Severin Schwan - CEO Daniel O'Day - CEO, Roche Pharmaceuticals Roland Diggelmann - CEO, Roche Diagnostics Alan Hippe - CFO and IT Officer

Luisa Hector - Exane Jo Walton - Credit Suisse Sachin Jain - Bank of America Jack Scannell - UBS Richard Vosser - JPMorgan Vince Meunier - Morgan Stanley Andrew Baum - Citi

Good afternoon, everybody. Pleasure to be back here in London. 2017 was a good year for Roche. We achieved our financial targets, but more importantly, we made significant progress with our launches and with our product portfolio. You've seen this morning the figures. Sales up by 5%. Core earnings up by 5%. Dividend increased for the 31st time in a row. Pharma, up by 5%, very much driven by the new launches. Diagnostics, also up by 5%, again, as in previous years, driven by Immunodiagnostics. You see an acceleration of the growth throughout the year in 2017 and again, very much driven by the new products, Ocrevus, in particular, which we have launched in the second quarter last year. If you look at the numbers from a regional perspective, you see very strong growth in the international business for diagnostics. This is driven by China, over 20% growth in China, now the second-biggest market for the Diagnostics Division. Conversely, if you look on the pharma side, the growth very much driven by the U.S. where we first launched our new products. On the other side, Europe with a decline. They now also see the first impacts of the biosimilars coming in, MabThera in particular, which has been affected by the entry of biosimilars in the first countries in Europe. Now, what I find very satisfying is that the vast majority of our growth on a group level and certainly in Pharma is driven by the newly-launched medicines. So if you just take Ocrevus, Tecentriq, Perjeta and Alecensa, that accounts for two-thirds of the costs on a group level and it accounts for 80% of the growth for the Pharma Division, and it's this growth of the newly launched medicines which we compensate for the increased entry of biosimilar as we go into 2018. Unprecedented advances in our pipeline, we had 25 approvals across Europe and the United States last year. I just like to emphasize again the APHINITY data. So finally it was confirmed by the FDA. We got accelerated approval. We got an excellent label. And as we keep saying, this will continue to drive our growth in the United States as of now and later in the year when we get the label in the European Union. As far as the trial readouts are concerned, there was a lot going on. Let me just highlight again, Tecentriq in cancer immunotherapies where we had the first two positive readouts from our combination trials in first-line lung cancer and renal cancer, two of eight studies which we are going to read out by the mid of this year. Not only the quantity of our entities in our pipeline, it’s also very much about the quality and the potential medical differentiation. I think this is also demonstrated by the number of breakthrough therapy designations we have been given by the FDA, which speaks to the quality of those potential medicines coming forward. Turning to the operating results, core operating profit at CHF 19 billion at a record level, again, core earnings growing and the strong improvement on the cash flow with over 20% as a base for continued dividend increases. Now, turning to the outlook, I just again remind us of the rich portfolio we have at this stage. The portfolio has never been stronger. Of course, it comes at the right time. Last year, we had 14 new molecular entities in late-stage development, over 30 important new indications, and it is this pipeline, combined with the early success we see with recently-launched medicines such as OCREVUS, which makes me confident that we will grow through this upcoming period of transition where we will see the entry of biosimilars for major established medicines. On this note, again, the outlook you can see that we intend to grow for 2018 in spite of the biosimilars entering now in a much broader sense. We’ll have the full-year effect from MabThera. We will have Herceptin on top of it in Europe. But we will still grow. I mean, this speaks to the strength of the new portfolio. And we will keep our margins stable before the U.S. takes benefits in spite of investing into the launches of these new products, and in spite of continued investment into our strong products pipeline. And, of course, the difference is the various productivity initiatives which are on the base. We do benefit from the U.S. tax reform. We have been one of the biggest taxpayers in the United States. So, we benefit from this reform overproportionately, which will help us to grow core earnings in the high-single digits for 2018. And with this, I’d like to hand over to Dan. Thank you. Daniel O'Day: A very exciting year for the Pharmaceuticals division over the past year, both in terms of the strength and the momentum of the launches which I'll go into a bit, but also, of course, the portfolio readouts that we have, really a stellar year in terms of portfolio readouts. Certainly, the fourth quarter is coming through with Tecentriq readouts that were very important as well as others to complement the portfolio. So let's start with the Pharma division sales, digging down a little bit deeper. You can see the strength of the new launches in the United States numbers, where you've got Ocrevus off to a strong start. You have clearly Tecentriq and you have Alecensa predominantly contributing to those and they were coming only at the very end of the year which I'll touch base on. But 10% solid, solid growth in the United States and of course, those launches have yet to occur in Europe. They will start a little bit in the beginning of this year, more towards at the end of this year and into 2019. Japan with a 3% growth overall, very strong, and international at 4% driven largely by our progress in accessing public markets in both Brazil and in China. You can see that we were efficient at dropping that down to the bottom line in an environment where we're significantly investing in new launches and in R&D. On the royalty and other operating income line, we had some effect of some disposal products in that line that caused a significant increase from 2016. But that’s also the line where we’ll start to see some of our royalty income from products like Venclexta coming in the years to come because, as you know, we don’t book sales on that. But that’s a very promising product and will contribute to our P&L. On the cost of sales line, it reflects the fact that we’ve had to invest significantly in our manufacturing network to keep up, if you like, with the demand on our biologics, and yet using efficiency programs to make sure that as we bring those new facilities online, that we're not disproportionately affecting the cost of sales line. And of course, those will get more efficient in the years to come. M&D is largely an effect of both our investment and the launches, as well as some of our support system on our new launches in the United States. R&D is just a very strong portfolio. We've had eight entrants from early-stage in our late-stage portfolio this year, which is as much as the last three years combined. So, a lot of innovation coming through, as well as comprehensive programs in place to expand our current medicine. And G&A is really more one-off. It's the PSI effect from 2016 that wasn't repeated in 2017. If you exclude, that G&A is essentially in line with sales. And there's rounding here, but basically the core operating profit moved in line with sales at a time of significant investment. The good news story is that the new launches are the products that are driving our growth. So if you take Ocrevus, Perjeta, and Tecentriq and Alecensa alone, it’s driving more than 80% of our growth and still in a very early stage of launch and ramp-up. I’ll get into it. The immunology portfolio also doing very well. And you can begin to see, as we'll see this accelerate into next year, the introduction of the biosimilars with MabThera. Digging a little deeper into our oncology portfolio, we've got 3% growth overall on oncology. I'll talk a little bit about HER2 and CD20 on slides coming up. Let me just touch on Avastin. Avastin is a very, very large product, has couple of dynamics going on there, continued growth in the international region opening up to public markets and then some stagnation, slight decline in the United States and Europe for two different reasons. In the United States because of the entry of the immunotherapies into the front line lung cancer setting, we've seen some erosion there. However, it stabilized a bit in quarter four, I should say. And then in Europe, we had a particular situation with reimbursement in France that basically explains the minus 1% to minus 2% with Avastin. Very encouraged by the synergistic effect and I'll get back to this of Avastin with Tecentriq, and certainly Avastin will continue to be an important medicine in combination as we continue to explore our portfolio. Tecentriq basically well penetrated and continuing to grow particularly in the bladder segment as we go into launching it Europe at this stage. Alecensa is just a true success story at 100% very, very high market share in second line and now penetrating the first line setting in a very significant way for a medicine that's highly differentiated from the competition and positive lung cancer. HER2 franchise. I mean, obviously you see the figures here on how we're growing with Perjeta at 22% accelerating as we went into the end of the year. Kadcyla is still growing at 12%. But I think the real story here is APHINITY and the earlier than expected approval and priority review in the United States with a label that we think really represents the data well. And there's a couple of things about the label I would just