization is operating as one team. We are leading the way in AI-powered simulations through physical AI. At Dotmatics, we are also seeing strong momentum and a robust synergy pipeline. Revenue is developing as planned, and profitability is even exceeding expectations. Lever three, grow verticals.
Our data center team is a prime example of how much impact a ONE Tech approach can have. In fiscal 2026, orders grew by triple digits to more than EUR 8.5 billion, while revenue was up around 60% to EUR 4.6 billion. Growth in this sector is super fast, and we are innovating just as fast through R&D and partnerships.
Just last quarter, we announced that we are teaming up with strong partners to develop solid-state transformers and supercapacitors. Why did a key company in the semiconductor industry choose Siemens to design next-generation AI factories? I am talking about NVIDIA here. Because we are leaders in direct current switching, because we are masters of industrial automation, and because we are the only company that can deliver the full digital thread for these factories. Lever four, grow regions.
Our customers think globally, and so do we. Every market is different, but the winning formula is the same everywhere. Technology wins, speed wins, resilience wins. The U.S. remains a key growth market in automation. We are differentiating ourselves through software-defined automation, Industrial Edge, integrated hardware and software offerings, and our strength in large-scale greenfield projects. Our U.S. automation business grew 12% in fiscal 2026, and we expect even faster growth in fiscal 2027. I said speed wins.
This aspect is of utmost importance in China. Our Chinese customers are fully embracing Siemens Xcelerator and industrial AI. Our local for local portfolio in automation is proving the power of speed, faster innovations and development, lower costs, and strong customer adoption. In China, revenue with this portfolio grew more than 25% in fiscal 2026, and we keep launching new products. Now, we're seizing the opportunity. We're going global with selected products in order to grow faster. All this is ONE Tech at work. Finally, our fifth growth lever, the people in our company.
They remain our strongest growth driver. That's why we're fostering their expertise and leadership abilities and developing our organization further so that they can actively drive growth and convert the potential of industrial AI into value added. Let me bring all this together. Our ONE Tech Company program is now in full swing.
Industrial AI is at the heart of our strategy. We scale industrial AI by combining the real and the digital worlds for our customers, for Siemens, for our shareholders. The targets of our ONE Tech Company program remain unchanged: stronger customer focus, faster innovations, and ultimately, higher profitable growth. We're entering this phase of our transformation from a position of strength, and the planned spin-off of Siemens Healthineers is fully on track. It's a decisive step to reinforce Siemens' leading role in industrial AI.
We see major opportunities ahead for our business. We see powerful growth trends and strong key markets. That's why we're raising our midterm targets. We now expect comparable revenue growth of 7%-11%, up from the previous 6%-9%. For earnings per share, or EPS, we're raising our growth target to double digits.
We've started the new fiscal year with strong momentum and a clear plan that we can build on. Based on our record backlog, structural demand for AI and digital infrastructure, public investment programs, and a further pickup in our industrial markets, we expect comparable revenue growth of 9%-11% in fiscal 2027. We expect EPS pre PPA in the range of EUR 10.20-EUR 10.80, up 13% at the midpoint. This guidance excludes Siemens Healthineers. With that, let me hand over to Veronika.
[Non-English content ] Thank you, Roland. Ladies and gentlemen, good morning, everyone. Allow me now to jump right into the details of our fourth quarter and our outlook for FY 2027. We closed fiscal 2026 with an outstanding quarter, setting records across multiple key metrics. Orders reached an all-time high of EUR 34.8 billion, driven by outstanding business momentum at both Smart Infrastructure and Mobility. Digital Industries finished the year on a strong note as well.
Revenue growth accelerated to 10% with healthy growth across all businesses, led by Smart Infrastructure. A strong top-line performance translated into a record profit of EUR 3.9 billion in our industrial business, reaching a profit margin of 16.8%. Free cash flow was another highlight in the fourth quarter, coming in at EUR 5.6 billion, and this result once again demonstrates the strong cash generation power of our businesses.
