Analog Devices, Inc. (ADI) Earnings Call Transcript & Summary

Back to this session. Really delighted and honored to have the team from Analog Devices join us this morning. Prashanth Mahendra-Rajah, the Chief Financial Officer; and Mike Lucarelli, the Vice President of Investor Relations and FP&A. And what we'll do is we'll start with a few of my questions, but please feel free to raise your hand if you have anything else that you want to bring up, and we'll try and have some time at the end for your questions as well. But welcome, Prashanth. I really appreciate you being here.

And maybe, Prashanth, just as a high level, maybe set the stage, a lot of macro crosscurrents. And there almost seems to be a disconnect in that semiconductor industry feeling very strong, investors feeling very nervous. So set us straight, what are you seeing from a demand perspective?

Yes. So maybe I'll start by reminding folks that Analog Devices has about 80% of our business that we say is tied to GDP and then 20% is really more specific opportunities that we believe operate with great level of independence and growth from GDP. So we are very mindful of the macro indicators, and we're watching them very closely along with a variety of other piece of information that we keep our eyes on. But I'm going to reiterate what we've been saying for the last couple weeks that what we see today is continued very, very strong orders well in excess of our ability to supply. We expect to be capacity constrained at least through the end of the year. We expect that we will post sequential revenue growth as we have for Q2 versus Q1, Q3 versus Q2 and then Q4 versus Q3. I'm a little reluctant to make predictions beyond that, given kind of the storm clouds that everyone else is talking about. But from our vantage point and what we see we have not seen anything at this point that we would noteworthily call as different customer engagement continues to be very strong. Vince was out at Davos just 2 weeks ago. 2 weeks ago, he met with a number of our industrial customers. I would say, he probably was relatively popular just because more of our customers are realizing how important semiconductors are to their supply chain and they're trying to build strategic relationships with us. So it was a very good event for him, but it also reinforced to him that we are solving the right problems and we're getting the attention of our customers because we are helping them continue to advance the technology platforms that they're driving for their products.

Got it. Prashanth, is it possible that just given supply chain shortages and turmoil, right, geopolitics, lockdowns, et cetera, that the semiconductor industry is just seeing -- there's a delay, there's a lag effect between where the end demand will eventually be versus the signals that you are seeing right now.

Well, Vivek, I'd go back to the comment that I started with them out, the correlation to GDP. So if the investor community has high confidence that there is going to be an impact to global GDP then they should have high confidence that it will, in some way, impact analog devices. But I guess I want to talk about 2 examples or 2 areas that I think to help folks think about us maybe a little bit differently. The first is that there's been a lot of ink spilled on what's been going on in the automotive industry. And the disconnect between how can the shipment of chips into auto industry be so disconnected from the actual production of vehicles and that there must be something going on in the supply chain build. And I cannot dispute that, that's absolutely a possibility. But I can kind of go back to our quarter results and say that in the second quarter, we posted automotive growth year-over-year, Mike, of 30 -- mid -- 30-plus. And if you adjust that 30% year-over-year revenue growth for a couple specific items, you adjust out pricing, which is up sequentially year-over-year because of all the supply chain price increases that were pushed through across the industry and ADI included in that. If you adjust out for the growth in electric vehicles, which we would see in our BMS product. If you adjust out platform wins that we have won where key OEMs are moving their architecture to either our A2B platform or to Maxim's GMSL, you can explain more than all of our growth. So clearly, the secular wins are there, you can see that tied to the growth that we report. The second item, I think, that's important for investors to understand is we have structurally changed the profitability of ADI. And I tried to highlight that in Investor Day where we talked about our confidence that the floor of this company really is 70% gross margin floor. So again, I want to emphasize that in a downturn, we have conviction that we bottom at 70% gross margins, which again is well above the average for the semiconductor industry and used to be the prior peak for ADI. We've changed that structurally with the emphasis we've had on a hybrid manufacturing model, which gives us the flexibility to scale up when we need to, for demand, but it also gives us a lot more resiliency on the downside in being able to bring production in-house out of our foundries and keep that utilization level high.

Got it. It's interesting though, Prashanth, that when you gave that floor, which by the way was, I think, the first in the industry to actually -- it helped provide that kind of outlook. Your gross margins were 71%, 72%. Now they are 74%, so is it time to maybe set the floor at a higher level then?