highlight. First of all, in the label, it's suggested for high-risk patients. It’s left at the discretion of the physician in terms of the definition of a high-risk patient. So we know it's certainly hormone receptor status and nodal status, but it can be other things clinically and meaningfully clinicians feel are high-risk patients. So that's important. As you know, 75% to 80% of Herceptin’s use today in the adjuvant setting is in the high-risk patients. So this really is a very consistent label of what we've seen with Herceptin. The other piece of the puzzle that I think that is really impressive with the label is it specifically recommends a full course of Herceptin therapy in the neoadjuvant setting or the adjuvant setting. And what does that mean? I mean, it means that a full year or 18 cycles is recommended in the neoadjuvant setting. And even though we had neoadjuvant, of course, in our label for some time now in the United States, we have quite a high penetration of around 90% in the neoadjuvant setting, a minority of patients actually complete the full cycle. So there's opportunity here in both obviously the expansion of the adjuvant setting, but also additional expansion in the neoadjuvant setting and it bodes well for, I think, our upcoming regulatory decisions in Europe which we expect in around the middle of this year and in other countries around the world. So clearly, advances in the curative setting are well regarded by the regulators, and that's been represented in the package insert, and we're off to a good start, I would say, already. The CD20 franchise is clearly entering into a period of transition. I’m talking on this slide around Gazyva, but later on, I talk about the complete portfolio we have within 10 medicines in hematology today. But just to characterize a couple of the dynamics here, as expected, by the way, it’s very much our expectations where the erosion in biosimilars in Europe increased in the fourth quarter to around 25%. That is - around 5% of that is due to normal price pressures that have nothing to do with biosimilars and around 20% of that decline is specifically due to biosimilars with both have price and volume mix. We will expect this to increase in terms of exposure as we go into the first and second quarter of next year because we still have about half the markets in Europe that have launched biosimilar. Others will launch. So we'll see an acceleration of this exposure, all built into our guidance, all expected in the first half of the year. And of course, the second half of the year then will have the base effect from this year. So the declines will obviously be less so in the second half of this year. But I would say what we've seen so far is very much in line with our expectations and the confidence that we have to be able to grow through this erosion. GAZYVA is really doing well. As you remember, Gazyva in the first-line indolent setting was first approved in Europe before the United States. Of course, it’s awaiting reimbursement in many European markets, but where it is launched, and in particular in Germany, very strong uptake of Gazyva in the first-line indolent setting. And, of course, the fact that the relevant study has also read out negative puts us in a good position to change the standard of care in the first-line indolent with Gazyva. It's just starting out in the United States. Good early signals. We received the approval in November, so we're just starting to get going with Gazyva, and I'll come back to the entire hematology story. But that gives you a sense of where we see this CD20 franchise. Immunology continues to be a really strong franchise for us now, an CHF 8 billion franchise growing in the double-digit range. More than CHF 640 million of our growth last year was coming from the immunology franchise. And you can see Esbriet 17%. We now have the lower pill burden product launch product launch and with a good uptake. Continuing to find it slower than we expected to get into the mild and moderate segment, but we are making progress. And over time, I think with our activities with the IPF community and with patients, we'll continue to make progress there. Again, I expect a good double-digit growth for Esbriet as we go into next year, and we continue to maintain our market leadership in that area. Xolair, plus 15%. The pediatric, the asthma claim is driving that growth. And if anything, I think some of the new entrants that have increased, if you like, the prescribing attitudes of physicians in this segment of asthma. Actemra, the number one product in monotherapy and in subcutaneous, continues to drive its growth. We've also launched the new indication of GCA, which is getting some good uptake and ongoing. And we expect this whole area of immunology to be a good growth driver for us next year as well. Without a doubt, unequivocally one of the best launches ever in the multiple sclerosis field I think is indicative of the clinical product profile of this medicine, but also our ability to make sure that we had a smooth launch and a good connection with patients and physicians. We have now more than 30,000 patients that have been treated with Ocrevus. More than a 5% market share after only three quarters last year, annualized rate of more than $1 billion in turnover and growing significantly with strong demand. You see the RMS/PPMS split out of around 60/40. As you know, over time, I expect that to increase on the RMS side because of the number of patients there. But right now, it's still around 60/40. And the good news is that we're getting usage across the entire spectrum of the disease. Around 30% of patients are naïve to therapy or have not been on therapy for more than a year. 70% are switches. Those switches are coming from early-, middle-, and late-stage patients. And the other feature that might be important to you is that 70% of prescribing physicians in the United States have prescribed Ocrevus at least once. We're also getting good return for the second doses. That's traveled a bit the year – at the end of last year with the launch only in the second quarter, but we're continuing to see very good demand for this and good commitment to this medicine as we move forward. So this will continue to grow in the United States this year. Happy to say we've launched it in Germany with the approval in Europe, and we'll be rolling it out in other European countries. Many, of course, have to wait towards the end of this year for reimbursement. But we can expect that uptake in 2018 in Europe and heading into 2019 as well which will be beneficial to our ongoing revenue expectations. So I hope you agree with me. It's been a really strong year both for the underlying products and for the launch products. But it's been an equally strong year for our innovation and for our product readouts. I think this summary, and it's just a snapshot of some of the more important readouts, certainly, not a comprehensive readout for us, demonstrates the type of progress we're making. And again, I think the breakthrough therapy designation now at 2019 gives you a sense that these are not incremental advances. These are significant advances in each one of these areas and certainly providing new standard of care in multiple sclerosis and hemophilia are to be complemented by the ongoing work we're also doing in cancer care. So, very, very strong year on the portfolio side. I wanted to try to articulate what's behind our confidence to be able to grow through the biosimilar erosion curve which I meant again just to put this into context happens over the next several years because you have patent for Herceptin, MabThera and Avastin going off at different times and in different regions. So the question is why we’re so confident we can outgrow this. This is just one piece of data and it just takes a look at the products that have been launched in the past several years. And if you annualize the fourth quarter revenue for these products alone - and by the way, Venclexta is removed from this because we don't book sales - these products contribute 17% of our sales in the fourth quarter, growing significantly. But if I just took a snapshot right at that point in time and annualized it, this is greater than a CHF 7 billion contribution. So you can do the math in terms of what you expect to lose - what we expect to lose from our biosimilars over that period of time, apply this and then of course add the new line extension. The fact that we will be launching these medicines in multiple regions [indiscernible] isn’t in here yet and many, many of the other readouts in cancer immunotherapy and others that we expect to read out this year are also not included. So this is why I say with confidence that the innovation will drive us through this period of time which is quite an extraordinary statement. No other company has been able to go through such a significant patent erosion and have the innovation to be able to stand up here and say that we will grow through it. The updates on the late-stage oncology pipeline, this is just the Phase 3 program so obviously it's a much more comprehensive program. 