泰瑞達 (TER) 2002 Q3 法說會逐字稿

Good morning.

My name is David (ph) and I will be your conference facilitator today.

At this time I would like to welcome everyone to the Teradyne third quarter earnings release conference call.

All lines have been placed on mute to prevent any background noise.

After the speaker's remarks, there will be a question and answer period.

If you would like to ask a question during this time, simply press star and then the number 1 on your telephone keypad.

If you'd like to withdraw your question, press star and then the number 2 on your telephone keypad.

Mr. Newman, you may begin your conference.

Tom Newman (ph): Thank you, David (ph).

Good morning, everyone.

And welcome to our discussion of Teradyne's financial results for the third quarter of 2002.

I'm joined this morning by Chairman and Chief Executive Officer, George Chamillard, and our Chief Financial Officer, Greg Beecher.

We will be providing you with details of our third quarter performance and with our outlook for the fourth quarter.

First, however, I'd like to address some administrative issues.

Teradyne's press release containing our financial results for the third quarter was sent out via business wire and was posted on our Web site yesterday evening.

If anyone needs a copy please call Teradyne's corporate relations office at 617-422-2221 and we'll provide you with a copy.

This call is being simultaneously being Webcast over our Web site at www.teradyne.com.

A replay of this call will be provided on our site starting at 1:00 p.m. today Eastern Time.

If it's more convenient, you can also access a replay of the call by dialing 1-800-642-1687 within the United States, or 706-645-9291 outside the U.S. and providing the conference ID number 5951188.

Replays from both sources will be available through October 30th.

It is our objective to use this call to comply with the requirements of SEC Regulation FD.

Therefore investors should accept the contents of this call as the official guidance from the company for the fourth quarter of 2002 and beyond.

If at any time we communicate any material changes to this guidance, it is our intent to do so simultaneously to all investors to the best of our ability.

Investors should note that only George Chamillard, Greg Beecher and myself are authorized to supply company guidance.

The matters that we discuss today, other than historical information, include forward-looking statements relating to future financial performance and other performance expectations, changes in the company's business, statements as to bookings, backlog, pricing, design-ins, and demand for our products and other opinions of management.

These forward-looking statements are made under Section 21-E of the Securities Exchange Act of 1934. nvestors are cautioned that forward-looking statements are neither promises nor guarantees but involve risks and uncertainties that may cause actual results to differ materially from expectations.

Some of those risks and uncertainties are detailed in our filings with the Securities and Exchange Commission, including but not limited to our Form 10-Q filed on August 14th, 2002.

We caution listeners not to place undue reliance on any forward-looking statements, which speak only as of the date they are made, and we incorporate here the discussion of those factors.

Teradyne disclaims any obligation to publicly update or revise any such statements to reflect expectations or events, conditions, or circumstances on which any such statements may be based or that may affect the likelihood that actual results will differ from those set forth in the forward-looking statements.

Now let's get on with the rest of the agenda.

First, our CEO and Chairman, George Chamillard, will review the status of the company and the industry, and will provide guidance for the fourth quarter of 2002.

Then our Vice President and Chief Financial Officer, Greg Beecher, will review the details of our performance in the third quarter and will provide additional details on our guidance for the fourth quarter.

We will then answer your questions.

For scheduling purposes, we intend to end this call after an hour -- George.

Thank you, Tom.

And good morning to everyone on the call.

Our results for the third quarter were sales of 331 million, a loss of 27 cents per share before special items, and a loss of 91 cents per share after special items.

Our total net orders in the third quarter were 231 million, about flat with the second quarter's 228 million.

Based on orders expected in the fourth quarter and our existing shippable (ph) backlog, we expect fourth quarter shipments to be about flat.

In addition, we expect our losses to be reduced due to the cost-cutting actions we have taken.

Consequently, our guidance for the fourth quarter is for sales of between 310 million and 340 million, with losses of 22 cents per share plus or minus three cents.

Currently our near-term outlook remains unclear.