Let's now walk through the key drivers behind these results. In the appendix to the presentation, you will find the full set of financial details in the familiar format. With strong order bookings at the end of the fiscal year in both automation and software business, order growth at Digital Industries accelerated to 15% and exceeded our expectations. Orders in DI's, Digital Industries, automation business were up 16% with a book-to-bill ratio of 1.02x.
Growth in discrete automation continued, supported by the aerospace and defense industry and by the electronics industry. Increasing investments in AI infrastructure and in semiconductors also drove additional industrial demand in machine building. From a regional perspective, the United States showed a particularly strong momentum. Driven by project business, orders in the automation business were up 61%. Also, Germany benefited from larger orders, for example, in life sciences.
Germany also saw improving investment sentiment in export-driven industries. These factors contributed to overall order growth of 19%. In China, order intake rose 8%. Following strong momentum in July and August, growth in order intake in DI's automation business in China softened in September, and this slowdown partly reflected order shift into October. In addition, in October, we saw strong customer activity ahead of announced price increases.
Roland already highlighted the continued strength in DI's software business. The fourth quarter marked an exceptionally strong finish to a successful fiscal year 2026 overall. With orders of EUR 2.9 billion, we set a new quarterly record. Several larger orders booked in product lifecycle management, or PLM, and in electronic design automation, or EDA, drove this performance. In addition, we successfully converted projects from our sales pipeline into orders, including early renewals with significant upselling.
Major wins included high-profile customers across semiconductors, electronics, and industrial end markets, such as aerospace and defense. Both the automation and the software businesses at DI successfully converted strong demand into revenue, exceeding our expectations. Automation business grew 8%, led mainly by the short-cycle businesses in China and the United States. Both revenue in software business and annual recurring revenue, or ARR for short, increased 12%. Digital Industries' profit margin improved 540 basis points year-over-year.
Excluding severance charges, its profit margin increased 200 basis points. High volumes in automation business and a better-than-expected business mix drove this improvement. A clearly net positive economic equation also supported Digital Industries' profit margin. In addition, there were some slight currency tailwinds. In its software business, Digital Industries continued to benefit from the smooth and successful integration of Altair and Dotmatics.
Integration-related costs weighed on Digital Industries' profit margin by 60 basis points, in line with our expectations. Digital Industries entered fiscal 2027 on a strong foundation with confidence in its trajectory to seize growth opportunities in a solid market environment. Therefore, for Digital Industries, we expect comparable revenue growth in the range of 8%-12%. Double-digit growth in DI's software business and accelerated growth in its automation business will drive this increase. We are building on our leading portfolio, enhanced by AI and by our optimized sales setup. At the same time, industry-wide pricing actions in the automation business may lead to some volatility in customer order phasing.
For Digital Industries, we expect a profit margin for fiscal 2027 between 20% and 22%. In DI's automation business, we will see benefits from higher volume, while we diligently offset cost headwinds with productivity measures and adequate pricing.
In DI's software business, the transition to the software as a service, or SaaS, model has now been completed. Furthermore, we expect increasing contributions from Altair and Dotmatics, also supported by fading integration-related costs. DI's software business continues to execute fully in line with our business plan and will strongly contribute to FY 2027 profit margin improvement. Now, let me just make one important point on phasing. FY 2027 is a strong renewal year for our EDA business, with the sales funnel again concentrated in the fourth quarter.
For the first quarter, we expect Digital Industries' revenue growth to be in line with our full-year guidance, primarily driven by its automation business, while we expect its software business to grow moderately. We expect a Q1 profit margin for Digital Industries to be around 20% in FY 2027.
Let's now turn to Smart Infrastructure, or SI, where the team once again delivered an outstanding performance. With growth of 39%, Smart Infrastructure reached record orders of EUR 8.7 billion. This increase was mainly due to sharp growth of 62% in the electrification business and 45% in the electrical products business. Both businesses benefited from data center orders, which again achieved triple-digit growth, driven by the United States and Europe. In addition, the buildings business posted double-digit growth in orders. In the fourth quarter, Smart Infrastructure's revenue growth accelerated to 15%, reflecting the successful ramp-up of production capacities and the extension of our supply chains.
