Waters Corporation (WAT) Earnings Call Transcript & Summary

Vijay Kumar, the Life Science Device Analyst here at Evercore. A pleasure to have with us the team from Waters. We have CEO, Dr. Udit Batra. And from Investor Relations, we have Caspar Tudor. Udit and Caspar, thanks for the time this morning.

Thank you for having us Vijay.

Fantastic. So I'll start with my customary question here on maybe trying to review on third quarter. It certainly came in better at the instruments. And most of the Street, I think, walked away feeling look, there's something different what we're seeing here with Waters versus the other tools players. So maybe just walk us through on 3Q. What came in better? What surprised you?

So firstly, thank you for having us, Vijay. Q3 came in ahead across the board. Right? If you just look at revenues across each and every customer segment, across all major geographies and of course, instruments came back to growth and recurring revenues, as I mentioned in our analyst call, are like Swiss clock. They're always mid- to high single digits, and that was no different for Q3. The most enduring thing, which I think everybody is latched onto is that instruments came back to growth, especially LC after 7 quarters of decline started to show growth. And especially in our large pharma customers, which is where we saw some nice growth with LC. So all in all, steady improvement through the year of the trends, right? So Q3 was 4% growth. And with improvement across each and every segment. And that sort of also gave us confidence to raise the Q4 guidance. So we're expecting Q4 to be roughly 6-ish percent at the midpoint on constant currency revenue growth. And again, a couple of words on that, and I'm sure you'd want to dig into it a bit more, but a couple of words on Q4. Look, we saw improvement in trends. So we said, look, let's raise the guidance a little bit here. But we equally are cautious in sort of keeping the ramp from Q3 to Q4, same as last year, right? So he said, look, let's wait for more data points before we start getting too excited. But feel very good about where we currently sit.

Fantastic. And just to give us some context, what was the sequential step-up, I think, mid-teens last year? And how does that compare to like normal seasonality?

So historically, go back for the last 15 years, Q3 to Q4 is a ramp of about 22%. Last year, we saw one of the lowest budget flushes in history, and so that was about 15%. Adjusted for days, we've said, let's just assume that, that is the case this year as well, right? So that's the underlying assumption.

Got you. Given pharma is acting a little bit better, when the budget flush typically, when do we see it? Shouldn't we have started seeing like late November, December? And are we seeing budget flush?

We don't talk about trends intra-quarter. Good try.

Maybe another stab at it post elections. No, it's elections, it's been topical, right? It's -- have you seen any change in customer behavior post elections?

Not to speak of, right? Lot of perturbations, right, a lot of sort of discussion of perturbations. But nothing from a customer perspective, the CapEx trend are the same. The budgeting is the same. The funnel strength is the same as we discussed. So nothing really post election that I would have said what we said in the Q3 call should be modifying.

Got you. And what is Waters exposure to NIH?

Less than 1%, right? So I think when you look at the various things that came out in the narrative and the election. So there's the NIH funding. So it's a de minimis impact for us. Second, there was a discussion on tariffs, which I suspect is the next sort of big question that people ask. From a tariff perspective, the Trump era tariffs from the last administration are still intact, right? And so we were able to offset them. That was a low to mid-single-digit impact on the P&L, especially as far as China is concerned. And so when you look at the China tariffs, in particular, even if the tariffs went up to 80% to 90%, we don't manufacture anything in China for export. So it would be components, and that's another low to mid-single-digit impact, right? So you're not expecting a significant impact that we can't offset in the P&L. Then if you look at sort of the purported tariffs with China and Mexico, we don't manufacture anything in -- sorry, in Canada and Mexico, we don't manufacture anything in Canada and Mexico. So there's no impact there. And I think the only meaningful impact that could come is with Europe, right? We do manufacture our mass spec and our columns in Europe. When that happens, we've already started to look at alternate manufacturing routes, right? So we could offshore -- we could move some manufacturing to the U.S. We could move some manufacturing to Singapore and be able to offset any impact that you would see in the U.K. or Ireland. So all in all, a lot of discussion forced us to do a whole bunch of scenario planning, but it's early days. There's no clear legislation at this stage.

