Kirk Pond, Fairchild Semiconductor President and CEO

After serving in management positions for Texas Instruments and Timex Corp., Kirk Pond joined Fairchild Semiconductor (South Portland, Maine) as vice president of Logic in 1984 and subsequently was named executive vice president of the Logic, Memory, Discrete and Analog groups. With the acquisition of Fairchild by National Semiconductor in 1987, he became vice president of Logic with responsibility for the combined National and Fairchild logic businesses. In 1991, Pond was named president of the Standard Products Group with responsibility for the Logic, Memory, Discrete and Analog groups, comprising 75% of National's operations. Later, he became executive vice president, COO and a member of the Office of the President of National.

SI: To say the least, 2001 was a disappointing year for the industry. However, Fairchild seems to have weathered the storm — you acquired Impala Linear and you've announced what essentially is a flat fourth quarter. What's your take on this?

Pond: We've all been through some unprecedented times. The industry's growth rate of 35% up in 2000, and 2001's industry average -35% down is a singular occurrence. In my 35 years in the industry, this has been the first time we've had a downturn of this magnitude. We were all running along pretty well up to October or November of 2000, then everything hit a brick wall. Companies without the experience of going through these cycles were unprepared for what hit.

SI: Fairchild wasn't?

	Kirk Pond (Source: Fairchild Semiconductor)

SI: How?

Pond: When we first structured Fairchild we decided to approach the industry differently. We decided to build the less complex, building-block chips used in most end applications. We call ourselves the multimarket company because we have components in all end markets — whether computers, automobiles, consumer goods, communications or industrial applications. We have a broad market exposure in all key end markets, and you can find our chips in everything from computers to TVs and washing machines to automobiles, satellites and coffeemakers.

SI: So your markets are diversified. What else?

Pond: We have a very broad geographic disbursement. Approximately 20% of our business is in North America, 15% in Europe, about 7% in Japan, 15% in Korea, 10% in Taiwan, 23% in the rest of Asia, etc. This gives us good worldwide disbursement, so when the wireline or cellphone telecom guys — the Nokias, Alcatels and Ericssons — in Europe started falling apart, we shifted our resources to those building power supplies in China. This capability has enabled us to smooth out some market swings. So while the market was down 35% last year, our revenue will probably drop 18 to 20% — still a lousy year, but not a wild swing. Our multimarket approach was one of the reasons we could outperform the industry. Cost structure is another advantage. When we started the company, we financed and structured it to withstand the industry's cyclical nature and the ups and downs of the world's economies.

SI: How?

Pond: We set up a 70% variable-cost and 30% fixed-cost structure. Our competitors — heavily capitalized companies — typically have a 60 to 70% fixed-cost structure, and only 30 or 40% variable. This means that we don't have to build multimillion-dollar fabs for our products because they're smaller devices mostly in the power domain, so they aren't lithography-driven. This makes front-end manufacturing less expensive and lets us eliminate heavy fixed costs in our fabs. Our assembly/test area is another factor: We outsource about 50% in assembly/test — so upside or downside we can respond quickly. In a downturn, we can in-source more back-end production, giving us flexibility in our capacity. We want to bring the percentage of outsourced assembly/test down somewhat, so we've begun work on a new back-end facility in Suzhou, China.

SI: Any penalties?

Pond: It's more expensive to build at a subcontractor's than it is in-house, but we don't have the overhead, the people, etc. Our fixed/variable cost arrangement is such that we've been able to adjust our overhead and structure as our revenue dropped this 18% or so — we've pulled cost in and avoided hundreds of millions in losses every quarter. While currently we're operating close to breakeven, we've reported better results than most in the industry and we're well-positioned competitively when things start coming back.

SI: What is your backlog, book-to-bill ratio?

Pond: Fairchild ran a 1:1 book-to-bill in third quarter 2001, possibly slightly more for the fourth. Our backlog has ceased decreasing and is inching up. We maintain a 26-week financial backlog, so once things get past 26 weeks they fall off our radar screen. Generally, people look at this in the zero to 13 weeks, and 14 to 26 weeks ranges. Because 13 weeks is a quarter, people pay considerable attention to 13-week backlog. Now, although the 26-week backlog has remained relatively constant in third and fourth quarters, the 13-week one has climbed continuously because people are pulling orders into the short term. This is called "terms bookings" — bookings delivered in the current quarter. So while that 13-week sector has increased, the total has remained constant. We're not losing backlog, which is why we've projected our fourth quarter to be flat to slightly down.

SI: How does China fit in with your overall plans?

Pond: We've been very interested in Southeast Asia's tremendous growth over the last few years and done very well there. Last quarter, we did about 70% of our revenue there — not too many semiconductor companies can say this. Next year, Asia is projected to be the highest growth market in the world, and we're already well-positioned there. We're already doing considerable business in China — via Hong Kong and our sales offices in the mainland — and shipping it to the mainland.

As we've grown, I've felt the need for a permanent facility in China, an assembly/test facility, which can also warehouse products. It'll eventually have applications and design support. This year, the Chinese market is supposed to be $10B to $12B. We'll be right in the middle of it with our first new facility since we became Fairchild. We expect that by 2004 it'll help us more than double our revenues there. Currently, we ship about $200M. We consider this a tremendous growth opportunity.

SI: You've had a presence there for a while?

Pond: Absolutely. You cannot do business in Asia, particularly China, without first making a long-time commitment and establishing relationships.

SI: Anything you're planning to change in the company?

