Interesting interview with TER CEO on TheStreet.com.
thestreet.com
subcsribers only
George Chamillard is the chief executive officer and chairman of Teradyne (TER:NYSE - news - boards), a Boston maker of testing equipment for the semiconductor industry. The nearly $7 billion company sells its testing equipment to everyone from microprocessor companies to communication chipmakers. The company also makes connectors, which are used in electronic systems.
Chamillard has worked for Teradyne for 32 years, working his way up from production engineer in 1969 to CEO in 1997. He has witnessed firsthand the cyclical nature of the semiconductor business, and now finds himself in yet another downturn. After banner growth, the semiconductor industry began contracting last fall with slower microprocessor sales, putting Teradyne's business under pressure as well. Chamillard recently spoke with TheStreet.com's Caroline Humer about where the semiconductor and semiconductor-testing businesses are headed.
--------------------------------------------------------------------------------
Caroline Humer: There's been a lot of talk about whether or not semiconductor equipment companies are nearing the bottom of the cycle. What have you seen at Teradyne or what do you think is going on in the industry right now?
George Chamillard: Well, it seems to me that we're near the bottom of the cycle. People buy equipment for two different reasons: They buy it because they need increased capacity, buying more machines so they can make more devices, or they buy it because the old equipment doesn't do the new job.
During a downside, you don't see capacity buying. What you see is what we call technology buying. We're still seeing technology buys, but I haven't seen any capacity buying come on, and I'd really have to see that before I would believe it has bottomed.
Caroline Humer: And what's the difference between a technology buy and a capacity buy?
George Chamillard: Well, on the front-end of capital equipment, people think about 300 millimeter ... etc. Technology buy means the old tester simply won't test a new device that the customer's designed and so you have to buy a tester, either with some new options or with new capability, and let them test the new device. [Editor's note: 300 millimeter refers to the process used to build semiconductors. The industry standard is currently to use a 200 millimeter wafer in the process, but this will change to 300 millimeters in the coming years.]
Caroline Humer: I see. And so the fact, then, that you've seen more technology buys but not as many capacity buys would indicate what?
George Chamillard: That people are buying for technology; they're not buying for capacity.
--------------------------------------------------------------------------------
"To look at this guy saying installations have stopped, this guy saying bookings have picked up, ... I just think there's so much noise that I'd be skeptical of anything I heard."
--------------------------------------------------------------------------------
Caroline Humer: And are there any other signs, or is that what constitutes a bottoming process in your sector?
George Chamillard: Well, to me, that would be the thing you would look at. I think in the front end, not just in our sector, it's a similar issue -- as I listen to everyone else's conference calls and they're talking about technology buying with relatively little capacity buying.
But the industry shipped a lot of capacity in 2000. The semiconductor capital equipment industry just grew 70% to 80% last year, while the customers we all sell to grew 35%. So we shipped, we as an industry shipped, an awful lot of capacity that is sitting idle, or is underutilized, at least.
Caroline Humer: I see. Let's move on to what you're seeing now. One of the things we've heard about in the past week or two is perhaps a slowing of order delays or cancellations a little further up the chain than the semiconductor companies themselves. Does that also hold true in the chip equipment sector?
George Chamillard: It's difficult to make an assessment from any single point and try to apply that to the overall industry. I think most of the projections I've heard on what the semiconductor manufacturers are going to do this year are that they will be down somewhere near $175 billion of semiconductor sales, compared to over $200 billion last year. And ... I would expect that the semiconductor equipment industry would be down as well.
To look at this guy saying installations have stopped, this guy saying bookings have picked up, this guy saying I've already purged my backlog and so there's no cancellations left ... I just think there's so much noise that I'd be skeptical of anything I heard.
In the end, when it's bottomed, you'll see that people won't be reducing their shipments, and when their shipments go up, it's starting to turn up. And I haven't seen anyone saying it's turning up because of shipments. Have you?
Caroline Humer: No, not yet. That's an interesting point. You've been in this business quite a while, how long does a bottom like that last? When you look back on previous cycles, what can we learn about where we are now?
George Chamillard: This is a double-edged sword. The danger of saying that history doesn't count anymore is pitching the information that history has. The danger in saying this is like 1972, or 1973, is that it may not be. So, having put the caveat out there, I think we've had down cycles that are only a couple of quarters long and we've had down cycles that are several quarters, you know, five, six, seven quarters.
I think, in general, that in the same way people would say feast follows famine, famine follows feast, the bigger the feast, the bigger the famine, the bigger the famine, the bigger the feast, and we had a wonderful feast in 2000 and we're having a very sharp, rapid downturn. I think that will correct faster, and we'll have a bigger feast when it goes up again.