26 active Phase 3 trials right now, 16 with Tecentriq. And six of those Tecentriq trials we’ll read out over the course of the next three quarters. We've already had two readouts at the end of last year, as you know, with renal and with lung with the Avastin combination. But many, many more readouts to come and indicative of really our second wave strategy with immunotherapy, which was to say we want to lead an important indication with combination therapy with TECENTRIQ and that’s exactly what we intend to do over the coming months. Of course, that’s a whole another phase. I don't have time to go in here today with things like the bispecific antibodies with our personalized cancer vaccine which is yet the next wave after the combination therapies that we’ll be advancing in our R&D portfolio over the coming year. I thought I would just tease out a little bit the lung cancer space because obviously it's a very important one and it's an important one for us and for the rest of the field. We're often speaking about this. Let me first make sure I characterize where the green is on this graph because those are areas where we will likely be first with data, a very strong comprehensive large Phase 3program and that's in small cell lung cancer and squamous cell lung cancer. Those, again, we’ll be reading out over the next several quarters. And then I want to spend a little bit of time on the rest of the circle which is where most of the attention has been focused actually over the past several months in terms of data readouts. It is important to note that EGFR and ALK mutations account for more than 10% of the entirety of the first line lung cancer circle. When I come to the IMpower150 trial results, that will also be important. But then if you look at the pure non-squamous, I've also split it out here by PDL1 status, it's approximate PDL1 status. You have the high, you have the medium, and you have the low. What's important to note in our very comprehensive program, we have now three trials focused on the non-squamous setting. Success with the IMpower150, the OS data to read out shortly, but we also have IMpower130 and IMpower132 coming across as well. So we have a very complete program to read out in this area. And I believe this, in addition to whatever readout the competition has, will dimensionalize the first-line lung cancer space in 2018. I'm convinced there will be multiple options for different patient types in this setting which accounts for around 45% of the lung cancer setting. I'm also convinced that we’ll play a very important role there. And my confidence is based upon, at least, the early readout to be seen so far with IMpower150. I remind you that this is the interim readout. We had powered this trial to show significant differences in subgroups, and we used our alpha on that knowing that we would have potentially OS reading out a little bit later, which it will, but it will come shortly this year. But look at these results already in the PFS subgroup. I mean, all you have to do is look at that line of hazard ratio 1 and see that no confidence interval in any subgroup crosses the line 1. And I think this is really important when we start to dimensionalize the space starting with the eGFR and ALK mutations. We know these are very challenging patients. These are patients that have progressed on targeted therapy, and you can see hazard ratios of down to 0.6 there. You see the TF factor high and lows also being effective. You see any combination of our PDL-1 status also being successful and not represented here in those patients with liver metastases that also showed a significant effect. So I think this is a powerful data source that will only be complemented by future readouts and one I think that will be very important as an armamentarium for physicians and patients in the first-line lung cancer setting. I will remind that today, in that non-small cell lung cancer setting, we have around 35, on a global basis, it’s a little bit more in the U.S. around 35% share of Avastin use in first-line lung cancer today. You know the Avastin data in first-line lung cancer because it has been some of the strongest data. To suggest that Tecentriq on top of Avastin can also be strong allows particularly those users that are comfortable with Avastin to add an immunotherapy. And potentially, depending on the competitive data, we'll see it could have people move more towards the Avastin armamentarium for certain patient types in the first-line lung cancer setting. Just a quick and only a late-stage view to the hematology program, I remind you that we have 10 medicines that are now been studied between early and late stage in our hematology portfolio. This is just those that are in Phase 3, but it represents how we intend to compete, if you like, in hematology in the future. So, the light green line is where Rituxan and MabThera place today. It’s around half of the hematology indications. You can see in the dark green where Gazyva has demonstrated an improvement upon Rituxan as a standard of care. And then, you also see that medicines like polatuzumab and Venclexta come in that space particularly in the aggressive setting where we were not successful in showing Gazyva as a standalone to be successful. You'll see polatuzumab is actually paving the way now for a new therapeutic regimen within the aggressive setting as well. But we're also getting into areas outside of what we have accustomed, being accustomed to including multiple myeloma and AML, and you see the number of breakthrough therapy designations that we have here. Just transitioning to the strength of the data that was presented at ASH, polatuzumab in the relapse refractory in DLBCL setting, no medicine has shown a survival benefit here before and it's hard to treat patients. This is also the same area where the CAR-Ts have shown, a similar benefit to what we're showing here today, has a ratio of around 0.35%, complete response rates of around 40%, really, really strong data here. And obviously, we'll be discussing these results with regulatory authorities to try to expedite offering this medicine to patients. Equally, if we look at the difficult-to-treat, relapsed refractory CLL setting, you see a hazard ratio of 0.17, incredible separation of the curves here, and this in a regimen that allows you to treat and then allow patients to go off of therapy and still maintain their benefit. So, extremely strong data. Our colleagues at AbbVie have also indicated that we will be approaching regulators with Venclexta information to try to get the approval for this additional indication already this year. The Hemlibra program has all readout positive. You've seen some of the data, of course, in the inhibitor setting. I'll show you some of the extension data. We will be presenting the non-inhibitor data in the first half of this year, and also the additional information on the ability to dose this medicine less frequently up to once a month has also read out positive. This is a medicine that is transforming the field of hemophilia for patients in terms of the reduction of bleeds, efficacy, in terms of the quality of life measures, and finally in terms of the convenience and the subcutaneous administration on an infrequent basis. Just to remind you of the data at ASH because the data gets better the longer we look at it. In the adult population, which is the left-hand side here on HAVEN 1, this is an additional 10 months of data. And the reduction in treated bleeds went from around 78% in the first cut; around 88% now. And on the right-hand side, you see the pediatric data which has been supported by an additional 40 patients, plus an additional six months. And again, the number of patients with zero bleeds is 95%. A couple of early-stage assets, just to raise your attention to, and this one, in particular, because there’s a IR event coming up at a meeting in Miami on February 13. It's actually a teleconference. But we will show you the data from our bi-specific antibodies tested now in a Phase 2 in DME. In particular in DME, the ANG2 signal combined with VEGF scientifically provides potentially a further reduction in inflammation within the eye and we'll be showing you those results coming up here soon, a promising potential addition to the field of ophthalmology. I would mention that we also have the port delivery study that would read out in the second half of this year which could provide dosing frequencies much less frequent than they are today up to two every six months and that trial is ongoing as we speak. Another early one that just got breakthrough therapy destination - so this is our 19th breakthrough therapy destination for Roche and Genentech - is in an area of high, high unmet medical need autism. The study itself did not meet the primary endpoint, but the signals were very positive in the adult population. We haven't seen the pediatric yet, but some of the metrics associated with autism was enough for the FDA to say, let's investigate this further, and we certainly will take this now into later stage trials as well. Of course, these are just two of many that we have coming into our portfolio, but because of the BTD and because of the IR event I wanted to highlight those two. And then finally I think quite characteristic of our M&A strategy is to look for bolt-on acquisitions where we think we could add significant value. And Ignyta and entrectinib in particular is one of those medicines. It's a medicine that has pan-tumor activity, fairly rare mutations, but I remind you that 50% of cancers are rare mutations, similar to Alecensa which is a very strong product for us. We think that ROS1 data that showed a 78% overall response rate and the NTRK and also the CNS penetration with medicine has the potential cohort to be best in class. And, of course, we'll be combining our portfolio of medicines and our approach to comprehensive genomic profiling with our colleagues, Roche Diagnostics and Foundation Medicine to continue to evolve the field of cancer care to make sure that every patient with a diagnosis has medicine with significant clinical benefits. So we're quite excited about moving this ahead. This has very good therapy designation. It has prime designation. And we're also pursuing a Fast Track in Japan at present. So all in all, a lot to do in 2018 continued progress on the launches. We have a lot of regulatory events, as you can see here, and very important readouts to report here as we go throughout 2018. But I remain very optimistic about our overall approach in 2018. And with that, I'll turn it over to Roland for Diagnostics.

