Illumina, Inc. (ILMN) Earnings Call Transcript & Summary

Hello, everyone. Thanks for joining us on day 3 of our Healthcare Conference. I'm Tejas Savant, and I cover the Life Science Tools and Diagnostics sector at Morgan Stanley. I'm delighted to have Illumina join us this afternoon, and representing the company is Francis deSouza, CEO, so welcome Francis. Before we begin, just some important disclosures. Please see the research disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your sales rep. And I believe, Francis, you have a safe harbor to read as well.

Yes. So first, thank you so much for having me. I really appreciate the opportunity to be here. Let's start with the safe harbor statement. I've been asked to remind you that my comments today could include forward-looking statements. You should refer to our SEC filings for a discussion of the risks and uncertainties that could cause results to differ materially from our current expectations. It is our intent that all forward-looking statements regarding our financial results and commercial activity made during today's discussion will be protected under the Private Securities Litigation Reform Act of 1995.

Perfect. So Francis, maybe just to kick things off with the topic du jour, which is your decision to close the GRAIL transaction. Can you just walk us through what you think were some of the key considerations as you decided to take that call? And as you think about potential repercussions that could result from closing of the acquisition, how are you thinking about those?

Yes, sure. So I'll start by saying that at a very top level, what we wanted to do was make sure that this deal got a chance to be reviewed and got to a decision by the regulatory authorities. We are under review, both in the U.S. by the FTC, but also in the European Commission. There were no obstacles to closing in the U.S. with the FTC. So we're in good shape there. We're still going through the administrative review process with them, but there was no obstacle for us closing in the U.S. So the only obstacle was in the European Commission as we were going through the European Commission review. And we are challenging in the European courts, the jurisdiction of the European Commission to review the deal, and we're also going through the review process. And until a few weeks ago, we were on a timeline where we would get a decision on both of those fronts in Q4 before the drop-dead date of the deal of December 20. But in the last few weeks, it became clear as those processes got delayed that we were not going to get a decision before December 20, and that a decision would come likely in Q1. We couldn't extend the deal sort of indefinitely into Q1 because the HSR filing we did in the U.S. would expire by March. And so, the way we felt we could actually get this deal fully reviewed was to close the deal, but then hold it separate until -- hold GRAIL separate until we got a decision out of Europe, and that's what we did. So no issues in the U.S. As I said, we have engaged with the European Commission on this path. They've opened an investigation into sort of that process. We are going to continue down the review process for Phase II as well as the case we have in courts there. So the only additional thing is there could be a fine associated with this action that gets imposed a couple of years from now, and it could be up to 10% of one year's royalty, so $400 million-ish, doesn't have to be that. That's the cap on what it would be. One of the deals we looked at that involved a Dutch telecoms company, that also got a fine. They got a fine of 1%. So something less than 10%. And in the end, we weighed that against the risk of this deal just expiring and us having to pay the $300 million breakup fee. And so we felt that was the right answer in terms of going ahead closing and making sure we got to a decision in Q1 on the deal. In addition to the considerations around shareholder value and making sure we're doing the move that long term maximizes shareholder value, we also felt a moral obligation to close the deal because the -- potentially life savings, the savings life associated with during this deal are so substantial. By accelerating GRAIL in terms of its global distribution and the accessibility of the test globally, we will save many thousands of lives by getting this test into the hands of more people and making it more affordable than GRAIL would do on their own. And so there was an amoral element here, too, about saying, in addition to creating long-term shareholder value, we have an obligation to have this deal reviewed and get to a decision because of the life-saving potential of doing this deal. And so all of those considerations came together, and that's how we made the decision.

Got it. Can you remind us of the appeals process timeline following an initial decision from the FTC and the European Commission? And as you think about the scenarios here that could play out over the next, potentially 4 to 5 years, what's the best possible outcome? What's the worst possible outcome? And what do you view as the most likely scenario at this stage?

