Good morning, and thank you all for joining us for our next Southwest Ideas Conference presentation. Our next presenting company is Photronics, which trades on NASDAQ under the symbol PLAB. Representing the company today is their Vice President, Investor Relations, Ted Moreau. Ted?
Thanks, William, and thanks for joining us this morning.
I'm just going to get rid of this.
For you.
Yeah.
There you go.
Little housekeeping items.
Yeah.
This is okay, we'll do it this way. No? All right, how do we advance it?
Space bar.
Space bar.
Sorry.
I see the clicker.
No, I had to take off a pot.
Is that good?
Yeah, okay.
Working now?
Yep.
Okay. Photronics has been in business since 1969. Throughout the time, the company had done okay as far as generating cash flow, but you'll notice over the last, say, five or six years, cash flow has really accelerated, and operating cash flow over the last—these figures are trailing 12 months—and so operating cash flow over the trailing 12 months has been approximately 25% or 26% of total revenue. We've done a much better job more recently on generating cash flow off of our revenue. Bottom left corner, you'd see our geographic mix. This is revenue based out of geographic location, originated out of geographic location. For example, we have three facilities in the United States, even though those facilities ship to customers largely in the United States, but also a couple of international customers. Those facilities generate 17% of revenue, right?
80% of our revenue originates out of Asia and largely actually stays in Asia, broken down between China, Taiwan, and Korea. On the lower right-hand side, we actually have two business segments. We have a semiconductor segment that we call IC and a flat panel display business, FPD. If you combine those, about half of our revenue is what we consider high-end, and then about the remaining half is obviously a little bit more legacy and mainstream. Here is our global footprint. As I said, we have three facilities here in the United States. We are expanding here in the United States. We announced that about a year ago. We can touch on that in a little bit. We have two facilities in Europe. Europe is not a very big percentage of our total revenue.
Obviously, given the semiconductor industry, Europe is just not a very large percentage of the total semiconductor industry, particularly when we get into the manufacturing aspects. Obviously, Asia, a lot of manufacturing occurs in the semiconductor industry occurs in Asia and in the flat panel display industry. You can see our facilities there. We have largely in the United States—so in the United States, we're semiconductor-focused, also in Europe. In Asia, that's where all of our FPD manufacturing occurs. It's a little bit more split between—Asia's a little bit more split between IC and FPD. What is a photomask? I don't know if you're familiar with a photomask, what it is. It's very critical to the production of semiconductors. If you don't have a photomask, you can't produce a semiconductor chip, or you can't manufacture a semiconductor chip.
We basically take a semiconductor chip design, etch it onto a blank piece of glass, and then we ship that off to the customer. Our customers tend to be the manufacturing, the fabs, the manufacturing companies, such as Samsung, TSMC, Intel, but also like a UMC, GlobalFoundries, those sorts of companies. Our customer base is not actually the semiconductor chip design companies, like an AMD doesn't manufacture in-house, right? AMD would not be a customer of ours, or not even be a targeted customer of ours. We ship our products only to companies that actually manufacture the semiconductor chips. Our photomask gets inserted into a lithography tool. I'm sure you've heard of a company called ASML, pretty big, or maybe you have.
Our photomask gets inserted into the lithography tool, light shines through the glass, and it essentially projects the semiconductor image onto the wafer, and then they go through that whole process. At the very beginning of the semiconductor manufacturing process, it is obviously very critical. A similar process occurs in the flat panel display market, where there is a pretty significant difference between the two products. One is the photomask for the semiconductor world is about 6 in by 6 in. The photomask for flat panel display is very large. As you can see in the top right side there, two individuals standing up. You are talking a couple of meters tall, maybe 2 meters tall, several meters wide, maybe 3 meters wide. It is very heavy glass in that instance.
Our cleanroom facilities, while they utilize a lot of the same technology and a lot of the same tools, are not identical. The cleanroom facilities for flat panel display are much more automated because you need machines to move around these large pieces of glass just due to the size and how heavy they are. When you think about Photronics, what are the trends that are occurring in the industry? We will talk about the semiconductor industry and what is going on there on the next slide. A couple of things are occurring within the industry that I think are really important for the next five years for our business. One is we think of regionalization or reshoring of semiconductor production. Coming out of COVID, the industry realized it was very dependent on the manufacturing side, very dependent on manufacturing, particularly in China.
