Ladies and gentlemen, thank you for standing by and welcome to the Photronics Investor Day 2020 conference call. At this time, all participant lines are in a listen-only mode. After the speaker's presentation, there will be a question-and-answer session. To ask a question during the session, you will need to press star one on your telephone. Please be advised that today's conference is being recorded. If you require any further assistance, please press star zero. I would now like to hand the conference over to your speaker today, Troy Dewar, Head of Investor Relations. Thank you. Please go ahead, sir.
Thank you, Gigi. Good morning, everyone. Welcome to Photronics 2020 Investor Analyst Day. Joining me this morning are Peter Kirlin, our CEO; John Jordan, our CFO; and Chris Progler, our CTO. For today's call, Peter will begin with an overview of our business and recent performance, as well as a close look at our strategy and where we are going from here. Next, Chris will speak about our technology and how we align with market trends to help our customers succeed. And finally, John will provide an update on our long-term financial model and how we plan to improve financial performance. After that, we will all be available to answer your questions. The presentation for today's event is available on our website.
Before we begin, I remind you that comments made by any participants on today's call may include forward-looking statements that include such words as anticipate, believe, estimate, expect, forecast, in our view. These forward-looking statements are based upon a number of risks, uncertainties, and other factors that are difficult to predict. Actual results may differ materially from those expressed or implied, and we assume no obligation to update any forward-looking information. Additional information may be found in our SEC filings. At this time, I will turn the call over to Peter.
Thank you, Troy, and good morning, everyone. Thank you for your attention today and your interest in Photronics. So moving on to slide four, the agenda for the day will begin with a statement regarding why we believe Photronics is a compelling investment opportunity. We'll then frame that with the progress we've made over the last three years, in fact, since the last analyst day, in improving and growing our business. We'll then take a three-year look forward into our plans. John will codify Chris and my statements with financial targets. And finally, there'll be a period where we will be pleased to answer any questions you have on our plans. So why invest in Photronics? First and foremost, you'll be invested or investing in a global market leader. Market leaders tend to outperform others in both good and bad times.
We undeniably now occupy this position in the photomask industry. We have gotten there by taking risks and investing in growth sectors as we identify them in our market. Finally, moving forward over the next three years, our investment model will sharpen to focus on two of three areas we have previously invested in to generate expanded margins. Of course, it's a business model. Our business model is one that can fund organic growth as well as return cash to shareholders. Let's look at our market position. Where do we stand? If you look back three years when we started on the next phase of growth for the company, there's effectively three large competitors separated by less than $50 million. Big three and then a trailing three. Our revenue corresponds to $450 million.
We've taken a little liberty here because obviously we're not done with the calendar year. Photronics at $610 million has a $160 million gap between itself and the next largest competitor. Following is a third competitor with another $50 million gap. We have more than $200 million of market or top-line separation between us and number three. Of course, when we get to ROIC, that's less than $200 million or approximately $200 million. This is a factor of three smaller. There's a big separation still between the top three and the trailing three. We're delivering this market position utilizing global footprint, which now includes 11 factories in the U.S., Europe, Taiwan, Korea, and of course, now China. Peeling back the onion a little further, to understand where you are, you need to frame it within the context of where you've come.
If you look at Photronics, we've been making photomasks for more than 50 years now. We really had four phases of growth. First, not quite 20 years of our life as a capital-constrained U.S.-only supplier that grew by being a low-cost producer. Second phase, 15 years, was really a globalization and M&A-driven uptick. I think we probably did have the exact number, but close to a dozen deals over that trajectory to reach high $300 million. And at that point, we were the number three, not the number one, but the number three global player. We then went through a 15-year period, what we call phase three, with considerable volatility that sequentially took advantage of three technology transitions. The first place we aspired to be the technology leader was FPD.
If you look at 2006, it represents that really the realization of that aspiration where we led the merchant market in the G6 to G8.5 technology shift in FPD. Then moved to 2011. That was Photronics with a JV with Micron entering the high-end memory market, another $50 million step up. And then the third peak in 2016 was the full elaboration of our product line, both FPD and IC, to penetrate the high-end logic market. Step by step, but industry downturns to 2008, 2009 financial crisis, etc., up and down. In 2017, we entered what we believe to be the second golden period of growth for the company where instead of one dislocation, we invested in three, basically the LCD to AMOLED transition to mobile displays, the G8.5 - G10.5 transition in LCD displays for TVs, and finally the China market.
So looking at what that has done for the company, I'm now moving on to slide number eight, Troy. You can see that over the period, we had a double-digit CAGR, 11%. That's approximately three times or better the industry growth rate. I'll point out that 2019 was an industry downturn, and nevertheless, we grew straight through it. So it's $160 million of top-line growth in a business that's 50 years old, which is not easy to do, but we did it nevertheless. And we did it by investing nearly $300 million or approximately $300 million in China while preserving the balance sheet with the operational cash flow that allows us to not just grow the business, but return cash to shareholders. So specifically, how did we do against what we said we would do? We had a top-line target of $630 million and an EPS target of $0.80.
At the time, we said that contemplated a downturn somewhere between here and there because we're in a cyclical business. When indeed that downturn materialized, but what we didn't forecast was a global pandemic or a trade war. I don't think anyone at that time would have had the foresight to expect either, actually. And in fact, if you take out our second quarter, which was COVID-damaged, you end up with a $635 million run rate if you annualize the other three. And by the way, that includes sucking up the imperial decrees from Washington, D.C., which no one would have expected. So we believe that we satisfied the intent, certainly, of what we set out to do. And despite these obstacles, we expect to hit the $630 million and $0.80 target within the current fiscal year. So it's well within our reach.
It's slightly delayed, but given the obstacles, I at least feel good about the company's progress. Regarding the three key tactical objectives over the period, key milestones or touchstones, one was to get China production started in the first half of 2019, and we did indeed do that in FPD. We pushed the IC plan back to the second half of 2019. That wasn't because we couldn't get there. It was because we were adjusting the timing of the ramp to correspond with the strength in the market. Again, I remind you all that 2019 was a down year for the semiconductor industry. We said we were going to enter the G10.5 market. We had no presence there. Well, in fact, today we believe we have more than 50% share. Not only do we enter the market, we are leading the market.
