All right. Hey, everyone, I'm Tejas Savant, and I cover the life science tools and diagnostics sector here at Morgan Stanley. It's my pleasure to host Illumina today, and from the company, we have Joydeep Goswami, CFO, and Salli Schwartz, Head of Investor Relations. Welcome, guys.
Thank you.
For important disclosures, please see the Morgan Stanley Research Disclosure website at morganstanley.com/researchdisclosures. If you have any questions, please reach out to your sales rep. So, maybe, to kick things off, Joydeep, you know, big news last week with Jacob Thaysen's appointment as CEO. His pick seems to signal, you know, a fresh start for Illumina with a renewed focus in the core. His track record and margins was a big point of emphasis in the press release. Would you agree with that characterization?
So look, first of all, we're thrilled to have, Jacob's announcement come through. I think, you know, previous conversations I've had with all of you, we had emphasized that we are, you know, prioritizing bringing someone with experience in the industry on as quickly as possible, right, to get rid of that uncertainty. So I'm glad that's behind us. Jacob obviously has an extensive experience in the industry, both on the diagnostic side, but also generally, life sciences, tools and diagnostics. As you said, right, we, we have emphasized, and you've reminded us, that we need to get back to, better margins and a stronger track record of profitable growth. I think Jacob definitely brings some of that with his background, especially on the, you know, stronger margins front at a scale company like Agilent, right?
So all of those are very positive pieces for us.
Got it. Let's start with the recent guidance cut on the sequencing side of things. So three pieces to it, China, the macro impact, and of course, the 6,000 consumables. So start with China. Can you just parse out for us how much of the impact on mid and low throughput from China is due to competitive pressures from BGI versus just macro weakness? And would it be possible to perhaps add a finer point on it in terms of, you know, how much of that competitive pressure is from a push from the government towards local sourcing versus just, you know, deeper discounting, perhaps then, you know, of course, the anti-corruption crackdown in the region as well?
Yeah, so we haven't given, you know, parsed out those quantitatively, if that's what you're asking for, but all of those points are factors, right? And again, to remind folks, you know, we had gone in this year expecting a recovery in China to normal levels of growth. There was no stimulus baked in and second half because we figured the first half would be consumed by just, you know, getting out of the COVID funk and getting labs back up and running. In the first half of China played out kind of how we had expected it, right? But as we looked into the second part of the second quarter and looked in the rest of the year, we just did not see that, you know, the recovery happening.
I think, since then, since we came out, and most companies have reported a very similar kind of outlook in China. And we don't see, again, looking at all the macroeconomic factors that, you know, the recovery occurring anytime, you know, in the near future that we can predict at this point. The enhanced competition and, of course, you know, the government preference for local and certain tenders, and of course, remember that we are not allowed to sell to all entities in China, most U.S. companies are not, you know, continue to be elements that are, you know, restrict the market in China for us a little bit. Included in our second half guidance was what we saw as a, as an increased competitive intensity in China from BGI.
You know, that's translating into more lost deals to them, but, you know, price actually tends to have an impact as well on that, right? We're pretty, you know, directive in terms of how we select customers that where we will compete on price and others that we won't. So that is playing a factor, and that we assume will continue to play a factor. What we did not include in our guidance in Q2 was anything that has played out a little bit more recently around some of the crackdowns that you're hearing on hospitals.
Again, just to be very clear, we are, you know, we're very selective in who we interact with in terms of our direct customers, and our distributors, and we follow all of the regulations of the Chinese government and the United States government, in terms of how we do business in China. So we're not concerned about that, but, you know, if our end customers get impacted, by something, we, of course, don't have any control over that, right? But that was not a factor, and I don't think we have an update on any of that, yet.
Got it. You know, it's tough to compete with free in any region in the world. You know, why is this competitive pressure showing up only in the mid throughput and not in the high throughput portfolio? And, you know, while the U.S. and possibly the Western European markets are likely, you know, challenging for BGI to crack into, why couldn't they start to see more traction in price-sensitive geographies like Asia, LATAM, et cetera?