Both electrification businesses executed flawlessly, growing by around 20%. All three businesses contributed to further operational margin expansion. The electrification and the electrical product businesses benefited from economies of scale and continued to operate at benchmark levels.
The buildings business again increased its profit margin, supported by stringent cost management and a strong product business. Smart Infrastructure's profit margin of 20.3% included a positive effect of 50 basis points from U.S. tariff refunds. Looking ahead, we continue to expect healthy demand in SI's end markets. Data centers and power utilities are and will remain key growth drivers. Data centers are now Smart Infrastructure's second-largest vertical, accounting for around 20% of revenue.
At more than EUR 25 billion, our order backlog at Smart Infrastructure provides excellent visibility. For FY 2027, we expect comparable revenue growth in the range of 11%-13%. We expect the buildings business, which accounts for roughly 35% of Smart Infrastructure's revenue, to grow at a rate in the mid-single-digit range. The focus in this business will continue to be on selectivity with regard to orders and on expanding the profit margin.
Meanwhile, we expect SI's electrical products and electrification businesses to grow by rates in the mid and high teens, respectively. Both businesses are further expanding capacity globally and are executing backlog rigorously, especially in business with data centers. For FY 2027, we expect a further increase in profitability for Smart Infrastructure, with a profit margin in the range of 19.5%-20.5%. Key drivers will be economies of scale and a net positive economic equation. At the same time, we continue to invest in additional capacity across our low and medium voltage businesses.
For the first quarter, Smart Infrastructure expects revenue growth around the upper end of its guidance range for FY 2027, supported by a low basis of comparison for the electrical products business from the first quarter of FY 2026.
We expect Smart Infrastructure's Q1 profit margin to be in the lower half of its fiscal year guidance range, and this is in line with the usual seasonal patterns. Let's now look at Mobility. Mobility achieved record orders, exceeding EUR 10 billion, and secured several major contracts in the rolling stock business. Attractive long-term service agreements accounted for more than EUR 4 billion. In the fourth quarter, Mobility's revenue grew 9%, supported by strong backlog execution in the rolling stock and customer services businesses.
A favorable business mix drove the profit margin to 9.5%. As expected, Mobility delivered a material catch-up in free cash flow, reaching EUR 1.4 billion in the fourth quarter. This development resulted in an excellent cash conversion rate of one point zero three for the full fiscal year 2026. In early November, Mobility successfully closed the acquisition of key businesses of the MERMEC Group.
This transaction further strengthened Mobility's rail portfolio, particularly in diagnostics and measurement technologies. We expect MERMEC to contribute around 3 percentage points to nominal revenue growth in fiscal 2027. Looking ahead, we see attractive market opportunities arising from decarbonization, digitalization, and modernization efforts on the part of rail operators. Mobility's project pipeline remains healthy. We again expect a book-to-bill ratio well above one. Mobility's order backlog stands at more than EUR 65 billion, again with an improved gross margin.
Due to the long-term nature of many projects, revenue conversion will occur progressively over an extended execution period. Many of the recently booked projects will only start converting beyond fiscal 2028, and revenue recognition in the service business will start even later. For fiscal 2027, we expect comparable revenue growth in the range of 6%-8% for Mobility. Furthermore, we expect Mobility's profit margin to be between 8% and 10%.
While we assume operational progress, the profit margin will be held back by integration-related costs for MERMEC and higher transformation costs. At the same time, we are optimizing our business structures and making them more efficient. For the first quarter, we expect Mobility's revenue to increase at a mid-single-digit rate. We also anticipate that the profit margin will be in line with Mobility's full-year guidance. Ladies and gentlemen, as Roland already mentioned, a key highlight for the full fiscal year 2026 was our record-high free cash flow at the group level. We achieved a free cash flow margin of 14.5%.