Thank you, Dan. Pleasure to be here and to talk about Diagnostics. It's been a good year for Diagnostics as well, in particular on the commercial side and on the portfolio side, where you see here sales numbers 5% overall and for the first time, more than CHF 12 billion in total sales. Some challenges, Diabetes Care was a minus 4%. We continue to see pressure here from reimbursement and from the pricing side. And then if you look at the totality of the Clinical Diagnostics business, so Centralized plus the Molecular side plus Tissue, so everything that happens in the clinic, 7% growth combined, certainly above the market growth again. And the largest business was Centralized and Point of Care really doing very, very well. And this is also actually shown here in a really good geographic distribution where we look at where sales are actually being achieved and where the growth comes from. As you would expect, high growth in emerging markets, so our large seven emerging markets countries growing at a combined 17%, Asia Pacific, the Middle East, Latin America all doing very, very well. And in particular China with a growth exceeding actually 21% has been growing very well for the last 10 years. When we then look at the more mature markets, Japan growing 2%; Europe growing at 2%; North America also growing at 2% actually, excluding the sales performance of diabetes. So overall, good growth, good distribution. And let me take you through a couple of the growth drivers here by segment. So in Centralized, it's all the story about integrated approaches towards more area, the combination of immunodiagnostics with clinical chemistry sales, so immunodiagnostics growing 13%, much above the market. I spent some time on Diabetes, minus 4% here. We have seen a spillover of the CMS reimbursement cuts from 2016 in the U.S. now into the private market into - in the United States. This is in large part a response for the sales performance. On the Molecular side, virology with a flat growth. You may remember that we have seen the advent of the direct antiviral analogs, which actually has driven a lot of HPV testing towards our markets in 2016. So we expected this to be rebasing here in 2017. We continue to see good growth in HPV testing and in particular in HIV testing, so virology in good shape. HPV doing very well, we continue to win accounts. We continue to win national screening tenders such as in the Netherlands and lately in Australia. And then on Tissue Diagnostics, it's actually an overall really good performance between advanced staining, primary staining, where we have new products in the market and also very strong performance on the companion side. Let me switch to the financials, and you see the sales line growing at 5%. And let me take you through the different line items here. Royalties with some tailwinds, some of these are one-off effects that we benefited from 2017, so a nice increase there. And then in particular, the cost of sales line, which is nominally of course the largest in our P&L, substantial investment in the pipeline and in placing instruments, so in future growth. Also some one-off investments and higher cash sales in some of new segments and new markets are really contributing to this higher growth here with 7%. And then, of course, some of the margin pressure from Diabetes Care. M&D pretty much in line with sales, continuing to support the sales growth. And then R&D, which has seen an uptick in the second half last year but was a relatively moderate growth here. And then G&A, I think it has been mentioned in the past, this is largely the PSI, the past service income impact, which is actually a rebasing from 2016. So with that 0% on the bottom line. But as I mentioned, with our investments and with some of the higher cash sales and mix shift from 2017, I do expect the margin to increase in 2018. How are we going to achieve this and what's the strategy behind? And here's a slide that depicts what laboratories today look at. They want to have a high degree of efficiency, automation, high quality and high standardization and this is what we deliver to them, starting with high-volume and high-quality instruments, then the ability to connect different disciplines in laboratories to really support the efficiency in the laboratories. We’re the only one who can combine serology or immunology with new capacity testing or molecular testing, if you so want. Of course, this needs to be fueled by the ever-increasing menu and our largest menu in the industry to really leverage the installed base. And then other areas, which are really critical for laboratories to operate efficiently, are laboratory IT solutions, how instruments are connected with one another, how results are being reported. And then finally, opportunities which I'd like to talk a little bit more about in the space of digital, more opportunities to add value with the data that we actually generate in and around laboratories. The most important launch that we had was actually from 2016 that you can see here a really nice leverage and coming through 900 instruments that have been installed all from the e 801 which is the large volume immunodiagnostics instrument. So really, what is driving our immuno sales of 13% is also very much supported by this new module. And we’re very, very optimistic about this uptake and that's why we'll also continue to really invest in this uptake in the launches and in installed. We haven't launched in China which is the largest market for us in immunology. So we're confident that will continue to grow there as well. And then in connecting instruments, this is the first time in Roche Diagnostics that we have our own coagulation instrument using a lot of the technology and the knowledge that we have also from the core team work area expertise. And in this case, the first cassette-based laboratory coagulation analyzers, lots of different features, but again looking at leveraging our position in the laboratories and in the hospital. Touch just briefly on an acquisition that we made which is a company called Viewics from the U.S. West Coast which actually provides data-driven business analytics, so not clinical or medical analytics, but business analytics, again with the same intent to helping laboratories to really retain their efficiencies and their productivity in the laboratory. And then finally the next step and we're looking at enhancing the data that we generate and helping data management to our customers is a collaboration, strategic collaboration with GE, with the goal, actually, to really advance analytics and provide integrated clinical decision support solutions. And that is really taking all the data that we generate, but integrating this data in a software solution that then provides the right data to physicians to take the right decisions at the right time for patients, and we see the complexity of data that’s being generated in hospitals. And with this, we look at two leaders in their distinct segments with Roche Diagnostics in vitro, with GE in in-vivo diagnostics, and combining the data streams and the strengths there to provide clinical decision support, and also then enhancing that, of course, with other applications and other data that’s being generated in hospitals around patients. So, all together, it's been a really successful year on the portfolio side as well. We've launched all the key products that we had planned in 2017. I mentioned the year 1, a couple of highlights also around liquid biopsy. The fact that in this very, very interesting area of cell-free DNA analysis, we have now [indiscernible] solutions, standardized solutions for customers for therapy decisions but also for surveillance, all the way to large gene panels with 197 genes. And then for 2018, for this year, we look to continuously expand on this portfolio, of course, on the instrument side as well as on the assay side. I won't take you through all of these but maybe just to mention the one that we actually have already launched, which is the so-called Plasma Separation Card, which is really designed and defined for low infrastructure for areas where testing and access to diagnostics is really a problem. With this Plasma Separation Card, with a drop of blood, you can store this plasma for up to 20 days. No refrigerator, no electricity needed to actually bring that plasma to a highly-sophisticated testing area. And this, of course, very important in Sub-Saharan Africa and other deprived markets for HIV testing. We’re working closely with WHO and means to establish those standards. And aside from the contribution that we make actually, I think this is also very good business because this is certainly some underfunded business opportunity that we're looking to capture in the future as well. And with that, I thank you for your attention, and I’ll hand over to Alan for the financials.

