Telefonaktiebolaget LM Ericsson (publ) (ERICB) Earnings Call Transcript & Summary

Good morning, good afternoon, good evening, wherever you are around the world, and welcome to this investor and analyst update. Today, as you can see, we are here in Kista, Stockholm, in our studio, instead of being in Barcelona. I'm coming back to that in a little bit later. With me today, I have our CFO, Carl Mellander.

Hello, thank you.

Welcome, Carl, as well. So before starting, though, I would like to read the following statement. This report contains forward-looking statements. Such statements are based on current expectations and are subject to risks and uncertainties that could materially affect our business and results. Please read our earnings report and our most recent annual report for better understanding the risks and the uncertainty, and please see the last page in the presentation for further information about forward-looking statements. Any forward-looking statements made during this presentation speaks only as of the date of this presentation, and Ericsson expressly disclaim a duty to provide updates to these forward-looking statements and these estimates, assumptions associated with them. So with that, let's start with this slide. You probably have seen this before, but let's just come back to -- we were supposed to be today, actually, in Barcelona. Unfortunately, that event was canceled due to the coronavirus. So instead, we are here in Stockholm. So we have a little bit of an unplugged Mobile World Congress here instead. So -- and the background is basically that we have done a lot of preparations before the event in Barcelona, and our thought was to reuse that today instead. But before going into the agenda for the day, let's just go through our investor cycle that we have throughout the year. So as you see here, we should have been in Barcelona today. As I understand, next year, the Barcelona event is supposed to be between the 1st and 4th of Apr -- or March, sorry. On Tuesday next week, we will present our annual report that we will publish at that date. And then, on the 31st of March, we will have our Annual General meeting. And for those who are interested, you could actually the really recorded of Börje's presentation and also subtitle for those who want to see into that. And then we have the Q1, Q2 and Q3 earnings report, and those are basically the same time as last year. We will end this year with a Capital Market Day, again, a full load Capital Market Day in New York City in New York then. So we're back to do the sort of normal procedure as we had in 2018. By that, I will walk into the agenda for today. And bear in mind, for the technology today, we will try actually to connect 1 person from U.S. and actually 1 person from northern part of Sweden when we get to the Q&A session. So it will be a challenge, but I think we will overcome that. We will start with our CFO, Carl Mellander. So Carl, you will introduce.

Yes.

And he's with me here, of course, in the studio. Carl will talk about the market development and our position in the market. And he will spend some time on OpEx, and particularly R&D as a value creator, and he will continue to dig down a little bit further into digital transformation.

Correct.

Second out is then Thomas' record, we prerecorded his presentation here last week. He will do a deep dive into 5G, present our platforms, but also do a reflection on our launches in South Korea and Switzerland. And then back to the studio again, Fredrik Jejdling will join, and he will present Networks portfolio and dig down a little bit to the acquired entity of Kathrein. Then after Fredrik, we will have Jan Karlsson, who will, among other topics, touch upon 5G Core and the importance for the customer as well as for Ericsson going forward. And before we enter into the Q&A session, we will have a discussion between Carl and Niklas, who will actually discuss the potential in North American market, but also discuss about the -- what's the impact of the merger between Sprint and T-Mobile. And again, that's recorded last week when Niklas was here in Kista. By that, I would like to hand over the word to you. So please, Carl.

Thank you, Peter. Thank you, and thanks, everyone, for joining us here in this live webcast from our studio in Kista. So we call this whole event, Ericsson on track, and we will spend the next couple of hours to show you why we believe that we are on track. In essence, we have delivered a strong 2019 and we see a promising future also for our business, and we are executing on the strategy. We will talk about that today. And we believe we are on track to the targets for 2020 as well as 2022. So a couple of items, just to kick off here. In essence, I would say, we see a healthy market environment to start with. And the RAN market, which we pay a lot of attention to, obviously, grew by 5% in 2019, as confirmed by Dell'Oro. This is 2 percentage points more than we expected when we met actually at the Mobile World Congress last year. So the market has taken off a bit faster. For when it comes to Ericsson, we set out to win footprint based on technology leadership here in a selective and disciplined way, obviously. And we have seen some share increases, more about that in a minute. But we also see strong fundamentals in the market, not least in North America, where we now have also more clarity on the future when it comes to the merger and more about that together with Niklas Heuveldop later today. I'd also say that through the investments we have made in technology and also, of course, staying very close to the customers, we have leadership in 5G now. We have recorded 81 commercial agreements with unique operators. And we were the first to launch live 5G commercially together with our customers in 4 continents. So now 5G Core is becoming more of a reality as well for our customers, and we believe we are well positioned also for the 5G Core investments that we'll start. So of course, when it comes to financial performance then, as we are here to create long-term value for our customers and for Ericsson and our shareholders, we are also not letting go of the present and the performance management of our company today. So I think 2019 performance is a testament to that, where we delivered an operating margin of 9.7%, so very close to the 2020 targets, and where we quadrupled the free cash flow before M&A as well, almost SEK 18 billion in free cash flow before M&A, which is, again, 4x the amount in 2018. All those numbers, just as I said, that excluding then SEC and the DOJ, obviously. But let's drill down a little bit more in some of these topics now, starting with the market data. And if you look in general at the market now, I would say the environment is beneficial. It's positive when it comes to 5G now. And 5G is in early stages still. Of course, we see some early adopters going very fast, including U.S., Korea, Japan, China to come as well, and accelerating faster than we would have thought a year ago or 2 years ago, even more. But the ramp-up will continue. And if we look at our own latest Mobility Report, we see that between 55% and 65% of the world population will have 5G coverage by 2025 already. And we really are positioned to capitalize on this rapid shift in the market towards 5G. And to date, we have shipped already 4 million 5G-enabled radios. The graph that you can see here to your right shows that the estimates for 5G revenue have increased all the time with new reports coming out. So it's apparent that this market is growing faster than what we and other analysts would have thought earlier. I would also say that when it comes to the use cases for 5G, an early use case will be mobile broadband. But the interesting part of 5G, the 5G case here will be on enterprise and industrial use cases. And we see this still in early phases, but with a big potential. And again, according to our own research, we believe that there is an addressable market for operators in the enterprise space connected to this of USD 700 billion by 2030. So we believe the enterprise side, the industrial cases and IoT will, of course, drive demand over an extended investment period. If we look at the market shares for Ericsson, we can see that, 2019, to start with, was a strong year in North America for us. We strengthened our market position with all the major carriers in North America during the year. Our market share increased 4 percentage points from 48% to 52% now in North America. And given the relations we have there with the carriers and, of course, our technology posture there or status, we are continuously confident about that market and our strength there in North America. But it's not only about North America, obviously. We also improved the market position in other geographies as well. So if you look at EMEA, for example, we put some customer names there where we had been able to demonstrate technology leadership and have early 5G wins. Also, if we look at a market like Middle East and Africa, we have gained good traction there, and we also, there, mentioned a couple of names here where we are early out with 5G in those regions as well. I find it interesting to note there now, when it comes to market share, that in the fourth quarter, if you look at market share outside North America and China, because those markets are large and a bit special, and so if you look at the rest of the world outside those 2 markets, Ericsson's market share grew from 27% in Q3 to 33% now in Q4, so quite a dramatic growth actually. So that's about the market share. Let's look at the financial development. If we move on here and I think this is a summary of the turnaround that we have seen in Ericsson, starting on the top line side. We've had 6 consecutive quarters of organic growth. And we ended 2019, as you know, SEK 227 billion. The ambition for 2020 is within the range of SEK 230 billion to SEK 240 billion there. And the gross margin, of course, you've seen this profile, a strong improvement early -- earlier. And now, since second quarter 2018, we have delivered gross margins above 36%. And what has driven that? It's really, of course, cost out, not least in service delivery, higher efficiency there, but also the increased penetration of the Ericsson radio system and a better cost position and competitiveness in that portfolio as well. We have addressed critical or nonperforming, nonstrategic contracts, both managed services and digital services as well, that has also contributed quite a lot. As you know, the ambition for gross margin for 2020 now is within the range of 37% to 39%. And finally, to your right side here, if you look at free cash flow before M&A, I mentioned the numbers before, SEK 18 billion, or SEK 17.7 billion to be more exact, in free cash flow before M&A generated, which is 4x higher than 2018. And here, I would say, aside from a better profit, of course, we've also delivered higher efficiency when it comes to working capital. So working capital days are down to 75 days now, down from 89 year-over-year, as 1 example, showing that the machinery is more efficient, the lead times are getting shorter now in working capital. The ambition on free cash flow before M&A is strong. And earlier, we have shown the bridge between operating income, and we took the example of an operating margin of 12% that should generate free cash flow before M&A of 8%. That's how we try to define what we mean by the strong free cash flow. We focus in Ericsson a lot on value creation. And of course, R&D is a big part of this. That's at the heart of Ericsson. It's fundamental for us for the technology leadership, which, in turn, gives both competitiveness and improved margins. And we think about R&D in 2 parts: One part, the smaller part, is for maintenance R&D, we could call it that, where it's about maintaining and improving on the current offering; and then the larger part is about creating the new offerings, new product development for customers. And yes, we increased the reported R&D now by SEK 3 billion to SEK 38.5 billion, as you can see in the graph here. So R&D key for value creation. We also mentioned in the Q4 report other investments we are doing now in digitalization, in compliance and in security. Of course, digitalization, I will speak a little bit more about that in a minute. It's about both customer experience and operational excellence as well; compliance, of course, to secure our processes and ensure that we do business in the right way; and then security, which both has to do with how we protect our enterprise but also, very importantly, product security, which is very high on the agenda now, obviously, as we enter into 5G also. In the graph here, you can see how OpEx has been reallocated from SG&A to R&D, which is an investment again for value creation. But you can also follow here the percentages, that as a percentage of net sales, the overall OpEx envelope has come down. Our ambition is to continue on that trajectory. Although the investments or the expense levels as such might go up, we expect that to go up a bit in 2020 for the reasons that we have talked about now. A bit further look into R&D, as I said, being at the heart of Ericsson and who we are as a tech company. And the purpose, of course, is to drive both long-term sales growth but also margins. And the proof points that we see so far is, of course, gained market share, grown top line 6 consecutive quarters and strongly improved margins as well. So if you look at the graph here, just to put things in perspective and try to address the return that we get on the R&D investments, if you look at the green bars here, you could see that R&D has increased quite dramatically over this period. And to put a bit of numbers on it, if you take a 4-quarter rolling basis until the first quarter in '17 and you compare that with the full year '19, we have increased R&D by SEK 8 billion. But during the same time period, gross income has improved by SEK 19 billion, so SEK 8 billion of R&D investment, SEK 19 billion of gross income improvement. And you see the gross margin underlying here is portrayed in the graph with this blue line. So clearly, we see that the idea of investing in R&D is bearing fruit also for margin generation as well as top line growth. I mentioned digitalization before. And I wanted to say just a few more words about the efforts we do in digitalization. And we do it to reach the next level of efficiencies for our company. We believe that the best result here, when we work in digitalization, is where we can combine both customer experience, an outstanding customer experience with internal efficiency, taking away manual process steps, for example. It's in the combination of those 2 where we really get good results. So that's how we have planned the digitalization efforts we're doing. We have identified 10 projects or lighthouses, and we have launched or will launch them. Now some examples you can see here, just to mention a few then, supply flow order tracking, which creates better visibility for our customers; you have preventive and automated customer support; we have automated handling of recurring billing, all to be faster, easier to work with, and again, as I said, take away manual steps in our own company. So this is an investment and with that comes, of course, the requirement to generate returns. And we are tough on the business cases here. And we want to see committed returns which can either be in the P&L, of course, cost saving, top line growth, but also in shortened lead time, which translates then into reduced working capital. So we are following up on both of these aspects. We intend to invest between $500 million and $1 billion per year in this, provided we see the returns, of course, per year, let's say, over the next couple of years. Payback is between 1 and 3 years depending on project. But I think it's also important to say that these projects are run in an independent way. This is not a big mammoth IT project, which can derail or get delayed, et cetera. We are running, in an agile form, at distinct projects, which we can scale globally after having been tested and tried. And if something doesn't work, it's easy to stop one part without jeopardizing the whole. So we believe this will generate value both for our customers, our suppliers and Ericsson as well. So I've talked a lot about value creation, and this is important for us in Ericsson. I would say we have now created a foundation for a bit new ways, enhanced ways of working with value creation. We have clear targets, both for profit and cash flow. We have set -- put incentive schemes in place that support this awareness and drive and decision-making in the whole company. For example, we work with economic profit, which is something, a metric that everyone with variable pay is measured on. And here, we look at both the P&L, of course, the profitability, but also cost of capital to create the link here and get it closer to actual value creation. And this has been key, I believe, for the reduction in working capital we have seen. And now, what we're doing now is to continue to embed value creation, modeling principles and thinking into both strategic choices, decision-making as well as operational, more day-to-day decisions, so much more cash flow focus thinking value -- through a value-creation lens when we make decisions, always making sure that we have options where we can compare value-creation potential in different scenarios, options. And what we want to achieve longer term here is, of course, that value creation gets completely embedded in our culture, in our DNA, if you wish. And that all decision-making on all levels is just penetrated by the value-creation thinking. So we think with the right incentives, right tools, right mindset and so on, we can do quite a lot here, including monitoring, of course, so that we can early identify both performing as well as nonperforming activities, so we can allocate capital and resources wisely or reallocate from nonperforming to performing or more promising. So that's how we think around value creation in Ericsson. Thank you for that and your attention to this piece. And now we're going to move on in the program. We are going to run a video with Thomas Norén, as Peter mentioned before.