Well, I think that the management team, ADI, from Vince down to all of us, we like to have a bit of conservatism in our outlook. And I think we're respected for that. So I'm not moving off that 70%, but our Head of FP&A here has told me many times, he would be surprised if we were actually able to touch 70% for a prolonged period. It would take a pretty cataclysmic event, I think, for us to be able to run at that level for 12 months.

Understood. So one thing I really wanted to discuss, Prashanth, is what has helped Analog Devices achieve these level of gross margin, but also importantly, sustain. And how do your acquisitions play a role in achieving this, what is kind of the best-in-class model?

This is for investors and really for our customers, this is the innovation premium that we bring, right? The -- as more investors have paid attention to the semiconductor industry, I believe, that their understanding is becoming more clear that there is digital, there is memory and there is analog. And then within Analog, there is high-performance analog, and that's our sweet spot. That high-performance analog allows us to solve very complex problems for our customers. It allows us to really move up the stack and bringing more solutions to our customers. That's reflected in our ASPs, which are multiples over the average for the industry. And those high ASPs and that innovation premium that we bring is reflected in our gross margins. And I've said this a few times, and I know some people find it a bit hard to believe. Today's gross margin levels are not impacted by the pricing that we pushed through because we went to our customers, we explained to them the cost increases that we were incurring, whether it was coming from foundry operations or internally, we told them we were solving for gross margin neutrality because that is reflective of the innovation premium that -- or the problems that we're solving for them. So these margin levels that you see today are really the reflection of a number of other factors that I can talk about, but not from pricing because pricing for us has really been to solve for margin neutrality.

Got it. And staying on the topic of the acquisitions that you made, what has been that grand strategy? Why Hittite? Why LTC, right? Why a Maxim?

Yes. It is -- so Vince talks about it in terms of scale and scope. And we have this level of intimate relations with our customers in which we are helping them solve very complex problems. And the more that we have in our toolkit, the better, a complete solution we can provide for them. So it has never been about consolidating the industry, which I know has gone -- undergone consolidation. It's been about finding technologies that when we add to the portfolio, we can provide our customers a more comprehensive solution. So Hittite brought us that great RF portfolio, Linear brought us broad market power. Maxim brings us GMSL and vertical power, particularly nanopower capability. So it has really been always about when we’ll hear the challenges that our customers are facing. And we think about what makes sense for us to do to be able to provide a more comprehensive solution, that's what's driven the acquisition strategy to date.

Got it. Is there a way to think about the portfolio not just by way of product, but also how many of those designs are sole sourced to ADI. Because I found that was a very interesting point at the Analyst Day as well.

Yes. I don't know -- I'm going to use one of Mike's expressions a lot. So we would...

Like more than 50% or more than 80%. What's a lot?

What do we say publicly, Mike?

Both.

Both.

Yes.

My choice.

Maybe I'll use an example. We, today, still have a small business, let's call it, about $100 million. So in the scheme of $12 billion, it's not relevant, but $100 million of airbag sensors that are sold to a group of automotive customers at prices that are far above what they could buy for a similar solution on the market today. And why are they buying from us because those sockets are so sticky. It is simply not worth the engineering effort that they have to undergo to design our center out to put someone else in to save $0.10, $0.20, $0.30 per part. And that is that stickiness that when you create a market and you get in there the first time, you hang on to that socket for 10, 20, 30 years. And there is no differentiation in that -- in this one example that I gave this airbag sensor versus what is needed in today's automotive industry. There is a clear opportunity for them to save money by choosing a different supplier. They continue to buy from ADI even though we have long stopped investing in that technology -- over a decade ago, we stopped investing in that technology and it continues to be a revenue stream for us because of that stickiness. And when a customer is designing a product and they have a set of components that work, and these components aren't a material cost or percentage of the bill of materials, there is 0 interest in trying to redesign to save money on that.

All right. Prashanth, on the -- so you closed Maxim not that long ago, but LTC was closed, right, well before that. Have you...

In 2017.

2017. Have you started to see any revenue synergy benefits from LTC, from the power side have started to realize any of...

Mike, do you want to share the numbers on that?

Yes. Sure. So if you look at -- so when we bought Linear, they were growing, call it, 3%, 4% over the past 5 years or double that. We've doubled that growth. A lot of it is power, high-performance power, bringing that portfolio to new customers and instrumentation, aerospace and defense and BMS. We really supercharge that BMS position they had. Both of the portfolio, we would double down investment in it and taking a lot of share globally. So we're the market leader in BMS. And we basically taking the technology and the customer list and putting it together and you look towards Maxim, GMSL sells the same story. Great technology. We'll double down the investment and go take share at customers.