Good. Welcome. It's great to be here. Let me make a couple of comments about the financials. But before I do so, let me also express what Severin has said at the beginning that we're really confident about the future of this company. And why do I say this? Certainly, we see the biosimilars coming. I think that's one element. But the other element is really the new launches that we have in place, and I think really Dan framed that well. That was the CHF 7 billion annualized addition of sales that they're having, and I think he has not incorporated any growth of this product. I think you've seen what Roland has said about the dynamics in the Diagnostics business which are just tremendous, and we will get the margin up, as he said, in 2018. And then, I would say there's another element coming to it which is the financial performance. I'm happy to lead you through this because I think we come here with a set of strong financials in 2017, and we have set the stage well for 2018. Couple of highlights. I think, well, Severin talked about the sales growth of 5%; core operating profit of 3%, I’ll come to that; core EPS up 5%. The dividend in Swiss francs further increased. And let me mention that it’s the 31st dividend increase in a row and the guidance even said that the 32nd is coming. So, okay, just to put the track record here into perspective. And then, another point which perhaps is a little bit overlooked this year is the operating free cash flow that we have generated which increased by CHF 3.7 billion. I think that number goes to U.K. You really look at this cash generation that we have had in 2017. I think it's quite outstanding and allowed us to bring that debt down by CHF 6.3 billion which, I think, is a significant contribution here. The core net financial results help to bring the core EPS up, no doubt about that. Improved by 25%; I will come to that. And then the IFRS net income went down by 9%, very much triggered by impairment basically from both divisions, so we had impairments in pharma as well as in diagnostics. Good. Let’s a get a little bit more into the details. Sales, we have talked about. You see the operating profit there. It’s a little bit slow in growth than the sales. And this is really just preparing well for 2018. So we really invested into the manufacturing piece that comes with that. We invested into M&A. So I think really here, setting the stage for 2018. You see the core net income then increasing to by 6%. So that’s the financial result. That’s the financial results which contributed to a good dynamic of the core net income. We get to the core EPS of 5% and here I can say just for the sake of detail, this is about the non-controlling interest that we're having in here. And the point is sure, I had a fantastic result in 2017, and we had to, if you like, cut out that portion that we don't own from July, that brought the dynamic a little bit down compared to the core net income. IFRS net income, I’ve touched upon. I have a slide on that. And then you look at this operating free cash flow that we've provided with the CHF 3.7 billion up, 26% better. And then look at the free cash flow at CHF 13.4, billion 47% up. Definitely outstanding. When you look at the P&L, sales, everything said already. Royalty is another operating income. Yes, we had higher income from divestment. So the divestment gains increased compared to last year by CHF 230 million. But at the same time as you know, we had that impact from last year with the past service income where we did a restructuring on our pension fund in Switzerland, and we had a CHF 426 million positive. It's not exactly offsetting here. But we would say, okay, this post effect and very much driven I have to say to Pharma. So if you take that out with the PSI effect in Diagnostics in my opinion justified and then Diagnostics profits would have grown by 4%. But I think really here on the total P&L, I think that's basically, very basically awash. Cost of sales. Really here, I think we've done a lot on our manufacturing network, yes and then Dan told you said already. But major point is we have now quite some capacity when it comes to antibodies, and we’re growing into this. Now, I think that's a major point here which will also help us in the future. And as Roland said, I think the SP investment that we have done, especially on the China side, that will pay off as well. M&D, well the new launch was very clear. We've done the right thing here for our new products. Then R&D is clear, when you invested into the pipeline. Then you'd see G&A. As I've said, here’s the past service income effect. If you cut it out, you get to 10%, which is still a significant number. Administration has grown by roughly 8% when you look into the details. But here, very much about legal costs that we had to bring in. And certainly, as we defend ourselves well against biosimilars, that's one element here, but also on the other elements with the new products which drove legal costs to a certain extent brings us to operating profit growth of 3%, which are qualified already. And, here, you see it in a little bit more detail, I think on the margin side, I would say give us credit, I think rather stable on the margin side at least for Group and Pharma. I admit on the Diagnostics division going a little bit down, yeah, core operating profit basically stable. But certainly here, the Diabetes Care business comes in, and that has a certain drag on the margin. The core net financial result has improved quite significantly, thanks to the debt restructuring that we've done in the past years. And you'd see really the impact in total roughly CHF 250 million, which is quite a boost on the core EPS. You see interest costs came down by CHF 100 million. We did a little bit less or had less impact from debt redemptions compared to last year. And then, you'd see the interest cost on pension plans and smaller things, which helped us really to bring that number even further down compared to the last year. And when I said about the last years, you see really here the dynamic that we have had. It's not just all lower interest costs, I think from 2015 to 2016 roughly CHF 200 million less interest, and from 2016 to 2017, as shown before, roughly CHF 100 million less interest. But you see the effective interest rates here below, from 3.7% down to 2.8% in 2017, I can even mention that in 2012 the interest costs have been about 5%. So I think really the debt restructuring in all these years paid off. Yeah. And that's certainly a big topic for us and it's a positive topic. First of all, let me let me start like-for-like, and I know and remember all the concerns about our tax rate, the discussions that we have had in this room here. In 2016, 27.2% or the group core tax rate went down slightly to 26.6%, basically driven by the fact that we have shifted inventories a little bit around in our network. So I would say it's rather stable. But it's rather stable here which I think is a good message. But certainly based on the U.S. tax reform, there’s an opportunity for us and it might be even more significant than for other companies. And we've always said it because, well, we are significant in the U.S. and we have a huge operation over there as we all know, and we generate quite some profit over in the U.S. So what we are expecting now is based on the 2017 numbers, that really the group core tax rate in 2018 should come down to the low-20s. That's what we expect. And please forgive us that this is not really so specific because, well, it's about product mix, it's about - it’s really about profit mix, and we will see how these things turn out. But based on 2017, I think, we feel confident that this will be the outcome and certainly this is a significant benefit for the company moving forward. As you know, you know what I said last year, I think the group core tax rate is a bit artificial. I think the major point to look at is the effective tax rate. And when you look at the effective tax rate, it increased from 25.2% last year to 27.9% which is not surprising because the impairments that we have taken, in fact, we've taken in the U.S., which brought then the profits down and then brings the effective tax rate up. So really from a cash-out point of view, no change here but, certainly, a little bit of a swing in the effective tax rate. Okay. Non-core items. As you see here, it starts with the core operating profit, up 3%. I've talked about that, and I've talked about the IFRS net income come down by 9%. And what you're seeing is here really all the line items, I would say, basically stable. One is a little bit of an exception, and that's the impairment of intangible assets, and I've talked about this. Basically both divisions drove that with some impairments that we took. And let me also mention the global restructuring plan. Here, the restructuring charges are basically on the same level as last year with CHF 1.2 billion, which I think underlying that we are really adjusting, our business really moving forward. I think that's not a small number, but I think really proactivity is on our minds, and we contribute to that. Good cash. Yeah, went up to CHF 17.8 billion. You see the cash rate here. Pharma with a huge chunk, but also look at Diagnostics. They improved quite significantly. So it's really, really across the board. A significant improvement. And just see really where it came from. I think, on one hand, it's the underlying business doing much better. The other piece is net working capital. Net working capital contributed with CHF 1.5 billion roughly to that improvement. And I think it shows it makes sense to work on the receivables, the inventories and the payables, and that paid off. A little bit less investment in PP&E. We had CHF 300 million last year, we made a large investment, so I would say, we’re basically stable and then investment intangible assets came a little bit down and that all contributed to that outstanding cash flow development. So what does that lead to? It led to a reduction in net debt from CHF 13.2 billion to CHF 7 billion. And you see really here the ingredients to see also what we pay in Texas normally. In the P&L, you'll find a much higher number for that which is clear because we have deferred taxes and all these things, deferred liabilities. But I think the CHF 3.9 billion that we brought the cash out. You see the dividend for the year 2016. And then you see the CHF 7 billion net debt. What does that lead to? It leads to a 9% ratio net debt to total assets as we can take from that slide. And you know that our target is to be between 4% and 15%. So we're now in the range. We made a huge step from 17% to 9%. I agree with that. 17% is the full year figure here. These are the half year figures baked in. But at the same time, I would like to remind everybody, we still have CHF 19 billion of gross debt on the balance sheet. It's not like that we are getting to zero position soon when it comes to cross that, we still have significant debt that we can repay. Quick look at the balance sheet and when you look at the balance sheet I think cash and marketable securities went up by CHF 3 billion, that’s a good cash generation that we have and it shows. It shows really in our funds, which is good, our liquid funds. Other current assets, basically what you can say, receivable went up a little bit. Inventories went down. So stable. Non-current asset is a really depreciation impairments. And when you look really at the liabilities, current liabilities and non-current liabilities went down because of debt repayment. So short term debt went down and long-term debt went down. And then you see what it results in equities of CHF 29 billion, 38% equity ratio, net debt on total assets of 9% Good outlook, currency. Yeah, let me make a quick comment on currencies here. 2017, I said this morning to the media, it looks a little bit boring as the table on the right-hand side. I have to say I'm pretty thankful for that. I think we had more lively pictures in the past. I agree with that. So, yeah, be careful what you ask for. So I think really we had a calm year when it came to currencies. Basically what happened is now the Swiss franc weakened against the euro and strengthened against the yen and remained stable when it came to the U.S. dollar. So that's basically the picture that we have seen in 2017, and it really balanced out for sales, core operating profit and core EPS. When we look for 2018, and you know what our underlying assumption all the time is we say, all the currencies at the end of the year remain stable, and then we make a projection, if you like. And then we think the impact could be up to plus one percentage point coming from currencies on sales, core operating profit and core EPS. But you all know how unlikely this is. So when I looked at today's picture, this plus 1% would change to a minus 1%, and that's what I can say roughly, very roughly. The major fluctuation came from the U.S. dollar as you know. So, okay, we will see. We’ll keep you updated in the course of the year what the currency impact will be. But so far, I would say still good. And Severin talked about the guidance already. And let me lead you very quickly through it. I think on one hand, we say group sales growth stable to low single-digit. And I said I think we intend to grow. It's very clear. We intend to grow. Core EPS growth, broadly in line with sales, excluding the tax reform benefit, and we really thought it might make sense to give you really an idea about the underlying performance, excluding the tax effects. And therefore, we said both in line with sales growth and as I said, for sales we intend to grow. Core EPS growth high single digits including the U.S. tax reform benefit, as we've said, we expect the group core tax rate to go to the low 20s. And based on that, we certainly expect that there will be quite some benefits when it comes to core EPS growth. And then, the dividends, I don't know any other company which announced the 32nd dividend increase in a row when the year starts already. But to set everybody on the same page because I also remember vividly that we had a discussion here in this room in the past, so I want to be very precise what the basis will be for the outlook 2018. And somebody even wrote in the report, we even make it a little bit tougher for us with that adjustment, which might be true. When you look really at the numbers, I think the CHF 15.34 core EPS 2017, as reported, you will find that on page 144 in the finance report. I don't want to give you now all the specific figures. You'll find it there. And then, what you have to take, you have to add back the currency losses. The currency losses you find in the finance report in note 3 on page 52, but you also find it in Roche’s. It will be available over there. The point that I have to make is it's CHF 115 million. So, take that CHF 115 million and you just really on the number of the weighted average number of shares and onboarding equity security issue. When you do that, you don't exactly get to the 11. This CHF 115 million have to be adjusted by tax, and that's the number you don't find. And I disclosed it now. It's CHF 12 million. So, you take the CHF 115 million, you deduct the CHF 12 million, then you take the rest and divide it, and it brings you to 11. That's what you'll have to adjust to get to the basis for the core EPS for your projection on 2018. And with that, we are looking forward to your questions. Thanks for your attention. A - Severin Schwan: Very good. I hope you all understood the last slide. What I got is even tougher for next year. Good. Shall we start here and go to the left, please?