Yes. Okay.

He is Head of 5G Commercialization within Ericsson, and he will talk about 5G from a both technical -- technological and commercial angle here. And although this is recorded, Thomas will join us for the Q&A later.

Exactly.

So please roll the film.

Good afternoon. I am Thomas Norén, I am Head of 5G Commercialization at Ericsson. I'm very happy to be with you over video today. We're going to talk about 3 things: First, about 5G adoption; then I'll give you a little bit of a technology recap; and then I'll show you how operators can build their networks in a much smarter way with Ericsson 5G. When we met last year, in Barcelona, at that time, then there were no smartphones, there were no mobile subscribers on 5G. And boy, we have had a big development since then. We have launched 25 live networks by now. We have been first on 4 continents in North America, in Europe, in Asia and in Australia. And we have over 80 commercial agreements or contracts with operators. So 5G has really started to take off since we met last time. If we look at the different markets, the country where we have most usage of 5G by far is Korea. So in Korea, there are more than 4 million subscribers by now. And we can really get good measurements of data. You can see that 5G is very appreciated by the Korean consumers. They consume between 2.5 and 3x more data when they are on 5G versus when they are on 4G. So you can see that the applications, the AR/VR applications that they have launched has really taken up and they really -- it's really used by consumers. But the outstanding question THEN is, of course, does it make financial sense for operators? Do they really make money? And if you again look at just official data from the Korean operators, you can see in this picture that since they launched 5G, the average revenue per user is increasing. And this is, overall, a trend that we see with 5G. Operators, they try to provide more, larger data buckets, faster speeds, new services, bundles with media to offer more; and consumers, they seem to be happy to pay more. We actually did a study a bit more than a year ago when we asked consumers, well before 5G was launched, how much they would be willing to pay extra, and on average, they said 20%. That was a number or average all over the world. Of course, you can -- you're wondering how can people say they want to pay more. And they might even not know what 5G is or most likely they didn't know what 5G was. But I think it shows that digitalization and the lifestyles we now have really require a good mobile connection. We do more and more with our connected devices, smartphones and others and that is really, really appreciated by consumers, and they want more, and with 5G, operators, obviously, can offer more. So we see that this is taking off. But Korea is not the only example. Korea has built, of course, a large network, 90% population coverage, but so has Switzerland, in Europe. And Switzerland has taken a very, very different approach that is maybe more similar to what other operators are doing and will do. Swisscom created its coverage by leveraging the installed base they have with Ericsson on 4G. So they were able to deploy 5G on their installed base of 4G radios, in their case, on the 2.1 gigahertz band. And in this way, they could complement the additional coverage -- the additional speeds they get with the new spectrum on 3.5 gigahertz, and they can also complement that then with the coverage solutions on 2.1. And I'll come more to the great benefits this has over time. But they are not the only one. Also, if you go to North America, T-Mobile has built on a new frequency band, getting a 200 million PoPs. coverage in North America. So they have also paved the way for greater coverage build up in North America. So we see that this is happening in Asia, it's happening in Europe, and it's happening in North America. Now let's talk a little bit about the fundamental technologies that are underlying this growth. So to start with, let's look at spectrum because much of what we do in our industry is determined what you can do with spectrum. And as you know, we have built different generations on different frequency bands. And we've started to migrate even some bands to lay their technologies. The color coding we used here is going to be the same throughout the presentation. So orange is 5G, green is 4G, violet is 3G and blue is 2G. So over time, we started to build on lower frequencies. The advantage of that is it has much better coverage than higher frequencies. That's why we started there. But as 4G and 5G came along, we needed to add capacity and we needed to have higher throughputs. Therefore, we had to also go for bands where you can create the larger bandwidths, and those are on higher frequencies. Now the problem with that is that it's more difficult to build coverage because the higher the frequency, the less radio propagation you get, that's physics. So we needed to invent new technologies to address this challenge. And some of those technologies are beamforming, where you direct the beam towards the users instead of spreading it out over multiple users, and Massive MIMO, where you can transmit over multiple streams at the same time. So this is a clear development that we had to do. But also in the regular macro frequency bands, there is also an evolution. We need to improve our performance there also. So we are introducing multi-band products and more transmission paths also in lower bands. So networks are going through modernization. Now if we look at what we've done in Korea, we have built this, of course, the 5G there on a 3.5-gigahertz band, it's a new 5G band. The high-capacity solution is what you see in the middle here is the 6488 from Ericsson, it's a 64TRX product. It has an enormous capacity. It's extremely well suited for highly dense areas, where you need to beamforming, both horizontally and vertically, for example, with high rise buildings. But most of the network is actually built with a 3239 product, a 32TRX product, this is more compact, it's more efficient. But again, it well suffices for absolutely most deployment scenarios. If you go out further in the network, more rural, then we upgrade to 4T4R regular macro radios. So you can actually use that technology, radio technology that we've used in previous generations also for 5G. This is, of course, a great advantage. Now Ericsson has some unique capabilities when it comes to Massive MIMO. And the trick is basically that we have an architecture that is better suited than many others' architectures for this usage. So Ericsson has purpose-built ASIC for baseband, it's called the EMCA architecture. Now what we've done is that we have also taken part of that architecture and part of that baseband and built it together with the Massive MIMO antenna, and that gives a really superior performance. Well, first of all, it means that the connection between the baseband and the radio and antenna doesn't need to be built out according -- and scale towards the number of antenna beams you run, you only need to scale it according to how much user data you get. And that makes the transmission between the baseband and the radio much cheaper. You don't need to have as much capacity in that link. The other thing it does is that by having an industry-leading platform, that we have designed ourselves, that is purpose built, lowers energy consumption massively. There is no cost off-the-shelf hardware general purpose that can match the power consumption that we achieve because it's purpose built. And the last thing is that we can get much better performance in networks because by placing the beamforming processing at the antenna, we can take feedback from the radio signals, from the different users and their radio conditions, and adopt our beamforming and beam tracking much faster and better to the users. And this has proven to be a very, very good solution. Because, first of all, it's easier to follow the users as they move, so we get superior mobility with our architecture versus if you had the full beamforming capability at the regular RAN Compute side. The other thing is that you get much better coverage. You get 90% better app coverage, so for -- in uplink. So this means that if you have -- without this functionality, you can't get as far with your apps as you can with this. And the last thing is you boost the cell-edge uplink speeds by a factor of 10. And to prove this point, let me share you some real drive test we did recently in Kista. So in this case, we run a drive test numerous times, and what we did, we turned on and we turned off this uplink booster capability, where you either had the beamforming performed at the antenna or at the regular baseband location. And as you can see here, you get the similar uplink speed much, much further out in the network. You get 90% better uplink coverage. An alternative way of looking at it, at the same coverage capability, you get 10x higher speeds at the same location. So this is, of course, a phenomenal benefit for operators when they build their network. I said it was hard to get the same coverage with the higher frequency bands. Well, with this technology, we really boost the coverage as much as you can even on those high frequencies, a tremendous benefit for our customers. But there's more. We are also making sure that you can reuse your already made investments in infrastructure as well as in spectrum with something we call Ericsson Spectrum Sharing. So this is how it works. So traditionally, the non-Ericsson Spectrum Sharing in the slide, you have to have dedicated bands for each technology or you need to have a forced hard split in the same frequency band between the 2 different technologies, for example, 4G and 5G, or 4G and 3G. Now when we introduced 5G, we realized that you don't want to be dependent only on high-frequency bands. So if there is a possibility to run 4G and 5G at the same time, that would be enormous benefit because then we can introduce 5G on the lower bands in the way that Swisscom has done and created their 90% population coverage, they can reuse the already made investments they have. So by doing this, we create a coverage layer that is very beneficial. This is possibly the cheapest way to build coverage. But the other thing is that you can combine with carrier aggregation, the new lower band and the higher band. So as you can see in this picture, when you do that to the right, you extend the coverage of the higher TDD band. The reason is that you can use uplink on the lower FDD band and you can increase power on the upper TDD band and thereby extended its downlink coverage. That has, of course, enormous benefits because it gives you better user experience in more places and for more users by being able to combine the 2 different bands. And you extend the capacity because you can use your new frequency band, in this case, the TDD 3.5-gigahertz band, much, much further out. And by doing this, you enable for the next big step for 5G stand-alone. So let me share with you how this works. So the first thing you do is that we can introduce, and we do this now in Q2, we introduce 4G and 5G on the same band, as you see to the left in this picture. This is a software on the Ericsson Radio System hardware. And you can totally dynamically change capacity, depending on how much usage you have in your band between 4G and 5G. W hat we will then do in Q2 this year is that we have a further software package where you can instantly share both for stand-alone operation and non-stand-alone operation and 4G. And this means that our base stations will