We're the other two global suppliers sharing what remains. And finally, we said we were going to grow market share, and it doesn't matter whose data you use. I think it's undeniable that we did indeed do that. And finally, as I mentioned earlier, we had a balance sheet/cash flow objective, and that was to come out of this first phase of investment with basically similar or the same amount of net cash. And looking at the numbers, we're $36 million less. But on the other hand, we repurchased $79 million in stock during the period that wasn't built into the plan. So in our mind, we're up net $40 million here. We did better than we promised. And then finally, our ROIC is low. The entire management team, board, employees are aware of that.
Unfortunately, we're kind of dragging around the weight of our balance sheet and improving this metric, and it's hard to make it sing in a day. It's a long-term project. We clearly have made progress on this project, and we clearly have a lot of work that remains behind to do. And we do expect the next three years to see further improvement on our ROIC. So I want to peel the onion back yet one more time and look at the growth sectors. Right? China, 2017, $43 million. 2020, $208 million. We told investors we will get $150 million in growth from China. Well, we actually did a little bit better than that. In fact, if you look at our entire top line, step up on our entire top line, it can contribute to China.
Now, beneath the scenes or underneath the curtain, so to speak, there's been some rotation of the business. But anyways, a 69% CAGR is not shabby for any business, no matter how you look at it. So we're quite pleased with our growth in China. We have a lot of momentum there. The customers clearly like our products and our technology. To use a baseball analogy, we might be in the third inning. So China still offers a lot of growth potential for Photronics. At least we believe it does. So now moving on to FPD. Again, when we started, what we said was expect to more or less double the business, and you can see that indeed we have done that. We've gone from $100 million- $191 million over the period. You can see that black bar is high-end. Right?
We've done it by growing a high-end of our business. That is both G10.5 and AMOLED, so we have a, and in fact, you can see the red bar has shrunk, and that's been deliberate because over this period, we have effectively, except during the ramp, run at capacity, so we've used our market position to rotate our business to a higher percent high-end, but also on an absolute dollar business. We're using our mainstream equipment, some of it anyways, to build the non-critical layers of high-end reticle sets. So we've been very successful in exploiting the technology transitions that we invested in, so looking at now where we've been versus where we're headed, we invested in G10.5. That market is now stable. There's not new customer investment there of any magnitude, so we're done. We're the market leader. We're not done competing, but we're done with investing.
The next place we invested was AMOLED. And AMOLED has generated good growth. And its footprint, both in the market as well as in the technologies used to manufacture displays, is broadening. So the market is growing. It happened to be polluted last quarter by imperial decrees from Washington. But as we said on our earnings call, the market has rebounded, and we're presently in an oversold condition in our display business yet again, which is obviously where we want to be because we have the next wave of investment starting to come online as we exit the first half of our fiscal year.
I would also say we've not announced this with press releases, but within the last two months, we've had two Supplier of the Year awards in our display business, one from Samsung, who is the undeniable market leader, and the other just two weeks ago, or maybe even less than that, from Visionox, who is one of the up-and-coming leaders in the China AMOLED market. So by any measure, no matter who you talk to, it's quite clear that we are the market leader in photomask for AMOLED displays. The other place where we have been focused and we intend to continue to focus is growing our China business, both FPD and IC.
The imperial decrees that have damaged or did damage the business in the short term, underneath that is nationalism, which is actually, we believe, will drive the growth in the China market even and accelerate clip moving forward. This continues to be rich waters for a mask maker to be investing in, whether those investments are in IC or FPD. What do we expect to go over the next three years? It's kind of hard to know what the growth rate in the business is going to be over that period. Looks like this year in front of us, right? There's a lot of enthusiasm for it. Right? Hopefully, the enthusiasm at the beginning of the year will be sustained throughout the year. Anyways, our objective is to double industry growth rates, whatever they might be. We picked a range, 5%-7%.
That assumes industry growth rates of 2.5%-3.5%. If they're better, we'll do better. If they're worse, we'll expect to do worse. But anyways, we've been about this for 50-plus years. These are typical growth rates for our business. So these are the growth rates we've used in our projections. But as I said, our goal is double, whatever it happens to be. Given that we're investing in back of new facilities, we don't have to build brick and mortar. We don't have to buy whole lines. What we have to add is lithography and inspection. So the capital investments we're making, we expect to be more efficient. We also expect them to generate higher levels of income. So our expectations are for growth and operating margins in the high 20s and high teens, respectively.
Finally, in the last 25 years, we've not seen a dollar share of Photronics, but this is clearly our target. This is, in summary, where we're headed. John will give you more color on these targets. Finally, in summary, we've outgrown the business by quite a lot, our market by quite a lot, maybe 3x over the last three years. We have $150 million in separation between Photronics and its next largest competitor. If we do what we expect to do, we expect that separation to expand. We're sharpening our focus to target investments that will drive accretive growth. We're doing it with a business model that can both fund that growth as well as return cash to shareholders. With that, I'd like to turn the presentation over to Chris. Chris.
Okay. Thank you, Peter. I am Chris Progler, CTO of Photronics.
I'm going to go through some of our technology trends and how we're reacting to those, so first slide, I think people are quite familiar with what's happening in the broad industry now, evolving from this kind of simple world with cloud and edge devices to really a much more connected ecosystem. What that's really doing for the purposes of ICs and displays, as well as pushing more and more compute power now out to the edge devices, that's kind of the next wave of technology integration that's coming to this sort of network, and underneath this network, a lot of algorithms, AI algorithms, machine learning, all kinds of things like that. They're also driving compute and memory resources, so the overall trend on where this is going is quite positive for the electronics and display industries in general.