That's helpful, Udit. And sorry, just China tariff, you said the impact, if it does go to 60%, it's low single-digits impact to [ EPS ].

Low to mid-single-digit millions impact on EBIT. right? And nothing we shouldn't be able to offset in the P&L ourselves, right? So that's not something that we ought to be too worried about.

Understood. On China, Q-on-Q, China revenues were stable. Has China bottomed out? What are you hearing from the ground?

Yes. So I think worthwhile going back a little bit, right? Q1 was down 30%, Q2 was down 14%, 15%. Q3 sales were down sort of mid-single digits, but orders were in line with last year itself, right? So yes, we're starting to see improvement. Some of that has to do with also the baseline improvement, right? So the baseline is much lower in the half of the year in China. So we start to see an improvement overall. Now what's more important is over the last 2 years, we've seen significant weakness in China across different customer segments starting in pharma and CDMOs moving more towards the industrial side, moving more towards academia given the baseline impact, right? So that has all sort of flushed out over the last year as well. And in that 2-year time frame as the weakness sort of permeated through, we took the opportunity to restructure our field force. We expanded our distribution quite dramatically. We also localized a whole bunch of our portfolio. So bulk of our instrument portfolio is now localized. Now this puts us in a very good place as the stimulus starts to get rolled out. And we're already seeing some impact of that, as I said earlier, low single-digit impact in already -- orders already in this year itself, maybe sales get realized later this year or early next year. But we're starting to see activity province by province, 2 provinces are further ahead. There are other provinces that are starting to think through how they're going to roll out the tender. So we feel very well positioned as stimulus starts to come in. So yes, you could say China has stabilized and we're starting to see the impact of the stimulus roll through.

Got you. And on stimulus, is there some historical analogy on what it means for Waters when you have these kinds of programs?

Yes, I think this one is a little bit -- so the previous stimulus was largely focused on high-end instruments in academia and government segments, right, especially in academia. And if you go visit those universities now what you find is they're chock full of the high-end instruments, some of which were ours, right? So they don't need any more high-end instruments. This stimulus is now targeted towards Tier 2, Tier 3, academic institutions, towards government institutions, and then some industrial segments. It's a longer duration, and it's a wider stimulus. It's roughly 3x in size. So we feel very well positioned, as I said, Vijay, to capitalize on what we're already seeing. And there's a ton of activity on the ground.

Got you. So we should start seeing at some point in fiscal '25 stimulus. Would that be like a reasonable?

Yes. I think throughout the year, right? I mean there's -- as I said, a couple of provinces are further ahead. and you'll start to see some of that order impact already come in, in Q1, but sales will likely get consummated later in the year, right? Or even some will move even to the year after, right? But orders will start to come through already.

Got you. So from a guide standpoint, I think -- actually, not a guide standpoint, excuse me, China. What is China normalized growth for Waters, right? Is that still like a high single double-digit growth market?

Yes. I think China -- again, some historical context, right? So from 2005 to 2019, China added about 100 basis points to Waters growth until 2019. From 2019 to '22, the CAGR for China was about 9%, but the rest of the world was also 9%. This is when we started our transformation. So China was no longer accretive. And '23 to '24, China has been dilutive. We expect I think it would be safe to assume low single-digit, mid-single-digit growth as a starting point without stimulus is a reasonable assumption as you look ahead. And as the stimulus comes in, that adds on to the top. Now what is more important to realize from a Waters perspective is that we have several other drivers that more than offset any impact of China, India being one, GLP-1s, PFAS and we can go on and talk about those as well. But from a China perspective, I think safe to assume low to mid-single-digit growth, but then on top, the stimulus, which is not something that we can, of course, predict the timing of.

That's helpful, Udit. The -- since you brought up India, I think the Street perceives that as a differentiated angle for Waters. What differentiates Waters, right? Why is Waters unique when it comes to positioning in India?

