Okay, thanks everyone for joining us this afternoon. A pleasure to have with us at Waters this afternoon. We have Dr. Udit Batra, President and CEO, with us. And I think in the background, perhaps Caspar Tudor from Investor Relations is in the background. Udit, thank you for taking the time to be with us this afternoon. Good afternoon, Vijay. Great to be with you. Maybe I want to start with, you know, most tools companies have reported third quarter by now. By and large across the group, really strong growth. I think, you know, and it's really hard to stand apart when everyone's beating numbers, but you guys really stood out. 15% organic was really, really strong. Sometimes it's hard for me, like when I look at the fourth quarter guide. I think 6%-8%, you know, we just did 15% and stepping back to 6%-8% seems a bit conservative. Maybe if you could just talk about what you're seeing on the macro, year-end budget flush, you know, China versus Europe, some context here heading into Q4. Yeah, yeah. Firstly, thanks for the question, Vijay. I definitely expected that as a question. Look, firstly, the 15% growth, right? I just want to calibrate, make a few calibrating comments upfront. We're very happy with what we're seeing. There's no question about it. It's nothing short of really, really astonishing sort of performance and a turnaround from the team. So very happy with it, very thankful for it. And we know there are macro challenges out there, but just to sort of level set on the growth over the last couple of years, right? I mean, last year we finished at 15, year to date we were at 14%, and this quarter was 15. So really nice sort of territory. But if you look on a three-year stack basis, the growth is high single digits to low double digits, right? 9%-10%, right? I think that's a more reasonable way to look at the overall environment and the business, right? Everybody's printing good numbers, but let's look at it on a three-year basis. On a three-year basis and a full year, let's not sort of take a quarter here and there because it's so turbulent. We're close to 10%, okay? Historically, the tools industry is what, between 4% and 6%, right? You add 300 basis points of pricing to it, that makes it 7%-9%, okay? I mean, that's what we're seeing. So the 10% or the numbers that you're seeing on a stack basis from the whole competitive set in the analytical instrument space are not that far away from the historical numbers if you add the pricing to it, right? Now, so if you just look at the various prints and we do our homework as well on that front and you've looked at it, on a stack basis, Waters does seem to sort of be at the top of the podium, right? And there are a few others who are between 7%-9%, which is where the industry is. That lead of 100 basis points or 200 basis points versus the industry is easily explained by the commercial pieces and the innovation that we've talked about, right? So nothing dramatic. Now, I want to use that as a jumping-off point for the 6%-8%, which you say, hey, why 6%-8% and why not 15% again? First, I'll remind you that there was what, $13 million-$14 million that moved from Q3 to Q4. So that raises the base. And so that makes it what, 7.5%-9% instead of 6%-8% if you adjust for the base. Second, on a two-year stack basis, the baseline is still pretty high, right? So we're looking at a Q4 where we have had sort of a higher baseline. And if you adjust for the baseline, 7.5%-9%. And then in that number, we also have, we've also assumed a year-end budget flush. And we are seeing signs of it, right? So, I mean, not to talk deeply about it, it's no different than other years, but that's still an assumption. A good portion of our business is in the last week of the year, right? And that's, I mean, really nerve-wracking for somebody who comes with a different sort of background. The last week of the year is nerve-wracking. I think with those sort of explanations, I think a 7.5%-9% on an adjusted basis is a pretty reasonable number. And I wouldn't even call it prudent. I would just call it reasonable. I love the qualitative context. It's reasonable. It's not prudent. But no, the way, thanks for explaining that walkthrough from industry of mid-singles and to that high single straight with pricing and your growth pillars, right? The five pillars, how do you get that bridge is helpful. Since you brought up pricing, you know, pricing was 350 basis points in third quarter. Should these pricing levels continue into fiscal 2023? How should we think of pricing in fiscal 2023? That's a really tough one, right? And there are probably three parts to it, right? One, and if I just think of sustainability of the pricing, right? One for Waters for sure, and I can speak about Waters. I mean, we built the muscle to sort of understand the inflationary impacts to our supply chain and pass on what we think is a reasonable price to our customers, right? And is it 300-350 basis points in an inflationary environment that seemed reasonable? That inflationary environment, if it sustains, then you should expect a