Photomask being one of the key linchpins to enable new designs to go into manufacturing, of course, should benefit from these trends over the next few years. The first bullet there, we say golden age of design and integration. That may be overstated, understated. It's hard to know. What we clearly see is trends towards companies differentiating more with design than just simply node shrinks to get performance, more complex SOCs and things like that. Even in mid-range nodes, we see a lot of increase in design activity as customers try to get their products into this network and compete effectively. Since photomask and designs are really, of course, intimately linked, that increase in design emphasis is a very strong trend, we believe, for the photomask industry at large. What are the six things that these global dynamics kind of bring to mask value?
Of course, technology has always been a key factor. Time to yield, the second bullet. Cost, particularly on the display side, cost down, very, very critical for the panel makers. The last three, I'd say, have evolved in importance more so over the last couple of years. And this is speed, how quickly you get products to market, security, which Peter touched on some of those points, the global dynamics driving security. And then generally the word trust. Since photomask contains design information, customers really have to trust photomask maker knows what they're doing with that data, knows how to protect it, knows how to handle it. Of course, we have the longest history in photomask, and we have outstanding reputation in this area, so it is helping us. And that also increases the barrier of entry for new players that might want to get into these fields.
So what about our leadership? We really look towards partnerships to try to do that. process of record, this means where we co-design with customers for their new processes, and we have multiple process of record relationships. Because of capacity constraints in the mask industry today and the importance of technology, we are actually seeing a customer base more agreeable to long-term purchase agreements to lock in capacity and capability. And then joint development, more forward-looking things we use as well in our constellation of technology programs. The market still continues to push more technology into masks. Of course, there's a reliance on the other components of lithography and the scanners and the resists and things like that. But the user base, particularly on the FPD side, continues to push more technology into masks, increases the value, increases the ASP.
The multi-node supply and demand trends look favorable for us. The next slide, this is from IC Insights. It just shows an example of how IC node, anyway, migration might happen over the next three or four years. For those that follow equipment companies and to some extent EDA companies, they really require that blue box at the bottom to continue to grow. That cutting edge is really critical for their business models. For Photronics, we actually have value propositions across all of these nodes, particularly where we see the trend of designs going in a healthy way across these nodes. It's a very positive indicator for masks. How do these different nodes or node segments bring value, or how do masks bring value to these?
At the highest end, it's really about getting chips to market, so-called chip enablement, and assisting customers in getting designs integrated into the wafer fab and the new fab ramps. On the mid-range, which I would say are maybe the center three blocks or segments in these bars, time to market is very critical. These are not mature nodes, but they're relatively healthy, high-yielding nodes in wafer fabs. So time to market is very critical. And the ability to have design flexibility, this is particularly important for interaction between masks and customers. And then the yield. And then legacy, kind of the third segment, these are the top bars in this chart. It's really cost, capacity, and continuity. And we are actually seeing capacity shortfalls, particularly in the mid-range and legacy for photomask.
This is really due to the strength of design and design activity we see around the industry. When you look across these left to right, the top four bars, the red going down to most of the orange, very, very strong value proposition for photomask in there. It makes it a little bit unique, perhaps, compared to some of the other players in the supply ecosystem for IC and display. What are some of the things when we focus on the high-tech or high-end part of the photomask spectrum that's driving demand? On the IC side, of course, EUV has been a big factor. The mask and lithography technology is still maturing there. It's consuming far more writing and process resource than I think the industry anticipated. Most of the EUV masks today are built in large captive mask operations.
But because of that, they tend to do more outsourcing on their mid-range nodes to commercial mask suppliers. And we are seeing that trend. And we also do have development relationships in EUV. So that's certainly an inflection point for mask technology that's evolving right now. Mask complexity continues to increase. This means customers putting more content on the mask, more complex layers, more difficult masks to build, and more masks per design. That has run its course probably over the last five to seven years, but still, mask complexity continues to creep up. On the device side, SOC or system-on-chip trends has resulted in a lot more layers per design. And interestingly, particularly non-critical or so-called semi-critical layers, for example, increasing the number of implants and SOCs for multi-mode devices is driving a lot of mask count.
Additional gate oxide thicknesses are driving more mask count for a given SOC. But at the same time with that, we're also starting to see a kind of SOC disaggregation trend. Some people call these chiplets, but not really appropriate here. We're starting to see companies start to break apart the SOCs into individual chips and then reintegrate them through advanced packaging. Also a very, very good trend for masks because it will drive design activity as that disaggregation continues. So on one side, we've got more complex masks for tuning things to make SOCs work. On the second side, we're starting to see a break apart into more design. So that's a healthy trend. And finally, I already talked about it, cloud-to-edge computing migrations pushing a lot of high-performance computing and higher-end displays actually out to the edges of the network. And that's driving design activity as well.
On the FPD side, similar to what we saw in IC lithography maybe 10 or 15 years ago, masks are starting to carry more of the lithography performance load. These lithoscanners and the resists and other things are not able to deliver the performance on panel that a lot of customers need. So they look more and more to the masks to deliver that capability. And that increases the value of the mask in the patterning step. As Peter mentioned, AMOLED, very, very critical segment for us, and we're definitely technology leaders there. It was originally very expensive displays, mostly targeted for the premium smartphones. Now, thanks to China's influence, starting to migrate down to more mid-range nodes, a lot of capacity coming online for AMOLED.
Also some interesting trends on the TV display side, LCD TVs in particular, kind of a sleeper segment for many years, except for the panel size increases. This is the G10.5 business for LCD. But recently, over the last couple of years, a lot of innovation coming on standard LCD TVs, how the lighting is done, quantum dot lighting and quantum nano rod lighting. These all require more complicated compensation circuitry, which tends to drive more mask demand. Higher resolution on the display that's supported by Mini-LED backlighting. When you have dynamic backlighting with Mini-LEDs, which requires circuitry to operate, there's value in pushing the resolution harder for LCDs. So some really interesting things happening on display. We expect we'll drive litho-complexity, mask complexity, and mask counts going forward.