The exposure, the risk was very high. The industry started to diversify. We started to see more manufacturing occurring in other regions, or at least the initiative to expand into other regions, such as Southeast Asia. Obviously, Taiwan is a big area of manufacturing for semiconductor chips. Korea is. We are seeing, and I think everybody might be aware of TSMC's initiatives on expanding into the Arizona region here in the United States. We are seeing this massive, massive shift in semiconductor manufacturing into the United States. Referring back to that, one of the slides previously, we have three cleanroom facilities here in the United States. Our facility in Boise, Idaho, is our high-end facility. That facility can produce down to seven nanometer chip designs.
For perspective, a lot of the very leading-edge semiconductor chip designs today are like an NVIDIA GPU for AI data center applications. That's probably at three nanometers today. We're, I'll say, in arm's reach away from being able to produce photomasks at three nanometers away. We're at seven nanometers at our Boise, Idaho facility. Our Allen, Texas facility, we're expanding capabilities and capacity in Allen, Texas, right up the street here. Historically, we've been able to produce masks at 90 nanometers, and we are adding capabilities to go down to 65 nanometers and 40 nanometers. That is because of customer demand. Many of our customers at that facility have been on 90 and older geometries, but they are asking us to move to 65 nanometers and to 40 nanometers. We are satisfying that demand.
Additionally, a number of projects that we produce in our Boise, Idaho facility are at, say, 40 and 65 nanometers. Because the toolset in Boise can manage or can produce photomasks down to 28 nanometers, 14, 10, seven, that's very efficient, inefficient usage of those tools. We are shifting some of those projects over to our Allen, Texas facility. Once we're up and running in Allen, Texas, we're going to shift some of those projects over to fill up the cleanroom, the additional capacity in the cleanrooms that are coming online in Allen, Texas. That frees up opportunity at the leading edge for us to target here in the United States, which we'll talk a little bit more about. Captives. There's an element in the market. The market's about $6 billion for total photomasks, and it's split between captives and merchant suppliers.
Obviously, we're a merchant supplier. The captives, historically, I think 10 years ago or 15 years ago was about, I think, about 60% of the total—or no, I'm sorry, was 50% of the total market. Captives today now represent about 63% of the total market. The reason for that change is because the cost of producing a photomask at the very, very leading edge has become very, very expensive. I'll show a slide on that shortly. It is more of a—the value of the masks have gone parabolic, and that's why they've captured more of the market. What captives want to do, they want to push very leading-edge chip designs, right? I mentioned the NVIDIA GPUs at three nanometers today, right? That's requiring EUV capabilities, extreme ultraviolet technology capabilities, which we don't have in our cleanroom facilities today.
If we were going to invest in equipment designed to produce photomasks incorporating EUV capabilities, it would probably cost us $200 million-$300 million. Given our—we're about $850 million in total revenue, that's a pretty significant capital burden. We are producing masks at a slightly older technology, but certainly no less valuable to the industry. What captives are doing is, okay, so they're at 63% of the total market today, right? They are actually starting to open up some of their captive production to merchant suppliers a little bit more. I said earlier, Taiwan Semiconductor is a customer. They're looking to outsource a little bit more. Intel is a customer. Samsung is actually a 10% customer.
They're looking—all those companies are looking to outsource a little bit more so they can focus more and more of their photomask internal development initiatives on the very leading edge and then outsource a little bit more of their older production. Node migration. We'll see this in a slide shortly. Node migration is basically, as I said with the Allen, Texas facility, moving from, say, 130 nanometers or 90 nanometers down to 65 nanometers, down to 40 nanometers. You'll see the increased mask value or mask set value as you get into the newer geometries or the more advanced geometries. On the flat panel display side, displays are becoming more complex. We're starting to see smartphones that are being designed that are foldable. That's creating additional complexities on how to develop glass screens for those smartphones.
Our flat panel display, it's all about creating the screen for consumer electronics, right? Whether it's a laptop or a TV or a smartwatch or a smartphone, our photomasks help create the display. As displays become more complex, the mask value is going up. Why do we win? It's geographic location, close proximity to the customer, because shipping times are very critical. Sometimes we have to ship a portion of our photomask set within 24 hours. If we're able to do that, we'll win the project. Location is very important. Also, obviously, operational excellence. You have to produce photomasks that are very clean, very pure.