simultaneously support UEs that either run on a non-stand-alone mode or a stand-alone mode. So maybe I should explain. Non-stand-alone mode means that the UE, the device, is connected to 4G and 5G at the same time; stand-alone mode means that it's only connected to 5G. But in order for that to work, you need to have coverage. And that is exactly what we create with Ericsson Spectrum Sharing on a lower FDD band. Now why is stand-alone important? So there are a number of reasons. So one reason is that we introduce a new architecture for 5G Core that makes it easier to deploy slicing. You can have multiple slices per device, for example. The other thing is that we introduce a service-based architecture. And that means that it's easier to develop new types of services because you can combine network capabilities in a much more flexible way. You also introduce a much less complex architecture. And eventually, you will have less complex devices as well because the device needs to be only connected to either 4G or 5G and not both at the same time and that lowers power consumption. It also makes -- this simplified architecture also makes it easier to speed up control signaling. So the access to a webpage becomes much more instant. So we have done some measurements of this and here, you can see the combination of when you have stand-alone, the orange; the non-stand-alone, the violet; and the green, which is LTE. And you see from time from left to right, you can see how quickly you get -- the UE gets from an idle mode, when it's not transmitting or receiving, to an active mode. So you can see, it starts to get data much faster, and it also gets to higher rates, if you're able to combine multiple bands for stand alone. So even for end users, this will have a big impact. So stand alone is a big thing. The way we introduced stand alone in our core network is through our dual mode 5G Core. So our dual mode 5G Core is 1 core network for both 4G and 5G. And the benefit of this is, of course, that, as an operator starts to invest in this architecture, it can smooth the migrate capacity and its investment from 4G to 5G depending on how fast 5G picks up. It's also an entirely cloud native architecture, which has a number of benefits. Well, first, it use -- fully uses its micro services. This means that you can more flexibly introduce new applications, you can do canary testing, you have a much more robust operation of the network. The other thing is that we introduce greater degrees of automation, which means that you can -- as the life cycle of the core network and its applications evolve, we can manage those applications in a much smoother way. Obviously, we have developed a continuous software delivery mechanism. And we see that if you constantly take in new software capabilities instead of taking large software leases, you can reduce OpEx for introducing new software in total with 60%, 70%. So for the operator, this becomes a much easier task for daily upgrade network -- as they go along, upgrade their network instead of once in a year do a big refresh of the entire network. And lastly, we fully support open. So we have been the driver of the service-based architecture in 3GPP. We have made more contributions than our 2 closest competitors combined. We can support any implementation of Kubernetes. We can support any vendor of x86 hardware, any virtual machine and, of course, we have a full stack implementation also. But we think this is very, very important because we're going to foresee, and I'll come to that, more and more new types of developments where you will -- where you can deploy our 5G network. And these will be in sometimes new environments compared to what we have today. Lastly, we see that we have a very, very high performance. So performance depends, obviously, on what hardware you run and what applications you run. So with one of our Asian customers, we run a real test. We run a test where 3 vendors, Ericsson and 2 others, had to perform the same application on the same hardware, so equal and apples to apple comparison. And as you can see, we were 60% to 80% better than our 2 closest competitors. So we have an incredibly efficient usage of our hardware, which is, of course, very important as capacity keeps on growing in the network. Lastly, to support operators to introduce 5G, we also introduced AI, and we do that in multiple layers. We, of course, do that for planning, for optimization purposes overall of the network. But we also do that in smart ways in the nodes themselves. So for example, we have functions for more intelligent traffic management, which increases 5G coverage enormously. By doing this, we secure that the device is always camping on the most efficient frequency as it moves in the network. We have also evolved carrier optimization. So we make sure that the device is operating on a combination of carriers that is optimized for uplink and downlink, depending on what the user actually is doing or is foreseen to be doing. So with these technologies and more introductions of AI in a network, we are able to squeeze out more capacity and performance of our 4G and 5G networks. Now let me last finish off by giving you an example how an operator typically can use these technologies to build 5G in the most efficient way. So in a very simplified manner, a network looks like this. To the left, you have the city center, and out to the right is more rural areas. And you have built multiple layers. In this case, 2 layers of 4G, the green, 1 layer of 3G and 1 layer of 2G GSM. And the higher layers, the higher frequencies have, obviously, a little bit less coverage, less capacity need. This is your starting point, in a simplified way. Many operators have more 4G bands, have maybe different 2G bands, but the principles will be the same. Now a traditional way of building the network, you have the starting point, again, to the left, the picture I showed before. And you start to build 5G, the orange piece. And you do that, for example, on the 3.5-gigahertz band and you build out the network. You start, obviously, from the left, where you have your dense urban environment. And you build out to the right, towards where you have less usage. And you start probably to add also some millimeter wave for places where you really need a lot of capacity. And then you continue to build out this and have ultimately good coverage. The challenge with this is, as we saw before, that you have still not as good coverage per site when you run on 3.5 gigahertz versus the lower bands. So you require more nodes to do this. And it also takes longer time. Now luckily, there is an alternative way. So the starting point is the same. But -- and you also start -- point number one, you start by building out 3.5, and we start to do that with non-stand-alone, so the device is connected to both 4G and 5G at the same time. But point number one, you don't need to do as much because, point number two, in the striped area, we have combined 4G and 5G with Ericsson Spectrum Sharing. So you have: A, created better coverage on the lower band; B, you can do the carrier aggregation between the bands and you extend the effective coverage of your 3.5-gigahertz 5G band. When you have done this, you have more room to invest or to invest in even higher frequencies, millimeter wave. So now you can start earlier to invest in millimeter wave for where you really need the capacity, be it for stadiums, other venues, being it for fixed wireless access or other applications where you -- in your mobile network has an enormous demand for capacity and speed. And as you go along, you continue in this fashion and, of course, build out more and more of your legacy band, point number four; you continue to build out your 3.5 and your millimeter wave bands; and you can also afford, point number six, to acquire a new frequency band. And in total, you can see as you get a lot more orange, you get a lot more 5G in this way compared to the traditional way of building networks, but it requires that you have access to the Ericsson Spectrum Sharing, that you have an installed base of hardware that can support it, and you can do carrier aggregation. Now what about those unfortunates that don't have Ericsson in their networks? Is there a way to migrate those networks? And yes, there are. So if you look at this picture, you start in a similar way to the left, you have here light green, which means you have somebody else's 4G, and now you want to create the same to the right, the ultimate solution with multiple 5G bands as you had in the previous picture, and you want to introduce Spectrum Sharing and so forth. So how do you get from the baseline, in this case, to the ultimate? Well, the reality is that operators are ongoingly modernizing their network. They need to go to site, if they don't have Ericsson equipment, to do baseband capacity increases. They need to typically modernize old radios. The old 3G radios were 2 transmit, 1 receive, now you want to do 4 transmit, 4 receive. You get much better coverage capacity that way. And it's very valuable also on these bands. You want to introduce multi-band radios instead of single-band radios. Multi-band radios have 40% lower power consumption than 2 single-band radios. And you want to -- as you modernize, you want to capitalize on that benefits also. And typically, you want to introduce new bands, so you need to introduce new antennas. And we have penta- and hexa-band antennas that we are coming out also from our -- from Kathrein, for example, that we now have required. So operators need to go to site and modernize their old equipment, anyway, for OpEx reasons and for performance reasons. And as they do that, there are several ways to introduce also Ericsson and get the benefits of Ericsson. One alternative way is that you start to modernize your 4G network and you swap that completely and then you have a starting point and then you modernize gradually your other band, your other bands as well 2G, 3G as you need to do that. Alternative B, then you start to reform an existing 2G or 3G band, often, it's a 2.1-gigahertz band in Europe and you say, "I move all that traffic with my legacy supplier down to 900, for example, and I'm going to modernize this band. I use that as the basis for my anchor point for 5G, and then I introduce Ericsson on those bands, possibly, I add a lower band also. And then I modernize the network, and I end up in the ultimate solution." So I -- we believe, and we have seen multiple cases now where operators have started to make changes in their network. They are modernizing and they are bringing in Ericsson extensively, several customer gains on that. Now to sum up, 5G is really here. This year will be the year we believe where you have much greater ramp up. You will see devices that come down from maybe $1,000 a piece to -- China Mobile's ambition is to have less than $300 a piece, that will enable us to scale up. All OEMs will also have 5Gs this year, I believe. We can see that you can build networks in a much smarter way, and you can do that by leveraging the net record investments you've already made and spectrum you already made, investments you already made, and you can combine that with new types of AI to squeeze out the most of the networks. And operators start to see positive trends on ARPU, at least the ones that have launched early and been proactive. And they do that by offering a greater service and more capabilities and a better user experience, and their consumers seem to be willing to be a little bit extra for that. So with that, I conclude. Thank you very much for your attention.