The forecast for the technology segment of the economy is still weak, although some other segments are arguably improving.

So what are we doing?

As you know, we have dramatically downsized the company while making critical investments to strengthen our position in the future.

We've funded selective investment in internal processes, in our manufacturing infrastructure, in our field resources, and most importantly, in our engineering.

Although those investments have achieved the productivity gains and the product objectives we had in mind, the growth we saw earlier in the year has slowed, and our short-term future remains uncertain.

Therefore, we feel it's necessary to downsize our business to break even, assuming little or no growth in sales.

Consequently, our target is to achieve break even in sales of about $350 million per quarter by the middle of next year.

This is how we're going to do that.

In general, we have two major approaches for lowering break-even.

The first approach involves what I'll call plain old cost-cutting.

The second approach is a series of more strategic actions, which will result in lowering our break-even.

Each of these approaches impacts people, materials, facilities, and products that obviously there's some overlap between them.

First in the area of plain old cost-cutting.

We've discussed in previous calls the actions we've taken and continue to take with our employees to reduce costs, including layoffs, salary reductions, furloughs, and so on.

In the materials area, we continue to work with our suppliers to lower our costs, and in fact some of the value engineering that we did with suppliers earlier this year is beginning to show up in our shipments.

With regard to our factories, the sharp reduction in our production has resulted in significantly underutilized facilities, which we will continue to consolidate.

As we continue our programs to reduce costs with plain old cost-cutting, we continue to take strategic actions aimed at improving our cost structure.

A few examples of these might be helpful.

In the last call we explained that we had moved the electromechanical assembly of our J 750 (ph) product to our contract manufacturing group in connection systems.

By the end of Q1 2003 our J 750 (ph) systems will be assembled, tested, and shipped worldwide from Asia.

We continue to rationalize our product lines to lower costs.

The acquisition of Genrad (ph) and the consolidation of engineering programs in semi-test (ph) has resulted in some overlapping products and programs, and some others that are not core to our future.

To some extent this may overlap with plain old cost-cutting, but we have significantly increased our efforts at some (ph) assembly and material sourcing from low-cost regions from around the world.

We also are moving our North American system assembly and test operations to a single site, and we are consolidating other functions in the company, such as supply line management and other aspects of our G&A activities.

So we're moving aggressively to break even around the middle of next year, assuming a sales level of about 350 million per quarter.

And if that level of sales is too high, we'll do what we must do to adjust to a break-even level that reflects business reality.

Now, obviously break even is not acceptable.

Teradyne must (ph) be asserting that it is in very good businesses, promising high returns in the future, and that Teradyne can win in those businesses.

Why do I believe that?

Basically Teradyne is an enabler to expanding high tech nature of electronics products, whether it's the testing of semiconductors, testing of printed (ph) circuit boards, or high speed Internet connections, or building high speed back planes (ph).

Despite that technology is out of favor right now with investors, we have a 40-year history of technology innovation that's helped make the world a better place.

And I for one just don't think that's going to change.

The question is not whether technology will come back, not whether Teradyne will win when it does -- the question is when.

One possible indicator of when is that device unit volume continues to grow and the installed base utilization appears to be pretty high.

In fact, on our products we do a monthly survey of test capacity at our major IDM and subcontract customer sites.

The most recent survey indicates that our systems are essentially fully utilized.

We think that this is due in no small part to the 450 different device types that our applications teams brought up on our machines during 2001.

Any significant increase in demand should result in the need for more test capacity.

Increased demand has resulted in the need for customers to buy more test equipment has been strongest in the low-cost high-volume device segments and image sensor testing and we have been a beneficiary in both.

So far this year we've shipped about 140 of our J 750 (ph) high-volume logic test (ph) systems and over 850 since its introduction.

Last year we introduced a new test system called the IT 750 (ph), which is an image sensor (inaudible) system and it's built on the same 750 (ph) architecture.

We've already shipped over 100.

Teradyne's newest J 750-based (ph) system, the Integra FLEX (ph), now has four IDM customers that have bought over a dozen systems.