Then, of course, gaming, all the remote work stuff has kind of put a real strong kick into the FPD, called FPD display side of the business. So what is one of the things we see as a trend or a technology area to focus on? We use this term Photomask Velocity. Because of the amount of design activity going on out there and the desire for the customers to get their products to market quickly, photomask has always been very, very cycle-time sensitive. Oftentimes, wafer output of new designs is waiting on photomask. Fab output lithography is usually the most expensive module in a wafer fab or a panel factory. So if anything's waiting on photomask, it limits the output of the entire fab and the ROI of the fab. So photomask is always very cycle-time sensitive.
But I think more than ever, at least since I've been working in this field, which is quite some years, the speed to get masks to market and the importance of cycle time is very critical. And we call it velocity because we're not just trying to get masks out there fast, trying to get them out in a particular direction. As Peter said, it's a focused investment in technology strategy to win market share on the most opportunistic customer opportunities in regions where we really want to excel. So speed of design turns taking center stage across both IC and FPD. Those things are difficult because when masks are complicated and there's a lot of masks per device and the yields are lower, it's difficult to get them out faster. And there also are capacity constraints for mask-making equipment at the mid and legacy nodes in particular.
So, some of the things we work on to enable Photomask Velocity. Customer collaboration is very critical, and we have a lot of tools that have helped us streamline process development. Yields in photomask are very, very important. Self-correcting tools. We have a lot of development programs on those and more efficient, simpler mask processes. Tuning masks very, very rapidly are critical now because the mask is so embedded in the performance of the device. Our ability to react to customer changes quickly, tune the mask, and get it back out, we call them respins, very, very important, so we've developed suites of technologies and kind of response models to allow us to do that effectively, and then getting processes fanned out around the network, localizing the technology, particularly relevant for China, is quite important in this area.
A lot of our investments, or at least thinking on the technology side, particularly for IC, are targeted at delivering this sort of world-class mask velocity. So how has technology kind of fed into the business? This is a so-called virtuous cycle here where we develop high-end technology and capability. And you'll notice on the right, it's had its positive impact, particularly on the FPD side, which has been our emphasis. We develop that technology. We use it to capture a share of growth. It could be regional. It could be across customers. Again, it's a focus strategy that flows into our business performance, and then we reinvest that into high-end technology. So somewhat standard sort of model, but we feel like we know how to do it, and it's been demonstrated.
And as the bar charts on the right show, particularly for FPD, which has been really our tech emphasis, it's paid off well in this model, has shown to be successful the last couple of years. So how about China's strategy overall? China is really changing the landscape in our sort of markets, both IC and display, quite significantly. So it's worth emphasizing and bringing it up. So the domestic investments add local capacity to China, and we know those investments are substantial. Now, the number one LCD display supplier in the world. The goal is to win share for the domestic producers or foreign nationals operating in China and support, of course, the national agenda as well. What that does, it puts pressure on the foreign incumbents to improve their products, sustain their advantage, and their share.
So both of those two tend to drive new product development, design activity, and fab expansion. So they are working together at the highest end, foreign competition respond. And inside China, it's really adding capacity and moving up the manufacturing and product curves. So the other factor we see, integrated circuits and displays, really have become center stage in economic security landscape, particularly integrated circuits. And photomask, because they're the first thing and only thing really that brings a design into a factory, really stand at the crossroads of this design and enablement trend. Some examples we've seen that show this: 70% of LCD, 50% of AMOLED capacity expected to be in China by 2021. Advanced TV shipments, which I talked about the importance of those for masks, expected to double by 2024. And the mask count per TV also on path to doubling.
Foundry capital spending, which is in response to this design activity increase, increased 35% in 2020. These are very strong trends, again, for photomask and for our strategy. The Made in China 2025 growth, we see that. I've been following it, a little bit less than expectations. As Peter said, we think that will only intensify the efforts of China to win share, global share in both IC and also to pivot more and more emphasis on AMOLED because the LCD side has really been done quite well. Not a lot more to do there. In summary, we see photomask enabling technology for IC and display. Key inflection points: China investment in global share goals. LCD is pivoting to mobile, which is AMOLED, and we're seeing strong design across multiple nodes.
Good sweet spot on mid-range IC nodes, which is important for masks, and also pushing leaders to innovate, driving mask content. And our investment strategy is aligned with these trends. We develop focused processes, very critical on time to market, mask velocity, cycle time. And as far as trust, we are the leading trusted partner in the marketplace today. So with that, I will now pass it off to John to talk about the financial model.
Thanks, Chris. Good morning, everyone. This is John Jordan, the Chief Financial Officer for Photronics. Just to start off with a few points to reprise 2020 and emphasize some of the highlights that we had during the year and discussed on our earnings call. We mentioned on the call, each of the past three years has set a new revenue record, and the repositioning of the business we did to anticipate the market growth a few years ago achieved the business run rate during the year that we targeted nearly three years ago. Although the two quarters in 2020, disrupted by the COVID virus and the Huawei limitation, kept us from hitting the $630 million total year target in 2020, we're expecting to be well in excess of that level in 2021, and in a few minutes, we'll review our target model that highlights that.
As Peter pointed out, overall revenue grew at a compound annual growth rate of 11% over the past three years, and FPD grew at a CAGR of 24% from $100 million - $191 million in 2020. Revenue for products shipped into China grew at a 69% CAGR to $208 million, a nearly five times multiple of what it was in 2017. Operating income of $64 million in 2020 improved 23% over the previous year, and it's nearly double the OI we had in 2017. As revenue continues to grow, operating leverage from the existing investments will continue to improve the operating margin, as we'll see when we review the target model. Cash generation has historically been strong, with operating cash averaging in excess of $100 million over the last three years.