Yes. I think there's a great analogy between those 2. You've been doing this a long time, Vivek, so you remember, when we closed Linear, that was probably one of our unfortunate surprises was one, that Linear's BMS products were probably a little higher on inventory in China than we wanted. So we were unwinding that and had some difficult explanations to investors the first quarter to explain why we were showing a decline in BMS because there had been a little bit of oversupply just prior to the deal close. But then we took that product. We've dramatically improved the quality of it using ADI's tech kind of manufacturing capabilities, which were superior to Linear. We decreased the cost of the product through our typical design activities. We added technology innovation by making it wireless, which is now widely adopted, General Motors, Lotus, another very large European OE. So that has proven to be a phenomenal platform for us, and we have tremendous excitement that Maxim's GMSL is going to follow a very similar ramp.

Got it. Could you give us kind of an early report card on what you have seen from the Maxim assets so far, positive or [ surprise ]?

Yes. So we learned a lot from the Linear acquisition. We learned a lot from things that we perhaps would have done different. So for Maxim, we have really focused on speed of integration. So the organizational integration is largely complete. What remains are purely the IT systems. But everything else is -- has been collapsed down to operating as one company. We have been very impressed with Maxim's manufacturing discipline, and that's why Vivek Jain, who was Maxim's Head of Global Manufacturing is now running manufacturing for the combined organization. We like the model that Maxim has brought us in really doing more of that resiliency through hybrid manufacturing where they work with their suppliers to be able to build products internally or externally and we've expanded that model now. And all the capital that's going in this year and next year is really going in to build additional capacity, but that is swing capacity that can be used for products that are built within ADI or externally. They have a 90-nanometer process that we're extremely excited about. It's really designed for Analog, and it has been optimized in a way that we think it gives us an incredibly good cost position that allows us to -- as we tape out new products on that will allow us to be targeting some sectors of the market today that are a bit more challenging for us. Because of price, we'll be able to target those segments and still maintain the extraordinary margins we have today.

Got it. One other thing, Prashanth, that I find fascinating is that in the high-end analog industry, right, we have the 2 leaders yourself and TI. And they have such a different approach in that lots of 300-millimeter fabs, they are expecting to spend $30 billion, $40 billion on CapEx over the next decade. But you have chosen a hybrid manufacturing model, why is that? And do you ever see a chance when ADI would consider your own 300-millimeter fab?

Sure. Yes, I would say that we absolutely have access, and we use 300-millimeter today. The difference, I think, that is important for folks to understand between the 2 companies is -- and it's reflected in our ASPs is we are a very high mix, low quantity shop. So the value that we bring to our customers, which is reflected in that innovation premium that comes from a very, very broad set of products. And because of that broad set of products and the diversity of technologies that's needed to deliver that product, we build everything from 7-micron all the way to 7-nanometer. So a really wide spectrum of process technologies is required. And then that 300-millimeter is used where it makes sense, where we have enough volume that it warrants it, but it simply doesn't warrant it for a lot of our products because, again, those are lower quantities produced to support it by a very high product mix. But again, the math is out there and you can do -- you can kind of interpolate. There's a pretty substantial difference between our ASPs, and that ASP difference is a reflection of where the 2 companies are focusing.

Is there a difference in mix also, right? I mean they have a higher consumer mix, right? So to your point that's higher volume business there as well. The other thing you touched on was a hybrid manufacturing model, right, which was -- you went through a lot of detail at the Analyst Day. The question I have is that if your point is that we can depend a lot on foundries and whenever there is some turmoil in the market, we can keep internal utilizations high. But when we talk with those foundries, they are saying they have very committed long-term supply agreement. So who is right?

Well, one of the benefits that we have now because of Linear and Maxim is we are the scale player in analog. And if you are a foundry, who is your -- sort of who is your premier customer, your premier customer is someone who doesn't require CapEx because they're running on processes that have been depreciated because they've been paid for by the NVIDIA, the AMDs, the Qualcomms, et cetera. You want a customer who -- when you're building something for them you know that you're going to be building that for a decade, not quarters, decades. So we're in a pretty sweet spot in those negotiations with our foundry partners because we are a very desirable customer for a foundry partner. And as a result of that, we have the most senior level relationships. And in the end, foundry partners, particularly the Asian foundries, they think about this in years versus quarters. So if it requires us to bring some production back in-house for a couple of quarters, it is still well worth it for them because they know they've got another decade worth of that same production coming back.