I think long history in India, right? And over the last 3 to 4 years, what you've seen is outsized growth, right? So India went from roughly 5% of our overall sales almost 3 years ago to now 8% of our sales. Year-to-date, India has grown in the mid-20s, right? So Q3 was 26% growth. And this is driven largely by generics for export, right, where Waters has an outsized presence. We've always had it. And as manufacturing got repatriated to India, as these companies started to benefit from the patent cliffs that from the past, they've started to grow nicely and rapidly. And we expect that to continue, right, based on two things. One -- or three things: one, our outsized presence, right? Two, the sheer mount of amount of genericization over the next 5 years is pretty significant, roughly USD 230 billion, USD 240 billion, about half of it, more than half of -- slightly more than half of it is small molecules -- likely going to India. Then finally, the third thing is sort of the idiosyncratic driver of GLP-1, semaglutides are going off patent in 2026 in India -- Indian generic companies are already preparing in 2025 to start building up capacity in preparation. And there, we have an outsized position. already with the 2 largest GLP-1 manufacturers, right? So that helps. So we feel very well positioned with our commercial excellence in India with the historic strength, plus now what we're seeing from a generic's perspective and GLP-1's perspective.

That is helpful color, Udit. So this India trends, it seems like it's sustainable here. Any -- when you look at GLP-1s is one you brought up the small molecules, like outside of that, is there any other opportunity for you when you look at India more from -- not perhaps from a top line perspective, but perhaps from a back office perspective.

Sure. Sure. So unlike many other companies, Waters had not used a capability center, right? So we used to do a lot of our IT work with individual vendors. And roughly 3 years ago, we made the decision to open up a center in India, firstly, to in-source all of our IT spend and then over time, use it as an offshore facility for marketing communications for procurement work, et cetera, right? So we've set up a center in Bangalore, roughly 300-or-so people already, and it's functioning extremely well. Now that, of course, impacts the efficiency in the company, and it also allows us to get sort of incredible software talent that helps us on the informatics side. So India, not just from a commercial perspective, back office perspective. And eventually, as things develop as the biosimilars industry evolves, and there are some players who are working pretty hard to come up with their own biosimilars that will get approved in the U.S. as the biosimilar industry evolves, we intend to create technology centers in India, like we have in the U.S. and in Europe.

Got you. And is there like a margin play as you move some of these functions?

Yes. So I think, again, to give some context, I mean, Waters' margins are already one of the highest in the industry. And this is not because we've launched massive efficiency initiatives, right? Our business model is rather simple. When you think of Waters, think of a company that is in high-volume compliant settings where the regulatory barriers create massive switching costs, right? So high volume, once you're in, the customer -- and then you don't mess too much with the customer, you're able to sort of simply have a repeat business, be it with instruments, which is a replacement business in itself over a 7- to 10-year cycle, be it with our columns, which is again, a replacement business almost every year, many, many times every year, informatics is a replacement business and the service business is quite repeatable. So a very simple business model, which calls for a low SG&A, right? So hence, the margins are pretty high. Now we've not launched efficiency initiatives in the past, right? And so over the last 2, 3 years, with Amol, our CFO, and his team, we basically came up with initiatives that will add roughly 300 basis points over the next 6 to 7 years. And you say, well, okay, that should all come to the bottom line, not so fast, right? So now to sort of set up the overall algorithm, if Waters grows roughly 5%, we expect 50 basis points from efficiency and 50 basis points from leverage. We've invested 70 to 80 basis points in organic initiatives to support our high-growth adjacencies, which we expect to start seeing the benefit of over the next 2 to 3 years, right? And we've already started to see some benefit there. So if you just fast forward that algorithm, if the growth increases and the efficiency keeps coming, you can take that algorithm forward and you can keep building on it, right? So 20 to 30 basis points of margin expansion as a baseline every year is pretty reasonable, right? And the GCC, the Global Capability Center is part of that efficiency push.

Got you. And I know at the Analyst Day, like you had laid out these margin ramp, right? And that investment -- are we still in investment phase when we think about reinvestments?

