2010 ISS Blue Chip Panel: A “Perfect Storm” (Part 2)
By Scott Landstrom, SCI Consulting
“The Perfect Storm” is a modern-day metaphor for a rare combination of events that create a rare, unusual or unique outcome. On January 12, 2010, at the annual SEMI Industry Strategy Symposium (ISS), attendees experienced a “Perfect Storm” as a unique Executive Panel session focused on “The Structural Evolution of the Semiconductor Equipment Industry.” Five of the top six revenue firms in the North American equipment industry participated. The panelists were:
Instead of posing the questions to the panelists directly, I recruited luminary thought leaders to pose key questions via recorded videos. Following the analytical approach championed by Harvard professor Michael Porter, these eight luminaries challenged the panel with key inquiries regarding the evolving equipment industry, and the many business model challenges to ongoing competitiveness in this sector. The luminaries who asked questions were:
- Jim Morgan – former CEO and COB of Applied Materials
- John Osborne – current Board member Amkor
- Paul O’ Neill – former Secretary of the U.S. Treasury
- G. Dan Hutcheson – CEO, VLSI Research
- Danny Ainge – president, Boston Celtics
- T.J. Rodgers – COB and CEO Cypress Semiconductor
- George Needham – COB, Needham and Co.
- Bob Boehlke – former CFO, KLA-Tencor
I organized and moderated the panel, and Michael Wright, CEO of AISI Inc., was the co-moderator. Highlights of the executives’ analyses are presented below. (Note: This article is Part 2 of a two-part series. To see Part I, please click here).
Luminary # 5: Mr. Danny Ainge
General Topic: Historically, the United States semiconductor industry, like the National Basketball Association, has benefitted from an international flow of skilled labor, relocating to the U.S. that has enriched the “talent” in its ranks. For the semiconductor industry, this “brain drain” in our direction from overseas is being forecasted by some experts to reverse over the next 3 to 5 years, with 200,000 skilled technically trained workers projected to relocate back to China and India alone. How will the semiconductor equipment sector deal with this “reverse brain drain” as talented management and technical ex-pat staff relocate back to their home countries, potentially damaging the competitiveness of the U.S. equipment sector in the process?
Overall Summary: All panelists seemed to be in agreement that this trend is real, and that the magnitude of the projected “reverse brain drain” was probably in the right ball park. It was brought up that while the bulk of the semiconductor device manufacturing has migrated to the Pacific Rim, accelerating this process for device manufacturing related workers, the country with the largest share of overall semiconductor equipment industry revenue is still the U.S., so the argument was raised that international talent that wants to be at the epicenter of this sector should be attracted to work here. The point was also made that as U.S.-based equipment firms become more international in their site strategies, which is clearly happening, their ability to compete for talent in those regions, and either immediately or eventually station them in their home nation, should inherently increase as a direct result. The role that Washington should be playing on the national competition to attract international talent, versus the role that Washington is playing, was mentioned several times in a critical light by a number of panelists. While the U.S. government has been making it increasingly difficult for skilled, educated immigrants to obtain work visa’s, it was stated that competitive nations like India are actively marketing/recruiting their foreign-based skilled workers to return to their native country, and offering incentives in some cases to do so. Despite these obstacles, several panelists commented on the issue of not underestimating the ongoing cultural allure of working in Silicon Valley, in particular, as a fairly unrivaled combination of technology assets, cultural awareness and sensitivity. Moreover, the compensation scale of the U.S. was brought up as being the most lucrative in the world, translating to the best locale for skilled international workers to build a “nest egg.”
Selected Panelist Quotes:
Randhir Thakur: “As someone who has immigrated to the U.S. to work here myself, let me start by saying I have worked in India, Europe, and the U.S., and find this to be unquestionably the best, most accepting environment for an ex-pat to work within. Secondly, we must recognize that countries such as China and India are actively putting in place incentives to attract these talented professionals to return to their cultural roots, so that translates to us needing Washington to respond in ways that make it easier for international talent to work here, not harder, which has been the recent history. Moreover, from an international "hire and retain" perspective, multinational firms such as Applied Materials, with both geographic and served markets diversity, are uniquely able to both recruit and relocate international technologists while keeping them within the firm, versus largely domestic competitors. This provides opportunities for multi-nationally sited firms to identify outstanding talent in-situ, recruit and position them for a work assignment phase within the U.S., and then relocate back to their native country— all without leaving the company."