Yeah. Good question, Teja. So again, you know, basis of competition, of course, is different in different throughputs. And again, you know, in the high throughput, there, price is one of the areas, right? But there is a whole host of other things in terms of, you know, the workflows aspect of it, the back-end translation and, you know, the data analysis piece of it, the compression sides of it, which are enabled by the X and the DRAGEN piece. And then there are other factors that certain customers value more than others, which include the backward compatibility of our bioinformatics platforms, include now the more ambient ship capabilities that we are providing. So basis of competition, again, is not only different in different throughputs, but also for different customers.
The second thing is, of course, that we've always had competition, right? So if it wasn't BGI, it was Thermo and, you know, for a while, Qiagen, et cetera. So we know how to work and play with competition and compete with them, so this is not something new to us. Third, and it's true in China as well, right? So while prices in China have been impacted, we still do extract a premium for the innovation and the quality of our products that we bring to it, bring to the market. And again, that's a factor of all those areas that I mentioned earlier. So none of those have changed. Now, on the mid-throughput side, I think you are seeing a bit more competition because there is, right?
When our patents, the original, you know, patents went offline, we do see a little bit more of companies that are trying to copy the technology and have come out with some solutions into the market. Again, we have to be cognizant of that price differential, and we have been, but we still expect to extract the premium. And of course, as you can see, right, we continue to innovate and continue to introduce more technologies, such as the XLEAP chemistry on the NextSeq 1K/2K as well, that's expected. So you should see that continue and see us continue to play, you know, a role in that market.
Maybe the last thing I'll say on that front, right, so I mentioned this in the second quarter earnings as well, that we have not seen a big change in the win rates, you know, in mid-throughput, everywhere, including Asia, except with the exception of China, where we did see an impact. Where we did see an impact, partly because of competition, partly because of macroeconomics, is slightly longer term to convert the funnel of opportunities, right? So that's that is something that has impacted, and we called that out.
Got it.
Maybe just to add on the high throughput specifically, we're also selling now the NovaSeq X and including into China, and that's just a unparalleled or unrivaled-
Mm-hmm
offering that no one else can compete with.
Got it. Really, are you seeing any sort of, you know, China contagion in export-oriented economies in Europe, like Germany, et cetera, or not really?
Not as of this point, right? Again, you know, Germany is a good example, right? Germany is actually one of our faster-growing markets in Europe. You know, where they have been following the lead of the UK a little bit, is more investment by the government on funding clinical applications and helping kind of transition from that translational research into the clinical side, and then following it up with reimbursement. So, right now, we see, you know, at least on the clinical side, strength in the market continue. We'll have to look at, in Germany, again, you know, you've seen a lot of the reports in the Wall Street Journal, right? They are experiencing some economic challenges, so it hasn't translated so far into research grants, but we will have to continue to monitor that very carefully.
Got it. Fair enough. And what is your confidence in that year-end budget flush, Joydeep, at this point? I mean, I think when you, when you framed the guide, you'd assumed regular sort of, you know, year-end seasonality. Is that still a fair way to think about it?
So far, yes, right? Again, we continue to monitor that very carefully, but, you know, we had not assumed a very high level of flush. And then, you know, partly it was year-end seasonality flush, so both on sequencing and microarrays, right, you do see an uptick in the last quarter, just again, even on the DTC side, for arrays. And then the other part, which was guiding towards our our bump in Q4 over Q3, really was related to the 25B flow cell launch on the X side, which is, you know, the highest capacity flow cell, and you you generally will see an uptick from from consumers adopting that.
Got it. Switching gears to the, to the portfolio, NovaSeq X, I mean, obviously, you know, off to a strong start. You noted about 10% penetration across those 1,000 sort of NovaSeq customers that you have today. Where are you in terms of penetrating the single unit, so 750 customers versus the 250 multi-unit customers?
So broadly, you know, X customers, we think they've purchased about 110 instruments, so about 10% penetration.
Mm-hmm.
For the single unit customers, right now we're at about 5%, right? So that's 750 or so. And that's again, it's logical because the higher throughput customers who are the most likely to switch early to something that is higher capacity, you know, you have a higher switch weight amongst the multi-unit-
Right
- customers than the single unit customers.
Got it. You noted some technical issues that emerged during the rollout. Where is, where is that in terms of, you know, those fixes have they been fully deployed across the user base now?