In this way, we again demonstrated the cash-generating strength of our businesses. We have now delivered double-digit cash flow returns for seven consecutive years. Our current share buyback program is also running very successfully. Since its start in July, we have already bought back EUR 1.7 billion worth of shares.
Our rock-solid balance sheet enables us to return substantial cash to our shareholders. In line with our progressive dividend policy, we will propose to our annual shareholders meeting a dividend of EUR 5.50 per share for fiscal 2026, up EUR 0.15 year-over-year. One thing is clear in this regard: even after the Siemens Healthineers spin-off, we will, on the basis of our current dividend proposal, remain firmly committed to our progressive dividend policy. Let me also give you a brief update on the planned Siemens Healthineers spin-off.
We plan to transfer a 34.1% stake in Siemens Healthineers to Siemens AG shareholders at an allocation ratio of one Siemens Healthineers share for every two Siemens shares. On December 10 this year, we will publish the comprehensive joint spin-off report in conjunction with our notice of the annual shareholders meeting.
Comparable key figures for Siemens AG, reflecting the classification of Siemens Healthineers as discontinued operations, will follow soon thereafter. We expect that after the annual shareholders meetings of both Siemens and Siemens Healthineers in February 2027, the spin-off will take effect in the late spring, subject to the necessary filing and registration process. We plan to report Siemens Healthineers as discontinued operations starting with the results for the first quarter of fiscal 2027.
Once the spin-off has taken effect and deconsolidation has taken place, our remaining stake will be accounted for using the at-equity method. In the medium term, we intend to reduce our shareholding in this company, Siemens Healthineers, to being solely a financial asset. Ladies and gentlemen, Roland already outlined our ambitious group outlook for fiscal 2027. As usual, the appendix to the presentation includes the details of our assumptions for this outlook.
Targeted investments in research and development and expanding our capacities are creating the prerequisites for scaling industrial AI and executing our growth plans. At the same time, we will stay focused on rigorously driving productivity efforts across our businesses and functions. As a result, Siemens is well-positioned for the next phase of accelerated profitable growth while maintaining strong cash generation. Thank you for your attention. We are now looking forward to your questions. With this, I hand back to Simon Krause.
[Non-English content ] Thank you, Veronika, and thank you, Roland. Veronika already said that we have time to field any questions you may have through to 9:00 A.M . For technical reasons, we will also be able to answer questions in English, but they will not be interpreted. Now we would like to hand over to our operator. Thank you very much. If you would like to ask a question, please press star followed by one on your keypad. We will wait a moment for the first question.
[Non-English content ]
The first question comes from Axel Höpner from Handelsblatt. Please go ahead, sir.
Thank you. Good morning. I have two questions. First of all, at NVIDIA, do you see a cluster risk for the entire tech industry against the backdrop of the current news? The second question I have is if you can mention further KPIs with respect to AI, such as how many AI solutions are already in use, how many customers are using them, and what the total share of AI would be in your business. Thank you.
With respect to the first question, of course, we do have a certain cluster risk, which has been in existence for a couple of years. Almost all GPUs that are installed today come from NVIDIA. This has been the case for years. This is mapped by the development of NVIDIA. That was the first point.
Moving on to the second point. No, it doesn't apply to us, because the number of GPUs that we buy or need to train our models is not really all that great. Another issue I'd like to mention is that if you look at the other offerings, such as the stacks and the GPUs, if you go farther up, Omniverse, CUDA, et cetera, are all solutions which we use. We have partnerships.
We also have other solutions in the making. It's a very good partnership, but we have other good partnerships that have proven themselves, one of which we presented in Erlangen when it comes to the automated fab, for example. Against that backdrop, for Siemens, that's not an issue. Now, of course, the world does depend on NVIDIA GPUs. That hasn't changed. There are other efforts.