Rick Hill: "There's one characteristic of the American culture which makes this a very different place to work. It's especially noticeable here in Silicon Valley. In our workplace environment, there's no inherent hierarchy in our business structure: it's ok to be different, or as in the Japanese expression, "to be the nail that sticks up". Rather than being "pounded down," Silicon Valley is really a meritocracy where it's the value that you bring to the table that matters, not the way or the style in which you do it. Most other countries are not that accepting of behaviors outside of the group's norm. So, I think it’s our tolerance in this area that will continue to be a competitive advantage and will draw the best and the brightest to work in the United States."
Rick Wallace: “This issue is more germane to our country’s competitiveness than our industry’s competitiveness, because this industry has learned to adapt and respond to this trend with a more international footprint and approach. Domestically, however, I think we have lack of focus on the long-term ramifications of this issue here at home. Congress is great at worrying about the next election, but not the next generation, meaning specialized and very important issues— such as how we continue to attract and retain international talent as productive and contributing members of the U.S. economy for the long term— often get little or no mindshare in Washington.
(Post ISS Addendum – Rick feels strongly about this specific subject, and wished to add the following: “Further, this ‘brain drain’ issue is a much bigger competitive threat to my colleagues in small- and medium-sized businesses here in the United States, who rely on highly educated, specialized talent for innovation. This is a big problem in the long term, because it presents a threat to our country's future competitiveness. Because of this, I am a proponent of efforts to reinvigorate science and engineering education here in the United States, which will enrich the technical depth of our own home-grown talent pool and foster future generations of innovators to sustain U.S. competitive leadership in the global economy.”)
Steve Newberry: “While it is clear that our country, and our specific industry, needs to do a better job of marketing against international competitors for global executive talent, I think there may be a larger issue that we may be missing here. On one hand, we are certainly competing with other international semiconductor industry firms for such talent, and face all the issues discussed. Conversely, we are presented with “cross-industry” competition for the “best and the brightest” that— at least in my mind— is an even bigger issue. Whether it is SEMI, or the SIA, we need to do a much better job of laying out the allure and technical challenges still remaining in this sector, versus other industries such as Clean-tech and Solar.”
Luminary # 6: Mr. T.J. Rodgers
General Topic: One of the key components to the economic aspects of Moore’s Law has been periodic shifts to larger wafer size. In the last transition to 300 mm, there was a combination of negative factors that collectively resulted in this transition being hailed as a “train wreck” for the equipment industry. Besides taking nine years of “stops and starts” to finish, resulting in tens of millions of dollars of wasted R&D, the net share of the value apportionment allocated to the equipment industry set a new historic low. Many process equipment firms maintained wafer-per-hour unit throughput in their 300 mm version tools, resulting in 2.6X the total square centimeters of silicon capacity delivered, yet were only able to achieve a 30% price premium in their ASP’s over the 200 mm tools, effectively cutting the CAPEX dollars per square cm of silicon in half. How is the equipment industry going to protect against a similar, draconian experience when the industry shift to a 450 mm wafer size?
Overall Summary: There was wide agreement that the “stop/start” nature of the 300 mm transition was brutal, and led to R&D inefficiencies, and that the value apportioned to the Equipment industry was a fraction of historical levels during past wafer size transition nodes. It was pointed out by several panelists that this value that was not allocated to the equipment industry was also not retained by the wafer manufacturing customer base, but was instead passed down the value chain to the end market consumer, so there were “value apportionment” issues that negatively effected both equipment and device sectors. Additionally, it was mentioned that the overall general problem is R&D efficiency, and that 450 mm is just one key cog of this larger, comprehensive business model problem facing equipment manufacturers: the skyrocketing costs of the R&D required to keep technological pace with “Moore’s Law.” Further, it was pointed out that this may be the first wafer size transition that the majority of the device manufacturing market can not afford to implement, which could create substantial inefficiencies in the pace of technological development of equipment, should it be necessary to support a bifurcated platform development effort for two wafer sizes concurrently. Finally, there was healthy skepticism by nearly all panelists that the current business model of “who will fund” the transition in a manner that allows equipment firms to derive a positive ROI for their investments in a 450 mm transition, is simply not either in place, or within view, at this juncture.