So, yeah. So we had two types of things that we saw in Q2. These were above the, you know, the norm of the tail of issues that come up, right? So we did deploy fixes on both of those-
Mm-hmm
Towards the end of Q2. As far as we can tell, you know, majority of those have already been deployed. Now, you know, some of that, the ones that have not yet been deployed are more customer related things, so vacations played a role in some of these, lab readiness has played a role in some of these. But we're well on our way to fixing those right now. Again, those issues, as I mentioned earlier, they're not unprecedented in terms of an instrument like this, although, you know, they were not we were not anticipating them for in Q2 for us. But, you know, regardless, they were fixed pretty quickly, and we intend to do the same with any other issues that come up on the X, right?
As we are deploying it to new customers, new applications, we have a team that is ready to actually, to address those issues very quickly.
... Got it. And then as you move past the power users to smaller customers in the X upgrade cycle, is sample access a concern at all, where, you know, customers have to choose between extending the turnaround time and batching the samples, or you no longer get the industry-leading economics of the X?
So, you know, different customers will make different choices on this, right? If you take a look at just the limiting example here, right? So if it's the same number of customer samples, your X, you know, at some level, like, makes your capacity 2.5 times at least larger. So if you have the same number of samples, you're gonna wait 2.5 times longer to batch your samples and run it. That's not what you typically see happen, right? So it plays out somewhere in the middle, where, you know, they'll maybe wait a little bit more to batch, but then they won't run it at full capacity.
And what that translates to is you don't get the pricing that people assume on a gigabase per sample or $ per gigabase. Sorry, $ per sample or $ per gigabase, right? You get a slightly higher effective price that we recover. So we've seen that with the 6,000 transition from the HiSeq, so we expect to see the same. We don't—we generally never see customers wait and start batching samples at the rate that you know that would be that they would need for a full flow cell load.
Got it. Switching to the NextSeq 1000/2000, you're bringing XLEAP there in the first half of 2024. In light of the greater competition in the benchtop sequencing side of things, you know, Element had a press release with 100 units, you know, for AVITI. How are you thinking about pricing, and is there a possibility where we could see you pull forward bringing XLEAP to the NextSeq 1000/2000, versus, you know, 2024? Could it be sort of a 1Q situation?
Look, I think we're on track for where we had planned for the NovaSeq X launch in early next year, right? So we'll, we'll come up in due course of time and communicate the exact date of launch. I wanna reiterate what I said earlier, right? So, first, I mean, again, we're not seeing any acceleration in our win rate or deceleration in our win rate in the mid throughput, right? So we track that very carefully, and at least, you know, at the end of Q2, we haven't seen any noticeable change on that front. It is interesting, you know, when you see releases from Element and others, right, they talk about placements, not orders or purchases. So, you know, that'd be interesting to keep in mind as well, right?
That they may be placing boxes where the customer has purchased a, an Illumina box, right? So, And then the last piece is, you know, we get reports on how our instruments and other, you know, other companies' instruments are actually performing in the field with applications that they want, right? And we feel fairly confident that, A, you know, customers still will want our, our technology and our boxes. And two, that, you know, given the value we bring holistically in a total, cost of operation kind of, mentality, we still will continue to win and will continue to be able to charge a premium. The last thing is, you know, and, and we've said this publicly, right, that the premium remains the same, but I think prices generally per dollar per gigabase will continue to come down.
Mm-hmm
... in the mid throughput, and we have kind of taken that account in our modeling.
Got it. You called out about, I think, 15%-20% of orders were from new to Illumina and new to high-throughput customers for the NextSeq. What is the replacement ratio for these NextSeq upgrades? And does this upgrade dynamic, just from a mathematical perspective, mean that there will be downward pressure on those pull-through ranges that you've put out at the moment for the NextSeq?
Yeah. So, you're talking about just mid-throughput-
Yeah
... conversion, right?
Yeah.