commensurate increase in price. We don't know that yet, right? But we assume that there is some amount of inflation, but not to the same extent. And that assumption is sort of variable across different segments. Second, as we have gone through this process, we are working on problems that are actually adding a ton of value for our customers, right? And we're no longer just saying, hey, you know, we're going to go replace an instrument for something that you've done for 10 years. We're going in and saying, guys, there's a problem with cell and gene therapy COGS. If we analyze these filled versus empty capsids, or if we do online and at-line testing, or if we help you to select better clones or better raw materials, you will save a significant amount in your COGS, maybe even get drugs to market faster. That is worth more than 50 basis points. So we have an argument to make to claim a higher price, right? Those are two factors that argue positively to say, well, you know, maybe it's not 350 basis points, maybe it's a bit lesser, but it's still, they argue positively that you're adding value and there's an inflationary environment. But the third piece is habit. The customers are not used to the tools industry passing on more than 50 basis points, and that will be a negotiation. That will be a discussion. I can't give you an exact answer, but those are three variables we have in mind, and those are deeply sort of embedded in the minds of our sales teams, in the minds of our general managers to really think it through where we're adding a lot of value. We're not afraid to claim additional pricing where we have inflation. We have systems to show customers, hey guys, this is what we are, this is what we're seeing, and it's a fair pass on to you. But the third piece is sort of the industry norm, which is not easy to change. Understood. No, that's an extremely helpful perspective. And just maybe pricing contribution in current fiscal 2022, can you give us a sense on how that pricing has progressed throughout the year? Did we start Q1 at a lower base and then has it stepped up throughout the year? Yeah, I mean, I think you would have seen with individual prints, 300, slightly less than 300 basis points, 300, 350, and largely because the orders that came later had a higher price built into them, hence the higher realization in sales. So it has built up. So you probably see a little bit of a similar impact going forward, but not for a long period of time, right? I mean, it's not like we have 12 months of orders sitting there that we have negotiated the price on and we are completely sure what that's going to be. That's not how it works. Understood. And since you brought up orders, right, I think that segues nicely into this instrumentation question. You know, 20%, 1% growth in third quarter. And I think your peer had like a 30% growth in LC-MS. Like these are really, really big numbers, would it? And again, like we all struggle with, is this pent-up demand, pull forward of demand? And I think you've explained in the past, look, this is not, you don't stock these instruments, right? So just explain to us like why are these end markets so strong? Because it doesn't seem like it's share gains. Everyone's growing pretty fast. But again, it helps to sort of look at the facts first, right? I mean, so, and even for instruments, I mean, instruments for Waters on a, and I mean, if you look at quarter by quarter, some quarters we are at 20%, other quarters we are at 21%, another quarter we are at 19%. Our peer group also varies quite a bit, right? And we look at each of our peers in the same space. The portfolios are similar, but not exactly the same across the peer group as well, right? So I think that's something one has to keep in mind. But put that aside for a minute. On a three-year stacked basis, the instrument growth for Waters is 10%. So use the same algorithm that we talked about earlier. 3%-4% is the general growth of instruments if you look at historic levels for Waters. At 300 basis points, that becomes 6%-7%, right? Assume, and now if you look at the reported numbers for our competitors, right? and only instruments, forget consumables for a minute, only instruments, those numbers range from 6%-9%, right? On a full year stacked basis over a three-year period, right? so Waters is again sort of on the bookend of it, right? and you can argue, is it exactly there? Is it a bit higher, a bit lower? I don't know, but it's pretty nice, right? Now, if you take that as a baseline and you say, okay, I understand that the logical growth of the industry is between 6% and 9%, that's not that far away from history if you've just added the pricing component, right? I think we should sort of move past the fact that this is euphoria and it's pent up and it's this and that. It's not that pent up, man. I mean, it's a 100-200 basis points difference in the parts. Now, versus the market, there are Waters-specific drivers. There's the commercial drivers. There's a replacement cycle, which of course we hadn't replaced our instruments and perhaps