I think you take -- again, take a step back, 4 years ago almost now, we said we have this simple business model. We want to take it to higher growth adjacencies like bioseparations, like bioanalytical characterization, like clinical diagnostics with mass spec and like battery testing. All of these have that simple business model of instruments that get replaced, chemistry that is sort of relevant for that sort of molecule, service that is critical and informatics that is key, right? So I just want to take each of these in turn for a minute and then talk about further investment required. For bioseparations in particular, this is our chemistry business, right? It's a little bit more than the chemistry business. But in the chemistry business itself, we've increased our R&D spend to now roughly 70% focusing on biologics and large molecules. And that has benefited us a lot, right? An exhibit 1 there is our MaxPeak Premier Column, where basically the technology was developed in particular to reduce adherence and adhesion of large molecules to different services to create inert services. And you'd say, well, why is that important? Customers usually take 18 hours or so to saturate the column before they run their experiments. And now with the MaxPeak Premier Column, you can run the experiment immediately. So significant benefit once you identify the unmet need for the customer, right? . And that has now shown up in the results, right? So MaxPeak Premier has grown roughly 40% a year. So very deliberate investment in a specific unmet need that has led to a benefit. Take bioanalytical case in point here is the acquisition of Wyatt, right, to add to armamentarium of bioanalytical instruments. Clinical has gone from roughly a low single-digit grower to a double-digit grower as we allocated more resources and battery testing continues to grow nicely, right? So each of these areas has benefited from our organic and inorganic focus. Now we expect that to continue for a little while. And once you sort of invested in the right parts organically and inorganically, we expect that to diminish. And yes, the margin could expand more than the 20 to 30 basis points over the midterm. And the ambition should be in the mid-30s, right, and not in the low 30s.

Got you.

So you can see line of sight. But what I also want to say is if we find that there are good opportunities like MaxPeak Premier, like Wyatt or like our clinical business, we will be very transparent, and we will invest because I think that's the best thing for the company.

Got you. And you brought up some of the other growth drivers, PFAS and GLP-1s. Is the algorithm each of those buckets are now, I think you've given a 30, 40 basis points contribution. How conservative are those numbers just given some of the numbers we hear on.

Early days, right? So I think you can do the estimate for GLP-1s as a case in point, top down, roughly 3% of overall volume in pharma will be GLP-1s conservatively. If you look at the math for the 30 basis points that we've been public about, that is roughly 1/3 of that amount. right? So yes, we have room to overachieve, yes. Now what's more interesting is the qualitative part behind it, right? F for GLP-1s, we have -- our columns are [ specked ] in for use for the 2 GLP-1 leaders. Now you say, well, why is that critical going forward? Of course, you -- for each batch that you run, you use columns. And so that becomes a nice annuity. But equally, as the Indian manufacturers come online for generics where the volume will be substantially higher, they basically look at originators and columns are rarely changed, right? So that's an added benefit. For instruments, every sort of large pharma company usually has 2 to 3 vendors qualified. And in our case, it's roughly a 60-40 versus us and another competitor for both of those players. And in the case of GLP-1s, the volume is so high that the 2 large GLP-1 manufacturers want to do in-process testing, and they use our patrol system where we have 100% share as well. They want to do in-process testing to alleviate downstream QA/QC needs and optimize the process to get better and better yield, right? And again, the secret sauce there is Empower, our software, right? So the building blocks for GLP-1s are pretty robust, right? And that's why the 30 basis points we were pretty comfortable in committing to.

That's helpful. And on PFAS, again, it looks like you guys are a market leader out there. Could there be upside here from a regulatory standpoint as you get more testing?

So early days, right? The current growth is largely coming from water testing as well as academia and public health laboratories coming up with testing protocols, right? And PFAS is not just instruments. PFAS is instruments, sample prep, software and service, the same business model that we have in late-stage pharma. It's a high-volume testing regime. Now with these end markets, so you have public health institutions that are standardizing on our Xevo TQ Absolute given it's the most sensitive instrument. The EPA is standardized on it. Other institutions like independent testing labs start following their lead, right? And so currently, what we've quantified is the water testing end market. As this moves into food, into consumer testing, into original manufacturers who need to remediate their products to remove PFAS, the market gets larger and larger. Now I would be a bit cautious, right? If you start doing the arithmetic that way, you come up with absurd numbers, right? And the 40 basis points on that basis is pretty conservative. But I would equally argue that economically, there has to be a different solution. You can't keep buying a $300,000, $400,000 instrument for high-volume use, right? You have to find more economically viable solutions. And we will come up with those in collaboration with our customers. There might be more point of use testing. There might be more centralized testing with a lot more automation. So this area will evolve. So I would not -- I would be cautious in just starting to build the market size based on bottom-up numbers and then saying, well, it's going to be -- the penetration that you see in water testing will be the penetration in each and every segment. I mean the countries will go bankrupt that way.