We have other GPUs that are being developed, TPUs that are being developed, and other alternatives. When inferencing is done, you can also move to CPUs. That's another trend. KPIs. As we said before, 5% of our software revenue is from AI. We also have native AI offerings. That's also encompassed in that figure. This share is not small, and it's growing. If you look at our own engineering agent, it was recently launched. It is growing at a very fast clip. It's called Eigen. We are demonstrating very high growth rates, especially in China, looking back at just the first two quarters. So momentum is building. The question is how these AI solutions can be delineated from the rest. We have the monetization, which is not always direct.
If it's a native product and engineering agent, that's one thing, or it can also be a part of a solution within your software portfolio. That leads to the 5% share, which is growing fast. Over time, we will also think about other monetization models, and then perhaps we'll be able to tell you some more specific figures. But it's still early days.
Thank you.
The next question comes from Michael Flämig from Börsen-Zeitung.
[Non-English content]
Ms. Bienert, Mr. Busch, a question on deconsolidation of Healthineers, major special items. Can you give me an order of magnitude, transaction costs? Could you give us some figures here? Secondly, on industrial net debt, it was reduced quite a lot, which reminds us of a bank with an electrics department attached to it. You gained some free space here, some leeway. How are you going to use it? Are you considering M&As? Third question, your medium-term planning, that you increase your revenue, the difference between 7% and 11%, it used to be 6%- 9%. Is it simply because of the deconsolidation of Healthineers, or what is the operational effect? The effect or the impact that means that the operational core business is now performing better since you deconsolidate Siemens Healthineers.
Hello?
[Non-English content]
[Non-English content]
I'm sorry, Mr. Flämig. Here I am. I'm sorry. Sorry. I didn't switch on my microphone, says Veronika Bienert. Your question was about transaction-related effects in the business year, fiscal year, whether they are included in the guidance or not. Okay. Right. The situation is as follows for Siemens Healthineers. In the reported actual figures for fiscal 2027, up to fiscal 2027, up to the spin-off, the resulting free cash flow of Siemens Healthineers up to 2027 will remain in discontinued activities, and this has a retroactive effect also for the reporting periods that we presented. After deconsolidation, Siemens Healthineers will be reported according to the at-equity method in our accounting, and there, the Siemens share will be in the result of Siemens Healthineers. It's going to be a separate item in continued activities.
Now, you also had a question on certain transaction-related effects, the deconsolidation gain, for instance. That's something which will mainly be on the basis of the market value, the fair value of Siemens Healthineers as compared to the book value, the value carried in the books in relation to the non-controlling interests. This is an assessment as of a certain effective date, and this is why it will be recorded at the time of deconsolidation. Because of those fair value assumptions, it's very difficult to give you any kind of forecast on the amount of this item.
On net debt, you can read it both ways. I mean, it's great to have a strong balance sheet, but it doesn't necessarily mean that you're a bank with an electronics department still attached to it, or a whole group producing electrics goods.
I think that's something that we have left far behind. Well, EUR 8 billion new orders, EUR 83 million revenue, growing at 8%, EUR 4.5 billion, well, additional revenue between the two years, and we deliver cash flow of EUR 12.1 billion. I mean, that really is a topic where you might say, "What are we going to do with all that money?" Now, of course, we will once again pay out a dividend.
We have a share buyback program, which we are going to accelerate, as you just saw, and we are also talking about M&As, smaller activities in the area of M&As. We are not going to go crazy simply because we have so much money. We are going to proceed with the same caution that we exercised in the past. But of course, we have plans for inorganic growth as well.
As far as our plan is concerned, for the fiscal year, we have a guidance of 9%- 11%. That's one part, the medium term, and that was your question, 7%- 11%. In fact, you saw that the previous 6%-9% , that was a spread of 3%, which has now been increased to 4%. The 7% at the bottom are just in case if in the medium term, if anything happens in the market where we say we are going to stay within a certain range. But of course, our ambition is to achieve the, well, upper half of this range derived from this. If you ask me whether it's a mix, has it got anything to do with the fact that Healthineers was now separated?