So yeah. So for those, no, we don't expect a downward pull, push on that, right? Because people usually are upgrading or coming into this because they expect a certain volume to come through that, y-you know, gets them the economics that they expect for the NextSeq 1K, 2Ks. So the range that we had given, the 120 to 170 or so, we still expect to be in that range. Now, you know, exactly where we are in the range in a particular year, really depends on the volume or the sequencing throughput growth, right? So again, it seems like it gets a little bit more difficult to predict it, because the same number of samples, but you intend to do more single cell or more spatial analysis, right?
That actually implies that you're running the instruments a lot harder, even without changing the underlying sample number there. So we still think that we, you know, we should stabilize within that $120,000-$170,000 pull-through per year.
Some percentage of that 15% of-
Mm-hmm
... new to high throughput or new to Illumina, is new to Illumina, so brand-new customers.
Right.
With the ambient shipping capabilities that the X has, we're able to sell into places that we hadn't historically placed boxes in. So that's been a piece of it as well, which is, you know, truly incremental.
Got it. I want to switch to the numbers a little bit, Joydeep, if I may. Before I get into, you know, the guide and margins and so on, you know, just philosophically, you know, you're a market leader in sequencing. Why not be a price leader as well? You know, essentially, you know, starve the emerging competition rather than take your pricing cues from them. That way, you can also have the best shot at sparking, you know, demand elasticity while you cut costs, you know, more aggressively and protect your margins.
... I think just to a large extent, that's what we have been doing, right? So again, you know, we, you would expect us to be fairly calculated and not just looking at price, but looking at value we are bringing with our investments in technology into the market, right? So our pricing is set with that level of, you know, what level of value do we bring to the customer, looking at their total cost of operation, right? And so if we make workflow improvements that either radically reduce a customer's need for data storage capabilities or, you know, back-end analysis capabilities, you would expect us not to just give away all the value of that capability in the price we set for our, you know, for our consumables or for our instruments, for that matter.
So that's, that's one level of, you know, of, thought that goes into our pricing. And, you know, it's still leading the market rather than following, but it isn't always just going down to bottom level of price in, in every workflow. And again, you know, how you measure price matters, and so what you will see from, a lot of our competitors is a pricing that doesn't take into account the full impact that we are, or the full value that we are providing to customers. You know, I think reactive versus proactive, I mean, obviously, as we think about innovation and how we price our newer, innovations, I think it's much more proactive. Sometimes you have to be reactive to the market, right?
If, you know, it's I can't sit out here and say, "I don't care what somebody else is pricing, and I'll never match their price, or I'll never adjust my price." We don't do that, but we are very selective in terms of which customers we extend what price to, from a point of view of, you know, how we believe that customer can drive greater sequencing throughput and value to our investors.
You know, that's an interesting point, Joydeep. Do you think collectively, the industry, I mean, you guys and everyone else who's selling sequencing hardware at the moment, is moving away from this focus on price to a focus on value? Because this is, you know, a point of concern, right?
Yeah.
Because a race to the bottom eventually may not be to anyone's benefit.
I, I think that will happen anyways, right? Partially because, you know, at some point, if you look at your end customer, sequencing becomes a smaller part of their overall cost base. So further compressing sequencing costs without providing value in some of the other pain points do they have, has limited utility to them, right? So again, if you don't address things like, How do I actually use my sequencer for more things that I want to do? So whether that's, you know, other multiomics, like proteomics and methylation, or long reads or RNA, if you just drop the price on DNA, it's not gonna, it's not gonna help them, right? So, so you will see a lot more there. If you, you know, that's why we made the investments we made into DRAGEN and the technology, because that was becoming the pain point.
Same thing with data compression, right? It was becoming a main, a pain point for especially the high throughput end of customers, right? And you start spending an inordinate amount of money just trying to process and compress the data that's there. So you are going to see more of the value, but there's a reason for that, right? It's just what customers are asking for in terms of their pain points, in terms of their overall cost or utility of the instrument, the utility of what they can use the instrument for. And you'll see naturally more kind of migrate to that side of things rather than not.
Fair enough. You know, I have to ask the obligatory question in 2024. You know, even before Jacob's announcement, you reiterated, you know, your commitment to that 25% op margin target for core Illumina, but you also called it sort of stretchy in light of the implied, you know, year-over-year improvement. Can you provide some of the pushes and pulls in achieving that target next year beyond real estate optimization, SG&A cuts, and that, you know, natural leverage that will come with,
Yeah
... with the X ramps maturing? Do you think R&D cuts could be on the table as well, or is that going to continue being a sacred cow for Illumina?