Understood. Understood. So the current PFAS, I think that 30, 40 basis points assumes water, but there could be some regulation on the food side.

That's right. On the food side, on the manufacturers themselves to remediate, additional testing by academic institutions to identify what other parts of our ecosystem contain these fluorocarbons.

And I know a few years ago, battery testing was all the range. Is battery testing still growing for Waters?

Absolutely. I think just from a small base, right? It's not where pharma QA/QC is, it's not where clinical is, and it's not where PFAS testing is, but it's surely heading in that direction. Just look at the sheer volume of electric vehicles that are being used, even though the market is slowing down, the penetration of QC test is still in its nascent stages. And let me give you a couple of examples. Our battery thermal recycler is used for detecting early signatures of runaway reactions in batteries, right? So we work with CATL in China to get that as an in-process tool. And the next step would be to get it into QA/QC. The second product that I'd highlight is our Rheo-IS rheometer, right? So very often, you have to test the flow properties of slurries that are used to make anodes and cathodes for battery testing. If you remember your electrochemistry from high school, there's an anode and a cathode, each of these is made by sputtering on coatings onto them of different types of metallic species. And they're processed in liquid, but you have to measure the thermal properties as well as physical properties of that material. With our Rheo-IS system, you can do both of them. You can measure the thermal properties as well as the flow properties of these species -- of these liquids, and that's been a pretty nice innovation. So the market is developing. I expect the penetration to start rising. And there is no reason why the regulatory constraints in that segment should be any different than what you see in food or environmental testing.

Got you. Got you. So maybe perhaps next Analyst Day, we might have a 30, 40 basis points from battery testing.

Too early. Too early.

Understood. Maybe switching gears to pharma, Udit. It's the largest piece of business for you guys and a lot of debate around large versus small. I think one of your peers brought up early customers on a recent call. So have you seen any change in large versus small? What is your exposure to CRO, CDMO?

Yes. So I think 2 different questions, right? So let's tackle the customer piece first, and then let's tackle the portfolio piece second, right? On the customer side, 60% of our business is pharma, right? And I think the highest level message you should take away is about 75% of our business in pharma is in late-stage high-volume testing, right? And this -- and let me now sort of break that down a little bit. So if you take pharma, 25-ish percent is large pharma and 80% of that 25% is late-stage development in QA/QC. Another 23% to 25% is generics, right? So roughly 50% is levered -- 45% to 50% is levered towards marketed compounds and late-stage development in pharma. The third segment is CXOs, which includes both CDMOs and CROs. CDMOs, the contract development and manufacturing organizations are roughly 70% of that business, right? And they follow exactly the same trends as marketed compounds for large pharma in the large part. CROs, which is roughly 1/3 of that 25-ish percent is basically a slower grower, right, because some of that work comes from early stage as well and not just late-stage marketed compounds. And then that leaves us with biotech. That's about 10%, 12% of our overall pharma business. And bulk of that, roughly 70-ish percent is commercial biotech, which again behaves like large pharma, right? And then finally, the balance of the pharma business, about 12-ish percent is clinical, right? So now sort of I've given you a lot of facts to again bring it back home, 75% is QA/QC, late-stage development, generics and marketed biotech, right? So this is all a function of the volume of consumption of molecules that are already in the market. right? And you can see that when you look at our recurring revenues, right? And this is where the divergence from many of our peers comes. Our recurring revenues are truly recurring. They are basically low to mid-single digits in a tough time and mid- to high single digits in a good time, but they're like a Swiss clock, right? And you can see that, especially in 2024 when early-stage recurring revenues have gone down, we've sort of been resilient. Biotech, especially pre-commercial biotech is still under pressure, even though funding has improved, CapEx has not sort of been relieved. Pharma research between these 2 segments is roughly 10% of our pharma business. Pharma research is also under pressure due to IRA, right? People have restructured and you've seen a lot of announcements on restructuring from large pharma, but that's largely focused on pharma research and did see funding come down there. So CapEx has not been relieved. And finally, on clinical, we're seeing high single digit growth there in that segment ever since we separated it. So that gives you a flavor of the customer view, right? From a product view, roughly 30% -- 35% of our business now is biologics related. This is up from about 20%, about 3, 3.5 years ago. And when you look at the growth drivers in both of them, of course, in the mid- to long term, almost 50% of the pharma pipeline is biologics and novel modalities. So that proportion is going to continue to rise, but that's been a very deliberate effort, right? As I mentioned earlier, when we identified bioseparations as a growth adjacency, we basically disproportionately invested R&D dollars in that space. 70% of our R&D spend in bioseparations is towards large molecules, and that's been showing a lot of benefit, be it in GLP-1s, be it with the MaxPeak Premier Columns. So I hope that gives you sort of an overall architecture of our Pharma business. And I think the #1 takeaway is we are late stage, less fluctuating really proportional to the volume of consumption of biologics or small molecules.