Selected Panelist Quotes:
Rick Wallace: “As it stands today, I don’t see any of the constituencies involved in a 450 mm transition who are willing to fund it. And, I don’t believe there is an equipment firm in the sector that can generate a reasonable ROI for investing in 450 mm organically given the current value proposition and environment. Additionally, the industry has to move as a complete tool set, or the transition makes no sense, so if even one critical equipment segment does not see the compelling value to invest in 450 mm, the whole transition falls apart. Additionally, from a risk perspective, given what happened in the 300 mm transition, an equipment company can either be early, on time, or late with a wafer size transition, and two out of three of those are costly outcomes. That said, I do think there is a compelling argument to be made for the industry to take a consortia approach in the 450 mm wafer size transition, which could enable companies to collaborate on roadmaps and transition timing as well as share the funding to help minimize the financial risk to any individual company. “
Steve Newberry: “I think there were some unique circumstances that resulted in the 300 mm transition being a productivity bonanza for our customers, and those factors are highly unlikely to be duplicated at 450 mm. Not only were many tools increasing raw silicon area processed by 2.6X, but equipment reliability and availability both were substantially increased at that time. Moreover, the ability of customers to more quickly “ramp to volume” was dramatically improved and shortened. Each of these three factors improved Fab capital asset productivity, and efficiency which had a negative influence on equipment SAM following that transition. 450 mm will obviously see a dramatic increase in silicon area processed per tool, but I think any improvements in those other areas will not be nearly as dramatic, and I would be surprised in the price premium for 450 mm over 300 mm was not closer to something in the 60 to 70 % range, so the negative SAM effects will likely not be nearly as great for this wafer size transition, whenever it occurs.”
Dave Dutton: “I contend that the larger threat to the equipment industry is not the specific issues around value apportionment for the 450 mm transition, but the overall threat of the steadily rising cost of overall R&D. Generating a sufficient ROI argument for taking on challenges that approach the limits of physics, while selling into a marketplace that has seen the number of new fabs drop to a fraction of historical levels is the “big picture” problem. Figuring out how to fairly divide the investment necessary for 450 mm, and how to subsequently apportion the value it brings, is certainly a huge issue, but it is a sub-set of the overall R&D cost-vs.-returns issue.”
Rick Hill: “I think people are tending to overlook some of the structural elements of this potential transition. If the equipment industry moves in lock-step to achieve 450 mm -- which is the only logical way it can work— then those customers who can’t afford to build a 450 mm fab (which is the majority of the market) will be left behind as the equipment suppliers re-target their R&D spending towards the new platforms. If companies try to “split” their development efforts while facing the unprecedented challenges of 22 nm technology and beyond, we could see the overall pace of progress of Moore’s Law decline. So I think there are some underlying issues that make this particular transition even harder to generate a sensible ROI while delivering an advanced technology roadmap.”
Luminary # 7: Mr. George Needham
General Topic: One of the key metrics that Wall Street applies to specific industries or stocks is called the “PEG Ratio”. Basically, it is the P/E ratio over the earnings growth rate, and one general, “ball park” rule of thumb has historically been that, in the long term, you deserve a P/E multiple equal to your earnings growth rate, in percentage terms. Consequently, a long-term earnings growth rate of 20% would merit a P/E of 20X earnings, while a similar earnings growth rate of 30% would merit a 30X P/E ratio. Consensus estimates on the end chip market growth CAGR have dropped from the mid-teens throughout most of the 1980’s and 1990’s, to something roughly half that level. Consequently, if the industry’s earnings growth rate dropped proportionally to the top line’s diminished pace, we would see P/E multiples at half historical levels seen in the historical “peak” cycles. Will such decreased stock valuations occur, due to concepts like the PEG ratio, or are there reasons investors should value the semiconductor equipment industry more highly than the above logic would indicate?