Well, growth was below average, but just look at our existing portfolio, look at those figures, and then you're going to see that maybe we are slightly better than the medium-term guidance that is published by our competitors.
Thank you.
Then we have the next question from Christof Rührmair from dpa. Go ahead, please.
Now you should hear me. Good morning. Excuse me. If I come back to NVIDIA, but now there could be some concern concerning their leadership crisis, and it's not about the purchasing of GPUs, but the real concern is really the cooperation and the individual projects you might have together with them. Are you worried? Are there any concerns if the focus in a company might be on something like this in that company, that it could actually also have an effect on you, the problem at NVIDIA?
We are talking about a relatively big company with a large team, is the answer. Our daily work happens together with people, somebody who is responsible for Omniverse or other people. The projects we have, they will continue independently of what is happening at NVIDIA. Of course, the question is, what is going to happen to the leadership at NVIDIA? That is a question for the board of directors of NVIDIA.
The company is going to continue in any case. I believe that we are too much focused on one individual, and that is, of course, result of all the hype in the media. Our work is with people, two, three levels below that on the shop floor. We have NVIDIA people working together with us on our sites. We have joint projects, and honestly, I do not see any kind of influence.
It is a partnership between companies.
Thank you.
The next question, Clara Thier from Handelsblatt Media Group. Go ahead.
Hello and good morning. Two questions. First of all, could you give us some figures that demonstrate that the ONE Tech Company program strategy has now created some initial effects, more cross-selling, high efficiency, for instance, or more synergies, initial success, and on Digital Industries? The AI developed very positively, and at the moment, there is the thesis in the industry that in automation, there will not be any lock-in effects anymore with AI, because nowadays it is so much easier to integrate models. You can migrate models. Using AI, it is easier to combine the worlds of different manufacturers. Is it an advantage or disadvantage for you? Because Siemens actually focus a lot on proprietary systems.
Was ONE Tech Company program already successful? Has success already materialized?
You can see the success when you look at the verticals. Let us start with the data center vertical, and here we are talking about electrification. What you can see already now is that in the area of automation, we are more and more moving towards data centers. They need, well, faster control, higher transients. Building automation is very important here, and industrial automation. This is where it starts, and we definitely see that we, as a ONE Tech Company program, just one company, we can cover that demand on a broad basis.
The second thing is digitalization, building up a digital twin of the manufacturing facility, simulating how the cooling facilities work, or building technology in general. It is a holistic approach. Same applies to semiconductors. There are lots of investments in semiconductors in the industry.
As far as the infrastructure is concerned, like buildings, technology, electrification, fire protection, and naturally also on the shop floor with automation, we really can contribute quite a lot, and a lot is happening in these areas. The lever of a technology platform ranges all the way from electrification, automation, digitalization, right into the verticals. It is very encouraging, and we can see a lot of momentum there. That is only the beginning.
I guess we are now going to go through one vertical after another. We have four verticals that we are addressing in particular, where the growth rates are high. I already mentioned some of them. What we see is more and more traction coming step by step. Also, because customers increasingly want to see solutions for their problems and not just purchase components.
That is something that you can always do, but the actual lever materializes if you actually can make a data center more efficient, or if you can build it a few months faster because you are working with a partner who has a holistic approach to everything. The AI lock-in effect. A couple of things here. First of all, and you also mentioned the proprietary platform, hardware, PROFINET. That is something where we say we also work with other bus systems. It is only linked to automation to a limited extent. Changing automation, most of this comes with embedded software. It is not software-defined hardware, where you could say, with an AI solution, I can just cover everything across the board. But operation software, that is a different matter.
The kind of software that controls an entire manufacturing facility, there AI is really quite important, and AI can then aggregate data across different silos. But that is exactly our home turf, and that is exactly where we are strong, where we have AI solutions, be it in simulations. This is where we can integrate them or in our operation software, Insights Hub, where using AI, we can do predictive maintenance, Senseye, so visual optical detection of flaws. That is something that is not a danger for us. It is not a risk.