Got you. In this concept of replacement cycle that comes up a lot. Maybe what visibility does Waters have with it? And what kind of customer conversations are you having? What gives you the confidence that customers are beginning to start this cycle?

I think two sort of quantitative aspects and one qualitative, right? The first quantitative aspect is historical, right? This is -- as you use these instruments, you have to replace them because parts are no longer available. Instruments have been used too long and you start having too many errors come out of the instruments, so you have to replace them over a certain period of time, be it 7, 8, 9, 10 years depending upon what type of instrument you're talking about, right? And if you just sort of project that to current set of facts, we've seen a negative trend in LC for about 7 quarters. And Q3, we started to see it go up. Historically, when that happens, you see a prolonged period of positive growth, roughly 2 to 3 years once you sort of come out again in the growth domain. So that's the first sort of arithmetic historical statistical analysis that one can do. The second is just looking at our funnels. Quantitatively, we look at orders from customers and the conversion rate. What is interesting is in the last 3 quarters, the conversion rate has become much more predictable, right? So when I first talked to you in March, I said, "Hey, guys, this funnel looks very different. We think the conversion rates are going to be higher, the type of conversations we're having make us believe that the conversion rates that show up in June sales will be way higher than we've seen in the past. " Exactly what happened. Same thing happened from June to September. So you have 2 data points that give you a very clear indication that the conversion rates have improved versus what we saw last year, which means that customers have more predictability on their processes, on their CapEx, more visibility on it and so do we. So those are 2 mathematical features that are different. The qualitative feature that's different is the nature of conversations you have when you are discussing the replacement of 1 instrument or 2 instruments where you can have a bilateral discussion with a lab head who's got approval from their boss to say, hey, I'm going to get about $60,000 or $100,000 or $200,000 to buy 2 instruments. Can we have a conversation? You have a conversation with an account leader and a service rep and well, you can place that order. But when you're looking at multiyear replacement cycles, there are several factors to consider, right? The CapEx amount -- so you're looking at hundreds of instruments sometimes, right? Take a company like Pfizer, so it would be a much longer period. The duration is longer, the number of instruments is longer, the composition has to be decided. So is it how many Alliance iS, what traction is going to be Alliance iS? What traction is going to be Arc HPLC? What traction is going to be Alliance. That has to be determined. You have to talk to the IT folks to see what version of Empower you have because this is an enterprise-level software and you're going to insert about 100 instruments into that, right? So you need IT to be talking to you and make sure that's seamless because it's a compliant software. So a whole bunch of folks are involved. And a meeting usually in that scenario consists of roughly 20 people, right? The last one I attended was in Indianapolis and you had the Head of Finance, you had the Head of Procurement. You had the Head of IT. You had the Lab Head, and God knows how many other folks that I couldn't identify. But these were at least 4 departments that I could easily identify. On our side, we had our service team, who were going to be responsible to install these over a multiyear process. We had our IT department. We had our informatics team which sits outside of IT to say, hey, this is the version of Empower. These are the changes. If you use Alliance iS, this is what's going to happen. If you use Alliance, this is what's going to happen, et cetera, et cetera. So you have a set of cross-functional teams. The nature of those discussions is completely different, right? And I think one thing that's worth clarifying here is, is the paradox. And again sort of very often we get stuck in generalities in pharma spending. Pharma spending and research is completely different in pharma spending in QA/QC and manufacturing. Companies that are restructuring research are equally spending at the same time in QA/QC. Some of them in the news are saying, look, we're restructuring AB and C because we have a runoff of COVID revenues, et cetera. Those same companies are having conversations with us on multiyear replacement cycles, right? I won't talk more about the specifics, but you will read about them as they become much more public. But those conversations are happening now. And as those mature, we then have visibility on what's going to happen with the replacement cycle and what resources we need in place. So very sort of very different 2 mathematical and 1 sort of qualitative descriptor that gives us confidence that we're in the midst of a replacement cycle. But don't ask me exactly the takeoff point, right? So the takeoff velocity. I can tell you the area under the curve. You'll see 2 to 3 years based on history, 2 to 3 years of outsized instrument growth versus history. I mean that's what history tells us. But in terms of the takeoff velocity, I don't know. I mean, it ranges anywhere between 0.5% to 11% when you come out of the negative territory in terms of growth.