CapEx payments in 2020 were primarily to finish up the payments on investments in China and to make deposits on the next phase of investments to be delivered in 2021, and we repurchased over 1.7 million shares of Photronics stock during the fourth quarter for total repurchases during the year of 3.2 million shares for $34 million. The new China operations achieved a profitable run rate in Q4, and we expect that run rate to remain profitable. If we go to our target model, the target model demonstrates the operating leverage that revenue increases provide. We anticipate that based on these revenue levels in 2023, we would achieve impressive growth in each of our operating metrics. Obviously, we can't anticipate whether there will be geopolitical disruptions or other effects beyond our control that would interfere with achieving these results.
But with the investments that we've made and commitments for tools to be delivered in the coming year, our knowledge of the markets and growing market shares, we believe these target ranges are well within our scope. The mix of business will obviously affect the ultimate result. And in a business where backlog is minimal, these projections represent our best estimates of the revenue range and operating results we can anticipate. As the gross margin improves with higher revenue, increasing technology progress, and progress on supply chain, our firm cost disciplines will ensure that that improvement drops to operating income. Our strong cash generation will continue from increasing net income and depreciation. And we anticipate continuing capital investment of $70 million -$75 million annually. Improving net income and continued share repurchase should put us solidly in the dollar per share range, plus the ranges shown on the chart.
Our capital deployment strategy is clear by its execution. We invest where required to meet demand and technology developments, organic growth. Our investment in China was right on target to address demand growth and was supported by several major customer commitments. We've established clear target hurdle rates for all investments to improve the ROIC. Investments are not made unless they clear the hurdle rates and are supported by customer commitments. We stated on our earnings call that we are "actively open to M&A." We maintain close contacts with our customers, suppliers, and competitors. We will always seek to expand our scale in the photomask sphere when a transaction is available that's economic. That is, the price is reasonable and the target would be accretive to our existing operations.
Since we began the share repurchase initiative in 2018, through October 31st, we've repurchased almost eight million shares for $79 million and repaid $115 million of convertible debt. We strengthened the balance sheet and reduced average outstanding diluted shares from more than 79 million to less than 65 million at October 31st. Our target model anticipates continued repurchases and diluted share reduction, so as we've outlined, our tight discipline, the cost management, and operating leverage retain the benefit of revenue growth to improve bottom line earnings. As we saw in the target model chart, that operating leverage should deliver strong margin improvement to potentially exceed our defined targets. Continued strong cash generation provides the capital needed to continue driving top line growth and increasing market share and continue returning cash to shareholders and improving EPS.
We target investments supported by customer commitments to address demand and technology trends and customer needs. By screening investments through our investment criteria, we will continue to improve ROIC. So in summary, we expect our revenue growth to continue at a 5%-7% compound annual growth rate, expanding market share. Our operating leverage will ensure that that revenue growth results in increasing earnings growth. And our strong cash flow generation will continue to fund the investment necessary to lead the market while returning cash to shareholders and increasing EPS. I'll return the call to Peter to discuss our compelling investment thesis.
So again, thank you, everyone, for your attention. We hope we've answered the question: why invest in Photronics? Well, you're investing in the market leader who is actively putting the right technology in the right place at the right time to help grow the market, doing it in a way with an accretive bottom line impact, with a business model that funds growth as well as returning cash to shareholders. This concludes our prepared remarks. And now I'd like to turn the call over to the operator for questions.
As a reminder, to ask a question, you will need to press star one on your telephone. To withdraw your question, press the pound key. Please stand by while we compile the Q&A roster. Our first question comes from the line of Tom Diffely from D.A. Davidson. Your line is now open.
Yes. Good morning, and thank you for all of the nice information today. Maybe Peter first, if you could talk a little bit about more of how the photomask market has changed over the last decade or so from just a low-cost provider situation to more of a partnership with the customers. And how do you think that bolsters your position going forward?
Yeah. Well, our market is a very dynamic one. There's lots of moving pieces. The number of—so first of all, we look at the IC business. The number of customers that are still driving Moore's Law is narrowing. We're really down to—we're down to three, arguably maybe five, but it's a very small number. But these customers, the captors are all cost centers. They don't have a high degree of focus on mask cost. But what they do need is, as Chris, I think, elaborated, support at what they see as trailing nodes, which right now would kind of drop in the 10 nanometer -28 nanometer bucket. And they don't have the capacity to build all of that. So we need to be there with a product offering and responsiveness that allows us to best support them.
As far as the remainder of the market goes, the mainstream is what it always has been. The most mature nodes really haven't changed much. But then there's the sector that Chris carved out as kind of like the mid-range, which is, I don't know, today it's the 45, 55, 65, 90 nanometer node where there's a lot of growth. The growth tends to be. It's really focused. The capacity that's being added is being added in Taiwan and China. And mass cycle time there is really paramount to success. The ability to respin designs is critical. Again, local proximity is necessary. So given our global footprint, we think we have the best position of all the merchants to exploit that trend. And the geopolitical nonsense and nationalism is really raising the bar on data security.
Again, given we've been doing this for more than 50 years, trust in Photronics is synonymous in the marketplace. So our core competencies, again, are serving the current dynamics in the market well. Which just gets to FPD as a tale of two markets. It's an LCD market that is really consolidating. The bulk of the value in it is at G10.5. Of course, as I elaborated, we've not been long in that market, but we're already the strongest player with the highest degree of market share because we put the technology where it needed to be to support the customer base, which is China. Then, of course, there's AMOLED, which is an unbelievably dynamic market. It's primarily used in mobile phones today, but it's broadening into laptops on one hand.
And the AMOLED technology is showing up in large TVs, the high-end of the TV market on the other hand. There is no doubt by any metric, technology, market share, customer relationships, that we are the leader and we're benefiting. The companies benefiting dramatically from that leadership. And if there was a way, and I think Huawei divesting their Honor brand, a way to get mobile displays out of the 5G food fight, this is clearly good for us. So in the end, it's good for us, period. But getting that market, it's not military, it's commercial. Mobile phones, everybody carries around in their pocket. And it doesn't compromise network security. So getting that separated from other concerns is also clearly happening. The election results in Washington are likely to improve that dynamic, although time will tell.