Yes. So I'll start with some of the extreme ends. The best possible outcome is that we get a favorable decision out of European either one of the two processes. Either we win in the courts in Brussels, or we pass Phase II approval in Q1. Immediately then, we start integrating the companies quickly and deeply to realize the acceleration and the benefits of bringing the companies together. So we do that, and we continue down the FTC process. So the best outcome, we start the integration in Q1 and we're off to the races in terms of accelerating, getting this life-saving testing to the hands of people around the world. In -- and then, the worst-case outcome is that in the 2025, 2026 timeframe, either the European courts or the FTC asks us to divest in this asset. What that looks like is at that time frame, we'd have to decide what the appropriate divestiture was, whether it's selling the asset, selling part of the asset, taking the company public. But at that point, we will have created so much value in that asset that we feel the asset will be worth multiples of what we thought it had. And so worst case is we have to spin out either part or the whole company at that time frame and realize that benefit in terms of the value created in the asset. The best case is we keep it, and that's just part of Illumina, and that continues to generate long-term shareholder value.

Got it. You've talked, Francis, about the need to keep GRAIL separate and independent until the European Commission finishes its review. What does this mean on a day-to-day basis for the 2 entities? And how do you ensure adherence to this agreement? There was some chatter among investors around the possibility of a fine of 5% of daily turnover in case you fail to remediate any violations. How are you thinking about the likelihood of that scenario playing out?

Yes. So the way it's working at the hold separate, and this could be -- this could all go away in Q1 if we get one of the 2 European decisions on or it could last some time longer. But the way it's working is we're keeping GRAIL as a separate entity. Hans -- it's a separate organization that's completely -- wholly owned. All the equity of it is owned by Illumina. Hans Bishop continues to be the CEO of that entity, and that entity can enter into transactions with Illumina, but on an arm's-length basis. So on terms similar to what we'd offer anyone else. That could be them buying sequencers and consumables, but it could -- them buying services, for example. And so that's what we're going to do. We are going to have a trustee that will oversee the hold separate and make sure that their transactions and the interactions between the companies continue to be at an arm's-length basis. I think there was some misunderstanding as you said about that 5% fee and what that could look like. That is a very unlikely scenario where the trustee would find that we had sort of violated the hold separate. And we had worked with the trustee to come up with a remediation plan and a remediation time frame. And at the end of that time frame, we had still not complied with the remediation plan we came up with ourselves with them. And in that case, it will be 5% of daily turnover that we would be penalized. Again, it's a very unlikely case because: one, we'd have to violate the hold separate agreement; two, we have to then ignore the -- our own remediation plan. So it's a possibility and there is -- they contemplate that scenario, but it's very unlikely.

Got it. Turning to the GRAIL sort of synergy forecast here as you think through integration. I mean, I know you're not going to integrate for a while here, and that might push up the dilution towards the high end of the range, as you've mentioned before in the near term. But longer term, once you integrate the asset, where do you see the biggest opportunity for synergies versus the base Illumina business? I mean, there's the customer channel aspect, potentially R&D and SG&A as well, how are you thinking about that?