Yeah. So it's interesting. R&D has not been a sacred cow, although we have been, you know, as you would expect us to do, right? We, when we cut costs last year and even this year, we made very surgical costs across the organization. Now, real estate and other things were maybe a broader tool, but you know, for us, if you don't get to labor-based costs, you don't get to the level of cost realignment that we needed. And again, you know, going back historically, look, we had not planned to be at the margins that we are right now. What happened was we invested ahead of the curve in terms of what we saw, in terms of growth in 2021 and then in early 2022.
You know, as the markets turned, it quickly became evident that our cost structure was not appropriate to where, you know, growth post-2022, at least for the short term, was. So we will make cuts across the organization as required, but we are not stepping back from our commitment to innovation, right? So that the programs that are necessary, both in terms of core instruments, but also the end-to-end assays, will be sacred cows or whatever you want to call them. It's hard to sit here as a Hindu and say sacred cow, maybe it's easy. But, you know, I think part of that does not preclude productivity initiatives within the company, right?
So we're, we told you when we came out, at the end of Q2, that, you know, we've done the cost reset, but we are also looking very broadly at productivity, at productivity and transformation, business transformation initiatives throughout the company. And those are the things that will provide the question that you're asking is: How do I deliver on innovation and, you know, commercial infrastructure expansion as you expect, in countries?... without necessarily, adding a lot of OpEx and or maybe even allowing us to reduce OpEx and to increase, you know, COGS productivity. So we have a lot of those plans actually going into place right now, and we feel confident that we can continue to deliver.
Last thing I will say, the stretch piece of it is just something I wanted to be honest with you guys, right? We had expected 7.5% growth this year, and you know, going into next year, and you know, at least 22% margins when we put out those numbers, right? We are at 20% and you know, close to flat for this year, so we have to actually go find some you know, additional uses to get to margin expansion, where we will get to margin expansion. And I also wanted to give Jacob a chance to come in and have his own thoughts about that before we come back to you again.
Got it. Makes sense. I want to switch to GRAIL here. I think you expect the EC divestiture order any time, a decision in the FTC case by year-end, and then the EU jurisdiction appeal is set for, you know, verdict year-end or early next year as well. You know, most investors, you know, we talk to are convinced that the new leadership in charge, a divestiture is the most likely outcome, although, you know, the terms and timelines of that divestiture could vary depending on what happens with the appeals process. Do you think mid-2024 is a fair way to think about when GRAIL will be off the books, or could it actually be later than that?
Look, at this point, Tejas, I'll just reiterate what we have said earlier, right? We're waiting for the EU divestiture order, which really gates the options that we look at. You know, we have done everything we can to provide them all the information they've asked for, so the EC is not waiting for anything from us to come out with a divestiture order. As soon as we get that, you know, we have all of our ducks in a row to begin the process of lining up, you know, what is the best way to divest and team all of that up, right? So that will depend a little bit on what we get in the divestiture order, but we have everything locked and loaded to go down that path.
And while we're doing that, as we've said before, we expect these two court cases to play out. So the Fifth Circuit Court of Appeals in the United States, which, you know, looks at the FTC objections to the deal, and then the European Commission, sorry, the European Court of Justice decision, which is different from the European Commission, where they're looking at the application of Article 22 and a few other factors, right?
And again, how we proceed really depends on if we win both those cases, then, you know, there's a further kind of, enables us to have a lot more freedom in how we think about GRAIL and the opportunities to maximize its value for our shareholders, including, you know, divestiture as an option, keeping it with a very different cost structure as an option. There's a whole host of things, as we have communicated earlier on, you know, licensing, IP, et cetera, that may be possible. And again, you know, these are not things that we have completely lined up here, but we do know that there are options there. On the other hand, if we happen to lose either one of those two cases, then divestiture is a certainty, right?
Then we have already begun the planning to go down that path. Now, timing about when it occurs, it does depend on which divestiture option you ultimately pick. And then remember, the second part that's gating is actually receiving the divestiture notice, right? Because, you know, that really is the start of the process. It's gating the start to the process.