Anyways, our market position, our geographic location, our technology all play very well into the AMOLED market, which is what will drive the display industry forward from rigid to flexible, now from flexible to embedded touch sensors to large screen TVs to laptops, etc., into cars. It's everywhere. And it will, because of all the superior performance attributes, increase its footprint in the market and in our daily lives. So that's kind of those are the trends, and that's how we see our position within them.
Okay. No, appreciate that. That's helpful. Maybe on the technical front for Chris, when you look at all the different types of displays that are in the funnel that we expect to see over the next few years, how does that impact what your capabilities are as far as serving the different types of displays that are coming out? And is it meaningful from a, either we need to buy new equipment or you have to develop new processes? And maybe potentially the potential revenue you can get from those new types of displays coming out?
Yeah. Thanks, Tom. I think on the equipment side, we're in good shape. There may be some point tools we need for capability over the next couple of years. But we were the first to invest in the P800, the Mycronic high-end writer. That was targeted at AMOLED in particular capability and resolution. And we've done some other incremental investments, particularly in our Tainan location where we do most of our FPD R&D. So I think we're pretty good on equipment. Most of the technology is now process development related: blanks, films, resists, and integrating those things in with the equipment to make a functioning process for the customers. That's where the majority of the emphasis is now for R&D. And then as far as the different products coming out, it is vast.
I mean, we have tables and tables of new potential display technologies from, I'd say, 15-18 different flavors of things and how LEDs and Micro-LEDs will be integrated, QNEDs, many, many different things. They all have different compensation circuitries, different floor plans. I think some of those will be successful. Some won't be. As always, there's a large sorting process now in display technology, but generally, what we're trying to do there is partner, collaborate with customers to accelerate those R&D learning cycles so they can see if these things are going to be viable, and cost is a major issue in IC and FPD, but I would say the pressures are stronger, actually, in FPD, flat panel display.
So one of the challenges with these complex new display technologies is when you get to 20, 25 different lithography layers to make a display, you really have to work closely with the customer to drive the integrated product cost down. So try to not overcomplicate things, try to do what's necessary to get yield, but really have to keep an eye on the lithography flow costs because that segment is extremely cost sensitive. So that's another area we focus a lot on with customers and also, of course, for our own business health.
Okay. Great. Appreciate that. And then final question for John. How much variability is there in the margin between products and customers? And to get to your 50% bogey, does that require a specific mix?
Tom, the mix will be all over the place. But we know there's variability in the product mix in both FPD and IC. And it's a strong guess to try to project out three years. We've got a pretty good idea where we'll be in 2021. But in 2023, it's a mix of a lot of different things. So yeah, there's variability in the margins and the mix.
Okay, well, thanks again for the presentation today.
Thank you, Tom.
Thank you. Our next question comes in the line of Patrick Ho from Stifel. Your line is now open.
Thank you very much, and nice presentations. Maybe for either Chris or Peter, following up on Tom's question, you talked about technology collaboration as an aspect in terms of the development work, particularly in the high-end where masks are becoming more complex. Can you provide a little more color on these collaborations and how much value you can extract in terms of pricing from your customers to get that value? And secondly, how much control on the IP front you can maintain, given that much of the leading edge IC technology development is on the captive front? As you guys know, that's an area where these customers tend to want and keep control as much of the IP as they can.
Yeah. So I think Chris and I should both take a stab at that, Patrick. And I think the way you've framed your question and the way it already anticipates the answer, right, where we are the undeniable technology leader whether it be merchant or captive is FPD. So this is the area of our business where our focus is using technology as a lever both to drive market share and to drive higher value in mask pricing. And I see there are challenges, a different one, which is basically making sure our product line is appropriately articulated and our service is what it needs to be to support the demand that the captives present to us. And we don't have any EUV scanner in the basement. So the thought process that we would drive the technology roadmap on IC is wrongheaded.
We don't need to do that, and we don't aspire to do it. We do have an active R&D program in EUV. I'm pretty sure we ship more EUV masks than the rest of the merchant combined. Again, these are the crown jewels of the largest players in the industry. And they will use us for specific projects, but from the standpoint of driving the technology roadmap forward, they do that themselves. So Chris, I think there's a lot in that question for you to answer, including how we protect IP, etc.
Sure. I can make a few comments on that. So most of our technology collaborations are governed by either formal joint development agreements or somewhat informal agreements. But they do have IP protection, confidentiality provisions, who owns what, sorts of language and verbiage in there. And so that's the kind of legal or contractual tool we try to use to protect our intellectual property and then also the intellectual property we get from customers that we require to develop the processes. On the captive side for IC, most of our job, as Peter alluded to, is matching their processes well. These are kind of sub-nanometer sorts of matching exercises because they do not want to see a different performance on wafer for one of their masks versus one of our masks.
Most of the technology we develop for captives is how to match effectively in the presence of limited information because, as you said, Patrick, they don't share everything. So we a little bit have to find our way there. We develop technology to help us match processes. It's not reverse engineering, but it's other tools we use to match their processes precisely without them needing to share their intellectual property with us. For companies that rely on us to co-develop with them when we are actually developing the process of record on those, as I said, they're usually covered under joint development agreements. We usually do not share the specifics of our mask process with the customer, only the characterization and measurement data that allows them to effectively use the masks. So that's kind of how we cover that.
On the FPD side, as Peter said, it's a different landscape. There's really only one captive, and then there's a few JVs. So in this case, I would say particularly protective of our technology. We do share some elements of things, but mostly we focus on providing the characterization and measurement data to the customers that they use in loop, closed loop with us to converge their processes. But we're not that forthcoming, frankly, with the nuances and details of the mask manufacturing with most of the customers on the FPD side.
Yeah. The area in IC, yeah, Patrick, I just want not to misconstrue. The area in the IC business where our technology position is really generating value for us is not with a large captive, but it's with the China market in general and some specific memory customers globally. So in this case, it's not EUV. It's still largely, with one or two exceptions, it's not EUV, but it's largely optical lithography. And we, as a merchant, have the best technology we believe. And our footprint in China, I think in particular, is creating a lot of value for us in this respect.