Thank you, Thomas Norén for that. I hope you all found that useful, a bit of a deep dive into 5G and then Ericsson's capabilities in 5G as well. Now on to the next guest here, my friend and colleague, Fredrik Jejdling, Head of Segment Networks, to give us an update on the Networks business and market. Please, Fredrik?

Okay. Thank you so much, Carl. So what I will do here over the next 25, 30 minutes, is to, as Carl said, look a little bit at the market update for Networks as well as a Networks update in that market; and thirdly, towards the end, lift a little bit on the lid to see what are we thinking beyond 5G. So those are the 3 key topics that we have here. It's great to be here. Let me start off a little bit where Thomas ended, and that's on the market side then. And if we would say that 2019 was the time of implementing and the first steps of 5G, we saw about 13 million subscriptions in place across key continents, key markets like Korea as well as the U.S. market largely. We now see, of course, China kicking in throughout 2020 with more subscriptions. We have high-end mobile phones with early versions of chipsets, early version of high-end telephones and it's all in all generations introduced. You see a high price point at the beginning, but ultimately, the uptake of any technology lies lesser in the technology itself, but more in the handsets that are offered onto the market. Like any technology as well, the initial capacity is in the big cities to offload existing mobile networks. The key exception for that would be Korea, where we see a more distributed coverage, possibly also Swisscom. Initially, it is the business that we all know. It's about smartphone, it's about faster smartphones. We managed ourselves to bring up the speeds up to 4 gigabit per second last week in our labs. In live deployments, we look at 500 megabit per second, so mobility 1 gig in stationary situation in Korea. And it's built on the version of core networks that we brought in to be able to launch 5G earlier, which is what we call the stand-alone version. Now if we look a little bit at what we think is going to happen over 2020, it's now time for 5G to start scaling up. And in that we see subscription number growing up to, let's say, around 100 million subscription by the end of the year. And that would then mean that it's a faster uptake than that of 4G as and when that happened. And as I said before, the key driver for that is the handset pricing. And as Thomas mentioned before, the $300 point seems to be where China Mobile and other major Chinese operators seems to be the price point where volume can start happening on the device side. We think also that the early movers of 5G will start deploying a wider coverage. Some of them lucky to have an Ericsson installed base can do that relatively quickly. But we believe that, that is likely disregarding event to happen, to be able to find a more ubiquitous coverage across the geographies as well. We think though that on the totality of the operators, about 1/4 of their operators will actually launch 5G by the end of 2020. While the consumer use cases are the ones that are likely to continue, where we see the most AR/VR type of applications, such as in Korea, we see a very rapid growth of the application of 5G into other areas, and that is what we originally built 5G for. And that was to connecting with things and things with people and people with things. Those use cases are starting to come into play. And we work very much in partnership. And as you know, we go to market with our operator customer, CSPs, but we work also with the industry to enable those use case to motivate that type of investment in relation to standard type of WiFi applications and the benefit of our solutions there. To make all this happen, we need stand-alone. My colleague, Jan will talk a lot more about that later on. But those data core architecture will enable network slicing, different type of use case to be enabled on latency and speed characteristics that are required. So we think that while we enable 5G during 2019, the scale-up is likely to happen then on 2020. In this environment here, I would say that we've had a pretty good start at the Ericsson. As a matter of fact, we were first to launch 5G across 4 continents. And that's also one of the biggest differences between 4G and 5G that it happens across all major continents across 3 frequency bands at the same time. And I will get back to the challenges and opportunities with that. We now have 81 commercial 5G agreements in place. And more profoundly, so we are live in around 24 networks. We see then that in those live networks, we have very a high performance related to our competitors. And that is due partly to what Thomas was talking about. Our Massive MIMO architecture is different to that of our competitors, providing a better performance. In the U.S. market, we have then launched on the millimeter wave side, been able to then capture 65% of the U.S. city deployments. And as I mentioned before, also, the operators, CSP, communication service providers, that have our footprint installed, can very rapidly deploy 5G on existing 4G hardware through basically a software upgrade. Our performance advantage is also very much linked to that. We build and integrate an open ecosystem of devices. We made sure, as we developed 5G early on, to be very -- to perform interoperability testing with whatever provider in the market to make sure that anyone stepping into an Ericsson network can enjoy a great experience. So it's early on in 5G and -- but you've got to be more happy with a good start, than, of course, a bad start. If I then go a little bit more on the Networks update and talk a little bit what we've done here, and I wanted to show a little bit the full year perspective here of 2019 over 2018. And we then have shown an organic sales growth with this portfolio that we've developed over the last 3 years of 6%, and that is a high-growth rate than the market. And we're quite pleased with that. The market perception or the market reaction to portfolio has been, I must say, very positive. And that has enabled us then to grow in key markets for us, like the U.S. market, South Korea, Italian market, Germany and Saudi Arabia. And I think the conscious decisions we take to accelerate the portfolio, the variance and build a more proliferated portfolio serving multiple bands, serving also new technology with 5G and Massive MIMO is really paying off in the market. Ultimately, the technology leadership we're providing is evaluated in the sense that if we get ranked highest in the techno commercial evaluation of our customers, that's technology leadership for us. And in many, many cases, that has become the case over the past year. Our gross margin has then increased sequentially over -- to almost 42%, 41.8%. That's a very important measure for us. It has to do with accelerating a better, newer technology into the portfolio. And that has been weighing higher than the impact of the strategic contracts that we put forward on the market, which is also important because our market share ambitions are to grow. And that is, in itself, important for the profitability because it offers scale in both production, purchasing and ultimately, the way we provide R&D. So that has given us a SEK 24.8 billion operating income. And again, the higher R&D -- the benefit in gross margin over the high R&D has paid off. So the net effect is positive for us as a company. So that's how we performed over 2019. I would say that with that, we delivered then a 16%, call it, gross operating income, operating margin, and that is then in range with the 15% to 17% that we set forth now for 2020, as we've communicated many times before. Now I want to recap a little bit and then go forward again. And recapping a bit, what did we say about 3 years back when we met at the time in New York for Capital Market Days for Networks? We said, ultimately, we had 3 strategic priorities that we want to execute on. First of all, so we wanted to make sure that we can invest in technology and cost leadership, again, becoming the partner of choice for our customers. In their techno commercial evaluations, we need to make sure that we have the technology that come out on top. That actually has also another added benefit that newer technology is also more cost efficient for us. With that extra gross margin benefit, we wanted to do 2 things. First of all, we wanted to selectively expand market share where it makes sense, where the customers, again, evaluate our technology and the cost competitiveness it provides in their techno commercial evaluation. That was step number one. Number two, we wanted to make sure that we could become the undisputed technology leader and leader in 5G. So those were the 2 things we wanted to do out of the better gross margin profile. Practically, what we've done, just to put a couple of proof points to that execution, we have since recruited then 6,500 R&D engineers over the last 3 years. And had we not done that, we would not have been able to, again, make that market capture over and above the market growth rate that we've been able to do. At the same time, we need to be highly competitive, cost-efficient machinery. And to that point, we also reduced our service engineers and G&A force with about 3,500 people. We plan for about 100 radios to be released in 2020, that is doubling to the levels we were at when we started this journey in 2017. We're introducing now Spectrum Sharing, which I will get back to the only economical way to introduce 5G onto the market. And we also now are working, and we have also concluded then 5G stand-alone call for an end-to-end data call. We did that, just about Q4 here. Market share is steadily growing. But what makes me more pleased, in a sense, is that key operators choose us for the right reasons, for our technology. And therefore, we've been able to capture key deals that I just mentioned a few here, like Verizon, SoftBank and Optus. There's a handful already, but a number of others beyond that, but that gives me confidence that we are on the right track in our techno commercial development. So if you looked at the Q4 '19 result, we grew then 7% quarter 4 over quarter 4, and we increased our gross margin by 1.2 percentage point. So, so far, so good in our execution. Now if we look at what we want to do then, it's really the strategic priorities versus 2020 and 2022 that we've been communicating before. Those priorities stand firm. It is around becoming an undisputable technology leader. In that sense, we talk about winning the technology race. So it is really, out in time, understand what the requirements would be from an operator perspective and being able to deliver that customer value and cost per gigabit per second, being able to do it with the cost structure, including both serviceability, [ TK ] and manufacturing cost, that means the requirements on profitability, and ultimately, being able to price it in a good way. So that is about delivering that portfolio, winning the technology race, but also making sure that we make the long-term bets on technologies that are going to be valid beyond 2023. World class development. We work a lot on ensuring that we can reinvent the scale in the portfolio, in the sense that we can have a marginal lower cost than the revenue per developed radio. And that we can do through investments in architecture like modularity, software/hardware modularity, product life-cycle management tools, et cetera. In next-gen supply and services, we are focusing on, ultimately, for the operator, reducing lead times in half. Through rebuilding our factories into an Industry 4.0 type of a platform, integrating that with our site materials as well as service delivery mean that we can actually significantly reduce lead times. All of this is enabled through digitization. We have a very targeted approach with a select number of customers where we're going and do this type of activities. Compete on value. There's no secret that the radios we build now are far more complex in nature. They deliver far more output, far better cost per gigabit per second. But it's also important that the price points we enter in the market enable us to make money over time on this extra R&D investments that those 6,500 people represent. So for us, it is important that we get the price point right as we manage the erosion in the market in a good way. Finally, a significant part of our business is sourcing. We work very actively both on enabling a more innovative ecosystem but also, of course, make sure that we can get the best deal and making sure that we can get a very attractive cost base from our third-party suppliers and partners as well. So in essence, these are the streams we've been working on for the past 3 years. What we're now trying to do is to intensify them against the wanted position 2022 that spells out the clear market leadership built on technology, business and commercial leadership. Just touching upon where we are competing now in the market. And naturally, so we are in the middle year. And this is our prime market, the radio access network market, where framed it different, we'll invest competing -- continue investing for competitiveness and performance in eyes of the customer, cost per gigabit per second. We'll continue doing that to make sure that we are continuing to be a market leader, technology leader in those areas. Last year, we announced, actually about exactly a year back, the intention to acquire Kathrein antenna-based systems. We've concluded that deal. We had a start, which was a little bit bumpy, considering that we had some initial supply issues due to permitting in certain factories. We now resolved that, and over the year, we will regain full production now. And over a year, we will continue improving the performance of the Kathrein business. The rationale for the Kathrein business is undisputed, in the sense that it provides a combining radios and antennas makes us -- enables us to regenerate, reinvent the site on behalf of the operator, significantly reducing their OpEx on-site, power, site rental, and it's also a critical component for us to be able to deliver next-generation antenna-integrated radios, Massive MIMO, the competence and capabilities we didn't have before. We continue building our own site and fronthaul router portfolio and continue the partnership we earlier announced there. And we are looking into fixed access in the sense that we build the alternative to fixed access being fixed wireless access through utilizing 5G radio access. And as I said before, the extension of 5G into industries is in use-case basis in partnership and through our CSPs. There are a couple of points here why we think we are -- have a better 5G offering than our competitors. Thomas talked to it a lot. It's about the way we've architected our Massive MIMO and the way we developed our algorithm and the way we allocate RAN Compute in the system, most cost and performance effectively. I just want to call out a few things that are -- a couple of them are unique for Ericsson and also the 2 on the left-hand side. The 2 on the right-hand side, we believe have a better application, but are not necessarily unique invention. But if we look at the left-hand side, what is really critical and what has enabled us to take business at this point in time, it has to do with the reusability of the 4G platform. In other words, if you invest in 4G, you don't compromise any 5G investment in the future. You can actually reutilize existing base bands and radios that have been bought beyond after 2015. So a lot of -- there's no choice between 4G and 5G anymore hardware-wise. You can actually reutilize the equipment. So it's a safe and solid investment. Naturally, so if you buy -- if you need to introduce new bands, you need to buy new radio equipment to serve those frequency band. But that goes without -- say, without any -- in any type of generation shift. The second point I want to highlight, and this is an important point, this has to do with possibly the -- or not -- probably the only way you can introduce economically 5G into the network. If you remember before, you used to have to reform a certain band of network capacity to serve either 4G or 3G. We've taken that problem away with Dynamic Spectrum Sharing. The network would sense what type of device is sending what type of signal and on a millisecond, 1-millisecond basis, be able to allocate capacity to 4G and 5G. That takes away the problem of taking away revenue-generating 4G spectrum for nongenerating 5G spectrum, and it's a unique Ericsson innovation. The other parts that I want to call out here, 5G deployment, has to do with carrier aggregation. The way we safely can merge different pipes and ensuring that we can get actually better coverage, utilizing the uplink or the low-band, combined with the downlink capacity on the mid- and high band. And that gives us about a 25% extra performance. And the last part, which I know that Jan will speak a lot more about later on, is the dual mode core and this is actually where the magic will happen with 5G. This future type of use case is related to connectivity between things and people with things are enabled, and the network slicing functionalities will be made available through the 5G core. I'll let Jan continue on that later on. So we think we have some unique strengths as we're developing this portfolio, and I'm happy to say that our customers have been reciprocated that with extended market share for us. And it is what we talk about, a very unique and flexible RAN Compute architecture, the way we can facilitate compute, which is the most sensitive function when it comes to performance in the network. And the way we can dynamically allocate that from a central to distributed functions, even all the way up into the radio where extreme use cases are required. In that sense, we are also looking into a virtualized RAN architectures in the future as well. And as you know, we announced a collaboration with NVIDIA to discover the opportunities of this vRAN-based architectures. I mentioned before, we have a very modular Ericsson Radio System that has been in place since 2015, reusable for 4G up to 5G. Through the acquisitions of antenna -- or Kathrein antenna systems, we have the opportunity then to re-engineer, re-innovate the site concept to make it far less -- or make it far more, I should say, tight and less -- much more cost effective. And the last point, which we announced basically about a week back, is our new approach to services where we use an intelligent network services based on an AI-powered system that look at preemptive services rather than preventive. So before it even happens, we have an opportunity to diagnose the networks and make the necessary patches, software-wise, before a problem is likely to happen. So if we try to lift a little bit on the lid here when it comes to beyond 5G, the way we sit here, then we think there is a concept around agile networks here, where we want to be able to utilize the existing network and build use-case driven extensions on the networks at the most cost-efficient manner. So with the lowest marginal cost, we will be the partner able to bring on new use cases, whether that's a dedicated network for a factory facility, whatever extension else you want to build from your current networks. We think energy optimization is fundamental for this piece, as we got to break the energy curve. And we're doing a lot of research and development to that end, partly higher powered to be able to understand how to dynamically allocate power in the network in the best possible way. Virtualization will be a quick key concept. It's a technology that really enables a far better utilization of the nodes in the network. And we are then, of course, participating both in 3GPP as well as in O-RAN, driving openness in throughout the system. So that's a little bit the way we see 5G beyond where we are today. So concluding, we believe we have a very competitive 5G and technology portfolio now. It's early days of 5G. We remain very focused and humble in terms of the future, making sure we get the right investment to be able to build -- continue building a better 5G today, but also lean forward and look at what comes beyond 5G. Thank you so much. And with that, I'll hand over to you, Carl.

Thank you, Fredrik. Excellent. I think, as you said, it's no walk in the park, but it certainly feels like we are acting from a position of strength now, thanks to all the investments made in technology.