Got it. So let's say, hypothetically, we get into a downturn next year, and semiconductor industry sales are down 5%, 10%, right? What would ADI be doing differently? How much is kind of your fixed cost versus variable? What are the levers that...

Yes. So starting at the top, right, in kind of in a situation where there is a meaningful impact from a demand standpoint that impacts ADI. Now again, for there to be an impact on demand, what do we all have to believe? We have to believe that the automotive industry will be down. So fewer cars produced, but such a drop in car production that is actually more than offsets the growth in EVs. You have to believe that for deployment of 5G systems into U.S. and Europe will also be contracting because capital spend by the -- or the wireless carriers comes down. You have to believe that factory automation comes down even though you continue to see all the efforts that are being made to bring production domestically, whether it be U.S., Europe or diversified supply chains out of China, all of that would have to come to some meaningful pause. The instrumentation and test business, which is driven by the growth in electronics technology, not necessarily quantity, but the technology and the testing of those would also need to contract. You'd have to believe that digital health care, which has been a trend that has been growing now for us for...

7 years.

7 years. 7 years would need to see some disruption, and that continued to grow even through COVID despite the mix of how activity happen. And what am I missing, like aerospace and defense, yes, in a world where conflicts were on the rise, that defense budget would fall. So if you believe all that, then demand comes down for us. First thing we do is with 90 days’ notice, we're able to bring production back in-house from our foundry partners. So that's our rough notice period is about a quarter to let them know we're not going to be building products for you at your foundry locations, we're going to be running them in-house. The second thing that happens is our compensation system is designed to be highly accordion. So as revenue falls, then the compensation system adjusts pretty dramatically. So Mike's done some modeling. I'm not sure what we've shared publicly, but I'll hand to you. The 10% scenario is one I think you've been a little bit more vocal about.

Yes, I would say, I mean, I think down 5 or 10, I think most investors sign up for them, would buy the stocks today at that point. I think most investors are thinking down 20, but let's use your scenario of down 5 or 10. In that scenario, gross margins moved some down. But I don't think they touched 70%. OpEx, like Prashanth laid out, a bunch is fixed. 80% of sales fixed, 20% is variable. Most of that is a bonus. So puts all that together, our op margins stay in the mid-40s, 45% plus. As a result, I think you're seeing strong through cycle earnings and high level of margins through the earnings as well. So I think that's how I think about it in a down -- about a down 10%. So we'll see what happens next year, but we're well positioned to do well in that scenario.

Got it. Make sense. No, good to see that you have thought through that planning. On the pricing side, that's been a very interesting new dynamic. I mean semiconductor industry, right, moves low. Prices always go down. So it's kind of strange to see pricing going up in the industry. How sustainable is it? And I ask that in the context of most companies, including Analog Devices as I’ve said that we are really passing along the cost inflation from our foundries. What happens if those foundries overinvest and they go through a period of cost deflation? How sticky will your pricing be then?

The -- so when we -- when you look at our pricing historically, our pricing into the broad market has had a high level of inelasticity and we continue to get traction on that pricing in the broad market. And some of it comes naturally from as we provide more solutions than those solutions come with different pricing thresholds and the higher margins where we've had more deflationary pricing has been for particular [ S&K ] customers. And in fairness, it tended to be the Tier 1 and Tier 2 auto suppliers who would look for some commitment to a learning curve of pricing that would be provided over time because once they designed us in, they knew we would be there for 7 years. As we think about going forward, again, I feel quite confident that because there is no pin-for-pin replacement opportunity. When a customer designs us in, they're sort of committed to that level of pricing. The opportunities for us to be more flexible on pricing -- where we would want to be more flexible on pricing is we continue to have products that were launched 5, 10, 15, even 20 years ago that are being designed in for the first time. So a brand-new customer is building a new product. And as part of that solution, there is an ADI component that is 15 years old that solves their problem. That's where we have to be mindful for those new designs to ensure that we are cost competitive because those new designs will go to revenue in 5 years. But that's the time frame you're talking about, is flexibility on pricing that we want to offer in a more competitive environment will not impact that period. It is to ensure that we are not losing share as it converts to revenue 5 years in the future.