That's helpful. And since you brought up Alliance iS, why -- what is different about Alliance iS?

Fantastic instrument, right? I mean, let me sort of give a plug for our R&D teams. Waters is nothing if we don't attack -- if we're not attacking difficult problems for our customers. right? And that's become totally apparent to me in the last 3 to 4 years. We don't -- we're not a good follower. We're a very good leader in creating categories, be it UPLC, be it now with the Alliance iS, be it with Xevo TQ Absolute, or be it with MaxPeak. We're setting standards of innovation. And that's what our team loves to do. It can be deeply frustrating because people are focused on technology and not as much on market penetration, and we've started to correct that over the last few years. Now to come back to the story of Alliance iS, this is a product that's been in development for close to a decade, right? It's a complete renewal of the HPLC experience. And it has just one sharp unmet need that is trying to satisfy. Our QA/QC customers get extremely frustrated when they see errors, especially when they see errors that are avoidable. And by that, I mean, they put in the wrong vial for testing or they've used a wrong solvent or they've used a wrong column and they've tested something overnight. The results come out and they say, oh shit, this is useless. Now the story doesn't end there. They have to report it to their lab head and having buddies who are now in analytical R&D. So I started my career at Merck Research Labs in West Point, PA and some of my buddies now 25 years later are heads of these labs, right? And so they get super frustrated with their teams when a change request has to be filed to them and then to the FDA to say, "Hey, this is a market product, this error, please ignore because this error occurred due to an avoidable mistake. " Those are the most frustrating ones for them. Alliance iS focuses on that problem, and it reduces those errors by 40%, a value proposition that has resonated deeply with our customers, right? And they've had a chance to trial it. They've been on a test drive for 18 months, right? Because nobody had CapEx, so they basically said, hey, give us a beta instrument, and it was beneficial for us as well, if I'm totally honest because when you launch a new product like this, especially during a pandemic where you had to sort of procure so many different things, you want to make sure that it works in their hands in QC and it doesn't fail in QC. So they've been trialing for 18 years, given us feedback. We've improved the software. We've improved the experience. No instrument in QC has a massive uptick. This is not Xevo TQ Absolute, where things have gone from 0 to 100 in a year. It takes that much time, right? So it's taken us -- we've had 18 months to get it sort of debugged and customers used to it. And now as they're initiating their replacement cycle, we are seeing a larger proportion of Alliance iS than we would have seen historically given the life cycle of the product, right? So roughly 10% to 20% of the LC replacement is likely to be Alliance iS. And from a value perspective, that's a significant price premium, right? And so very happy with what's happened, but it's a deliberate process that's taken close to a decade. And if somebody says, you know what, we can retrofit something and come up with a solution. I think they're kidding themselves. This is a very conservative customer set that needs time to get a new product penetrated.