FLEX (ph) provides an opportunity for customers to make dramatic improvements in the way they manage their test operations while achieving major reductions in costs to test.

So we're growing (ph)pretty well at the high volume low cost portion of the market.

At the high performance end of the market, our target test system, aimed at devices that should go in volume production in 2003, has now more than a dozen customers who have bought about 40 systems among them.

Tiger (ph) has become the product of choice for high performance device testing in the SOC (ph) and microprocessor market.

In the mainstream of the semi-test (ph) market the catalyst is the clear worldwide standards for the OC test (ph) with over 1,100 systems shipped to date.

Every quarter this year we've gained major new design ins and new customer converts to catalyst, a testament to its flexibility.

Connection systems, we continue to extend the frontiers of technology in network performance.

Our design-in record has been a good one, with 42 new design-ins this quarter.

As described in an article in last week's "EE Times" (ph), we have practical approaches available today to allow back planes (ph) to be built that run over six gigahertz.

The newest connector product, GBX (ph), was designed with 10-gigabit network performance in mind, and it's gaining new design-ins every quarter, where higher performance networks are deployed.

They'll be continued to be built with Teradyne's connection's systems products.

In circuit test and inspection, the Genrad (ph) acquisition has allowed us to field the broadest range of test and inspection products to the assemblers of PC boards.

We also continue to see growth from spending for tests on military programs that will likely be part of future government budgets.

In broadband test, our ability to help customers deploy DSL or high speed access on cable has long been proven.

In spite of some significant trial activity, there's not been much capital equipment spending by these customers in the past quarter.

Teradyne's recent announcement of a partnership agreement allows us to strengthen our customer care product portfolio with cable operators and therefore expands our available market.

In our diagnostic solutions business, we remain tightly focused on providing diagnostic and text solutions to the major automotive OEMs for use in their manufacturing and service operations.

We recently introduced a new tablet (ph) PC-based test system which will begin shipping in Q4 to a major global OEM.

And the list goes on.

But I think you get my point.

The restructuring we are going through is not pleasant, but it's necessary in the current environment.

The aftermath of these actions will be a Teradyne that's sized to make money when growth resumes, a Teradyne that's in some great businesses, and a Teradyne that's gaining momentum in those businesses.

On the one hand, these times are the most difficult we've seen in our 42 years of business.

On the other hand, we believe we're reshaping the company to be an even greater success to the future than it has been in the past.

That's pretty exciting.

Now I'll turn it over to Greg to provide you with some more details.

Thanks, George.

Good morning everyone.

Our sales of 331 million were up seven percent from Q2 and up 33 percent year-over-year.

We had a net loss of 49.9 million, or 27 cents per share on 183 million shares before special items.

In the quarter we had special items amounting to 64 cents per share.

Our loss after special items was therefore 91 cents per share or 166.8 million.

The special items, which include goodwill and manufacturing facility impairments, along with severances and other smaller one-time items, have been detailed in the earnings release.

The impact of all these net charges is a reduction in our quarterly costs going forward by about 12 million.

From a product point of view, semi-test (ph) sales were 51 percent of the total, connection systems were 29 percent, assembly test was 14 percent, and our other test segment, which consists of broadband tests and diagnostic solutions, was six percent.

Our gross bookings were 247 million, down nine percent from Q2 and up about 17 percent from a year ago.

Net bookings were up about one percent from Q2 and up 104 percent from Q3 of 2001.

From a product point of view semi-test (ph) net bookings were 53 percent of total, connection systems were 20 percent, assembly tests 17 percent, and our other test segment was 10 percent.

Despite the environment, we continue to experience backlog pull-ups (ph) in our semi-test (ph) business in the third quarter.

The total was 15 million and the pull-ins (ph) were all from subcontract customers, and all for our Catalyst (ph) and J 750 (ph) products.

Cancellations in the third quarter were 15.7 million, with less than $1 million in assembly tests and around 15 million in connection systems.