Great. That's helpful. Maybe as my second question, maybe for you, Peter, in terms of your comments and Chris's comments about the Photomask velocity and how that's a differentiator for Photronics, I do believe in that. I do believe in the technology leadership you guys have. But given that the mid-range or slash mainstream market has seen a bit of a lift in the near term and probably over the next several years as we look at automotive devices, industrial types of devices growing in nature, the velocity helps getting that incremental business initially because you're able to deliver technology at a faster rate than the competition. How do you position yourself when the rest of the marketplace, particularly on the merchant side and particularly in China, when they start catching up?
How do you defend some of those positions and maintain your "market standing" with those customers that you achieved early on with velocity? But as the competition catches up, how do you maintain that leadership?
Yeah. It's a really simple answer, and that is leading in the mainstream market. That's our roots. I go back to the growth of Photronics chart. We became the number three global producer based on our ability to service the merchant mainstream market better than anybody else. That's us. That was us until 2010 in the IC business or 2011. So we have business processes in place and tools we've developed internally, software tools that allow us to do a better job of servicing customers than anybody else does. That's who we are. That's still how the industry sees us. So we're quite confident that as the mainstream market, particularly in China, grows, that we're going to be able to have an outsized footprint in it.
The challenge really is getting the tools in place, which in our case will have to be a blend of new and old in a way that we can approach that market cost-effectively so that the financial performance is additive. But getting the share, don't worry about that. Getting the share at the right price point is our challenge. And it will be a blend of what's old and a blend of what's new to do that most efficiently.
Yeah. Patrick, and I can just add the comments. We do think pretty hard about so-called kind of glue we can put around photomasks to make them, let's say, harder to duplicate because eventually, as you correctly point out, competitors and others will figure out how to match the mask, and the same way we match the captive, they are going to figure out how to match the signature and profiles of our masks and be able to provide them. So we look at other, let's call it, applications and technology services glue around those masks that even if they match the signature, our mask still will provide better yield and better performance on panel or on wafer than what the competitor can provide. So that's an active area of development for us and focus.
Great. That's helpful. And final question from me. In terms of your operating model, maybe for John, you gave the outline of the $725 million and $750 million revenue goals. The operating margins look very, very healthy. You guys have shown a consistent track record of managing expenses and even cutting costs. At the same time, it seems like the biggest lever in that is gross margins. If you can give a little more color on the key variables on the gross margin line, and I know it's a lot of variables of mix, volume, pricing, things of that nature, but what are some of the key items there? Are there additional cost cuts on the gross margin line you can do to get it potentially even higher than the targets you put out there?
Or is this simply just getting revenues to those metrics to get to those gross margin numbers?
Yeah. Good question, Patrick. And as you know, we've got a high level of fixed costs. And once we pass that fixed cost threshold, the variable costs are pretty limited, primarily material and freight. So the contribution margin, as you're familiar with, is pretty high. So it's really incumbent on us to get our revenue levels to those to get our revenue to those levels. And with the lower variable costs, those margins will develop. So yeah, there's upside to it. We've tried to be conservative with our estimates of the material cost. And with the work we're doing on supply chain, we've got potential to improve on those as well.
Yeah. So I feel compelled to chime in on that answer, right? There's not a day that goes by that I don't work on cost reduction, me personally. It's what this company does. There is always, every year, a tug-of-war between price down and cost out. Normally, that tug-of-war is a draw. There's a little more, a little less of one weight on the rope in one year versus the next. Beyond having that tug-of-war be a draw, as John correctly stated, right, the leverage in the business is on volume.
Great. Thank you very much.
Thank you. Our next question comes from the line of Gus Richard from Northland. Your line is now open.
Yes. Thanks for taking my questions. It's always tough after so many good questions. I did want to talk a little bit about the industry and capacity utilization at the high-end, mid-range, and legacy. Could you talk a little bit about that and what you're seeing currently?
Yeah. So I'll start with my favorite topic. But aside from last quarter and the month of November, for the last several years, our FPD business has been sold out. So we're adding capacity always into a sold-out business model. That's sort of kind of besides the qualification ramp period in Hefei. That's the ideal situation for our business, right, because the likelihood that you can load an incremental capacity, the best indicator of that is, are you full? Because if you're full, there's more demand for what you sell than what you can make. So this has been our situation in our FPD business, which is wonderful to attribute to the entire team in that part of our organization working hard to make sure we stay out in front of the competition.
As far as our IC business goes, as I mentioned earlier today on the earnings call, we've seen a rotation of demand in our mainstream business from the U.S. and Europe to Taiwan and China. We do cross-qualify customers across our global network so that if we have more demand in one location than we have capacity to address it, we can respond to variation in demand. The challenge for us comes when there's a rotation in demand like there has been and is, and likely will continue to be in the mainstream. In order to address that, we are sold out, right, right now in Taiwan and China in the mainstream part of our business, right? We're sold out. We're building some of that business elsewhere.
The challenge for us is rotating the capacity now to where the demand continues to grow. So sold out mainstream in Asia, not elsewhere. At the high end, if you look at the year-over-year trends, right, we hit a soft spot this particular last quarter, right? But as we move into next year, we have a high-end tool coming on board in Hefei. We'll see how the market goes, right? But clearly, with the three-year targets we have out there, that looks like a sold-out business model at the high end as well without capacity addition. So that's kind of how we are.
So I was specifically trying to ask about the IC business and not just your capacity, but the overall industry capacity. At the high end, it appears as if EUV masks are absorbing a lot of the write capacity at the captives, which would suggest that they have to outsource what you mentioned. And at the low, there's just, we're running out of eight-inch capacity in the fab and wouldn't imply that there's a lot of activity in the more legacy products for power, etc., etc. So when you think about the IC market as it pertains to high-end, mid-range, and legacy for the whole industry, what do you see? And is there a potential to absorb all the capacity that's out there and create, particularly in the legacy, an increase in pricing?