others had, but we hadn't, and we have that as a unique driver. We have as a unique driver our service attachment rates. We have as a unique driver on an instrument, from an instrument perspective, new customer segments like contract manufacturing where we hadn't had equal weight. So if you just add those up, that gives you a bit of the explanation for the cushion that we have versus the market. And second is new products. This I'm absolutely ecstatic about. Across the instrument portfolio in small molecule LC, large molecule LC, meaning UPLC, in different portions mass spec, high-res mass spec tandem quads for pharma, tandem quads for food and environmental, the QTofs and software across the board, we have new products and they're contributing meaningfully and they should continue to contribute, right? You can explain it logically. This is not, this is not magic, right? I mean, the numbers are heady and they're high, but if you just sort of calm down and say, okay, let's look at it at a three-year basis, let's account for the pricing, let's look at the overall market, do the math. When you do the math, the numbers are 6%-9% and Waters has a bit of a lead and here are the reasons why there is a lead. And then the bigger question is, does this sustain, which is I think your question under the question. And I would argue exactly the same drivers, right? I would say, okay, is the underlying demand 3%-4% volume demand? I think that's believable because there are sort of salient drivers for our business across the board. Is the pricing going to be 300 basis points? We just debated that, right? So you can keep that as a separate block. And then are the Waters-specific drivers going to sustain? And I would argue that we have confidence that they will, right? So I'm not saying that we are at 9%-10% instrument growth next year. You subtract pricing, it becomes a bit lesser. You add pricing, it becomes a bit more, but that's sort of the math, yeah? That's extremely helpful, Udit, and I know in the past, I guess when you look at the replacement cycle, the service attachment rate, innovation, CROs, right? I think for overall corporate, it was about 100 basis points incremental growth. Is that the right number? You know, when I think about on the instrumentation side, because I feel like it's been well above on this new product and innovation, you know, what its contribution has been on the instrumentation side. I mean, again, I take a three-year basis, right? I don't sort of take the middle years because of COVID and all the dynamics, right? On a three-year basis, it has been higher than 100 basis points. I think we can agree on that, including new products. Between new products and our commercial initiatives, the market for instruments is about 8-ish%. We are at 10%, right? So it's between 150-200 basis points. Now, don't ask me exactly what contributed what, but that's sort of the overall sort of number. I think it's a safe assumption to say next year again, if the market is X, we are X plus 100 basis points because of the commercial pieces that we still have line of sight on, right? And new products are also starting to contribute. So I think it's a similar assumption. It's been a bit higher, you're right. Now, I'm not brave enough to say it's going to be exactly that because new product adoption goes through ups and downs and it's a long-term game. But about 100 basis points is not a bad assumption, which is what we said at the JP Morgan conference last year and again this year, right? So. That's pretty remarkable, and in the midst of all these geopolitical challenges that you guys have delivered on the core drivers, I think one related to that, Udit, is how cyclical is pharma? I know pharma is acyclical, but your customers, when they're looking at these headlines, right, they do make purchasing decisions or maybe there's some cautiousness, right? Like how, I wouldn't call it how worried I should be. Maybe what's the best way to characterize how pharma customers behave in a cycle? So I mean, think of it in two dimensions, right? Pharma is not pharma is not pharma, right? It's not all the same, right? Think of it in two dimensions. Think of the value chain of pharma, right? And think of the type of customer, right? So the value chain of pharma is research, development, and let's say manufacturing. Waters is heavily focused on late-stage development and manufacturing or QA/QC, right? And that's the least cyclical, right? Irrespective of what's going on from a CapEx standpoint, people don't take away. And having been a pharma person myself, you don't sort of pull away funding from your only late-stage drug. You don't pull away funding from marketed compounds, right? So I think that's a safer real estate. Research, early-stage development, yes, can come under pressure if there are CapEx constraints. So that's one dimension to look at, right, for large pharma. But then there are different types of customers, right? So there's large pharma, there are smaller biotechs, and there are contract manufacturers. All three face different