Yes. I'll start by saying that I'm actually really happy with how the GRAIL team is executing on their own on the plan that they had. So they had committed to launching a product in the first half of this year. They launched the product, it's now available in the U.S., and I've done it as of June. They committed to the results of their studies in the first half of the year. They published those on track. They committed to the NHS deal, starting the rollout this year, and you saw the announcements earlier today that the NHS is starting to roll out to 140,000 people as part of the largest clinical trial done. So everything they've said is on track. Having said that, we believe there are substantial synergies and very significant acceleration that can be achieved once we are allowed to integrate on a number of fronts. One, the GRAIL plan on their own calls for them in the next 5 years to really be in 3 countries: the U.S., Canada and the U.K. Illumina has products in 140 countries around the world. And so we have a very substantial ability to introduce the GRAIL product into health systems where we have relationships in all of those countries around the world. And so we can substantially accelerate the commercial reach of GRAIL around the world. Next, we have a very substantial market access capability to help drive reimbursement. And this is something we have built up about over a decade now. And when we first bought Verinata in 2013, you will remember, Verinata was the first clinical test for genomics or the first clinical genomic test, and there was no reimbursement at all. It was a self-pay test when it came out. Over the last decade, our teams have worked with payers, whether they're public or private, across all 50 states, at the federal level and around the world. And today, over 1 billion people around the world have coverage for genomic tests, for some kind of genomic tests, NIPT or therapy selection for cancer patients or genetic disease testing. And we have teams that now have deep expertise and experience and tools in getting reimbursement. So we saw that even recently, where last week, Michigan became the first state in the U.S. to have Medicaid coverage for critically-ill children in the NICU for rapid whole genome sequencing. That work, on a project called Project Baby Bear, was something that we'd worked on for a couple of years to make that happen. And so we have deep expertise, and so we can accelerate reimbursement for the GRAIL test around the world. And that has very dramatic, say, implications on the revenue ramp and the number of lives that can be saved. So that's market access. We also have deep regulatory expertise. Illumina has cleared products in dozens of countries around the world. And so we can accelerate getting GRAIL cleared in countries around the world. Next, there are operational synergies. We can scale up their lab infrastructure. Today, they run their tests out of their development lab in Menlo Park. Illumina has large-scale production labs, both in the U.S. and around the world. We run the testing, for example, for the government of England as part of their GEL project. Our tests have run, have this capacity to run to millions of tests a year. That means we have the real estate, the facilities, we have the equipment, we have the trained lab personnel, and we can scale up GRAIL's production lab capabilities very dramatically, very quickly. In addition to that, we can deliver more cost savings to GRAIL because our supply chain is much larger than theirs, and we buy from vendors at a much higher volume for some of the same raw materials that they do. And so they can plug into our supply chain, increase the resilience of their supply chain and get lower costs. So those are some of the synergies that we can bring to GRAIL once we integrate more fully.

Got it. Does the closing of the transaction and the ongoing legal process, Francis, complicate your ability to roll out additional pilots along the lines of the NHS one in the rest of Europe? I mean, is that a concern that you're worried about?

As of now, GRAIL is going to continue down the path it has, where it had studies that it was launching and that it had already launched, and some new ones coming in the U.S., and the NHS was the other big pilot that they were going to do. Then the next steps for them are to actually sign up health systems. So they can continue down that path. What we are going to do once we can integrate is accelerate their -- expand their budget and accelerate their ability to do more than they had initially planned to take on.

Got it. Down the road, how do you see the dynamics playing out between a multi-cancer test and a targeted single cancer early detection test working together? I mean, do you view these as complementary tests? Or do you think that they would be essentially competing for the same reimbursement dollars? And then I was just curious to get your thoughts around some of these multi-modality approaches coming around, including introducing AI-based imaging and so on as an orthogonal marker, if you will, to increase the performance specs of early detection tests?

Yes. So even before we did the deal, we did a pretty significant due diligence that included speaking to what would be prescribing doctors for this test. And what we found is that the Galleri test is unique and doesn't really compete in the same space as the single cancer tests that are being developed. And let me give you an example. And I did some of those calls myself. If you are a GI doctor and you are -- your focus is to look for colorectal cancers, then you'll want the single cancer test. And in fact, you actually don't want a 50-cancer test because what are you going to do with all of those other findings, right? And so that's a market that's very built for a single cancer test, or it could be a small group of cancers that are in your specialty area. Now a 50-cancer test plays very nicely to general practitioners, to -- and they do want this to be as part of your annual physical, for example, if you're above 50 years old or if you're a family history. And for them, they wouldn't replace that with a single cancer test. So they're saying, look, if I had the choice, I'd much, much rather go with the 50 cancer test. And they -- and in the multi-cancer world, tissue of origin is very important. They're also saying it's not a substitute if you show up with another multi-cancer test that doesn't tell me tissue of origin, because that will create more stress in the system that we don't want. And so they're very complementary. There is a single cancer test market, and then there is the 50-test market, of which GRAIL is really the only participant in that market.

Got it. And then Illumina, obviously, has a very large base business outside of GRAIL, and I want to spend a little bit of time on that as well. As you think about how you frame the guide for 2021, you spoke about a bunch of nonrecurring items in the first half versus the second half. Despite that, I mean, how much sort of conservatism do you think is baked into your guide at this stage? I mean, we've had a little bit of a spike in the COVID resurgence angle and yet, I mean, you kind of like shrugged that off much earlier in the year than some of your other life science peers. Walk us through the puts and takes in terms of the second half guide and the momentum into '22.