Yes. No, I can conclude and say that we don't stand still. We recognize it's a very competitive market, technology-wise, and we need to stay on top of this. So we need to lean forward even more. There's nothing about that.

Exactly. I have one question though, Fredrik, because then it's a lot in the press obviously about coronavirus. And I thought we could just address that, while Jan is coming. Welcome, Jan. We'll come back in a second. Obviously, we need to be humble and monitor the situation very carefully and so on. But can we talk a bit about the supply chain so far?

We can do that, and I think it's possible I should have brought it up, but I think it's a very good question. Naturally so, is that we monitor this more than daily, I would say. We -- I should start by saying, we don't see an immediate impact in quarter 1 right now. But it doesn't mean that we doesn't monitor in detail. And I -- we have a fairly distributed supply chain, and we work with ensuring that we have a dual vendor for the critical components, meaning that we can be more flexible in the way we manufacture the equipment. We also have different manufacturing sites. We got 2 main Ericsson sites, one of them being Nanjing and one of them being Italian, and we've got a select number of partners. So we actively make sure that we can distribute out manufacturing to where it's needed to meet customer demand. We started up our manufacturing plant a couple of weeks back post people went back to work in China. And so the Nanjing factory is up and running. We got 161 of 170 subcritical suppliers working with us. The key is to ensure that we get lanes booked because the transport and logistics in China is slightly complex. So we make sure we can preemptively book those type of lanes to make sure that we can get our equipment across. Again, we don't see any material impact now in Q1, but we are closely monitoring this. And if there should be a different message, of course, we will go out with the others sooner, so that would -- if that would happen.

Absolutely. Okay. Now, we're all over that issue, for sure, from all kinds of aspects, including health and safety of our employees and partners and so on, and the supply chain serving customers.

Of course. Yes.

Okay. Thanks a lot, Fredrik.

Thank you.

It's very good to hear about Networks.

Thank you.

And now I welcome Jan Karlsson, who is Heading then of our Digital Services business in Ericsson. And will be great to hear about the transformation journey and where we are when it comes to Digital Services. Please, Jan?

Thank you very much. Really, happy to be here, to share some of the things which are on top of the Digital Services agenda. So what we'll focus on today is 5G core stand-alone, also known as 5GC. Why? Because it's a core, new core architecture from 3GPP, needed to fully realize 5G. And it's a superimportant shift for some of our customers. Some of our customers are making the shift already now in 2020. And this shift goes hand-in-hand with the shift to cloud native and automation. Cloud native and automation will change fundamentally how we develop, how we sell and how we deliver Networks in the future. But first, a few minutes on our strategy execution progress and our financial performance. So don't please read the slide as an exact time line. That is not the purpose. But I think as we all realize, a transformation of a business area spans across various dimensions. And here you can see customers, portfolio, commercial and operations. Continued good strategy execution has taken us to where we are. And an example of that is the portfolio rationalization we started, kicked off in the beginning of 2018, which allowed us to accelerate the development of our cloud-native products or generation of products. So if I start with the customer dimension first. We've addressed 75% of the 45 critical bad bank projects by the end of 2019. In the customer dimension as well, it's very encouraging to see how our competitiveness, the competitiveness of our offering, has allowed us to increase our net sales. And I want to emphasize as well that on this customer dimension, the move now to cloud native and automation is a move that will take a long time, many, many years to come. If I move on to the product portfolio, we are investing, continue to invest in the portfolio to make all targeted software cloud native. And 5G Core is the lead product in that portfolio. Commercially, we continue to work across develop, sell and deliver to capture the full value of our software. Commercially as well, we are implementing a beachhead sales strategy, where we make sure that we land with beachheads with anchor offerings and then have an articulation of the rest of our portfolio in attachment to those beachheads. And if I move on to operation, we're, of course, capitalizing on this technology shift with cloud native and automation, increased industrialization of our service delivery to make our operational efficiency increasingly more competitive. On this next slide, you can see how we've systematically improved our performance with a top line increase and a significant improvement in our bottom line. Yes, we have sales seasonality. You can see this very clearly on this slide. And this will, of course, continue going forward with a very high portion of revenue in the fourth quarter. So earnings will continue to vary between quarters. You can also see how our gross margin has been stable since the past 6 quarters. Going forward, we continue to address the critical projects and improve with regards to industrialization and automation. We see that gross margin increasing. Now a glimpse into 5G Core stand-alone or 5GC. What is it? As we all know, our customers have a challenge to grow their traditional business around mobile broadband. For most, it's a very healthy business where the winner is one that delivers the best customer experience at the lowest cost, and the evolution from 4G to 5G supports that. It enables operators to provide their customers with a better network performance at a lower cost per gigabit. And the first step to build or to start is to build 5G radio network and upgrade the core network to something that is called 5G EPC, Evolved Packet Core. It's the same architecture, the same interfaces as for 4G, but with the added ability to connect to the 5G base stations or 5G NR, new radio. This first step is called non-standalone, or NSA, as you can see on this slide. NSA increases mobile broadband speed significantly and the cost per byte is lower compared to 4G. But it does not have the features and functions that will allow CSPs, operators to address new growth opportunities. Fixed wireless access? Yes, absolutely. New revenue streams from enterprises? Not so much. So the full business potential can only be realized with 5G stand-alone. It makes the networks even faster, more programmable for new use cases. And it makes also the networks more open for innovation. Needless to say, as Fredrik has said, we are working and developing our offering across radio and core to be the most competitive on the market across non-standalone and standalone. And on this slide, what you can see highlighted in blue are the products in the Digital Services portfolio. Our portfolio consists essentially of software-based solutions with corresponding service delivery and support services. The most important offering in the evolution to stand-alone 5G is our cloud core -- within cloud core, our 5GC. This is really the anchor in this evolution, and it's also a core platform for our business. So what is core all about? The core network provides a number of functions which are distributed across several software products. The core network acts as a super router of the entire network. All user traffic passes through the core network with different priority, with different security classes. Nothing can be connected to the network without the core network allowing it. It keeps track of all the connected devices. It provides, of course, also the gateways that allows data from external networks, like the Internet, to be -- to pass-through and be consumed by devices. It also provides the charging information in order for operators to monetize the traffic. The interfaces between the different functions are standardized by 3GPP, and it is very common that an operator chooses different vendors for different functions, different products in the core network. And even if the interfaces are standardized, there is a systems integration job, a service delivery job to deploy the software into the customer environment and to set up, configure the specific features that the operators require. To go beyond the mobile broadband and traditional services, the core network needs to be much more programmable and software defined. It must be possible to more flexibly make use of capacity, of resources and of the functionality, and a good example of this is slicing. I think many of us have heard about network slicing before. The slicing support in 5GC is much more advanced compared to previous generations. So the one and the same network can be configured for different characteristics serving different purposes, for instance: one slice for mission-critical communication with guaranteed voice and data between first responders and emergency control; one for consumer gaming; another one for field service staff supported by AR equipment. So these additions in 5G standalone allow the slices to be kept isolated, isolated from each other, making it easier to guarantee security and quality. And it's also possible to have one device connected to various slices, for instance, one for voice and another one for data. In standalone, the possibility for external applications to be connected are also improved through network exposure. In the past, we focused quite a lot -- or operators have focused quite a lot on the retail opportunity, business-to-consumer and business-to-business. With standalone, through exposure capabilities, which are standardized, it opens up for new business opportunities with regards to wholesale. So deploying 5GC is really the only way to fully reap the benefits of 5G, including the radio benefits, performance, low latency and simplified architecture. So non-standalone, the non-standalone option has from the start really only been an intermediate step towards the stand-alone architecture. 5GC is a completely new architecture. It's -- most of the equivalent network functions from the 4G standard have been changed, new functions have been added and the interfaces between the functions are completely new. It's a new architecture. It's a must to provide the programmability and the flexibility, which are needed to address new business. And just to give you a feeling of how big the difference is in an RFP for 5G EPC non-standalone, we typically receive, in a request for proposal then from an operator, roughly 1,000 questions. With 5GC, with standalone, it's more like 5,000 questions or more. It's a completely different level of complexity. The market window for 5GC is open now. The leading operators around the world are looking to launch 5G standalone within the next 12 to 18 months. And we address this with our dual mode 5G Core, supporting both EPC and 5GC. So if I go back to this slide as well. In our 5G Core, in our dual mode core, the EPC functions have been redeveloped from scratch, so that we can provide the TCO efficiency that our customers are asking for. And I'm really, really proud of what our employees have been able to do across flexibility, across performance, robustness, really, really good work. And it's not really only about the products within the 5G Core offering. As I said before, when I mentioned about the beachhead and the anchor offerings, you can see 5G Core as an anchor offering, which is pulling the rest of the portfolio with it, for instance, the cloud container distribution and our cloud infrastructure; for instance, network managers, service assurance, expert analytics and management and orchestration; for instance, charging and mediation and BSS, 5G voice and lawful intercept in communication, to name a few. Now I've said cloud native a number of times, but not really explained what it is and how it will change the way that we and our customers work. So let's watch a video to take a closer look. [Presentation]

So I think a pretty good video of what cloud native and automation is about. And this, for many of you watching here, you might not realize how big a difference this represents for the telecom industry. Typically, we upgrade a product once a year. With this technology, we can go to updates, upgrades once a month, even once a week. And it pulls the vendors and the customers much, much closer to each other, allowing for this automation that was explained in the video. So it allows us to do upgrades or the deployment of new functions without maintenance windows, night time and without risking ongoing service. So is this for real? Or is this here and now? Or is this for tomorrow? And on this slide, you can see 2 press releases that were published not so long ago, one with Ericsson and Telstra, where we are first out with the cloud native core in a production environment. And Telstra are fundamentally looking to change their ways of working and really capitalize on the benefits of cloud native in terms of agility, flexibility, efficiency. On the right, you have a press release with KDDI in Japan. Together, we created a cross-organization of end-to-end CI/CD software pipeline for 5G standalone. And this pipeline seamlessly deploys software from Ericsson's R&D organization into KDDI's environment without human intervention. So what does this mean for us? How does this affect our business mix? We will remain a business where software and the corresponding services together create value for our customers. We see that our software portion will increase slightly in the coming years, partly driven by the automation of the service delivery and the industrialization of the service delivery, but also, as we reduce the hardware component and focus more on the software value. Systems integration is and will remain a significant portion of our business. You can see to the right that the ratio of our recurring revenue has materially increased over the past 2 years, and it will continue to do so, driven by the increased ratio of software in the sales mix, driven by the continued work to shift the licensing model from perpetual licenses to subscription licenses, and driven by CI/CD with seamless upgrades supporting the move to software subscriptions. As an organization, we are really embracing the new technology: cloud native, automation and industrialization. We've made significant progress and we have our first cloud native offering available today. But we have a long way to go compared to where we are to be where -- to achieve the targeted operating model. And it's not only about the Ericsson, it's really the balance between the Ericsson and our customers. This is not a change we do on our own to fully capitalize on the benefits, for instance, of CI/CD. It's a joint work between our customers and us. Some products are ahead, some products are little bit behind. So again, a journey in our transformation. To conclude, my final slide. So 5GC is a must to fully realize the benefits of 5G. I think that point, hopefully, is quite clear. 5G allows for services which are more mobile-broadband oriented, but also services which are oriented to enterprises and opens up for wider wholesale opportunities. Cloud native and automation are the technology for the future and in particular, with the cloud native core networks. Cloud native and automation can make us and our customers more efficient. And my mission is, of course, to continue to execute on our strategy, focusing on our customers and supporting the organization to progress towards our target and achieve profitability in 2020. And with that, thank you very much.