types of challenges. Large pharma we just went through. For smaller-stage biotechs, which are early-stage, focused on early-stage research, yes, that can be volatile, but that's less than 5% of our business, right? Contract manufacturers, yes, can be exposed to overcapacity and undercapacity. And we're seeing such a cycle now, right? A lot of euphoria, a lot of repatriation, a lot of capacity being put forth, and you're seeing sort of that cycle of building up of bricks and mortar. And that probably is going to slow down a little bit now, right? That said, again, our growth is proportional to the volume that the contract manufacturers are producing, not their bricks and mortar, right? It's not like they're purchasing a bioreactor and they're going to produce more mAbs. Whatever mAbs they produce and they have to test, that's what our business scales, right? So these are three, so that's sort of a two-by-two way of thinking about it. Let's think about the value chain and let's think about the type of pharma customers. And we feel, especially because we are more late-stage and development-focused, we are in a much sort of resilient real estate, even within pharma, right? And if you just look at the numbers, I mean, 12% year to date on a stacked base is also double digits, right? So I'm not worried about pharma. And even within pharma, our composition has changed, right? So we've gone from 80-20 small to large, now 70-30 small to large molecule. In the large molecule space, there are credible new products that are solving pretty significant problems, right? So don't ask me if this is going to become 40% or 60%, I don't know, but we are solving problems on the small molecule side and on the large molecule side for our customers. Gotcha. And sorry, when did it go from 20% to 30%, Udit? What was the timeframe? Over the last two years, right? We've seen a fairly significant uptake of products MaxPeak Premier Columns, the ACQUITY Premier instruments. From mass spec perspective, I mean, BioAccord has been talked about a lot. There are several other products on mass spec side that have led to that transition. Step by step, the applications for UPLC, the applications for our Columns, the applications for mass spec have been more and more focused on larger molecules. The pipeline, our pipeline is also focused more and more in that direction for good reason. Gotcha. And just because you brought up BioAccord and MaxPeak Premier Column, let's start with MaxPeak Column, right? I know on the calls you give a lot of details on the ability to look at oligonucleotides or higher sensitivity. What percentage of your customers have adopted MaxPeak Premier Column? What cycle are we, and are there any other new applications for that specific product that could be incremental? I can't give you specifics on that front. I can talk about the types of molecules where it's relevant, right? I mean, it was designed to reduce the binding of metal binding compounds to our Columns first and then to our instruments, right? It's a coating on the surface of our Columns and our instruments. Now, it basically reduces the experiment time from about 18 hours to right out of the box. Usually our customers said they need 18 hours or what is called passivation to sort of saturate all the binding surfaces that are present in these Columns, and then you could do the experiment so that the molecule elutes through the column, right? With MaxPeak, it doesn't stick out of the box, so you can do the experiment right away. It's a very significant benefit, right? And without having to reoptimize the method, and that's the most important thing, without having to reoptimize the method, many of the customers have transitioned to MaxPeak, right? It's easier to do it for newer molecules than for existing molecules, but we've seen a significant number of customers transition, right? And it's of course more relevant for larger molecules because they are more prone to binding to surfaces, but even small molecules are prone to binding to different surfaces at different conditions. So I can't give you more specifics than that. All I can tell you, it's been the best launch. Each time we raise the target for the team, they exceed it, right? So we haven't seen the end of the uptake of MaxPeak. That's fantastic. And then on BioAccord, similar question. I know it sounds at least from an external perspective that there's incremental data points, right, relative to BioAccord, whether it was the Sartorius announcement or I think AstraZeneca was cited recently. Are we at a tipping point on BioAccord now? Why is this different? For me, it's mass spec is mass spec, right? So what is special about BioAccord? Mass spec is not mass spec, doctor. Even I'm learning, right? So there are different types mass specs. there are mass specs, there are tandem mass specs, there are TOF instruments and different types of instruments for different users. But in a simple way, just to talk about BioAccord, right? BioAccord was designed for the QA/QC space, right? The target product profile was to come