Yes. I'll start by saying that we are seeing real strength in the core business. I mean, it really -- the year started strong and then just accelerated. So Q1 was strong and Q1 was -- Q2 was really strong, too. And what's driving that is primarily sort of the reimbursement we saw put into place last year across all of our clinical markets. So in genetic disease testing, the number of covered lives in the U.S. for whole genome, whole exome sequencing, for genetic disease diagnosis increased by a factor of 10 last year. In NIPT, for example, we saw very large payers like United and Aetna reimburse for average risk. In cancer, not only do we see an expansion in the covered lives that can take advantage of genomic panel testing for therapy selection. We actually saw an expansion in large panels. So not only do more lives have access to panels. Those panels are now larger panels that have -- that means more sequencing and plays very well to our TSO 500. And so all of that reimbursement got put into place last year, and we're seeing the beginnings of that tailwind this year. That's a durable tailwind because we still need to get penetration into that expanded market opportunity, and that's what we're seeing play out. So first point I'll make is really, really good strength in the core markets as we come into the second half. Having said that, there are some benefits we saw in the first half that won't play back again in the second half. Some of it is the COVID instrument placements that we saw in the first half, about 55 million of those placements. And we don't expect more instrument placements. We expect to see consumable pull-through, but not more instrument placements for COVID. So that's one factor that won't be repeated. The U.K. Biobank deal ends its sequencing in the third quarter of this year. So that's about $20 million a quarter that you won't see in Q4 that you saw in the previous quarters. We saw $20 million in NIPT royalties that came out in Q2 as a result of selling some patents, a little bit of stocking, about $20 million in stocking. So those are some of the unique factors that played out in Q1 that we won't see in the second half. And that's why the guidance looks more conservative than you'd expect. But the core business is really strong. And that strength we expect to continue playing out in the second half of the year as well. We've also built into our guidance, the reality that we're in a pandemic, and that, that could present unexpected challenges, whether it's people not going into labs as much if there are outbreaks or disruption in the supply chain, for example. So we have factored some of that into our guide for the second half.

Got it. I want to pick your brains a little bit on the competitive landscape, Francis. I mean, some of your older patents have started to roll off here. There are a number of emerging short-read technologies, there's Singular, Element, GenapSys, Omniome, et cetera, looking to make inroads into the market. How do you view your competitive moat against these emerging technologies? And then I have a quick follow-up as well.

Yes. I think when we have a market as big and as rapidly growing as the market we're in, you will continue to see -- and we've seen that over the years. You will see venture-funded companies come into the space. We are seeing sort of the SBS companies that you named, looking to come into the market as the first wave of SBS patents roll off in 2023. We continue to compete with some larger companies, too. Obviously, Thermo is in our space and BGI is in our space. Our approach is going to continue to be what it has always been, which is continue to innovate and lead in terms of superior value delivered to our customers through the innovations we bring to the market. We have a fantastic team. We are deeply committed to innovation. We spend 18% of our revenue on innovation. And the pipeline we have in front of us in terms of innovations is stronger than it's ever been, in terms of better, faster, cheaper sequencers, end-to-end workflows, deeper AI capability. And I am very optimistic about how that positions us competitively against the players that are coming into this space.

Got it. The other question that comes up ever (sic) [ every ] so often with Illumina is that as you look to introduce a higher throughput box, what drives that decision, right? Is it essentially sort of feedback from customers that there would be enough volume to more than offset a lower per genome price point? And if so, I mean, where do we stand in terms of that evolution today?