Great. Thank you, Jan or Jean-Jacques as we call you, internally.

Some of you.

Some of us do that, exactly, but thanks a lot, Jan. I think the transformation journey in Digital Services had a couple of key components. And it's about, I would say, putting the house in order and getting into, like, a high-quality unit here, managing the 45 contracts well on track. It's about cost efficiency, well on track, really a big achievement standard already. But then, of course, growth in the new portfolio because of the fact that legacy has been declining over quite some time, and we talked about that. Now you sounded very confident when it comes to 5GC and how that will be a beachhead for the new portfolio. What more can you say there about the growth you see in front of you?

I think cloud native is a big bet. And like I said, it's not just one product, it's across our portfolio. And we've started, we embarked on this, I think, back in 2016, definitely in 2017. And it's so encouraging to see that such an investment, such a big bet is really paying off. The feedback we get now from customers is very, very encouraging in terms of our -- the competitiveness of our product. And I want to emphasize as well that it's not just about the product. If we didn't have those processes, automation, CI/CD, together with the product, we would not be able to -- we would not be able and our customers will not be able to capitalize on the benefits. So it's -- I'm optimistic in terms of the impact of 5G already now in 2020 and definitely beyond as well.

It's all coming together now. Okay, great. But that's good to hear. And thanks a lot, Jan.

Thank you.

You will be back for the Q&A a bit later as well.

Absolutely. Thank you.

Perfect. Thank you so much. And now we will move on in the program here and listen to Niklas Heuveldop, who is the Head of our North American operation. It's also a video segment here, but also Niklas will be available for the Q&A after this one. So let's roll the film. Hi, Niklas Heuveldop, Ericsson's Head of North America. It's great to have you here and have a chat about the North American market. And I must say, 2019 was a great year for North America and Ericsson in North America as well. Now the carriers that have been working on the 5G build-out for about a year now and how has the market played out would you say?

It's been an interesting year for sure. I mean, North America is now up and running in the high band, the millimeter wave, the mid-band and the low band. Started almost -- it's soon going to be 1.5 years ago. So already in Q4 -- early Q4 of last year, Verizon started experimenting with fixed wireless in 5G. Then AT&T came out a bit later with mobile wireless, focused on the enterprise segment. And during the year, both of them have gone into consumer mobility play as well as more devices have become available. Sprint started in the mid-band, entirely consumer focused in a couple of markets. And then T-Mobile started with millimeter wave first, but has now, just before year-end, launched nationwide 200 million PoPs in the low band. So we're up and running in about 85 cities across North America, 55 of those supplied by us, so it gives us about 2/3 of the cities deployed, and we feel really proud about that. We had a very fortunate early start into the market. So the first 6 months of the year, we lit up most of the cities. Lots of learning. So I mean, I think there is a lot more conviction, passion in the operators now that millimeter wave is here to stay, plays an important role in the build-out. But it is also clear that you need to work with all bands to really get the compelling consumer proposition put out. And that's, I think, what we're going to see more of now going forward.

Interesting. And interesting to see how they select different strategies. And I think it's great to see that we have also been able to serve those different strategies from our side as well in a good way, building market share. You mentioned enterprise and the enterprise opportunity in 5G. That's one sector that many people have great hopes for, of course. And how do you see that playing out so far and into the future?

Well, you need to remember, all of our customers have been active in the enterprise play so far in 4G. But I really think, based on what we've seen so far, during 2019, that there is enough evidence to suggest that 5G really gives a whole new opportunity. Combine that with the Edge Compute capabilities, artificial intelligence that operators have been investing in, it seems pretty clear that there is a whole new business opportunity for our customers to seize in the enterprise space. We have done research with ADL, as you know. And it's pretty clear that there is 20%, 30% potential new upsides for our customers in enabling industry enterprise transformation. Early experiments during the year, in the health sector, so we have seen, for instance, processes in hospitals and devices being tracked, patients monitored, early experiments with robots; I mean, we've done some of that as well; smart manufacturing, of course, asset condition monitoring, AGVs, autonomous guided vehicles. So a lot of experimentation going on that really stresses the network. So it will depend on 5G-type capabilities to really enable those at scale, but we don't have the devices yet, right? So the ecosystem is still nascent. But I would say the early experiments are really encouraging to see. I think there's a lot of value to be had out there.

That's interesting. And I mean, our own research talks about the USD 700 billion by 2030. So I mean, definitely a big addressable market for the carriers, for the operators. So it's going to be interesting to follow how the -- actually, the U.S. or the North American carriers, they take charge of this opportunity and make it come real. What about spectrum, Niklas? Do the carriers have the spectrum they need? And what's going to happen there?

They have plenty of spectrum. I mean, the FCC has auctioned out more spectrum in the last 4 years than in all previous years combined. So there has been a lot of spectrum coming into the market. It's primarily been millimeter wave recently. As a matter of fact, there is an auction going on now, 39- -- 37-, 39- and 47-gig spectrum. We had 24 just before new year. So there's lot of millimeter wave coming into the market. But of course, what everybody is really excited about now is the mid-band auctions during 2020. So first now, there's going to be the CBRS auction. That's about 70 megahertz of licensed spectrum in the CBRS band. And then finally, there is now a motion put in place to also auction 280 meg of the C-band in 2 tranches, 100 meg first, 180 meg in the second tranche. A lot of expectations in the market that, that spectrum will come into the market with an auction end of this year, first availability late next year, September time frame. And that will, of course, then allow our operators to go a lot faster by building nationwide coverage. I mean, that's one of the lessons learned during '19. The advantage of having a mid-band spectrum or working in the lower bands, it allows you to work on the existing grid and go a lot faster. When you're working with millimeter wave, there's a lot of zoning and permitting that needs to happen, and it takes time.

That's right. Right, so do we -- without going into specifics on the customers and so on, but do we expect a CapEx boost then following additional spectrum coming out? Do you think that, that will be visible in investment levels?

Not the spectrum per se because we have seen -- again, as I mentioned, we have seen a lot of spectrum come into the market. I think that is part of our run business. I mean, there's always been new spectrum added to the market and there is build going on. I mean just because we have all this millimeter wave spectrum doesn't mean that you're going to see nationwide deployments. Millimeter wave is going to be more for citywide coverage kind of scenarios, right? So you will see a different profile on that spend. I think the one question and T-Mobile had -- the new T-Mobile has taken a position on that is that they intend to spend more CapEx with the merged entity. So that in itself, I think, we all have expectations for, and to be seen how fast they can activate and how fast they can go really.

Exactly. And I mean, that's a big and very interesting topic in itself, of course, that the news on the merger no, finally coming through. How will that impact, do you think the markets, not only for the merged entity but also the other players?

Right. So I mean, everybody has known for 2 years that this may very well happen. So everybody, of course, understands what that would mean. It's been 2 years since their first announcement. I don't think that there is going to be any change in strategy by any of the other players. They are executing on their strategies, which are and have been under execution for quite some time now. So I don't think we'll see a change there. Yes, they have been very active in the market as evidenced by media activities and others. So the big question is, what is T-Mobile, the new T-Mobile is going to end up doing now with the merged entity, with Sprint, because that -- there has been commitments about an additional $5 billion spend over a 3- year period. How much of that is going to be in the earlier cycle versus the later cycle, how fast can you go because we still have restrictions to deal with. There is zoning, permitting and there is work on tower crew capacity that needs to be done. So we're following that, of course, very closely. We're super excited about the opportunities with that. And then again, we have been looking at this dynamic of consolidation around the world, a consolidation from 4 to 3 has typically shown that the 3 remaining operators have more resources to spend and will invest more, which is also for the benefit of consumers because it builds a better service, a better experience. What is new here now is that we see Dish coming into the market as well, and obviously, going to try to come in with a disruptive play. And my expectation is that that's going to create a whole new dynamic and maybe open up new market segments altogether. So maybe what we've seen as we're going from 4 to 3 is good, bringing in a disruptor is bad may not actually play out that way in North America. I'm optimistic that the way we're going into 5G now, and Dish basically doing a stand-alone 5G, but in a different way, can actually create new market segments and net new revenues for the industry. And that's one we need to stay close to. It's super exciting. And by the way, Charlie has his earnings report here in a couple of hours. So by Monday, we'll know more about it.

We will know more, for sure, about that, exactly. Now but -- okay, so -- but that's interesting. I mean, it's going to be good for consumers, for industries and for the general economy, of course, that actually, this has now finally happened. That's the way I read what you say. And clearly, we are open for business with all these customers.

Big time.

Big time, yes absolutely. But you mentioned the word disruption here. So I have to ask about O-RAN. How do you see that from the North American context: threat, opportunity?

Natural evolution. I mean, our networks will continue to evolve towards more open architectures. Medium to long term, I would say, primarily positive. I mean, I see opportunities to lower R&D costs, bringing third-party solutions in a different way. Again, I don't see that any of our customers are really pressing us to reduce the spend. They're trying to get into new spaces. I mean, again, thinking of the enterprise opportunity, I think O-RAN may lower the price points, so that we can actually go after bigger market. And again, I mean, we are, of course, betting on being leaders in O-RAN as well. So that should be good for us as well. We should participate in that uplift. Short term, there is a bit of an overhype right now in the market, some confusion. So it's not going to be without challenges to get to the end state. But I think, again, medium to long term, I'm convinced it's going to be positive.

Good. And I guess an important part is for us to be sit at the table and be part of this development as well.

Absolutely.

And like we are with the virtualization of different kinds and so. We talked a bit before about the investment levels and so on. And how would you describe now, again, if you summarize this year and into the future, about what's the investment level? And also, on our side, we have talked about shortage of tower crews as well. You mentioned it, I mean, a few minutes ago, top of mind, obviously, to get big deployment capacity on our side as well. But how do you see if you summarize these factors playing out now in terms of investments and our capabilities?

I mean, generally, we subscribe to the notion. I mean, we've had 2 years of good growth. I mean, double digit last year going into this year. I mean, we're coming in at a pretty high build ratio, you could say. I mean, again, everybody has been busy building out the underlying 4G infrastructure, the fiber networks to then put 5G on top. I think there is a little bit more to be had. I mean [ Dolores ] talking about the 4% year-over-year growth. I think we'll agree with that. I think there is a bit more growth to be had, in particular, now with Sprint, T-Mobile happening, that should drive an additional spending in the market. I mean, we've done a lot of work on the tower crews. We have certified, what is it, 842 climbers last year. So we're trying to add capacity into the market. So that's going to help. And then also now, as we go into millimeter wave and, call it, city urban deployments, we don't need to send people up a 600-foot tower. So we can work with bucket trucks and different form factors. So I'm optimistic as we see our customers shoring up build permits that we can actually pick up some pace there as well. So I think 2020 and going forward, it's going to be a bit more balanced between macro sites and street environments, rooftops and street-side. So I'm hoping that, that will also allow us to go at a higher pace.

Fantastic. Niklas, you're oozing with confidence here. And of course, as you say, not without challenges, but definitely I think you've done a tremendous job there last year in building an even stronger position. Let's continue with that.

We feel good but never complacent about where we are, and it's a super exciting market. So much momentum, so much that's going to happen. Also, really leading the way, hopefully, then on the enterprise transformation, which is the big bet for all of us. I'm super excited about that.

Which could have global ripple effects as well, of course.

Yes, exactly.

Exactly. All right. Thanks a lot, Niklas. Thank you.

Thank you, Carl and Niklas, recorded a week ago.

Yes, exactly.

About a week ago. You wore the same suit as you did there.