The fact that there is no longer a lock-in, it is an opportunity because it offers so many more possibilities for customers where they can improve their own productivity. I understand why you are asking this question, but actually, what you are suggesting here is not true. Lock-in, let me just say this as a last point, it is all about platforms.
In addition to our automation, we have our TIA Portal. It is an engineering platform that is combined with our automation. So there, with this platform, we are accelerating it using AI. Eigen Engineering Agent is a perfect example where we have the opposite effect. We have a huge installed base, and on this base, we can even do much more with AI, and this is where major growth is going to come from. But we are not entirely relying on the lock-in effect of we are just TIA Portal customers, but we expanded it.
We added electrical planning, and we are going to add in November, we are going to have a launch of process automation. So more and more, we are also addressing other verticals because we can actually build their specifics into our Eigen Engineering Agent. It can be trained to cover more and more vertical-specific specifications.
We can expand this gradually. I hope that is an answer to your question and gives you a sufficient explanation, right?
[Non-English content]
The next question comes from Markus Frühauf from FAZ. Please go ahead.
Good morning, Ms. Bienert. Good morning, Mr. Busch. I have a question with respect to the debate around the dangers of artificial intelligence. People around the tech scene have been very worried about this. Mr. Busch, do you fear that this debate will lead to a stronger regulation of AI? If so, how would that impact on your growth opportunities? Thank you.
That's a very good question. Let us start here. I don't believe that in three to five years, AI will have destroyed humanity or humankind. These are scenarios that I don't really understand. I don't get where they're coming from. What we do see is that AI is becoming more powerful day in and day out.
We started by talking about copilots, now we're talking about agents, and now we're moving on to swarms of agents that can be deployed in order to destroy firewalls. This is something that in the cybersecurity sector is an increasing danger. But time and again, we can combat that with AI itself. This is a danger to a certain degree that curtails security, that is becoming increasingly fierce. But with every flank that opens up, we have people that can take counteractions.
The second is regulation. If AI becomes rogue and does certain things that it's not supposed to do, well, that can happen, but I think that people are exaggerating the reality. If you take a top model and tell it to break a firewall and put it in a sandbox, what is it going to do?
It'll work with thousands of agents day in and night out in order to break the firewall. In an industrial environment, if we deploy AI, we give it a task to solve a problem. We don't tell it to go rogue or do anything of the like. These dangers only arise if AI is deployed with an objective that can provoke dangers. That doesn't happen in the industrial sector. What we are increasingly seeing is that more and more people are looking at regulation and trying to figure out how AI can be deployed within guardrails so that certain limits are not exceeded.
This is already done today, by the way. If you try and ask our model how to build a hydrogen bomb, you will not receive a response from the AI. Guardrails are already built in.
I don't believe that this is that much of a danger. There's one more important thing. Should, for any reason, the new training of models decelerate due to regulation and restrictions, well, for industrial AI, for applications in the industrial sector that we see today, the models are so good that the limitations are elsewhere. They're in the training of the models using specific data, the provision of certain data when it comes to integrating AI into a process, where a process may have to be redesigned instead of AI just being integrated into an existing design. That can be a lever. But if you look at the quality of the models that we have today, the top models, the maps, and the open models that we know from other regions like China, this works.
It goes so far that with simple models on the shop floor, you can solve a lot of problems that need to be resolved.
The next question comes from Marilen Martin from Bloomberg. Please go ahead.
Good morning. I have a question with respect to the regions. Could you go into further depth? With Eigen, we've seen that revenue and orders have grown substantially. In China, the orders have increased less than revenue and vice versa. Now, I thought the Chinese products would be more affordable, and therefore I anticipated this effect being the other way around. Perhaps you could give us some commentary on that. If we look at AI, we've seen an audit in China, maybe there was a revenue increase. What was the driver of these effects? Could you give us some insights into that?