Got it. You know, following PacBio and now GRAIL and, you know, sort of the tangle with the regulators on both sides of the pond, I feel like, you know, M&A might be a bit of a lightning rod in your boardroom at the moment. So that said, you know, how open are you to considering a push into multiomics via proteomics or perhaps spatial biology? You know, adjacent seeds where you wouldn't step on anyone's, you know, any customer's toes, at least, there's clear synergies to be had, and the two technologies actually play well with each other.
Yeah, so we're already doing that, right? So remember, so I wouldn't say all M&A is out, right? I mean, we have been doing tuck-ins, you know, through all of this. So I think the larger M&As are going to be a little bit more difficult in the short term for us and, you know, honestly, for most companies, given where the regulators are these days.
So we're looking at tuck-ins where, you know, we have a very experienced BD team that looks at technology licensing or other kinds of partnerships to be able to bring in technologies and then, you know, to be able to optimize it on an Illumina workflow, an Illumina bench, et cetera, and add value so that, you know, we're bringing a product to market that's really, you know, distinguished and differentiated from whatever else is out there. So you'll see us continue to do that, and I think it's a good point you're raising, right? Because for us, if we invest in something in an end-to-end workflow, it has to improve the utility of the sequencer and has to bring the customer value over what they can already get in the market.
We further gate those applications by whether they are large enough, so the market has to be large enough and has to be growing faster, right? So a lot of what you just stated in terms of, you know, proteomics or spatial or other things, right, there are many opportunities for us to play a part in either directly or indirectly helping the growth in those areas.
Got it. To wrap up, Joydeep, one final question for you. So, you know, there's competing visions for Illumina's future at stake here. On the one hand, you know, you have a company that excels at selling, you know, valuable hardware. On the other, you know, you have a company that could push into select applications and potentially capture much more of the value that it creates. Obviously, you know, involves incurring some costs in the near term, as we know, with the GRAIL dilution. Is the organization culturally ready today to buy into a more circumscribed, formal vision after the last couple of years, where you really sold hard on the value of, you know, getting into screening, et cetera?
So, you know, I think GRAIL was a very different kind of bet, right? Where it was a, you know, slightly different set of customers, the time frame in which that bet would play out. We knew from the beginning that it was a long-term bet, but it was such a huge bet that you know, if it played out right, right, it created a market which would rival the size of our business in the outer years. So let's keep aside GRAIL, because it's a very specific type of play, and it required capabilities that were different from what Illumina had, right, in that sense.
We could add value to those, but it required, it was a build versus buy decision, and the buy decision was, you know, for that side of things, would play out in the long term and would add a lot to our capabilities. Now, everything else, though, that you have talked about is not an or issue, right? It's an and issue, because for us, the way we're thinking about it is allocating for the core business, right? Allocating capital so that we continue to maintain our lead as the best putting out the best sequencers and not necessarily always just focused on, oh, we got to drop the price per gigabase, and that's the only thing, right? We're much more sophisticated, as you've seen with the X and the XLEAP chemistry, right?
That it brings other points of value, which have now become pain points for the customer. So you continue to see us doing that in all levels of throughput, right, and appropriately gating launches that we have that bring that level of step change in value, and I will define value more than just price to our customers. At the same time, we're taking some of that as we get productivity in some of our things. And again, this is something we've said before, right? We're invested in modular capabilities in some of these innovations, which then allow us to bring the next set of innovations to you at a much lower cost overall.
Some of that, then, capital that we now can allocate to things that are an end-to-end workflows, which, A, improve our share of wallet, improve the ability then for customers to, to then, go after these applications on the installed base of sequencers, which then has a synergy flow, synergistic flow, in not only, you know, making money for us in, in the library prep and the, the back-end analysis, but they actually drive a lot more sequencing as well, right? So to me, that's not an or, that's an and. And, and, you know, honestly, we have started, we've pivoted into that direction. It's been more than 2-3 years now, so you should see some of that come through, as we get into 2024 and 2025, in some of what we bring to the market.
Awesome. Joydeep, Carrie, thanks so much for the time. Covered a lot of ground.
Thank you.
Thank you.
Appreciate it.