You can't even have like a small speck of dust, right, on the glass, because otherwise that could show up on the wafer as the image is projected onto the wafer, and that would be a nightmare scenario. Obviously, we've been generating a lot of cash flow. The semiconductor industry, everybody is about, I think, about $600 billion today, or maybe give or take. Everybody seems to believe that the industry is moving to or going to grow to about $1 trillion in market size by the year 2030. The photomask market historically has been about 1% of the total semiconductor market. Obviously, the market size is $6 billion today. As I said earlier, 63% captive, 37% merchant.
Assuming that the industry forecasts are true and the semiconductor industry does grow to about a trillion, that would suggest that the market size for photomasks would be about $10 billion, right, in the year 2030. If we also assume that the merchant suppliers, maybe we can, because captives are outsourcing more, maybe we can expand market share as a group from 37% up to, say, 40%. That would say our addressable market is $4 billion. Today, this is semiconductor only, not flat panel display. This is semiconductor only. Our market share in semiconductor IC is about a third of the merchant portion of the market, right? By 2030, conceivably, if all those numbers hold true, we could be looking at a semiconductor IC business of about that probably doubles in size in the next five years, right?
We're about $600 million today. We could be about $1.2 billion if all those things hold true. Now, with that said, we do have to go out and capture business at the leading edge, which is why we are investing in our U.S. facilities. Actually, I did not mention that we are investing in a facility in Asia to capture more of that captive opportunity. Yeah, I touched on the U.S. investment that's occurring. We are also investing in one of our facilities in Asia. Again, we have facilities in Korea, Taiwan, and China. It's important to understand that we actually, on the semiconductor IC side only, have two joint venture facilities. One is in Taiwan and one is in China. The China facility we opened towards the end of 2018 became a first mover in the China market.
Therefore, we've developed great relationships in China. China has grown from 3% of total revenue in 2019 up to about a quarter of total revenue today. China has been a very significant contributor to our revenue growth over the last five or six years. With that said, because the semiconductor facility in Xiamen, China, is a joint venture, we are actually focusing on our Asia investment in a non-JV-related facility. Since I told you that our JVs are in China and a JV in Taiwan, that kind of gives you a pretty significant hint as to where our facility, where we're expanding. Our facility in Korea currently has been operating, producing masks up to 28 nanometers. We are extending that down to eight nanometers. We announced this on our last earnings call, which occurred at the end of August.
We're extending that down to eight nanometers initially and eventually down to six nanometers. This project does not include EUV capabilities, but it's still more leading advanced nodes. It is really at the suggestion, I'll call it suggestion, of a very significant semiconductor company based in the region, right? They are suggesting that we, which is a current customer of ours, expand capacity and extend our capabilities because they want to outsource more to us. We're very encouraged by that. What's really, as I think about the stock and everybody looks at our valuation and it seems to be a little bit handcuffed, why is that? I think one of the reasons is investors are concerned about the China exposure being 25% of total revenue. There's two comments there.
One is, okay, yes, China is 25% of total revenue, but it's split between IC and FPD, right? In the last quarter, most recent quarter, which was the July quarter that we reported at the end of August, our China revenue was a little bit more in favor of IC versus FPD. IC had come down. FPD remained pretty steady. This is China specifically. Excuse me, it's China specifically, right? Of the 25% approximate of revenue that we generate out of China, about half of it, it's split about pretty close to 50/50, right? That gives you a perspective on what the total exposure is on IC, China and IC or FPD for China.
Because of the valuation that we get as a company and the exposure we have to China, the investments that are occurring both in the United States and in Korea are very significant to help us more, I guess, better diversify our revenue mix across geographies. We said that we're going to start to generate revenue out of the U.S. facility in about a year from now. It is towards the end of our fiscal 2026. We have an October fiscal year end. Our Korea facility will be about a year or so after that, in the fiscal 2028 timeframe. Those two efforts are really designed, obviously, to take advantage of opportunities that we see, but are also very significant in diversifying the geographic revenue mix even further than what we currently realize today.
This slide really talks about node migration and the value of masks going into the more leading edge, right? Historically, say 130 nanometers, 90 nanometers, I talked about our Allen, Texas facility producing masks there. A mask set, which for one semiconductor chip design, it's not a one mask per chip design. It's actually more like 25 or 30 masks in a 130 nanometer chip design per semiconductor chip design, right? So 25 or 30 masks. As you get into higher and higher geometries, like today at 10 nanometers, seven nanometers, it's actually more like 80 masks. There are many more masks to be produced because of the complexities of the semiconductor chips at what we call high end. Our high end business is 28 nanometers down to seven nanometers today.