up with an LC-MS Workflow, LC-MS instrument that could do for large molecules what LC-UV does for small. So foolproof, simple to use. You put in a molecule on one side, on the other side you get its mass, but also you get its identity, but you also then get its mass and you can quantify the mass and the component and mass of the component. So you can get the ID and another critical quality attribute in one simple instrument, and you can do that with mass spec in research. But people said, "Why can't we do it in QA/QC? Because I'm testing 60 different things. Why can't I mass spec to substitute for 40 or 50 of those critical quality attributes? And I can trust it more, it's automated," etc., right? So that was the dream. Now, over the last two years, several customers have adopted or have tested mass spec for QA/QC. Some are ahead and some are a bit behind. So all top 20 pharma have actually tried it, right? There is at least one customer who has already qualified, and this is a big deal in QA/QC lingo, qualified the BioAccord for use in quality assurance and quality control. That has a whole bunch of different requirements, including software requirements, simplicity of use, robustness, da da da, and you have to have discussions with regulators on how, if you have actually qualified the instrument. So that's step one, and it's a very significant step. Then they want to use it for identification of the molecule. So is this HUMIRA or is this some other molecule? They can already do that. Then they say, "Well, how much HUMIRA and how many degradates of HUMIRA in my sample?" That's the second step. That's what's being done now, right? Sorry to take you into the weeds. All I'm trying to illustrate is that irrespective of your facility mass spec and your sophistication, this is a slow induction period, right? And a slow induction process. It's a long induction process for LC-MS and QA/QC, which is where we designed the product for. What's happened in the meantime over the last two years, we said, "Well, a simple mass spec," and several of our process engineering colleagues, including myself, said, "Well, I could use this in other applications. Why don't I use it for clone selection?" So we called up Sartorius, and they said, "Yep, great idea. Let's try it." And that's going really well. People saw it being used in clone selection. They said, "Well, why can't I use it for detecting mAbs and yields in perfusion?" This is what AstraZeneca did independently, right? And of course, collaborating with us, but independently said, "We want to do an online. We want to develop an online tester that is more sophisticated than we have in process development. That requires qualification, but in process development." Now, that of course is a step towards going into QA/QC. For clone selection and process development, high volume application, but engineers using it or physicians using it who have no deep understanding mass spec, but want a box that gives me five answers, and I don't have to wait six to eight weeks for it, so that application came up. A third application was in raw material testing that Janssen piloted. They basically said, "You know what? We're really worried about the variability in our raw materials, especially cell culture media. And we want a simple instrument that operators can use day in, day out to detect minute changes in cell culture media because minute changes can lead to significant variability in titers and significant variability in output," so they wanted to make sure that they were testing it, and they've qualified that process. What started off as a, and I would say arguably the toughest problem to solve, which is QA/QC, a simple instrument in QA/QC, has now morphed into two other types of applications. We're very optimistic about a simple LC-MS called BioAccord. And of course, that yields then a whole bunch of other things, right? That yields a focus on Columns, that yields a focus on better software, on better service, and other types of instruments and sensors that you can develop, right? You can see how rich that area is. That's why we struggle to quantify how big this whole space can be, right? I will remind you that I had told you that we would start quantifying it in 2024, and that has not stopped you from asking the question, "Give me a number sooner." So we will try to give you some sort of idea of the overall space in the early part of next year, but 2024 is when you will start to see sort of a meaningful impact. No, that's really helpful explanation, Udit, because we're trying to do the math, as you were saying, some ratio of bioreactor to BioAccord, but this is bigger, right? You're going upstream and there is downstream. It looks like the opportunity could be far bigger than how we thought about it. I think the issue is timing, right? So QA/QC takes a bit longer. It's a very conservative industry, and I come from the industry, right? So in my tenure, I would not want somebody to test a new instrument for a marketed product, right? It's a great, "I love you," but not that much, right? So it's