These cancellations were due to a combination of a loss of some lower margin contract manufacturing business and a slackening of demand across several of our customers.

On a broad basis, our business continues to be driven by mobile, wireless, broadband, and consumer applications, with networking and telecom remaining weak.

In addition, there continues to be a trend toward the center of gravity of all our business shifting to Asia.

This is reflected in both our orders and our shipments, and we are investing (inaudible).

Based on net bookings, our approximate book-to-bill ratios by segments were .74 in semi- test (ph), .89 in assembly test, and 10 (ph) in our other test businesses.

In connection systems, the book-to-bill on a gross basis was .64.

At the end of the quarter our backlog stood at 545 million, of which 71% is shippable within six months.

Our ending headcount stood at 8,364, including 7,600 regular employees.

That total also includes the addition of 86 people that were part of an asset purchase that we made during the quarter.

The result was a net reduction of 268 people from the prior quarter.

Gross margins in the quarter were 23 percent, R&D expenses were 78 million, or 24 percent of sales.

SG&A expenses were 74 million or 23 percent of sales.

From a geographic point of view, our net bookings in the third quarter showed a decrease in orders from Asia that we discussed in last quarter's call.

Despite the fall off in Asian orders in Q3, on a year to year basis, South Asia is up 85 percent compared to this time last year.

The distribution of orders in the third quarter was as follows -- U.S. was 45 percent, Europe 17, Japan eight, Korea 3, South Asia 19, Taiwan six, and rest of world two percent.

The geographic distribution of our sales for the quarter was as follows -- U.S., 42 percent;

Europe 18;

Japan, seven;

Korea 3;

South Asia 17;

Taiwan, 11; and rest of world was two percent.

Moving to the balance sheet.

We ended the quarter with cash and marketable securities of $543 million, down about 37 million from Q2.

Receivables were 223 million or 17 percent of sales, an increase of 16 million over Q2 levels.

This represents 62 days sales outstanding, up one day from the previous quarter.

Inventory was 327 million or 25 percent of sales, down 16 million from the second quarter.

Capital spending totaled 24 million in Q3 while depreciation was 32.

Going forward, as George said, we expect sales for the fourth quarter to be between 310 million and 340 million, a loss of 22 cents per share, plus or minus three cents, before any possible special items.

Gross margins are expected to run between 25% to 28%.

R&D should run between 22% and 25% , and SG&A should be between 21% and 23%.

Our tax rate in the fourth quarter will be 36 percent for (ph) shipping.

We plan to reduce inventory by 15 million or more.

Capital spending will be around 20 million or less, and depreciation is expected to be around 33 million.

We expect our cash balances at the end of year to be around 510 million.

Now I'll turn it back to Tom (ph) for questions.

Tom Newman (ph): Thanks, Greg.

David (ph), we'd now like to open the discussion for questions.

All right, sir.

At this time I would like to remind everyone if would you like to ask a question, press star then the number 1 on your telephone keypad.

We'll pause a moment to compile the Q&A roster.

Your first question comes from John Pitzer (ph) of Credit Suisse First Boston.

John Pitzer (ph): Good morning.

I was hoping you'd give a little more granularity on what the current break even level is today and walk me through sort of on a quarter to quarter basis what kind of targets you're putting out there for fourth quarter break even and first quarter break even and what kind of cash burn do you anticipate before you reach the break-even level -- maybe what a cash flow break even level is for the company.

Greg Beecher (?): I'll answer part of the question.

In the fourth quarter, cash break even will be around 475.

Our GAAP -- I'm sorry, 375.

Our GAAP break even is about 340.

And that will be coming down each quarter in step.

Let me just take -- let me take it back.

We started last quarter at 479 break even.

We expect in the fourth quarter our break even to be 440 on a GAAP basis.

And we're going to take it down in steps thereafter to get to about 350 around the middle of next year.

The actions we're taking affect a number of items in our P&L.

It might helpful if I describe what the P& L, we expect to look like when we get to the 350 in terms of gross margin, as well as R&D and SG&A.