Well, when it comes to Germany, the 19% growth, the situation is as follows. On the one hand, we benefited from a bigger order in the life sciences business. That was a project. Above and beyond that, however, we do see that, especially when it comes to customers that are active in exports, especially in mechanical engineering and machine building, we had positive effects. We look at the current ifo index and similar barometers, it is safe to say that in the German market, there was a certain increase.
But above and beyond that, for instance, in chemicals, pharmaceuticals, and automotive, the market environment remains under pressure, and the customers are very reticent to place orders as before. So, the picture is very diverse. But if you look at machine building and export-oriented industries, we benefited greatly from them.
To give you an outlook on the first quarter of a new fiscal year, well, we hope that we can continue benefiting from this in Germany. You also broached the topic of the U.S. and China. As far as China is concerned, the situation was the one that we outlined before. In September, we were a little weaker, and this, to a certain degree, was due to the fact that distributors executed certain orders or postponed certain orders to the next quarter.
You can see that in the beginning of October, things picked up significantly again. Having said that, we see that there have been some spillover effects. As a matter of principle, issues such as price increases also drove certain behaviors in China and the U.S. The U.S. is still strongly affected by the data center and utility business.
Here again, if you look at the outlook to the new fiscal year, we still see strong demand. We believe that the end markets are set up quite well in the area of data centers. In the next weeks and months, as far as approval procedures are concerned and the like, we will see a certain degree of volatility. But over the medium term, we see very strong demand in the U.S. as well.
We do not see any further questions in the German channel. I see John Revill is still in the English channel, so we are going to move into the English conference. Please bear with us for a minute. Thank you.
If you would like to ask a question in English, please press star key followed by one onto your phone. One moment for the next question, please. The next question comes from John Revill from Thomson Reuters. Please go ahead.
Hello, everybody. Thanks for taking my questions. I was just wondering, on the data center question, you guys said orders increased by triple digits in the fourth quarter. Was that an acceleration from the previous quarter when it previous nine months when it was also triple digits? Was it a higher rate or not? Secondly, do you expect any kind of slowdown? Because there is a lot of concerns about data centers at the moment, like the funding for them, but also increased regulation in the U.S., especially after the midterms. So, do you think there could be sort of that big fuel growth at a slower rate moving forward? Or do you see any kind of concerns of slowing construction of these buildings? Thank you. Did you hear me?
Yes, we could hear you. One moment.
Oh, yeah. Sorry.
We had a quite strong demand in the U.S., and the expectation is that in the upcoming quarter, we still have a continued demand in data center. Definitely, we can see in October that, and in this, the outlook in the first quarter, that the demand will kind of continue. But in terms of the translation into revenue, we need to execute really on the order backlog. Therefore, with the increase of capacity in production and the supply of material, we expect that we will kind of catch up in the course of the first quarter. Yeah.
Excellent. Thank you. But do you see sort of midterm a kind of slowing down perhaps as more regulation comes in or.
Um.
There's less funding for these people?
You know, there's a discussion that also back to the elections, that there are discussions also that people are standing up. They don't want to have a data center in their backyard for various reasons, water, noise, and the like. So, there are discussions on the one side. On the other side, it's bipartisan acknowledgment that a big driver for the economic growth in the U.S. is really building out data centers. Therefore, it could be that an approval process can take longer. Would it mean that maybe one or the other data center might not be built? Could be, or maybe built later. But I would say, having that dynamics in mind, there's nothing what worries us, at least not for the next couple of years.
Thank you.
Not at all.
All right. Thank you very much. We are already a bit over time, so we have to close the conference call now. Our conference call for analysts with Roland and Veronika will begin shortly at 9:30 A.M. And the analyst call will be broadcast live at siemens.com/analystcall. And you will hear from us again at the latest on February 11th, 2027, when we'll release our first quarter. Thank you very much and goodbye.
Ladies and gentlemen, this concludes our conference call. A recording will be posted at siemens.com/conferencecall. We say thank you and auf Wiedersehen.