newer in development. But then the other applications give us comfort that we will start to see uptake a bit sooner in those applications, right? So we've got, of course, our internal ramp-up curves that we keep testing left, right, and center and pushing our teams for more. But I'd say the QA/QC ramp-up is a bit longer, probably the largest market. The process development ramp-up is a bit faster because as a few process engineers get comfortable, the workflows get comfortable, they're sort of adopting, and you're saving them sort of six to eight weeks getting the product to market much faster. The adoption becomes faster, and raw materials should be rather straightforward if once we have qualified a few customers who are well-respected in using it, right, so I think that gives you sort of the input variables. Timing is the biggest topic, right? I mean, I think the end state, we can all calculate, but I think it's a ramp-up that one has to think through. That's extremely helpful. And then sticking on to this topic of innovation, mass spec up 40%. I know BioAccord goes into that, but beyond BioAccord, I know you have Cyclic and mass spec parameters, right? What's driving that strength? Is Is this against some order timing that benefited or anything happening from your customers? No, Vijay, I mean, it has a lot to do with Waters-specific topics. Of course, the demand is very good, but it also has to do with Waters-specific topics. Remember, I think two years ago, two years and change ago when I joined, a lot of folks came to me and said, "You guys are nowhere mass spec anymore. I mean, high-res, you should really exit." And I mean, and I said, "Look, we need some time to study this." And we talked a lot about LC. We didn't talk mass spec has, I mean, and in the meanwhile, thanks to Waters' technical focus, we never mass spec, right? I mean, we were innovating mass spec all along, right? We came up with BioAccord. The SYNAPT was there, but then Cyclic had already been launched before I got to the company. I think what we've done that is different. It's not that the pipeline didn't have all these instruments in it. The pipeline had these instruments. I think what is different is they're being launched with a completely different discipline. We're very focused on different segments, just like I explained with BioAccord, right? We said we had this segment, but there were two other segments we went after, right? With Cyclic also, right? We always knew that this is the only instrument in the industry that can distinguish shape as well as size of molecules and very relevant for spike proteins for the SARS-CoV-2 virus, for example, right? We knew the application. It just, we had not sort of built the commercial plan to say, "Okay, we'll go to these customers first, these one second, these third." That's very different. The pipeline is rich with new applications for these advanced instruments, right? So Cyclic is just one example in the high-res space. MRT is another one, which is even in an earlier stage. It has been launched. The applications are being developed. The software is being developed. And I think that's going to be a fantastic tool in the armory of people doing early-stage research in proteomics, right? Or other types of omics. But as you move into high-volume applications, I mean, each area has different types of innovations, right? So we've talked about the advancement with the Xevo TQ Absolute, right? The most sensitive instrument. And now the application for PFAS testing has come exactly at the same time. I mean, you can call that sort of a big coincidence. And it is exactly the right instrument at the right time for the right customer segment. And we're seeing massive back orders on that one, right? And we want to get clarity on demand as much as you do. You want to know, "Okay, so how big is this going to be?" I also want to know because I want to know how fast should we be producing, right? And right now, we can't produce fast enough, right? Then if you move on to other types mass spec, i mean, the Xevo G3 QTof, this is sort of the workhorse in development where you take each protein and you analyze with a significantly higher mass spec, faster than, but a bit less powerful than the Orbitrap, but much faster, right? And you can analyze a lot of attributes. The advancement there has been the software. We've put that instrument on waters_connect. Now you say, "Well, okay, so big deal." Huge deal because what you now do in development is easily transferable to the BioAccord in QA/QC. That's a significant advancement and very sort of logically thought through by the teams. Across mass spec portfolio, going from high-res with Cyclic and with MRT to the high-throughput tandem quads to the QToF in late-stage development or early-stage development to the Xevo TQ Absolute, you see sort of a renewal of mass spec portfolio. It's no mystery that there's an uptake. Yes, the demand in the market is also higher. There's no question. Again, I would go back to the same logic, right? It's a 10% growth, and others have