Yeah. Yeah. So Gus, thank you for reframing that. Where the growth looks like it has the potential, I mean, historically, mask makers and the industry in general added capacity at the high end as the industry moved down nodes. And it created a situation where, as the customers migrated, obviously, the demand from the last node would diminish. But at the same time, you'd use some of the capacity you had installed for that prior node to build the semi-critical and then ultimately non-critical layers. So this is kind of like the dynamic. For the first time in my career, anyways, and I've been doing this for 35 years, it appears that the high end, the mid-range, is going to have more demand than there is industry capacity. And I'm answering that not from a Photronics perspective, but a global perspective.
And that's problematic because it's very difficult, as I mentioned earlier, to buy new tools to address that market segment because you can't do it cost-effectively. So in this world, should it materialize as we think and as I think you're suggesting, and I think others have suggested it as well, it'll create a circumstance where the demand-supply curve will invert, which will give us pricing leverage, which is a rare thing in the photomask business.
Do you see that price leverage in the mainstream and legacy rearing its head in the current fiscal year, or is it too difficult to predict?
Yeah. I think depending on, I mean, the driver of this is China, right, on one hand, and some of the applications you elaborated on the other. We could easily, I think, start seeing it the middle of mid this year, right? There's a tenuous balance right now globally. And if China grows the way most anticipate, it's not going to be at the high end. It's also, by the way, not in the real legacy business. It is the mid-range, the way Chris articulated in his chart. That's where the growth is. And that is where a supply-demand inversion is in the works. You can see it. You can see it coming. So absent any global disruptions, we could easily see it before the year is out.
And then sort of how much of your revenue on the IC side is from the captives outsourcing some of their demand because the write times for masks have gone through the roof, and they just don't have the resources to build what they need?
It's funny. I think we had a chart on that that we removed from the presentation. That part of our business is between 10% and 15% of the total, depending on the quarter. That IC total revenue, total revenue, 10%-15% of total revenue.
If this price stability plays out, can the industry and the mainstream add capacity cost-effectively given I would imagine that a lot of the mainstream tools at this point are fully depreciated that are installed? Can you add new capacity and be competitive at current pricing?
Very tough. Very tough to do. They're very, very tough to do. Have to add a point tool or two, perhaps, and then the numbers still work. But new lines, no, absolutely not.
Got it. Got it. And then I think that's it for me. Thank you so much.
Thank you. Our next question comes from the line of Quinn Bolton from Needham. Your line is now open.
Hey, guys. Thank you for the presentation. I wanted to just sort of, I guess, follow up on Gus's sort of line of question, and maybe this is more of an industry question than a Photronics question, but I guess I'm a little surprised with your longer-term growth CAGR to the 2023 of 5%-7% and your comments you think you can grow twice the rate of the industry. That would sort of imply that the mask industry grows only sort of 2.5%-3.5%. I guess I look at IC unit growth or revenue growth forecast over the next three years, and I would think that forecasters are probably projecting 5% or 6%. So my first question is, am I right thinking that you're looking at kind of a low single-digit growth rate for the overall photomask industry?
And if that's the case, what are the factors that cause it to grow less than semiconductor revenue? Is it the pricing dynamic that you were just discussing, or are there other factors going on there? And then I've got a follow-up question.
No. It shouldn't, right? I think to clarify again or reinforce, we expect our growth to be double semiconductor industry rates, whatever they might be. But I'm a skeptic of what forecasters have to say, right, personally, right? We look at historical semiconductor industry growth rates, and we circle what we think will happen based on what's been happening over the last five to 10 years. If they jump up more, we'll see more growth. The photomask industry tends to grow more or less in lockstep with the semiconductor business. So we don't know what that's going to be, but if the more optimistic numbers turn out to materialize, then we should see double that rate, whatever it happens to be. When we gave our last three-year targets, we openly stated we put a downturn into our projections.
And people told us the industry had changed and there was not going to be downturns anymore. Well, that turned out not to be the case. So we don't know, but we're just looking back at recent historical numbers and trying to peg those. But yeah, if it's better, we'll do better.
Got it. Okay. Thanks for that clarification. I guess the second question I have is, I look at the operating leverage in the model, and it's very significant. You highlighted the high fixed costs, the low variable costs. My question would be, let's assume semis grow 5% or 6%, which hopefully, based on everything you've outlined today, would say your CAGR might be closer to 10%. Do you have the capacity today in place to support upside to, say, above the $700-$750 model, or would you need to bring in significant additional mask writers and inspection equipment, which would then raise the fixed cost? I mean, it's a tremendous operating model to get to $700-$750. I'm just wondering, at what point above $750 would you have to meaningfully increase your fixed costs to support higher revenue?
Yeah. If the CAGR was running along at that rate, we would be adding lithography and inspection. But again, right, the last wave of investment, we had to build factories. We had to buy and equip entire we had to build lines, right? If the industry is growing along at a 5% or 6% clip, it would give us the luxury of, yeah, making investments, but not needing to make them full-boat. So we think we can preserve that business model in a higher growth scenario. But clearly, if we're at 10% or 11%, we're going to be bucking up against the type of margin profile that's at the top of our projections, which is pretty good for this company. Haven't seen margins looking like that for as long as I've been here, which is 13 years. But we wouldn't be satisfied, right?
We'd love to see 30s at gross margin. That would be great, right? I'd like to see that. The scenario you're describing might be able to get there.
Great. Thank you for those additional comments, Peter.
Thank you. At this time, I am showing no further questions. I would like to turn the call back over to Troy Dewar for closing remarks.
Yeah. Thank you again for joining us today. I really appreciate your time and interest. I hope you found this time helpful for you. Just to let you know, I'm available for any follow-up questions you may have. So once again, thank you and have a nice day.
Ladies and gentlemen, this concludes today's conference call. Thanks for participating. You may now disconnect.