Pond: (Laughing) Sure! At any time, I could come up with 50 things I'd like to change. That's key to a dynamic, growing company. But our fundamentals are strong and constant. We're continuing the strategy we've followed from the beginning — that won't change. We'll continue managing our balance sheet and aggressively approaching the marketplace, both with new products and acquisitions that'll continue to help us grow faster than the industry. The SIA projects the market to grow 6% over 2002. If that's the case, we'll grow faster because we have our two-prong strategy of growing internally through new product development, and externally through acquisitions. We won't change the fundamentals. Certainly, operationally, we'd like to continue to increase efficiency — we've got about 6% of the total multimarket segment and we're working toward a 10% share of the total multimarket segment. This means identifying customers and applications and getting design wins.

SI: What are your market gains?

Pond: Our power market share has gone up about 8%. Our share of the total multimarket is about 6%. Over the last four quarters we've picked a half a point market share per quarter.

SI: You're expanding aggressively into the flat-panel-display area and commercial products. Where are your next growth areas and how are you planning for them?

Pond: About 30% of our production goes into consumer markets, including displays. About 25% or so of our revenue goes into the industrial and automotive market. About 30% goes to computer and peripherals, and we ship about 20% into the communications market. These are our big markets — we're well-balanced. Currently the consumer and industrial markets are growing the fastest for us. Last year it was communications and computers. Again, our business model gives us the flexibility to respond to market opportunities.

Over this year I expect the consumer market to be strong, computers will begin picking up more steam, and wireless communications will be taking off as well. The wired communications market — the infrastructure segment — will remain soft until second or third quarter 2002. So I'd say, consumer, computer, to a certain degree industrial, and again, a strong demand across all segments for power. We want to be the power supplier from the wall to the board. When you plug a cord into the wall, we want to control, condition and distribute that power to and through the board. Power is about one-half of this total multimarket segment we've identified — standard linear, discrete, standard logic, optoelectronics — those segments are about a $25B part of the semiconductor market and power is about one-half of it. This is primarily what the China facility will support.

SI: The PC has become a commodity, and now many look to the Internet for major growth. What's your perspective?

Pond: The Internet's a phenomenal tool, and we're only beginning to learn how to use it. Those who proposed foolish uses have gone bust and serious developers have arrived. Applications are driving computational needs increasingly higher. When my son sends me a streaming video of my grandchild, it requires enormous computing power. PCs may be commodities, but there's a lot of technology in those boxes — operating systems and the fundamental system-level chips may be commoditized, but manufacturers of sound and video cards, CD-R and CD-W drives strive to differentiate, customize and add value to their products. All this requires semiconductors.

SI: You are one of the few manufacturers that still owns fabs. How do you see the movement toward the fabless model?

Pond: It's primarily an economic issue. Building deep submicron fabs isn't inexpensive. Only a few companies can afford it. When Morris Chang builds six brand new, multimillion-dollar wafer fabs, it's good for the industry. It allows many who otherwise couldn't bring their designs to fruition to do so. It does much for the industry's growth. Fairchild isn't doing more foundry work because we don't need that type of processing. With our types of products, we cannot afford to pay TSMC a 30% profit. The $1B that TSMC made last quarter came from somebody's pocket. I don't want it to come out of mine. Our needs are different from those of a company building a 40-million transistor network processor that requires TSMC's deep submicron capabilities — they can afford to pay a margin for it. We don't require multimillion-dollar fabs; we have cost-effective manufacturing operations that put out high-quality products for our 50,000 customers worldwide.

SI: What trends should we be paying more attention to?

Pond: From a financial perspective, I think that the inventory burn that happened over 2000 is an important issue. Manufacturers came into the year with lots of inventory that has been largely used. In spite of the modest view that most economists seem to have, GDP growth this year will be much better from a semiconductor perspective because that inventory must be replaced. People need to continue building their products. As an industry we may have a better year than forecasted.

Another trend is that, increasingly, the production and manufacturing of goods using semiconductors is taking place in Asia. The U.S., Europe and Japan are becoming the design houses, and manufacturing is being done in places like Taiwan and China. People must prepare for that and structure their business to handle it. If you're primarily a U.S.-centric semiconductor company, you're suddenly going to find that many of your customers are in Shanghai.

SI: Is there anything we can do to minimize industry cycles?

Pond: Over the last few years, the industry has patted itself on the back over its "sophisticated" supply chain. Let's follow it backward: A company like Compaq sells PCs but buys its boards from FIC in Taiwan or Solectron in Guadalajara. They tell their subcontractors to carry the inventory because, as a company, they don't want to. However, they add, if you don't respond instantly, you lose my business. So subcontractors hedge their bets by buying more components or building more boards and warehousing them because they don't want to lose the business. In turn, they come to someone like us and tell us they want to buy a million parts a week, but they don't want them shipped — they want us to hold them — and if you can't do it, we'll go across the street to your competitors. So we, in turn, put a little extra product on the shelf to be able to respond. Everyone in this supply chain is making a judgment at every step. The end customers — the computer manufacturers, the cellphone companies, all the providers — all say, my customers are buying a lot of product, I don't want to be out and lose the business, so I'll carry a little extra in finished goods.

While everything grows, this works. The moment anything stops, there's inventory backed up at the customer, at the box manufacturer, at the subcontractor, at the distributor and at the chip manufacturer. Our "sophisticated" supply chain insulates us from the end consumer — not a good situation. We've learned that lesson. Over a year ago, when things began softening up — regardless of what our manufacturing customers told us — we looked at what the end consumer was buying, adjusted our production accordingly and didn't get burned. The current supply chain has not served the industry well.

SI: Any advice to the industry?

Pond: The intellectual capability of a few makes a large difference. As an industry, we must not lose sight of the fact that individuals create great innovations. Even in multimillion-dollar companies we must nurture that handful of creative people who drive the industry. We must let them be creative, give them maneuvering space, support them. We mustn't lose sight of the fact that it is just a few creative individuals who come up with new ways to build things smaller, faster, cheaper. We must care for them.