Thank you. Luisa Hector from Exane. A couple of questions. Could you comment on the range of your sales growth guidance, please? Particularly interested to hear what it would take to actually be at the lower end of that range, the stable growth of sales, because it's over about - it looks like it’s about CHF 1.5 billion of sales, which is quite a reasonable size. So, what would it take to hit the lower end of that range? And then, secondly, just on the restructuring charges, because we can see in the annual reports that you did take some charges preparing for the biosimilar erosion and perhaps some field force reductions. So, just any color you can give around that and possibly what the benefits might be in 2018 and beyond. Thank you.

So, first, on the guidance, what does range of stable to low-single digits mean? Really, the stable is we put in a floor here, but read it in a way that we intend to grow. It's a more cautious growth than in 2017. We see an acceleration of the biosimilar effect, but clearly the intent is to grow. As the second question is concerned, yes, it also concerns the commercial organizations, but it really goes throughout the various functions. One of the major items we have been working on already in the past is the restructuring of our manufacturing network. We had over capacities from the small molecules. We had to invest for large molecules. As a consequence, we had to divest certain sites, we had to close sites but we were building up in other areas. And on the portfolio side as we have this mix entering into new areas like MS or hemophilia, we also have to adjust the sales force accordingly. And with that the kind of adaptations, we continuously I should say undertake to adapt to a changing environment and to adapt to our portfolio.

Jo Walton from Credit Suisse. I'm afraid I'm going to push you further on restructuring. So you've taken over CHF 1 billion a year for three years. Can you help us on the sort of returns that you get on that? How much of that is cash? And are you close to the end of CHF 1 billion a year of restructuring, because they come down materially over the next few years and what might the cash flow elements could that be? And could you also talk not so much about the manufacturing, but how we should be thinking of your SG&A going forward? So in the Pharma business, that's growing at about 5% or 6% for the last couple of years. You're entering a period where your sales growth is going to be less than that. When you launch everything that you want to do on sort of stable SG&A or could we actually see that ratio come down? Thank you. Daniel O'Day: So as far as the restructuring is concerned, what we do is we look individually at each project which we undertake, so take manufacturing. So we literally do an NPV calculation. What would it mean if we take these capacities out, what are the costs associated to taking a factory out of our network. Typically it's difficult to sell our factories because there are overcapacities in the market overall. So, you have costs if you take out a factory. And then you look you know what are the savings over the coming years. And if we believe this is NPV positive, then we would go for it. So typically you have lasting effects when you abate these factors. And when it comes to the commercial organization, again, you know that we need to execute well on the launches. I mean, we have to fund those sufficiently. I mean it would be unforgivable not to fund an Ocrevus launch as we did in the United States or as we do, as we speak in the European Union now as we rolled out Ocrevus beyond the United States. But we don't simply want to increase the M&D costs, so we have to get it somewhere. And whether you take it from, you take it from the mature portfolio and you look at all your infrastructure, your back office functions, et cetera. So it's always very, how should I say it, specific for the respective county, for the respective function, and if it is NPV positive, if it makes sense from a business, from a strategic point of view, we’ll go for it. And the same is true across support functions, for example, where we streamline, where implement new systems, et cetera, and all of these kind of parts eventually end up in the restructuring cost. You also have to see it helps. From an accounting point of view, its industry practice actually that of a certain magnitude, you take it as non-core cost. And it helps the organization to really go for it and not to postpone those things. Psychologically from an internal management point of view, actually it is also very useful because if you always have the deposit aligned then people delay it, they delay it into the next year. They wait for it. If your book is separate - if you take this out of the performance management, then people have a real incentive to do the right thing right away. So that's, I guess, the color I can give to you. Perhaps, Dan and Roland, do you want to add for your respective divisions. Now, G&A, at the end of the day, it’s the same story. I mean we continuously have to work on getting more efficient. Alan, for example, we are looking a lot at procurement now. We are consolidating our indirect procurement function to get savings out of that. And that, of course, will then eventually fall down to bottom line. Now how does it all - come all together? If you look at the guidance, what we have is investments into the product launches. We continue to invest into the product portfolio because we have a rich portfolio, a rich pipeline we want to achieve, and we make up for the difference to keep margins stable by these various productivity measures. I hope this gives you a bit of a feeling of how it all comes together.

How much is cash. Daniel O'Day: Sorry?

When some of your restructuring is writing down assets and some of it is cash flow which we will eventually see in the operating - in the cash flow, so maybe it's a question for Alan, but of the more than… Daniel O'Day: Alan, do you know how much it is …

I would say - well, we can't just say that very roughly, Jo because all these programs are very different. I think the manufacturing certainly is higher. I would say it’s roughly 20%. In other areas, it's lower. So when we go for the sales, we basically have no write-off here. So I'd say the vast majority is cash-effective. So when we have a cash-out all coming from it, I would say roughly 20% non-cash, 80% cash.

You want to add on the various productivity measures. Fine. Good. Let’s go to the next question. Sachin?

Sachin Jain, Bank of America. Three questions, please. Maybe just a high level one on midterm. So you talked very constantly about the outlook, launches annualizing to CHF 7 billion, offsetting biosimilars. It seems you're trying to work back some of the cautiously interpreted comments in the press room a couple of weeks ago. Is that fair or is flat to low single-digit the new normal that we should be expectedly excited about, noting that 2019 is a tough year as well given [indiscernible] U.S. biosimilars? Second question, on the dividend, given you’ve mentioned a few times accepting again this is the second year of growth. But the last couple of years, it’s only growing 1%. Given the tax windfall and how vocal you’ve been on sort of 30% to 40% cash flow growth, just any perspective on that rate to dividend growth, what it reflects on M&A aspiration. And then the final question is for Dan on Rituxan, the 20% biosimliar erosion. I wonder if you could split that out for price and volume. And I wanted to check I understood the volume decline should annualize into the second half of the year, and what gives you confidence that should happen? Thank you.

Let me start. On the tone of my communication, I guess you are referring to the FT article which came out earlier this week or earlier this year, I should say. To tell you a bit of background story of the FT article, it might help you to put it into context. Actually, this interview was done right after Q3. So, I had the interview in - with FT after the Q3. And I was just coming out from one-on-one meetings with you and investors, and everybody in this meeting asked me, what is the margin expansion that you expect in 2018? And then right after these meeting, I was getting into this FT interview. And around this time, this is really the high season for budgeting. So, we were in the midst of our internal process of what kind of productivity measures can we still implement, how do we squeeze out unnecessary costs, et cetera, of the organization to be able to reinvest that into our portfolio where we see so many opportunities to be able to invest and reinvest into the new launches as the more mature products come off. So, think of that. I come from battle, I have these budget discussions, everybody is telling me that they need more money. I aggregated, it doesn't work on a group level, pushed it out. Again we have hot debates. I get into the Q3 Investor Roadshow and kind of - I get this message like how - not do the margin expand in 2018. How much are the margins going to expand in 2018? And then on the same afternoon, I have the FT interview which has been published now at the beginning of this year. I hope this gives you a bit of a flavor. Now, what I said, however, in those interview, I still fully stand behind that. I mean, it is tough in times that you have such a portfolio transition where we have such a rich pipeline which we need to nurture. There is, as I said, pressure in the system to work on the efficiency and to work on the productivity. I think I also said it's a good thing because it keeps us fit, and it's the right thing to invest into the new portfolio and the upcoming opportunities. But I hope this gives you a bit of context. At the same time, I am extremely confident and I'm more confident than ever. Just look what works out in the fourth quarter. I mean, it was a stellar quarter in terms of readouts, in terms of accelerated approvals. Think of APHINITY. When we got an early approval, we had a superb label for Perjeta. We got the Hemlibra approval. We had positive readout for Tecentriq, the 150, the 151, et cetera. So I'm more confident than ever that the pipeline has never been stronger. We have a year with stellar launches, in particular Ocrevus, which continues to grow strongly. And that gives me the confidence that we will grow through this biosimilar transition. I hope that puts it a bit into perspective. Now on the second question, dividend gap, I mean, we'll see how this develops for the years to come. But no doubt, after so many years of as dividend increases, we want to increase it further and we want to keep an attractive dividend policy. And it's a good problem to have. We are now year end at 9% percent debt versus assets. This is now in the middle of this range which we gave you. If the leverage is not ideal anymore, then we'll find ways to return the money to the shareholders. Don't worry. And for Rituxan, I think…

Biosimilars.