I try to be -- stay consistent.

Good. So now we're going to start the most complicated part, I guess, of the whole exercise is trying to dial in to Niklas and Thomas. So I will actually start there to see do we have Niklas on the line from the U.S.? Can we hear you?

I'm here. Yes.

Great, Niklas. And I will do the same check with Thomas from the northern part of Sweden. Thomas, are you with us?

I am. I am. Can you hear me?

Yes, perfect.

Great, Thomas. So stand by now for the Q&A session. So ladies and gentlemen, we will now start the Q&A session.

[Operator Instructions] So with those instructions, I would like the operator actually to open up for the questions. So please, operator?

Our first question is from Sandeep Deshpande from JP Morgan.

Two questions, if I may. I mean, firstly, regarding the U.S., I mean, given what you've seen in the recent change in the market in the U.S., I mean, how are you seeing the spending progressing? I mean, earlier on in the year, it looked like there was a sort of a pause in the U.S. And so do you expect that to change through the year? And then secondly, based on your presentation, you are very bullish on this Dynamic Spectrum Sharing. But overall, Dynamic Spectrum Sharing essentially would mean that you will bring 5G into the existing 4G bands, which may be already fully congested. I'm not entirely sure how that brings a positive experience to the consumer, though it may bring the ability for an operator to indicate that they have 5G already in store?

Thank you, Sandeep. As we have Niklas on the line, I would ask you to take the first question on the U.S., of course, and then Fredrik, when it comes to Dynamic Spectrum Sharing. Niklas, please go ahead.

Okay. So yes, as we had communicated earlier, we saw a bit of a slowdown in Q4, but still a very strong full year, 11% up for us, so still running at a very high rate. And now with the T-Mobile merger being on track to conclude in a favorable way, we expect that new entity to be up and running in Q2, April 1 is the target date. So that should, of course, lead to a pickup in spending. Then the question is how fast. I mean, they have, of course, been planning for quite some time. I don't have full visibility as to the sequencing and how fast they can start consolidating and then actually putting new equipment on the towers. So it may take some time before we get into, call it, a good cadence, but I will absolutely expect to see the total spending throughout the year to start picking up. Again, I think that that the lower of 4% year-over-year growth is a good guideline that we also stick to.

You're talking about the U.S. growth of 4%?

Yes, sir.

Thank you, Niklas. And Fredrik, what about the Dynamic Spectrum Sharing and the consumer experience?

Yes, Dynamic Spectrum Sharing. Let's look at the alternative to Dynamic Spectrum Sharing, and that is reforming a predictable bandwidth, let's say, 5 to 10 to whatever in partitions of 5. So I think the alternative is to do that. Then I think typically, 5G will be introduced with more mid-band or high band. What Dynamic Spectrum band does -- Dynamic Spectrum Sharing does on the lower band, it allows for coverage across the whole country by not having to allocate the full band, but actually, dynamically allocate it. So I think it is probably the only way to do it, both from a user experience perspective, but also from a financial perspective.

Okay. Sandeep, are you happy with those 2 answers?

Okay.

Thank you, Sandeep. And then we are ready for the next question. Operator?

And our next question is from Achal Sultania from Crédit Suisse.

So first question, Carl, is on the OpEx side. I just wanted to make sure that I heard it correctly. Like, are you saying in your remarks that OpEx-to-sales ratio will be up from 28.1% last year, as we go into 2020? And is that dependent on the sales being at the midpoint of your guidance range? Or is it going to be up either way, whether we hit the lower end or the higher end of the sales target?

Thank you, Achal. No, what we're saying is consistent with what we said in the Q4 report as well, that we do some investments now in OpEx in Digitalization, Security and Compliance, and that will increase the absolute amount. But our ambition is still to reduce the percentage of net sales on the total OpEx. So that's our ambition.

Okay, understood. And maybe one on the Digital Services side: can you give us some more color around how much is legacy now in DSS as a percentage of sales? And any color around how much money those legacy businesses are losing? Just trying to understand, like, how should we think about the cadence of profitability improvement as we go into 2020 and 2021?

Thanks for the question. So if I take -- we divide our portfolio into a growth portfolio and a legacy portfolio. And as you can imagine, our sales over the past years have declined mostly because of the decline of the legacy portfolio. Where do we stand in 2019? 30% of our sales are with the legacy portfolio. 70% are with the growth portfolio. And we see this really flattening out. Now we don't see a continued steep decline. And even if there was a big decline, the impact on our total sales is going to be marginal, given that it's only 30%. In terms of 5G, it doesn't yet really move the needle. But I think it will in 2020 because, like I said on the call earlier, we see customers wanting to deploy standalone, probably faster than we had anticipated because radio is there, terminals are there and the core networks are there. In terms of R&D spend, another data point, I'm not sure how useful it is. But at packet core, we're developing, we're spending roughly 50-50 of our R&D on 4G and on 5G. Why? Because as you can imagine, 4G is still paying out salaries given that 5G is not yet moving the needle. Not sure if that was an answer to your question, but some data points at least.

Okay.

Thanks, Achal. So next question.

And our next question is from Johanna Ahlqvist from SEB.

Three questions, if I may. The first one relates to China, if you can give any update on when you expect the Chinese 5G rollout to kick off? And then second question to you, Jan. I see your slide on the increasing recurring revenue and I'm just wondering, do you expect that this transition to increased recurring revenues will not be on behalf of sort of decreased top line? And then the third question to you, Carl. I'm just -- I saw some numbers on the digitalization initiatives you have, which will cost SEK 0.5 billion to SEK 1 billion per year with positive impact from 2021. I just want to confirm that this is included in the OpEx guidance that you left already.

Yes, maybe I can take the last one first, Johanna. Yes, it is. Yes, it is. So we're not providing any new information. It's -- we're maybe detailing a little bit that component, but it's in line with what we have said before. So Fredrik, do you want to take the China question?

Yes, I can take the China, and I can do it fairly quickly. I mean, to this point, we've done large-scale field trials and evaluations linked to those for 5G for the major operators. That, we're quite pleased with the outcome of those results, that's good. They are quite demanding. However, the larger scale rollout is a bit depending on the market share allocation, which is likely to happen -- or the decision on that is likely to happen end of Q1, early Q2, a little bit difficult to forecast. So I'll have to get back on that, Johanna, because it depends a lot on that ultimate, call it, allocation then.

Yes. And of course, this applies across both radio and core, what Fredrik just said. So we have quite encouraging results as well in the evaluations that the Chinese operators are doing for 5G standalone. And regarding recurring revenue, to answer your question, Johanna, no, we don't see this impacting top line. We're actually very, very happy with the development of the recurring portion of our revenue. And it gives, you can imagine, so much more stability, so much more predictability compared to a lower portion of recurring revenue. So I think it's one of the -- over the past 2 years, I think one of the most powerful metrics of the success of our transformation. So thanks for the question, Johanna.

Okay, Johanna. You happy with that?

Yes, I am.

Thank you, Johanna. So we'll move into the next question.

Our next question is from Frank Maaø from DNB.

So a couple of questions. So you mentioned on the strategy execution part that you have a work stream on how to compete on value. My question goes to how you work, if you could give some more color on how you work actually to capture the value of upgrading the 5 million radios and attach base bands by software upgrades basically. They capture the commercial value of that to support 5G, DSS, AG RAN, finally standalone over the next 3 years or so? I would assume that some -- you made some giveaways to some features, but yes, if you could elaborate a little bit on potential upside from having that kind of installed base on what -- on which to upsell those features?

Sure, we can do that. Fredrik?

So I can put a bit of color to that. Typically, we sell our hardware with what we call capacities that grow over time. As and when the capacity needs to be utilized, you deploy capacity to the network. On top of that, we have software packages. Now I can't get into the exact commercial detail, but of course, the software upgrade enabling 5G is part of that software sales package that we also have forecasted a part of the revenue range that we've given and the growth of 3% to 6%. So that's very much there. Ultimately, when you introduce 5G, what is important, both from a radio and capacity perspective, hardware and capacity perspective, is that the price points reflect the increasing richness of the radio. So I mean, the 64TR radio is far -- fairly quite a bit more complicated, complex than a 2TR radio. So our ability to work both with the feature of upgraded 5G network that we mentioned, but also the way we put new radios into the network and for that matter, baseband, the capacity related to that, those initial price points are very important to get right for us. So those are just some of the -- a flavor of what the parameters that we're working on to be competitive. Now needless to say, it's a highly competitive market and technology -- advances in technology enable us to gain a temporary advantage around certain aspects.

Okay. Frank, did you have a second question? Or?

Yes. So if I may, you made some pretty bold claims when it comes to your infield performance based on your MIMO architecture and also the 5G core performance, which is great. But while we have this product launch in -- I think it was in London last week, where they also made some pretty strong claims reiterating that they were 12 to 18 months ahead of the competition, they launched a multiband radio with 400-megahertz of ultra-wideband bandwidth, I think it was 400-megahertz per carrier and so on. So how would you help us reconcile who is ahead here?

No. Of course, I mean, it's a good question. I think ultimately, our customer decide in their evaluations to who would they prefer as a vendor, needless to say. I think what we spoke about specifically on the Massive MIMO radio system, which is used in a bit of a different architecture, enabling us to be more cost-efficient in the deployment as well as moving algorithms and software up into the radio, meaning that you have a shorter response time. And that is why we're seeing that uplink performance advantage that Thomas can talk a lot more about. Now ultimately, I mean, if you really look at it, we were the first operator to launch across 4 continents with 23 contracts. And it's really from that perspective, hard to see -- hard to be behind when you don't really see anyone ahead. It's always going to be in a very proliferated demand or bands across many operators. Sometimes, 1 competitor might be ahead on certain bands; other times, we will be ahead. But we feel comfortable about the way we have developed the portfolio. We feel comfortable about being first on 4 continents with our networks and the 81 commercial contracts. I think that needs to sort of speak for itself.

Okay, Frank?

Yes.

You're good?

And finally, a question to Fredrik regarding open RAN. Could you talk a little bit about the limitations that the device ecosystem basically places on the expected short-term development of open RAN? I think you talked about that last week in an industrial event according to Light Reading. If you could elaborate a little bit, Fredrik, on the device side limitations that will be there for open RAN?

Yes. I mean, if we look, I mean, starting in the end question there, of course, I mean, our interoperability testing across multiple devices enable a rapid deployment that otherwise would have to be done more a system integration basis if we stand on a one-on-one unilateral or bilateral basis. We do it across the network vendor system. I think just to back up a little bit on the -- not talking maybe open RAN, but maybe O-RAN then, or openness in general, I think we need to understand also that 3GPP as such is an open environment in such a way that you can participate off RAN basis. Anyone is allowed to participate and as an open standardized environment that has connected 8 billion subscriptions over the last 20 years, that's a starting point. The second point on O-RAN, we've been a member for a little bit over a year now. And we have actually been part of in bilateral and multilateral agreements being able to standardize and open interfaces like noncritical sort of network management functionalities called A1. We have been standardizing open up the interfaces across the central unit and the distributed unit inside the new 5G architecture. So we're working a lot with these type of factors. Now when it comes to certain interfaces, one is to realistically look at the cost performance benefits of what has been developed in the baseband architecture that Thomas talked about. And that gives to the way we see it, significant performance advantages and, thereby, cost advantage as and when deployed in the networks. So we see potentially, as also Niklas alluded to, there could be use cases over time where the performance requirements are limited in certain environments. And if -- when and if and when those happen, of course, if there are O-RAN standards defining that, we will have to be the first player playing in those areas as well. So that's a little bit to give you a flavor of how we see on O-RAN. So we work a lot on openness. Then it comes down to cost performance analysis on the existing bands where we have developed bespoke compute environments that are required for these extreme use cases around 100- to 400-megahertz of mid-band compute performance. That takes a lot of the acceleration capabilities and software capabilities that we developed that we spoke about. Again, parallel to that, if use cases develop where O-RAN specification is sufficient, we need to be first there as well.