grown similarly, probably a little bit lesser. You do the same algorithm. It's a 2%-3%, 3%-4% sort of grower. You add pricing. It's not that far away from these. That's extremely helpful, Udit. Just one on PFAS. Can you quickly remind us on, are there some regulatory changes that we need to be aware of for PFAS testing? Yeah. I mean, they're all in evolution, right? So the EPA is demanding much more sensitive testing for different types of fluorinated carbons, right? So earlier, they said, "You know what? It's a chain length of eight to 10 carbons," and then fluorine's attached to it, different amounts of fluorine attached to them. Now they're saying, "Well, even if there's a trace amount of fluorine with the smallest amount of carbon chain, I want to see it." So the demand for sensitivity has increased dramatically. And again, I would sort of segment it. There are three different segments for the application. One is water testing, right? So drinking water. And this is universal. This is global, right? So public health authorities in the U.S., public health authorities in Europe, and now in Asia are asking for the same application. Second is in food testing, right? We've seen testing for infant formula in parts of Asia. And I don't need to tell you why that part of the world is so sensitive to that segment. So in foods, people are looking for trace amounts of PFAS. And then finally, the third and probably a very significant segment is effluents from industrial customers. So the effluent of manufacturers in the industrial segment are being asked to sort of reduce the amount of PFAS that they are eluting to the point where the countries are asking them to shut down their plants if they can't meet those requirements. This is one of our big customers who is a US-based customer but has a plant in Scandinavia, right? So three different segments. And again, it's like the BioAccord where you have a very sensitive tool. You have a very sensitive requirement, but it's relevant for several different segments. That's extremely helpful. And then switching to sort of your non-BioPharma, the more cyclical part of the business, industrial, I think has 30% of Waters revenues. Can you talk about, in a compare and contrast versus 2008, 2009, what was the business mix then? Industrial as a percentage of revenues. And even within that segment, I think your exposure has changed. So how has the customer mix changed within industrial? Quite different, Vijay, right? So industrial, I mean, it's an unfortunate sort of labeling because industrial includes, it's 30% of our business, right? Back in 2008, 2009, it was still 30%, but the composition has changed dramatically, right? So now two-thirds of the industrial segment is, I would call, non-cyclical, right? So food and environmental, it's roughly half of it, right? So 15% of our overall business is food and environmental. And we just went through the discussion on PFAS. We could go through with trace quantities for additives and foods. That's not cyclical. That is proportional to the GDP and population growth, right? So that's half of the 30%, 15%. Then 10% of the 30% is TA. Roughly 40% of the TA business is now serving batteries and electronics customers. That's also not as cyclical, right? That is also a segment that has secular drivers, right? If you add those two up, 15% + 4%, 19% plus 60%, which is pharma, you're almost 80%. Almost 80% of our overall business has secular drivers that have an upward slope. Now, don't ask me if it's 3%, 4%, 5%, or 10% underlying growth; it's going up. Now, what we didn't talk about is academic and government, right? That's 10% of the business. And that's dependent upon government funding. And we saw incredible growth in that segment as well. And I was asked in different groups, "Hey, is academic and government back? And are you guys going to grow 20% in that segment?" No. On a three-year stack basis, that's a low single-digit grower. Mass spec portfolio is very well received, especially Cyclic, MRT, and some of the tandem quads in the academic segment. We have not yet finished with our turnaround plan and transformation plan in academia. That is the last segment we addressed. There's a lot more to come there, but I wouldn't expect 20%-30% growth in that segment. That one is definitely dependent upon government funding, right? And that can be cyclical. Overall, I mean, long story short, 80% of the overall portfolio is, I would call, non-cyclical, and two-thirds of the industrial segment is non-cyclical. That's extremely helpful. And then China, recent headlines on lockdowns, etc., some unrest out there. Any change in China, Udit, what you're hearing from the ground, how we should think about longer term? So Vijay, yesterday, I was with our general manager from China. She's visiting the U.S., and God bless her that she came, and she's going to spend a significant amount of time with us here before she goes back just to meet folks and reconnect. The demand side is, it's terrific, right? You go across the board in pharma, in different segments of industrial, in