Biosimilars. Right. Daniel O'Day: For additional color, thanks for the question. I’m going to give you some confidence I think then quite realistic and quite aggressive around the biosimilar road in financial year built into our guidance and just to MabThera because its one way or another data so far as you know between quarter three we had about a decline of around 16% , quarter 4 around 25% that reflects more countries coming on board and a greater adoption of biosimilars. Of that 25% in quarter 4, as I said, around 5 percentage is normal effects on pricing, 20% is biosimilars. And in that biosimilar network, it’s about a third of price, two-thirds on volume. We expect MabThera to accelerate for the first two quarters of 2018, obviously, bringing on new countries on both, roughly probably the same price-volume mix, we’ll see. But as we head into the second half of next year, then you’ll have the base effect on the launch of MabThera. For Herceptin, we expect the entry of biosimilars already sometime in this first quarter. The dynamics with Herceptin are that we can have up to four biosimilars, possibly five if you include the very end of the year with Herceptin next year. So, the dynamics are slightly different than we have with MabThera with two. On the other hand, I think to counterbalance that a bit, we've got a higher percentage of subcutaneous penetration with Herceptin. With Herceptin, we have - the country has launched 50%, 60%. We have about half that with MabThera. And we've seen, at least in the initial entrants of biosimilar entry, that the subcutaneous is a very sustainable advantage for healthcare systems or for patients, at least within a certain price span, and we're not seeing that price band be violated so far by the competitors at this stage of the launch. At some point in time, I think they will be. So, I think that should give you a little bit of a color for what we expect next year. It's all been factored into – I don’t think we're far off, by the way, your consensus estimates on biosimilars in terms of what we have. And again, we expect the launches to more than offset that on a global basis.

We’ll take the next question, please. Yes. Daniel O'Day: Maybe if we go to the left.

Thank you very much. Jack Scannell, UBS. Two questions, the first around sort of R&D and R&D investment prioritization. In some therapy areas, for example, Alzheimer’s, there’s a huge amount of technical risk, but arguably not a huge amount of commercial risk because so far nothing works. In other therapy areas, particularly oncology now, a lot more stuff seems to be working. But arguably things may be very, very crowded. So, for example, there’s about example, there’s about a thousand immuno-oncology drugs in clinical development, a thousand clinical trials testing various drugs with PD1, PDL-1s. How do your portfolio management process think about that commercial risk, the idea there may be lots of things that technically work and work quite well, but actually the commercials don’t because given the competitive intensity? That's one question. The second question is just, is there anything you can tell us about discussions you've had or thoughts you have about the acceptability to European health systems of adjuvant Perjeta and the sort of budget impact implications?

Let me start on the R&D question and the allocation within R&D. I very much share your view. We have areas where the technical risk is relatively high, even if you are in pivotal trials, if you're in late stage trials and Alzheimer is certainly a good example for that. And I would say this applies in general for CNS where you need bigger trials to really have the proof of concept really. And there are other areas like in oncology where you can very often see early on if you have a good response whether you have a drug or not and it becomes much more of a question, are you faster than the competition and is it better than the competition. So there are different dynamics. I would only say from Roche point of view, the way we look at it is always what is the degree of medical differentiation. That’s the starting point. That’s the key hurdle. We wouldn’t go for me-too opportunities just to kind of bet on we are faster than others. And if we wouldn’t have a really strong hypothesis, we believe in areas like CNS, we wouldn’t go for it either. So it’s really a question of setting the medical and the scientific hurdles high. That is the key element when we transition opportunities through the pipeline. And then depending on what is the competitive situation, you might take different strategies. So, for example, for Alzheimer’s, for gantenerumab, we have put in some gating. We look at interim results to not kind of have too much cost because actually the technical risks materialize. In an area like oncology, with the competition you lay out, it is very much about how do we take out the white space, how can we benefit also from regulatory changes that the FDA is now much more open for accelerated approval? How can we potentially shift from early trial right into pivotal trial to not be second, which is so important in a field like oncology. So the key hurdle to progress is medical differentiation. What difference is it potentially making for the patient? And then asset by asset, we take it from there. Right. There was a second question. Yes, adjuvant growth. Daniel O'Day: I mean, let me just add a bit as we move ahead, but I think the other piece of the puzzle because I'm responsible for managing the late-stage [indiscernible] we obviously look at return on investments, risk ratio projects, and that’s going to be different in different therapeutic areas. We essentially try to look at a balanced portfolio, at the end of the day. I would say very importantly on oncology to what Severin said, I mean, speed matters, and two elements of things that are accelerating speed in oncology are biomarkers and real-world data. And those are significant new applications of some newer applications, if you like, to our clinical development design and our clinical development execution which we are putting to work very, very actively. I mean, as you know, we're really looking at Phase 2s as registration trials there, and anything you can do to accelerate those or to supplement those or replace those if real-world data becomes extremely, extremely important. So a longer discussion. But on adjuvant Perjeta, I mean, I think this will go very much along the lines of how the neoadjuvant indication went, which is, as you know in Europe, as you expand patient populations in those countries, you negotiate a certain discount as you go into those expansions. But I have every reason to believe that the level of evidence that we have in APHINITY, as it was in the United States, will be supported by the regulatory authorities and the payers within a normal course of discussion and negotiation. So we look forward to bringing that medicine to patients in Europe as soon as possible as curative. Remember, it’s curative.

Thanks very much. Richard Vosser from JPMorgan. Just going back to the guidance. You seem to be highlighting many times about the growth. So what do you actually need to see that the growth will be or you would be able to think about low single digits toward the high end of your guidance, what do you need to see in the terms of the launches to get more comfortable with that higher end of the guidance? Second question just on Perjeta. We saw a slight uptick in U.S. sales growth of Perjeta in the fourth quarter, so just early impressions of the adjuvant launch and expectations there. And then thirdly, on Hemlibra, just early feedback around the inhibitor launch, what you're seeing in the U.S. and when can you file for non-inhibitors there, too? Thanks very much.

Very good. Dan, you want to take all three questions? Daniel O'Day: Yes. What's the first one again? Sorry.

What makes you confident to… Daniel O'Day: The sales, right, right, right. Well, I think it's – look, it's a dual dynamics, right? It's the – every day, there's some new news about biosimilars, right? I mean, you all read it. It’s – either they're getting approved by CHMP or they’re not. They're either getting ready to launch or they’re not. So there’s a certain degree of confidence intervals around biosimilars. Again, I'm confident we've taken a good approach in the budget, but that could be slightly more or slightly less, and I think that has an impact on our overall sales budget. I think there less uncertainty around the launch projections because the launch projections are solid and we know where Ocrevus is going. We have a good feeling. I would just put into context that with Hemlibra - and we’ll get to that for a second - it launches in a small percentage of the patient population to begin with. It’s only around 4% or 5% of non-inhibitor. The larger patient population comes later this year. But I would say in general, the launch products we have – the confidence intervals are more narrow. So it’s really around – some of the uncertainties around which biosimilars hit when, what our pricing strategies are. And quarter-to-quarter, if we see revenue move, we’ll give you updates on how we’re doing relative to our guidance. So I think we have a certain range that we’re working as we see the world right now on Perjeta. Yes. We have seen - based upon the APHINITY results prior to review in the fourth quarter, we’ve seen pickup in the United States that’s predominantly adjuvant, more use on the adjuvant setting. I would say it’s not yet reflecting as far as I can tell the extended use in the neoadjuvant setting. I think that will happen more this year. But good initial uptake in dynamics, NCCN guidelines, reimbursement coming through as expected. So that will be a key growth driver for us in the United States next year. And then finally, Hemlibra, yeah, it’s only been about a month. I would say a couple of things with that CHF 3 million which is in reported results. There really isn’t any stocking in there. As you know, we’re still waiting on a final CMS decision. So some of those patients are anxiously awaiting the output of that. Private payers seem to be adopting very well, our pricing strategies, certainly, in the inhibitor setting and demand. I mean, demand looks very good again for a small population. In terms of the filing, as you can imagine, we’re actively putting the files together until now, which will include both Haven 3, which is the non-inhibitor data, and also Haven 4, which will allow us to look at different dosing intervals. We intend to file as soon as we can this year. And our hope is that depending on how the FDA in particular receives the file that we could bring that approval still this year, and we'll give you updates as the year goes on. Does that help?

Yes.

How are we doing with timeline?