Great. Thanks, Fredrik. You're happy now, Frank?

Yes. Thanks.

Just wanted to break in and say, there was a number, which I'm not sure if it came out correctly earlier in the discussion here, and it was about the growth in North America according to Dell'Oro. It's 1% estimated for 2020. Not sure, however, how they have factored in the Sprint and T-Mobile merger in those numbers, but it's 1% as per Dell'Oro. The world growth is estimated at 4%. I just wanted to clarify that again. Back to you, Peter.

Okay. Thanks, Carl. Then we are open for the next question.

And our next question is from Simon Leopold from Raymond James.

I appreciate you guys doing this call in light of Mobile World Congress being canceled, so good to get this briefing. I wanted to follow-up on your comments regarding your production in China, so a couple of aspects there. One, help us understand what portion of your revenue is based on production out of China? And I believe you said you didn't expect much of an impact on your production due to the coronavirus. I'm wondering if, one, you've got some thoughts on what the sort of variance is around your confidence in terms of supply chain checks for you? And two, does this give you some competitive advantage where perhaps your competitors are facing stiffer challenges, getting factories up and running or getting components relative to your abilities?

Shall I take that? I guess, yes, Simon, I think a very relevant question. As we said, this is something that we monitor daily, if not a lot more than that. We can't get into any distribution between the major sites we have and the proportion of revenue allocated to each one of them because we have our 2 main manufacturing sites that we own ourselves. Those are in Tallinn, one of them, and Nanjing being the second one. Then we got a range of various, what we call, manufacturing services company partners, which we have across the world. And we try to make sure that we can be resilient both from a component supply perspective as well as a manufacturing site perspective as far as we possibly can to serve all our customers in the various markets we serve under. Then we got a third factory, as we said, but that is likely to go -- that will go live here during H1. Now we said limited no customer impact with the visibility we have today. Of course, it is a -- and we did start a manufacturing site in Nanjing, started that up in February 10 at midnight. We have been able to onboard people in a sufficient way so far. As I said, we've got about 160 or 170 of the critical suppliers up and running. So again, with the information of the data at hand right now, we see limited impact on Q1. But of course, it's an extraordinary sensitive situation, should it hit one of our manufacturing sites or should one of our suppliers -- key suppliers be impacted. So we work around the clock to try to mitigate any impact and, of course, as soon as we have contrary information, we'll be sharing that as soon as we can. The challenge is more so from the perspective of logistics in China where travel is fairly limited between the various regions. So it's important for us to proactively ensure communication and travel for those unique components, et cetera, to be able to assemble in Nanjing. So that's what we monitor as closely as we can.

The final point was about competitive advantage. I think...

That is hard to say. I can't really say. I look after our own supply chain and try to find contingencies for that at any scenario in a very challenging situation. So that's what I put focus on. I can't really answer to our competitors in that.

Simon, was that coronavirus update good enough?

Yes. No, I appreciate it. I have to imagine, you do have some competitive advantage. It's just very difficult to assess what seems to be a fairly fluid situation. But thank you for taking the question.

Thank you, Simon. We move into next question.

And our next question is from David Mulholland from UBS.

I just wanted to ask one on the Digital Services business. And obviously, you've been working for a few years to move towards a cloud-native environment, kind of visibility to continuously innovate. I think it was in the last, what, 18 months, we had a software glitch in one of your updates that caused some issues. So just as you look to move to this type of an environment, what procedures have you put in place to make sure that as you potentially have this direct link to deploy solutions to your customers' networks, that it is robust, that we don't have those risks going forward?

I think risks always exist in software. That's just the nature of the beast. But I think since way before -- yes, way before the 6th of December event in 2018, we had procedures in place of securing good quality in our products. Unfortunately, we -- there was a fault in one of our processes, which caused this incident. And since then, we've doubled down even more with regards to processes in R&D, in verification. And like you're hinting, CI/CD goes hand-in-hand with quality. You can't do CI/CD with insufficient quality. And I'm really proud with the evolution of our product quality across the portfolio. It has steadily improved and just keeps improving. And our sites are set very high going forward. So super relevant question. You can't guarantee 100% quality in software. But I think another point to this which is really important is the way that our 5GC is designed in terms of resiliency. Already in 4G with the concept of having MME improve is improved even more now with 5G. So hot topic, very important topic in the discussion between us and the customers, when you move to a cloud-native Kubernetes how are we ensuring security, how we're ensuring quality. So more than happy to take this off-line if you want where we are and where we're going. But I feel confident with the progress and the plans we have going forward. Thanks for the question.

Maybe just 1 quick follow-up on that then, if it's all right, just the -- obviously, your customers are going to remember that December event. So is that still something that comes up in discussions today as you start to move towards 5GC? And are they comfortable that you've put enough measures in place to still make this type of a transition?

Yes. No, I would say so. I think, as we all know, we've been around for so long, it's how you deal with a situation that defines you. If you say no, no, no, but if you say yes, we made a mistake, and this is how we're going to correct it. So I've been in numerous discussions with the affected customers, but also with other customers about how we're improving our processes to make sure that this doesn't happen again. So again, to answer your question, no, I do not see a business impact because of the December 6, 2018 event. And like I said previously, I feel, if anything, that we're -- I would say that we're actually increasing market share in our portfolio and, in particular, in the packet core area, very confident about the opportunity in 2020.

Thank you, Jan. David, you're good?

Yes.

Thanks, David. We will move into next question.

And our next question is from Alex Duval from Goldman Sachs.

You talked today about the incremental investment in digitalization, which, of course, is included in your targets. Clearly, you've also recommitted to the market to your EBIT targets driven by the revenue range and 10% margin targets. So I just wondered to what extent that can be scoped to dial that down or defer investment through the year should the scenarios evolve a bit differently for regions like China in terms of pricing or market share? And then my second question is, you've talked today about several tangible ways in which you're on track for your 2020 targets. But also, you clearly have these 2022 targets. So as we look forward to that sort of longer-term range, I wondered if you could just talk a bit and potentially rank some of the biggest drivers that will help with that progression, if we think about things like growth in 5G Core which, obviously, you're quite bullish about, but also digitalization across the business and any other key factors?

Sure. Thanks, Alex. Yes, on digitalization, I would say, because of the way we plan and execute on these digitalization projects, they theoretically, absolutely can be dialed down, as you say, if we need to. But let's say we are definitely not there yet. We are committed to the 2020 target as well as 2022. And we believe that the investment in digitalization will give a return that clearly justifies doing that. But theoretically, on your question, of course, they are -- we can scope them up and down if need be, for sure. When it comes to 2022, we are firmly committed to those targets as well, obviously. And maybe the most useful thing is to hear a bit from Jan and Fredrik, when it comes to your thoughts on what is going to take us towards the 2022 levels.

No, I can say, I mean, to start with, the next milestone is, of course, this year, and we've said, given a range, 2020 of 3% to 6% growth on the revenue side and a range of 15% to 17% then on the operating income, in that scenario, there is, of course -- you alluded to the China situation. There are various scenarios there on China in relation to the U.S., that gives us, to our best knowledge that, that is the square, so to speak that we will land into. Our ability to continue to capture market share in line or faster than the market on RAN is, of course, a critical component. And our ability then to turn our R&D into productivity that I talked about and being able to meet this demand with more output on the cost side is, of course, a critical component on meeting the long-term targets. So that's why we focus on the combination of a competitive portfolio by 2022, 2023 as I said and match that with a development force that can sort of break that curve. Those are the levers, I would say, from network side.

Good. And Jan?

Yes, a couple of comments. One is, I think, we've shown our capability to execute on our strategy, lay out the strategy and execute on that strategy. If you look back on where we were in 2017, a very, very complicated situation. We could not even talk about 10%, 12% in operating margin. But based on what we have achieved so far and, of course, confident that we will continue to execute. And it's about continuing to execute on the cost side, not just off-shoring relatively low-hanging fruit, but automation, industrialization, working much smarter compared to how we've been doing in the past. But it's also about the success of the portfolio, the competitiveness of the portfolio. And a couple of points there is it's all about -- maybe I'm repeating myself here, but it's all about 5GC. Of course, the evolution to non-stand-alone, but then the stand-alone opportunity. And there, if you see that 5GC is an opportunity itself we need to capture, imagine the upsell opportunity with orchestration, with assurance, with a cause. The suite of products is really articulated around those anchor offerings and allowing us or capturing that anchor position will give us a very strong position to continue to generate the top line that we have in our plans. So like Fredrik said, it's all about 2020, stepping stone into -- for the plan here to reach the 2022 target. Thanks for the question.

So thanks, Alex. We can take 1 more question as we're getting closer to the half hour here. So please, operator, open up for the last question.

And this question is from Jörgen Wetterberg from Nordea.

I'll limit myself to one since we're short on time. I have a question on North America maybe for Niklas. So wondering about the software-hardware services mix shifts from -- in 2020 versus 2019. We've understood that a lot of the major customers have deployed a lot of hardware, antenna sites, et cetera, and that there will be more software upgrades in 2020, 5G hardware activation codes, Dynamic Spectrum Sharing functions, et cetera. And then you've also mentioned that services will increase overall. So how will the mix shift here? And how should we think about gross margins?

Did you get that, Niklas?

Thanks for the question. Yes, thanks for the question, Jörgen. No significant change in the mix. We saw an uptick on the services mix in our business during 2019. If anything, we'll probably see a little bit more of that. We're going to be very busy rolling networks out, and we have taken some important market share. As for the rest of the business, I don't expect any fundamental change. We're not done with the 5G rollout. So don't expect this to go into software-only for quite some time. We're going to be busy building and installing base stations. So no significant shift, maybe a bit more on the services side, but we managed that pretty well in 2019. So nothing to add I think. I hope that clarifies the question.

Thanks, Niklas.

Yes, absolutely. Maybe one more question, if I may. Jan, on the Digital Services, you said that the 5G core launches are being planned for by most major operators and to be operational in 12 to 18 months. When will we see impact on your sales from that? Will it be on the date of launch? Or will it be before that?

So a few operators are -- and the leading operators, yes, are planning to launch standalone in 2020, in South Korea, in China, in the U.S. and in some European countries. To answer your question, when do we see the impact, when do we see -- expect to see the impact? Yes, when they go live, definitely, definitely. And -- but again, it won't be a big impact in 2020. There will be 5G-related revenue in 2020, but it won't be big needle mover. That change will come in 2021 and going forward. But we're -- it's so important for us to be selected in those early launches by those leading operators. They are really showing the way, leading the way to the rest of the world. Yes, so good question. Thanks.

Thanks Jörgen. So before closing this event, I actually would like Carl to finalize with some closing remarks. So please, Carl.

Yes, I will. Very briefly then the heading of the event here was Ericsson on track. And hopefully, we have demonstrated that we are on track. We feel good about the development we've had so far and, of course, never ever complacent. But we think we have a good position now to capture the opportunities provided by the technology shift into 5G and to create additional value there. So I hope you found it useful to get a bit of a drill down into the Networks side, the 5G, Digital Services and North America together with us this afternoon. And I would like to thank you for your participation and for joining us here. So thanks a lot, colleagues. Thanks, everyone, on the webcast as well. Thank you.

Thank you.