But I don't think you can get a lot out of history. Why not? In terms of semiconductor equipment ... there's been a series of major changes, structural changes in the industry. For example, if you go back five years, the subcontract, test-house side of the business and even foundries, compared to the integrative device manufacturers, were very small.
Today there's a significant percentage, I don't know, 30%, possibly, of assembly and tests being done at subcontract houses and foundries -- I would guess about the same as wafer manufacturers. And then you have the semiconductor companies who are competing against the integrated device manufacturers, Intel, Motorola, National Semiconductor and so on.
So you have a very different structure to the industry, which makes me suspicious when you hear, "This was the history we had last time, etc. ...
--------------------------------------------------------------------------------
"We had a wonderful feast in 2000 and we're having a very sharp, rapid downturn. I think that will correct faster, and we'll have a bigger feast when it goes up again."
--------------------------------------------------------------------------------
Caroline Humer: OK. Now let's turn to Teradyne. Has visibility improved since the company last discussed financial results, or from when you last filed your 10-Q in mid-May, in which you described visibility as limited?
George Chamillard: Well, I don't think this is like you wake up in the morning and there's fog, and at 10 o'clock it burns off. I think this takes weeks to months to burn off, until you really know if you're looking at something clear or not, and there just hasn't been enough time gone by for me to think there's any significant change.
Caroline Humer: You're a supplier to the major semiconductor sectors, including communications areas like networking and telecommunications, as well as personal computers. How do you see these markets developing? Do you see a difference in terms of signs of recovery in any of these areas or in terms of going forward? Which areas will you look to for growth?
George Chamillard: Well, let me comment on the first part of your question, which is the different segments. I think I read the same stuff you do, which is that people are starting to say they're seeing PC demand -- I don't know if it's picking up -- but the inventory has been burned off to a great extent. For that to flow down to chips, I think will be a while yet.
In terms of what's going to drive it on the upside, in the end you've got to say do you believe that the Internet is fully built-out, or do you believe we're in the early stages of a decades-long deployment?
Caroline Humer: Do you think the communications sector or the Internet with its PCs or handhelds will drive growth for Teradyne in those areas?
George Chamillard: I think the whole Internet appliance is going to be ... it's just nowhere near saturated. Never mind being saturated in New York City and Boston, but once you get outside of the United States and once you get outside of the developed nations, it hasn't even started.
I think also the opportunities for the industry are around medical applications, and particularly around image processing...
Caroline Humer: And I'm curious about how something like that would affect Teradyne?
George Chamillard: Well, first it would affect us two ways. Teradyne has really two large businesses. The most famous, and the one we've talked about most so far, is the semiconductor test equipment business.
We also have a business that's in the connector backplane EMS [electronics manufacturing services] industry, and basically what we're doing there is building products that help people package their electronics.
But if there are more semiconductors built, which I think there will be, no question, if there are more semiconductors built, more testers will need to be built, number one. Number two, if there are more semiconductors built they have to be packaged, they have to be put on circuit boards, they have to be put on the cabinets, they have to be interconnected with cables, and so on and so on. And that's the other business we're in, which is focused around building backplanes and cabinets that package all those electronics.
--------------------------------------------------------------------------------
"I don't think this is like you wake up in the morning and there's fog, and at 10 o'clock it burns off. I think this takes weeks to months to burn off."
--------------------------------------------------------------------------------
Caroline Humer: In terms of a recovery, is it still too hard to say whether something could start in the second half of this year or whether it will be in the first half of next or the second half of next?
George Chamillard: Well, the second half of this year starts three weeks and two days away.
Caroline Humer: Starting to rule that out, I guess. But beyond that, is there anything you could say in terms of when you would expect a recovery to start or anything like that based on what you've seen so far?
George Chamillard: The industry, in the early '90s ... we said we wished we had more applications so that we weren't tied just to computers. And what we really were saying was that ... 80% of the buying decisions were made by the management of large insurance companies, large brokerage companies, people who bought large computers, and 20% of the buying decisions were made by consumers.
Today, probably half of the buying decisions that consume the semiconductor are made by consumers. That's great, except that consumers are somewhat more fickle than the management of the large insurance companies making the capital decisions on computer systems.
Caroline Humer: Is there anything else you want to add?
George Chamillard: I think it's easy ... for investors to get discouraged when they see what's happened to their portfolios over time. I know I get discouraged when I look at mine. On the other hand, I think that speaking for a guy involved with running a high-technology company, the investments I have in Teradyne are two: One is a financial investment; the other is a career investment.
I can't think of a better place to be than the one I'm in right now.
Caroline Humer: Great. Well, thank you so much for taking the time to talk to TSC.
George Chamillard: Thanks a lot, Caroline. |