Understood. So there is a notion of stickiness.

Yes, absolutely.

Got it. The next thing, Prashanth, is we see all the CapEx announcements, right? I think your CapEx intensity has been fairly measured, right, part of it because of Maxim coming in as well. But we have seen very large CapEx announcements, right, from your peers. Why shouldn't investors be concerned that the industry seems to be overinvesting at a time when the macro cycle seems to be decelerating.

I think it is where that capacity is being built, right? I'm not sure that I should -- that I would agree that investors should not be concerned, but I think it is doing the homework. And I'm sure you're happy to provide them some of that. The investments that are going into capacity, what nodes are being built and where could that potentially end up with some level of future supply in excess of demand. I can tell you where it's not going. It's not going into Analog. It's not going into the legacy nodes, which are at capacity today. Those are really being -- the commitments in that spend is really coming from a small handful of companies that are investing in the legacy nodes and ADI being one of them.

Got it. So is there a simple way to think about how much does it cost to build that trailing edge capacity today versus what it -- because in the past, a lot of the digital guys, right, equipment was depreciated, you would buy that refurb from what I've heard, there isn't even refurb equipment, right, to be had. So it is just a lot more expensive to build that same level of capacity.

Yes. That's a hard question to answer at a general level. But if I was to give you the ADI specific example that we have clean room capacity in Oregon, which came from the Maxim acquisition and in Ireland from a legacy ADI facility. And because the concrete steel and walls are there, capital that we spend on tooling in those 2 locations has a very, very high ROI because it goes to revenue production almost immediately. So we've said publicly, I think, Mike, correct me if I'm going off script here, that we are doubling our internal capacity in the next 2 years. So that increase in internal capacity to 2x is really coming from investments that we're making at those 2 -- primarily at those 2 facilities, but it's a very high ROI return for us. So as you look at companies that are spending capital in this industry, a lot of different parameters. And if you have to break ground and start with a shovel, then it does become a much more challenging payback model.

Got it. What are you doing from a kind of a due diligence perspective on your backlog and the profile? So what are lead times today. And what are the steps that you have put in place to make sure that the demand signals are for real?

Yes. Yes. So lead times are stable. They are extended, but they are stable, which is a good sign. Our backlog is routinely reviewed and we actually have had -- we've gone through a couple periods where we've asked customers to recommit to their backlog. So we've sort of flushed it and said recommit to the backlog. Having said that, I don't know that there's a semiconductor company out there that would tell you that some portion of their backlog is phantom. So I expect that when I look at the amount of backlog we have, there is clearly a period of that -- a portion of that should be sort of excess. But even adjusting for that, I have enormous confidence on the ability to grow revenue sequentially for the next 2 quarters, Q3 and Q4. The other piece that goes into that is we monitor cancellations. And our cancellation activities have been -- they've essentially been flat, very, very modest tick up, but it's not even -- it's almost not even worth mentioning. We're talking fractional amounts. So we are monitoring cancellations. And while there's been that modest tick up in cancellations, orders have increased even more so. So all indicator for us are that today, customers continue to be in need of our products. Escalation calls to the executive team remain high and it is likely to be a period of strength, I would say, through the balance of this year. Now we cannot fight the macro. So there is a period where that is absolutely going to impact the whole industry. I just don't see it happening in 2022.

Is there a way, Prashanth. What's your gut on this phantom? Is it 5%, 10%, 20%, 30%? Like where -- as an industry, do you think it's a smallish number or medium size, large size, extra large?

I hate to put a range on it, but if I had to -- I'd call it between 15% and 30%.

I would say to give some context on backlog and what the coverage is. Historically, pre-COVID our backlog coverage is about 80%, 85% of the quarter.

Of the current quarter.

Of the current quarter. Right now...

We would need 15% book and ship just to hit the number.

Right now, our backlog goes out into mid-'23. So Prashanth's point, let's say, it's 30%, about 1/4 of it. We still have backlog covered 2x normal, 3x normal.

Is that clear because that's an important distinction for folks to understand. In a normal time, we would start the quarter with enough backlog that would get us through 85% of the quarter. So maybe 10 out of the 13 weeks, and we would need book and ship the orders that would come in in the quarter that want to be delivered in the quarter to be able to hit our full quarter guide. Now we have backlog that is running into next year.