The value of the mask set increases very significantly as you go down to more advanced geometries. In the 130 nanometer range, there's actually a report that came out from Morgan Stanley, the Japan team of Morgan Stanley. It was two days ago, and they said a 130 nanometer chip design, the mask set, the value of the mask set there is, say, $50,000. When you get into like a 14 nanometer or 28 to 14 nanometer chip design, you're talking about maybe about low millions for your mask set. As you get into the very leading edge, which is where Taiwan Semiconductor produces their own masks and the area that we're trying to crack into a little bit, it's actually in the $10 million-$15 million range per mask set. Very significant increase in value on mask sets.
That is why you can see on this chart, on the lower left-hand side, the kind of the yellowish or goldish region, that is where we play today. We are trying to get into that more kind of brownish shade area in the future. Again, FPD, I do not touch upon it a whole lot because it is 25% of the business versus 75% on the IC side. Again, it is the foldable phone designs, consumer electronics, anything with a screen. Those are really the drivers. Also, it should be noted, a lot of people ask us, well, what is the influence that wafer starts have on your business? What is the influence that wafer fab equipment has on your business? There is some correlation because obviously, as the semiconductor industry grows, the photomask industry also grows.
Really think about it in terms of chip designs or smartphone designs, new screens, because a photomask set from, say, five or 10 years ago could still be utilized today to produce a semiconductor chip in a manufacturing facility. It is really about the new chip designs because it is all about printing and etching onto the glass that new chip design. Why do we win? Again, I mentioned being close to the customer, but then also high quality masks, very clean, no defects, very critical. That is how you become a trusted supplier. Again, we are about a third of the total market today. That has grown very significantly compared with, it was a fairly fragmented industry throughout the 1980s, 1990s, and the first part of the 2000s.
It has, through technological advances and then scale, a lot of the peers actually kind of fell out of the market. There has been some consolidation as well. It has kind of turned into a closer to a duopoly situation on the merchant side. Again, we're about a third of the IC market. On the FPD side, we're about 28% of the total market. On the IC side, there's one player that's a little bit larger than we are, Texscend. They just IPO'd in Japan at a $2 billion valuation, which is very interesting because their total revenue is about $800 million. Our total revenue, when you include, and they're IC only, when you include flat panel display, we're about $850 million and we're at a $1.2 billion-$1.3 billion valuation, right?
Pretty significant discrepancy in valuation kind of scratches our head a little bit, but we are going to try to close the gap on that one. Historically, here are some of our financials. Again, October fiscal year end. We are going to report, have our next earnings call on Wednesday, December 10th, pre-market. These numbers are for through fiscal year 2024 because fiscal year 2025 has not been reported yet. You can kind of see in the 2022 timeframe, a step function up in revenue and profitability. That was because the customer base historically had operated under the premise of very fragmented industry. We can play the suppliers off each other and they had all the pricing power. Pricing power really shifted into our favor. I think we have more to go, but we will see if we can make any headway there.
I'm not saying anything about the future there, but we've improved our pricing power coming out of COVID. The customer base realized if they don't have a photomask, they cannot produce manufactured chips. Our lead times during COVID went from one to three weeks up to four to five months. They were in panic mode. We were able to increase pricing. The whole industry, semiconductor industry, raised pricing during that timeframe. That really helped us improve our margin profile from the gross margins in the low 20s into the mid 30s over the last several quarters. That obviously really helped with cash flow. Operating costs as a percentage of total business tend to operate around about 10% or 11% of total revenue. You can pretty quickly move your gross profit dollars into cash flow.
Capital intensity ratio historically, we're about 15%, I would say over the last five or six years, 15%. We're going through a pretty CapEx intense cycle over the next couple of years. This was kind of, this was an estimate of 24% of capital intensity ratio for 2025. This is estimated based off of consensus numbers from a couple of months ago. That accounts for the expansion here in the United States. We're also replacing a number of end-of-life tools globally because we have a lot of tools that have been in our cleanroom facilities for quite a long time. They've been fully depreciated and they're no longer being supported by our supplier base. We need to upgrade some of those tools. We're going through that process over the next several years. I mentioned we're investing in our Korea facility.