Yeah. There are a number of factors that drive a launch of a big product. One, for the research market specifically, is exactly as you said. We spend a lot of time listening to our customers to understand from them what projected demand would be at different price points. And what we're hearing right now is that there is strong demand for doing much, much larger cohorts. And a number of things are driving that demand. One, the availability of high-throughput single-cell tools, where people really now want to understand the differentiation of cell expression across the body. And there are trillions of cells across your body. So to do even a representative sample, you're talking about experiments in the tens of millions of cells that you would want to do. That's way bigger than anything that can be done today. And so people want to continue to drive the price down, so they can do much larger single-cell experiments. They want to do much larger spatial experiments. They're also telling us that they want to take on much larger cohorts, so they can understand the genomic drivers of complex diseases like cardiovascular disease, like chronic lung disease. But also, neurological conditions like Alzheimer's and Parkinson's and autism and schizophrenia. It's really clear if you look at schizophrenia running in families, but there is a genomic component to it. But unlike monogenic diseases, like single cell, it's a much more complicated cause and effect. And so the only way you'll get to understand that is by doing cohorts that are measured not in the thousands or tens of thousands, but in the hundreds of thousands and the millions. And so, we're now getting much more visibility and much more granularity around the elasticity in the market associated with going from a $600 genome to a $100 genome and even beyond that. On the clinical side of the market, the bigger driver of launches is, when is the segment ready for a cleared end-to-end solution. And so what that requires is that the market is opened up, so there has to be the discovery. And then the first wave of genomic markets is typically driven by LDTs. And so you've got to get to a maturity phase in the market where the product -- the test has stabilized, where the content is more stable. And at that point, the market's ready for an IVD, and that's when we'll launch an end-to-end IVD into a market. So that's what we look for in terms of those launches in the clinical market.

The other question we get, Francis, often is PacBio, for example, has now bought Omniome, they're looking to launch an integrated sort of short-read, long-read solution at some point down the road. How trivial or not is that challenge in your mind? What are your plans for the long-read market? I mean, clearly, it's the same sort of core point in terms of customers, and it's an important and it's growing sort of like niche within the broader sequencing markets. How are you thinking about that opportunity?

Yes. We believe long reads, as we always have, are a complementary part of the market, and they represent, maybe now slightly less, but 5% of the overall sequencing market are served by long reads. Now the reason I say slightly less is short reads have continued to expand their capability. 2 years ago, I would have said that maybe a gene like CYP2D6 is better suited to long-read. But today, with the advancements we've made, short reads can read it really well. Two years ago, I might have said paralogs are something that long reads could do, and short reads weren't as good at, but today now with the advancements we can do that. So we're kind of eating into that 5%, but it still [ represents of ] a complementary part of the market. There was a hypothesis that people were sort of asking about maybe 9 months ago about, wait a minute, if short reads get cheaper and get more performant, could they just compete more in the short-read market? And what we found is that's not -- we didn't believe that was going to happen, knowing the technology. And then you saw PacBio buy Omniome. So it's clear that long reads couldn't get to the performance short reads, and that short reads are going to be the way that the vast majority of the market is served. And so we've believed that the whole time. In terms of stitching together data, we've been doing that now at customer sites for a very long time. If you go into some of our large customers and some of the population programs, you'll see a fleet of Illumina machines, and then you'll see some long-read machines. And we've always been committed to an open data format. And so customers can take our data, sort of combine it in terms of an analysis with data from other sequencers. There's not a lot of use cases for deep integration. Typically, what you do is you'll come up with a reference genome on the long-read and then you'll use that for short reads. And that's something that customers have been able to do for a number of years now.

Got it. And I mean just pulling on that string a little bit more. I mean, one of the things that PacBio has sort of talked about in terms of their Invitae collaboration is this sort of the dream in sequencing, right? So you have long-read data, but you have a short-read price point and throughput. Do you think -- do you worry about that at all? Or is that something which you feel as, if and when it happens, Illumina has levers to pull in terms of your pipeline that you can respond effectively?

Yes. I think the way we get to the dream is short reads continue to expand what they do. And I'll tell you why. If you look at the vast majority of the markets, if you look at NIPT or cell-free DNA for -- in liquid biopsies, for example. The fragment size you're reading, in NIPT it's 130 base pairs. Even in liquid biopsy, it's less than a couple of hundred base pairs. And so being able to read 10,000 base pairs or 1 million base pairs doesn't really bring anything to these clinical applications. What it does bring are restrictions. The sample input you require for something like a PacBio is so much larger than you need for an Illumina machine. The cost for the sequencing is so much higher. And then if you look at ONT, you give up so much in accuracy. And so there's a lot of penalties you pay with no added benefit in terms of the applications we talked about. And so short reads are going to continue, I believe, to set the gold standard in terms of accuracy, in terms of price performance. And with the advancements in machine learning and AI, you will continue to get a more perfect view of the genome as we just methodically eat away at the rest of the genome. And so that's, I think, how you'll see it continue to play out.