Got it. So you still have enough cushion, even if, let's say, 20%, 30% is whatever -- correct. Is there a certain end market where you think there is more of the phantom?

Right now, it's just very broad. I mean the demand remains very, very broad. Our revenue production -- our revenue for Q2 -- for the next 2 quarters, it's almost not relevant to think about our revenue as demand driven because our revenue is which markets we're choosing to supply, both geographically and end markets. We are trying to be incredibly fair, making sure that our small customers as well as our large customers are getting product, but it's not really a function of demand. So -- but if you ask about where am I perhaps a little concerned about backlog being perhaps overinflated. I think every CFO is going to point to China and say that's probably the area of highest risk.

How much is the China exposure?

We don't share.

About China, what was the question?

China backlog.

Or exposure. Historical sales. Like kind of domestic.

Yes, yes. So China represents 20% of our revenue on a design basis, meaning that their products that are designed there, but on a ship basis, it's 30%.

Understood. And then quickly on the topic of cash returns. I think ADI did a very effective job of announcing an accelerated buybacks and then consistent buybacks. What is a better use of cash now to continue on the path of buybacks or to see that, look, there's a lot of maybe M&A opportunities given the valuations in this space?

Our current commitment is all of this cash that we are generating is going back to our shareholders. So I'll just go through our capital allocation model again that starts with dividend, which is the cornerstone of our return policy, whether you look back over the last 4 years or the last 10 years, we've grown the dividend [ sort of ] about an average 10% rate. And our commitment is, on a go-forward basis, look for approximately a 10% CAGR. So that may not be every year, but through cycle, our goal is continue to grow that dividend at 10%. We're very comfortable with our debt structure. Don't look us -- don't look for us to meaningfully reduce our debt that we have today. We'll continue to roll it forward. We may actually add some modest incremental debt in the current year and possibly next year simply as a way to help insulate investors from the higher capital that we're spending to expand Ireland and Oregon. And the remainder of that cash is really coming back to shareholders through the repo program. We are running at $800 million a quarter now, which is for a company with an $80 billion to $90 billion market cap, we're basically taking a percent out the quarter.

Right. Make sense. On the topic of tools, we had a chance to host Applied Materials and Lam Research earlier in the conference and they are seeing a lot of supply disruptions, right, impact their ability.

I had dinner with Doug from Lam last night.

Right. So does that impact your growth objectives in any way if their availability is getting disrupted.

When we call for a higher capital spend in 2022, we had always planned for it to be back-end loaded. And so we are still holding to the expectation that the tools we ordered and the delivery slots that were confirmed to us are expected to come in over the course of third and fourth quarter which are necessary for us to hit that sequential revenue growth that I spoke about. It is a very integrated supply chain. So there are constraints on the equipment providers because they're not getting enough chips. There are constraints on the chip manufacturers because there's not enough equipment to build it from. So that interplay allows both of us to have more adult conversations about how we help the industry to ensure that where it make sense, we're prioritizing each other.

I think I’d also like to add one thing to Prashanth is the benefit of a hybrid model is you can also get additional capacity externally. We don't -- we’re not fully reliant on those tools and the CapEx dollars to get additional supply, additional capacity.

Got it. Yes, that would have actually been my next question, which is, right, does the flexible model or hybrid model actually gives you that effect. But do you have to recognize that very early in advance? Like what do you have to tell your foundries today be prepared for this lately?

It's a 90-day cycle.

90-day cycle.

That's really the way to think about it. It's a 90-day cycle, assuming that they have the wafers to get it, right. I mean in -- earlier in the year, that was not the case, but I'm sure as folks have been observing, there has been some contraction in PCs and mobile, which require a broad set of technologies. We don't have very much exposure to PC or mobile, but that's likely to open up some slots.

Have you seen any benefits from the PC and mobile market not doing as well?

We have seen more willingness from certain suppliers to have conversations about pricing increases and whether we are going to take those pricing increases and whether they really want to push them in this environment.

I see. But those are discussion stages or they have actually happened?

We have held off on a second round of cost increases. I don't think we're going to be able to hold off forever. But the more those other market soften, the greater our negotiating position.

Understood. Terrific. Thank you so much for taking time. Thank you, Prashanth. Thank you, Mike. Thank you.

Thank you.

Thank you.