Overall Summary: There was general acknowledgement that, absent other business model changes, a top-line revenue growth rate dropping in half would flow to the earnings line, and produce a much more sobering valuation model for the industry. However, the panelists were fairly united that the specific metric represented by the PEG ratio was, if not inapplicable, certainly not as influential a metric as it may have been in previous chapters of the industry’s development. Moreover, the problems represented by the “time constant” of looking at a long term earnings growth rate in an industry that is prone to hyper-growth, followed by massive downward “resets” was challenged, with an assertion being made that such long term metrics are inherently over-optimistic during a “down cycle”, and over-pessimistic during an “up cycle”. Moreover, a number of points were made that the “absent other business model changes” clause is demonstrably false, as the industry as a whole has become more asset-light, more operationally efficient, and better at supply chain management. The cumulative result of these fundamental business model changes has substantially increased the level of cash generation realized by this sector at a given, “top line” revenue level. Additionally, several panelists pointed out that in this day and age of much more constrictive accounting standards, with many non-cash expenses being assessed against earnings, that a “cash is king” mentality has emerged over a focus on either GAAP or non-GAAP earnings levels, as investors value free-cash-flow as the superior metric of corporate performance over EPS.
Selected Panelist Quotes:
Rick Hill: “First of all, you have to determine the “risk-free rate of return” before you can set performance metrics, such as P/E ratio, for the entire market. Everything in the market is determined on a basis relative to alternative investment vehicles. If interest rates are very low, as they are today, then the emphasis towards the equity markets will drive "acceptable" P/E ratios up, regardless of the steepness of the earnings growth curves. Secondly, I think that comparisons within an industry of faster vs. slower earnings growth play a central role in determining which firm deserves a higher P/E multiple. However, I think that comparing metrics such as PEG ratios across industries is much more problematic.”
Steve Newberry: “The modern day sophisticated investor is much more focused on what the net-free-cash-flow is, what the return on invested capital is, and how well a company can “harvest” it’s position during an up cycle to continue to add cash to the balance sheet, then they are concerned about EPS. The reason is that the earnings metric has several non-cash expenses inherently associated with it with today’s financial accounting standards, and expense protocols. If you analyze the performance of the top firms in the industry on a cash generation basis, with their new asset-light business models, you will find they were able to do a much more efficient job of generating net-free-cash in the last peak cycle than in any previous cycle. Finally metrics like the PEG ratio which focus on long term growth rates can be misleading in a cyclical industry that experiences hyper-growth phases, followed by revenue ‘reset’ phases, inherently being either too optimistic or too pessimistic depending on the phase of the cycle, diminishing the usefulness of such a ratio as an investment metric.”
Rick Wallace: “I think comparing where we are as an industry versus the situation in the 1990’s is unfair, because I would argue that “Tech,” in general, had a different investor profile in that time, and investor analytical metrics that were very useful in that period may not be nearly as applicable to a more mature industry. We are more disciplined in terms of what we will invest in, because there is always a new “bubble” to chase, if you are not careful. We are also sounder in terms of our operational business models, and I agree that generating cash is where the rubber meets the road in terms of what is of vital importance on a going forward basis.”
Luminary # 8: Mr. Bob Boehlke
General Topic: Over the past decade, what Michael Porter defines as the “concentration ratio,” constituting the share of the buying volume of the top four customers collectively, has gone from 27% to over 65% in the semiconductor equipment industry. What this means is the top four customers in the equipment buying market now constitute almost two-thirds of the equipment CAPEX. Porter argues that there is an inherent shift in pricing power away from the producer (equipment industry), and toward the buyer (chip industry) with the development of such large, powerful, and increasingly shrewd buying entities. How will the semiconductor equipment industry maintain their gross margins and achieve EPS objectives in the face of such a dramatic shift in price negotiation power towards these increasingly powerful mega-customers?