Got it. I just got a question on e-mail on the 30 different sort of PopGen initiatives you talked about in the last earnings call, Francis. How confident are you that we are at that sort of inflection point in population sequencing? There has been a couple of sort of false starts in the past and oftentimes headlines are followed by an uncertain wait before samples start to flow. So as you look at that landscape today, are you confident that POPSEQ is here to stay and should grow off that base on a year-over-year basis over the next few years?

Yes. How far we've come in the last couple of years, Tejas. Where in 2019, we were talking about the first ever rollouts of a health system doing population sequencing. And this was the NHS doing genetic disease testing and cancer therapy selection. And because it was the world's first, there was so much that we learned, both the NHS and us, in terms of how you roll out genomic testing into a health system. I'll give you a couple of examples. In cancer, what we learned was you really need to do fresh frozen sample accession rather than -- or sample collection rather than FFPE. And that required a pretty dramatic change in the workflow in the NHS, associated with getting tissue samples from cancer patients. There was a lot of work we had to do in integrating into the IT infrastructure in the NHS. It turns out although it's 1 system, there are a lot of different IT infrastructures you have to integrate with. And that took us a lot of work to get to a point where an NHS doctor who's just going into their existing workflow and order an Illumina test. Now that we've done that, though, we have this terrific experience, the NHS has and so does Illumina, in terms of what it actually takes. And I think we're uniquely placed in the world in terms of just having that bolus of experience, which helps us in terms of forecasting and projecting, but it helps our customers because we have a lot of experience now in terms of how you do it more globally. And so I think that gives us the confidence to say, okay, here's where we expect the 30 to come in. And just the fact that there are 30, gives you sort of more redundancy in the projection model to say things could slip in a couple of places, but you have a couple of places that could pick up the slack too. And so, we are much more strongly positioned than we were a couple of years ago when we were doing the first ever population programs.

Got it. Another one that just came in over e-mail here on, do you need to launch a higher throughput sequencer to go after some of those single cell and spatial sort of high throughput opportunities that you talked about? Or do you think you have what you need today to serve that market with the NovaSeq and the new flow cells?

Long term, if you listen to our customers, they will say absolutely, they would love a machine that was faster, cheaper, did higher throughput. So no doubt, and I think there is demand for that at some point, to say look, can you get even more samples going. And they love the NovaSeq, but they're already looking past the NovaSeq to say, now I want to do order of magnitude more samples. I want the price to go down even further. And that will require new instruments in flow cells over time. So I think there's many, many years of innovation still in front of us.

Got it. One final one before I let you hop. M&A, I mean, your hands are a little bit tied in terms of acquiring competitors or customers. So as you think about sort of these feeder technologies, single cell and proteomics and spatial biology, et cetera, is that a possible area where we could see Illumina make an inorganic sort of tuck-in move?

What we look for are -- we do look for absolutely technology tuck-ins. We've done a number, some we've announced, some just we are integrating into our technology. So we are looking for ways to enhance the competitiveness of our products, to expand the capabilities and deliver more value to our customers. So absolutely, you should expect us to continue to look for technologies to tuck in. And then from time to time, if we see a big opportunity where there is an opportunity to create a differentiated offering, because there's a strong IP position, for example, or some other reasons like that. And if there's an opportunity where we feel at Illumina, we can deliver it better than anyone else in the world. If those 3 things come together, then from time to time, like happened with GRAIL, you'll see us make a bigger move, but that's rare. But you will see much more of those technology acquisitions that we talked about.

Got it. Fair enough. This was a great overview, Francis. So thank you for joining us. We appreciate your time this afternoon, and I hope you have a productive rest of the conference.

Thank you, Tejas.