academic and government, the demand is excellent, right? And I can go through the details. It's the same thing that you've heard me talk about before. The commercial execution has been excellent, right? I mean, we saw China as a first mover for our CDMO initiative. For the instrument replacement with the Arc HPLC, China did better than any other geography, and the others followed, right? So from a commercial execution perspective, from a new product perspective, the traction is no different. And the demand side is no different, right? So I don't see fundamental demand changing. And there's a lot more penetration to be had with the pharma customers and other customers for our Waters in China. I think what is different, Vijay, is how long the sporadic closing and opening of the different cities has been, right? I mean, at any point in time, one in five of our colleagues in China is subject to a quarantine, right? So meaning in the morning, you're about to leave for work, and they say, "Nope, the city is shut down. You stay home." One in five. So 20% of the population at any point in time across China is subject to immediate change in their routine, right? And that takes a toll. And I'm not worried about the demand in China. I'm not worried about our execution. I'm not worried about the new products. I'm worried about the mental health of colleagues in China, right, and after speaking to Rebecca yesterday at length, we're going to try and help and think through how we can, whatever we can do from a distance, we will try to do and try to make sure that colleagues are not feeling undue pressure. I'm not worried about the demand. Is it 15% or 20% in the year, or is it 25%? I mean, it's in that real estate. It's been in that real estate for the last two years. I don't expect it to be any different, but I'm not worried if it's lower or higher in one quarter. Understood. And then maybe in the last minute or so, Udit, one on diagnostics. I think our friends from Roche, they're making a little bit more noise in diagnostics. I'm kind of surprised that a traditional diagnostic company is launching an mass spec-based diagnostic instrument. Where's Waters on your diagnostic for it? So out of all the adjacencies, this is the one that I feel requires the most induction period. And when you break the problem down, it's sample prep, it's the instrument, and it's the software, right? And our focus has been on specialty diagnostics, right? I mean, basically, drugs of abuse, endocrinology, and specific types of cancers. We're not looking to sort of capture the whole universe. We believe that's where we can specialize and do well. Now, there is a lot of work to be done on those spaces. So I don't know if somebody is going to be able to overnight come up with a simple mass spec that can go into high-volume applications. I'm as curious as you are to see if somebody can do that quickly. We feel we've made a lot of progress on the sample prep side. We feel the instruments have been simplified. There's work to do on the software side, which we're doing, but it's a journey, right? So I'm not going to sit here and say we're very close to success here. Out of the five adjacencies, this is the one where I feel the degree of difficulty is the highest, but it has the highest payback if you come up with a simple workflow and you can get into routine diagnostics. And yeah, I mean, if Roche is able to do it, great. I mean, that opens the doors for others. I mean, we're a mass spec player, and we would probably partner with somebody who's in these large high-volume labs. We don't intend to get there ourselves. There's no question. And just on the other adjacencies, we feel extremely good, right? I mean, we feel very good about bioanalytical. We feel very good about bioseparations, about battery testing. And they have come out of the gates even faster, right? So I think it's a portfolio game, but on the diagnostic side, work to be done, good progress so far. I don't think it is as simple as just putting a box in a high-volume area. And I could be wrong. I get back to mass spec is not mass spec, right? It's different for different customers. Maybe last question here, Udit. I think Amol on the third quarter noted underlying margin should be 100 basis points in fiscal 2023, where you will continue to make that 70 basis points of investments. Is that all, or are there any other variables like FX or inflation I need to think about? I think you should think about FX, right? I mean, especially for the first half of the year. I think it would be reasonable to assume, at least as we sit today, that part is that headwind is not gone, right? So 100 basis points of expansion at a 5%-6% growth, and that's a 20-30 basis points of organic expansion. But I would today say that gets eaten up by FX. That would be a reasonable assumption. I'm not using the word prudent again. I'm saying it's a reasonable assumption for now. Fantastic. I think with that, we're out of time, Udit. This is extremely helpful. Thank you for the time. Thank you, Vijay, for having us. Bye.