Overall Summary: There was zero dissent from the validity of Porter’s model that projects that a larger, more concentrated buyer base will, in fact, have significantly more price negotiation power. The concept of seeking meaningfully differentiated IP positions which uniquely solve customer problems, and then to continue to evolve these differentiated positions was repeatedly mentioned. The clear “first line of defense” against unreasonable pricing power is obviously meaningful and high-value differentiation against all other competitors, and the IP and “know how” necessary to maintain such differentiated positions. There was a fairly visceral dialogue around the concept of customer’s attempting to “diffuse” unique equipment technology and “know how” to other equipment suppliers, reducing the differentiated position and bargaining power of the firm being “diffused,” and the ethics of whether such practices were “fair game,” or not. The consensus was that such practices are distinctly not “fair game,” and cross over the line from “hardball negotiation” into unethical practices. Significant dialogue centered around the concept of the ability to self-assess a firm’s value proposition, as it stacks up against competitors, and knowing when to “hold the line” against unreasonable pricing pressures not consistent with the value being delivered.
Selected Panelist Quotes:
Mark Jagiela: “We don’t face nearly the same type of concentration ratio in the ‘back end’ segment that the ‘front end’ does, with our top ten worldwide customers only constituting roughly 30% of the market collectively. Having stated that, one of the key issues that arises with the evolution of such powerful customers is the attempt to ‘commoditize’ your product and technology, and we saw Intel try to do that in test some years ago, where they attempted to create an open standard architecture for test, essentially trying to make test a ‘fungible’ commodity. Needless to say, from an innovation perspective, and from a perspective of a company that has IP to protect, it is not possible to participate in such a strategy and still maintain a proprietary position.”
Rick Wallace: “The only viable strategy to combat these pricing pressures is to be a ‘moving target.’ The most defensible position you have is to continue to evolve your technology, capability, or value proposition, and as long as that matters in the industry, to me, these areas represent the only sustainable differentiation. The moment such unique capabilities lose emphasis, if the design node advancement pace was to diminish, for instance, then the industry is clearly at risk, and the weak ‘members of the herd’ will definitely get picked off. So the question becomes: ‘How do you continue to go to ‘higher ground’ and innovate in areas that make a tangible, high 'value-added' difference to your customers, and be simultaneously very shrewd about not undertaking tasks that they can do more effectively themselves?”
Steve Newberry: “Sophisticated customers in any mature industry are usually pretty good at attempting to deny that suppliers have unique and meaningfully differentiated positions, which aids them in attempting to effect price erosion. One of the keys to combating this power is through understanding the totality of the comprehensive value of your product and service offering through competitive data gathering and truly understanding your customer’s needs and priorities. You have to know your value components, and how they add up against competitive alternatives, whether those value terms consist of technological capability, speed to ramp yield, ability to provide unique field support, the strength of your technical team to quickly and thoroughly understand customer problems and challenges, etc. Only through understanding the totality of your value, how it stacks up against competitors, and how it meshes with the customer’s key requirements can you hope to “hold the line” against unreasonable pricing demands.”
Randhir Thakur: "We have to look at this as an overall "supply chain" issue. In situations of aggressive customer bargaining strategies, the best leverage tactics are built around clear and advantageous product differentiation. Product differentiation can be amplified through process integration learning. Further, early and very intimate collaboration with customers to fully comprehend their future requirements and to align our technology roadmaps with their needs can help our industry become more R&D efficient, which as mentioned earlier, is an increasingly large cost component at 22 nm and beyond. While customers may be pressuring us at the ‘top line,’ we need to continue to become more efficient in managing our cost structures so that reasonable profit margins can be achieved.”
To read Part 1 of this article, please click here.
(Author Scott Landstrom runs a strategic consulting practice that specializes in advising CEOs on strategy formation, revenue growth engines, and liquidity events. His email is: firstname.lastname@example.org.)
March 2, 2010