Vince Meunier from Morgan Stanley. I have follow-up questions, the first one is on immuno-oncology and Tecentriq in first-line, non-squamous NSCLC. I mean, you said as you time that time to market is more than ever a crucial, very important. So, should we understand that in the segment of first-line non-squamous NSCLC, you consider that KEYTRUDA is here. They are the first and so it will be very challenging for you to grab market share. And this is the reason you believe that the squamous and the NSCLC segments are better for you. So, is it true? Is it a good reading? And then, what is in your view the time lead you will have in these two segments? I have also a question on biosimilars and the adjuvant setting. Perjeta is infused. You say that subcutaneous is a good protection for biosimilars of Herceptin. And so for the combination of Perjeta infused and Herceptin, how can you protect your franchise against the possibility to use Perjeta IV and biosimilar Herceptin IV which is potentially very good for the patient and for the payers? The very last question is on diagnostics, I mean, I understand that there will be operating margin improvement next year, I mean, in 2018, with other drivers. Is it the diabetes deterioration which is no more important because it’s smaller or is it because there is a key driver in the unit which is driving that? Thank you.

Thank you, Vincent. So let’s start with Tecentriq. I want to say no to your question about whether or not I feel that – I think you’ve been, the tone of your question was do we think we’re going to be less competitive in non-squamous. Absolutely not. I’d point out squamous and small cell to begin with because I think we have the first trials that will show out there, and I think there’s significant potential space. But clearly, the non-squamous place is an area we think we can be very competitive. You’re right, time matters. But in this case, time is a matter of months, not a matter of years. So I think all of these readouts in the non-squamous setting, at least, are 150 overall survival or 130 or 132 will happen within months of when we expect to see the data from the competition, both the KEYTRUDA data and also some of the CTLA-4 and PD-1 data. So, in that context, I would argue that time is not a significant advantage and anyway, what makes the bigger difference is of course, the level of efficacy both across the entire population you’re studying. And I think more and more, we’ll be subsetting patient populations in these trials to determine who can be assigned to what trial regimen and get the most effective response, which is why I emphasized a lot with the 150. There's this very consistent efficacy across every single subset. And again, we don't have insight into how every other competitor is powering their trials, where they're using their alpha. But I think at the end of the day, in terms of uptake, in terms of physician preference and patient preference, this will matter. Because I think physicians will have a precise approach at the end of the day in first-line non-squamous lung cancer setting, and I think we'll play very well there with our three trials at least. We have every chance to play there well with the competition. And then quickly on the biosimilar, Herceptin, and I think you know your question really centered around, well, I think the way I read your question is there's still an advantage to Herceptin subcu when you have to get Perjeta IV, and I think the answer is yes. Because these infusions are not given simultaneously, they’re still given sequentially. So you still save significant infusion chair time. Yes, you’re in the adjuvant setting in metastatic session. You also have an IV infusion of chemotherapy and also of Perjeta. But it's still a time saving. It’s still robust and it's helpful. Now what I would say, in the midterm, as you know, as we articulated, we have a whole formulation of Perjeta and Herceptin that we're studying that could be obviously administered simultaneously and this could be a significant benefit but it comes only in a few years’ time. So just put that into context, it’s not something that bridged the initial gap of biosimilars, but it is something in the mid to longer term that we're looking out to increase efficiency and to tie these two products together.

That was the diagnostics question. Hold on. Daniel O'Day: Yes. Let me try to give some color here, and I think your first one, you hinted at diabetes care and you're right. The relative margin contribution is smaller. You're absolutely right. There are some elements that are being analyzed. Some of the pricing reimbursement cuts, so we'll see how that plays out against a certain shift towards continuous glucose monitoring. So I want to be somewhat cautious there. We are investing in the large instrument settings, and these are large programs, right? These 900 instruments that we’ve placed so far, these are multi-month programs where you have to win it and then you have to bring in new engineers to install. Once you're in, you know that you're going to be in an account anywhere from 5 to 7, maybe 8 and 10 years. So that gives us a certain plan ability there. I also mentioned we had a higher rate of cash sales. Typically, we try to go for the reagent rental model, but in some new segments, initially there is a high in cash sales, right, because of the smaller instruments. So this is the other thing. And then finally, of course, just like for the rest of the group, I mean, we're working on operational measures. We're looking at the end-to-end solutions around support functions. That will also, of course, add to the margin.

Before we break out into the sessions, I would take one more question, and then I would ask you to bring your questions into the session. Thank you.

Andrew Baum, one question. So you raised precision medicine as an important driver going forward, and your role as custodians of Roche is to future-proof the oncology strategy for the next decade. So with the focus on precision medicine, how quickly do you think that that is going to transform the commercial landscape with baseline NGS or liquid biopsies? And how does that impact how you think about M&A given your Diagnostics business and the costs or the relative valuations of therapeutic versus diagnostic assets? And then one final point, same theme, to what concern is although the regulatory environment lends itself to precision medicine approaches? The commercial landscape, maybe the good is the enemy of the great, i.e., KEYTRUDA plus chemo. Just it’s so easy for the physician to use that even if there is a subpopulation with a biomarker which is addressed by two small molecules or whatever it happens to be, the commercial application provides friction where it actually prevents the precision medicine approach translating into a monetized bill opportunity?

We could spend a whole dinner, I think, speaking over the joy of that someday. But let me just say that I have appreciation for a variety of aspects of your question. I mean, as I tried to indicate and I really do feel this, I think there's a bit of a chicken-and-egg thing going on here with precision medicine and particularly when you look at rare mutations. Let's park immunotherapy for a second because we'll get back to that. But there's no doubt that there's more and more medicines targeted at rarer mutations. And the dynamics we've always seen at the physician office is, if I'm going to use NGS testing, then I need more actionable mutations, more medicines that allow me to treat those rare mutations if I'm going to do it. And I personally feel that given what's in our pipeline, what's in the competitors’ pipeline and also what's already been approved, that we're very much at that tipping point now. So there's that, there's clinical utility, clinical action ability, if you like, to be able to help these patients with rare mutations that’s at a scale that I haven't seen before. And also aiding this is the acceptance, the adoption, and the reimbursement of NGS. I think there was a very important moment in time here in November last year when Foundation Medicine received a companion diagnostic approval for their panel in solid tumors. And jointly, there was also CMS approval. So, FDA and CMS work together on this filing. And one of the big barriers - I'll just take the United States right now - has been who’s paying for NGS. And as we know when CMS covers NGS, payers often follow. And I think we have a real tipping point in the whole NGS field that Foundation Medicine had led and other companies will benefit from. But I really think that's another dynamic that will occur as well. So if you reduce the barrier to using NGF upfront because of reimbursement and because of reimbursement and because of its acceptance, and by the way the CDx and FMI will continue to grow in terms of companion diagnostic marker. It's a very flexible regulatory approval that allows every time you have a new mutation if could be added to the front page of FMI. Plus a CMS approval for reimbursement. I think this will increase the adoption. The number of medicines will increase the adoption. And you said how quickly, I mean I think it's going to be quick. I think in five years from now we'll see this routinely used in the frontline setting and with the advent of blood-based assays becoming more and more predictable, more and more correlative to the tissue-based assays, that will also expand the population of patients that can be tested that don't have good tumor samples to begin with and they could also be tested routinely throughout the course of their disease to look for mutations before they begin, so you can adjust therapy. So this is not the future anymore. It’s the present and will only increase, I think, in the years to come. In cancer immunotherapy, I think there will always be a debate around. Is it easier to have a therapy that can be an all commerce which was the whole reason by the way that we initiated the stance into the front-line lung cancer setting with chemo combos in Avastin and why we had such a comprehensive program there is because we thought that would certainly be the first wave. And I agree with you, as long as you have good results in chemo, combos or in other combos that go across mutations like we do in IMpower150, I think that's not going to encourage immediate testing in immunotherapies. However, I think over time there's going to be more rational combinations that will come to use and we're going to see a greater division of that. What we're likely to see in the early phase of this is physician accountability, physician use, so just use Avastin, as example. If you have 35% efficiency Avastin as the preferred treatment in the first-line lung cancer study and you’d tell them that on top of that if you Tecentriq, you get a synergistic, significantly additional benefit in PFS and we'll see OS coming out. Then you have physician preferences that will drive this and say, fine, you know I'll add Tecentriq. You have others that may be more keen to a non-Avastin regimen and those patients – those physicians will look for data in the non-Avastin space perhaps and look for those. So there's a lot of dynamics. Some will be biomarker driven, some will be physician preference driven and some will be the personal relationship between physician and patient, their characteristics, how fit are they, how much can they tolerate in terms of a regimen. All these things will come into account if you’re like in a clinical manifestation. But I do believe that we'll get smarter on the biomarker side in the next five years in cancer immunotherapy with RNA therapeutics and others. And we'll get to a more rational approach over time.

Thank you very much.