You can assume that since we're still investing in the United States in fiscal 2026, we're also expanding into Asia and we have the end-of-life tool replacement costs. You can assume that fiscal 2026 CapEx could be pretty elevated as well. I don't know if you want, I mean, our numbers are pretty stale by now because it was about almost three months ago that we reported earnings. I'm not going to really touch on that. We have about six minutes left. I figured I'll open it up for questions if anybody has any. Yes, Bill?
Would you please go back to slide 10? It's the encourage here. Graphs can be deceiving.
I'm going to ask the question, is the price increasing, adding increasing rate as the mask set cost, as that graph would indicate, or there's some visual dynamics there that changed?
Yeah, obviously, I had mentioned, say, a 90 nanometer chip design a lot of times can be 25-30 masks for the mask set, right? There are actually some projects that we work on where even at a 190 nanometer or 130 nanometer, it requires at least 100 photomasks for that project or for that chip design for whatever reason the complexity is there. We're still probably talking $50,000 to maybe $100,000 for that chip design, right? Obviously, given what I said about at, say, seven nanometer or at 14 and 28 nanometers, 80 masks will produce a million or $2 million or $3 million worth of revenue.
There's increased complexity that's occurring with the semiconductor chip design. That's causing each individual mask, yes, is going up in value, but you're also having more masks, right, relative to the average that you would see at the older geometry. It's kind of a combination, right? I hope I've given you enough detail where you could actually probably run through that calculation.
The graph is showing an increasing rate of growth or accelerating rate of growth.
Yes, right, because of the factory. Because, yeah, again, at like 14 or 28 nanometers, you're talking about $1 million, $2 million, $3 million for that mask set. At three nanometers, you're probably talking $15 million, right? It's a pretty significant parabolic move. Yeah.
You walk through that math, you're going into a trillion dollar industry.
How much of that is in the more complex geometry where you do not currently play?
Okay, yeah.
Is that growth, is it equal or how is that distributed?
Yeah, that is an excellent question. At the beginning of the presentation, I did talk a little bit about as you get into the higher end of the geometries, the mask value goes up and that is where TSMC produces their own masks. Again, they are producing masks at three, four, five nanometers today. We are doing masks at seven, just a few at seven and a few at 10, mostly 14, 22, 28 nanometers. What have we seen in the industry over the last, say, year or so, right? All the growth has come in the semiconductor industry has come at AI hyperscale data center investments, right? That is all or largely very leading edge chip designs.
Again, NVIDIA GPUs at three nanometers, right? What's interesting though is, yes, your point well taken that that's where all the growth is occurring and conceivably could be occurring over the course of the next five years. Obviously, we're trying to capture a little bit more of that, right, with the investments that are occurring. There's an important piece to a NVIDIA GPU design that's not as well known. That is, okay, three nanometer chip design, 80 masks approximately. A third of those masks are actually produced using EUV capabilities at, say, three nanometers. The remaining two thirds of the photomasks for that mask set, you can produce at 28 nanometers. As Taiwan Semiconductor has come to the United States, yes, they've got what, five or six cleanroom facilities that they're putting up in Arizona. They've already got one up and running.
They are building a photomask operations here in the United States designed to satisfy the three nanometer, four nanometer, five nanometer chip designs, right? Masks at those chip designs. It is our, through conversations, it is our understanding that they are looking to outsource the 28 nanometer portion of that three nanometer NVIDIA GPU to companies such as Photronics and our other major customer that is also based here. Yes, we think that all the growth in the industry is going to largely occur, or a lot of the growth is going to occur AI hyperscale data centers. At some point, we do see AI proliferating more and more at the edge devices. That'll open up opportunities for us because you do not put NVIDIA GPUs in a smartphone or in glasses or smartwatches and whatnot.
As that occurs, we do think that that is a catalyst for the industry. Hopefully, PCs, autos, the rest of the semiconductor market start to improve. It has been pretty sluggish. You have seen numbers out of a lot of the analog chip suppliers. Hopefully, that improves at some point, but it sounds like that is going to take a little bit more time. We have 30 seconds if anybody wants to slide in. Otherwise, hopefully I will see you the rest of today. My contact information is on every press release we issue, which is not all that many. Feel free to reach out. Happy to talk anytime except for the two weeks leading up to the earnings call. That is about it. Thank you very much. Appreciate it.