Got you. That's helpful color on Alliance iS. Since you brought up area under the curve is -- I think you've said instrument CAGR versus '19 as low singles, maybe 1% is the area under the curve, we just have to do a 5% CAGR, and that's the math and what the opportunity.

Yes. I think I would do something like that, right? I would basically say, look, I would draw a line, say it's 5%. I try and figure out how far under we were. And I would say, what does it take to catch up and spread it out over a 2- to 3-year period, and there you have it, right? It's not tricky and there are some few scenarios you can do around it, but it's not tricky. The area under the curve is the area under the curve, you'll have to catch up to 5%. But what is more important is the new drivers on top of it. right? The 5% will come like clockwork, right? There, we can almost guarantee it because people have to replace instruments. But what is more interesting is you have GLP-1s now, which is an additional driver. You have PFAS, which is an additional driver. You have the India generics penetration in that market that's opened up very differently than the past. And of course, the headwind is China that you can offset. And then finally, you have pricing, which is better largely due to the fact that we have new products that are differentiated in the market, plus the operational excellence that the team has built during this tough time in the last 2 years.

That's helpful, Udit. On pricing, just given the change in business mix, what is the go-forward pricing? Is pricing plus 50 to 100 go forward?

Yes. I think historically, we have done 50 basis points. In the future, you should expect at least 150 basis points, so 100 basis points incremental. And why do I say that with so much confidence, in '23, we were able to do it, right? When things slowed down in '24 as the volume continues -- volume has continued to be challenging, we've still been able to pass on pricing, and we're still between 150 to 200 basis points year-to-date, right? So that, I think, you can be reasonably confident that, that's going to happen largely on the back of new products and the operational excellence we've already seen with the team. Equally, what's important is to see what happens when you pass on price to your gross margin and your operating margin, right? I mean everybody talks a good game on pricing and people did in 2023. I think we were one of the few companies that expanded the operating margin and the price fell all the way through and not -- did not get spent in the P&L.

Got you. And obviously, a lot of positives here, Udit. The 2 things we haven't touched about. One is share gains. Is Waters gaining share versus competition? And two, are there any headwinds we need to be aware of for fiscal '24. Is it industrial?

I think the share gain piece is always tricky to sort of pin down, right? When I joined the company, I was probably the first CEO in this sector to come and say we've lost share, right? And I calculated it based on the fact that we were not replacing instruments but others were, and we were falling behind, and we had not introduced new products in a while. So it was a pretty straightforward math. Now equally, when I sit on the other side now, I'm reluctant to say there is massive share gain in replacement business. In the replacement business, which is 70% of our instruments business, customers are highly conservative, unless you really abuse them, they will not change from one vendor to the other, right? And we benefited from that when I came into the company, and we were able to see a nice growth because we were catching up on replacement that we had missed. Now there is an advantage to having introduced products sooner than others in a category which benefits from replacement. So yes, you might see some one-offs where we have an increase in share even in replacement, but 30% of the business that is new every year, new instruments, that's for sure, an area where we've gained share, right, be it with the Xevo TQ Absolute, be it with the Alliance iS, be it with the Xevo G3 QTof in the mass spec space, or with the MRT now, right? So we feel extremely good about where we are with our portfolio. And in terms of headwinds, I think I feel very good about where we sit now given the fact that we have come from 2 years of serious austerity on the instrument side. But equally, I mean, there are variables, right? I mean, it's going to -- while the replacement cycle is underway, you want to look for more data points before you have confidence. While the China stimulus is underway, you want to look for more data points until it's underway. And then there are perturbations, like we saw with the U.S. election and those things have to emerge. Right? And those are the uncertainties that we don't necessarily control. What we control is good innovation and solid execution and that you'll continue to see.

Fantastic. With that, I think we're out of time. Udit, thank you so much for the time this morning.

Super. Always a pleasure. Thank you.