The gross margin in the 350 is expected to be 37 percent.

R&D will be 20 and SG&A will be about 18 percent.

One thing you got to keep in mind when you're going through a restructuring of this magnitude this past quarter, a lot of the actions affected people.

The subsequent quarters there's a lot of actions that come from material costs down.

And as you might imagine, that takes some time to roll through your P&L -- as you get into competitive shootouts often you adjust your pricing before you can get lower prices from your vendors, and that gets reflected in your P&L later.

I haven't scheduled out the break-even by quarter for public consumption, but we have plans internally as to how we're going to get there by the middle part of next year.

It's about 350.

John Pitzer (ph): Just as a quick follow-up, can you give us a sense of which divisions have the most to go to get to break even?

And George, specifically on the back plane (ph) and connector business, a lot of price competition coming out of the Far East there.

What's your view of that business going forward from a profitability perspective?

Well, I think in TCS's (ph) case specifically, they have the more difficult challenge in terms of pricing.

And of course the fundamental principle is we're not going to take business that we can't make some money at.

We're not going to make full profitability on many of the large buys that are going down these days, but we're not going to take business that doesn't make sense.

But as that business is under price pressure, of course, that impacts the mix we have when we get to 350.

If we have a much larger portion of the total company that's coming from TCS (ph), which is product with (ph) business under a lot of price pressure, we've (inaudible) 350 (ph) is going to be difficult to meet.

But on the other hand, if we have about the same mix we have now, that would be okay.

And I think we're going to see that TCS will be close to break even a little more than what they're shipping now.

In the $100 million range, they'll be break even, probably plus a little.

But that is a group that has the biggest ball, but that's of course the group that -- while we have excess capacity in many of our facilities around the world, that is a group that has the most difficult problem with underutilized assets such they tend to be printed (ph) circuit board fabrication plants which are not very flexible in terms of cost structure.

So that's a group that has the biggest problem.

John Pitzer (ph): Thanks, guys.

Your next question comes from Shakar Pramanick (ph) with Prudential.

Shakar Pramanick (ph): One more question on break even.

If you could give us a little more light why going to break even at 350 in the middle of the year versus maybe accelerated somewhat closer, maybe Q1, what's the thought process behind it?

What are the impediments in achieving that early on?

Let me take that one.

Keeping in mind we started at 479 last quarter and to go to 350, some of the actions, one of many actions is getting lower material costs from our vendors.

And that involves getting lower prices immediately from vendors as well as sourcing in low cost regions (ph).

And as you get those lower material components, you have to first sell (ph) up your old inventory.

That's at a higher cost.

It's not just based on how the economy works -- it takes longer to show up in your P&L.

George mentioned we're going to consolidate a number of facilities into one site.

That's not something that you can do immediately.

We're going to go as fast as we can.

But we need to make sure that we don't adversely affect customers or engineering programs as we downsize the business.

And we think a 350, given where we're at now, is an aggressive target with the timeline we set forth.

And the 350 is the number that we picked, and you could say it's arbitrary, but we think it's an aggressive number given where we are now.

Shakar Pramanick (ph): Great.

Also, on the Q4 orders, given the revenue guidance, should we expect basically the net orders down are about 10 percent or so?

You mean bookings?

Shakar Pramanick (ph): Yeah..

Well, here's how I'm thinking about it.

You have a pot that you can ship out of.

And that pot is made up of your backlog in new orders coming in.

Now, the question is -- isn't so much where that comes from, the question is, is the demand on the outside of that pot, which is currently around 330 million, does that demand hold up?

Now, we said we're going to ship 330.

So we think it will.

The rest becomes arithmetic.

If you start with backlog at about 540 million, and we're going to ship 330, you got to say what bookings would you expect to come in to keep the -- to support that demand going into Q1?

If you follow the math, I think it suggests that certainly Q4, Q1 bookings have to start coming up.

And so the real question isn't so much focused on, are you worried about backlog or what will your bookings be, but will the end demand of 330 million will now hold up?

If it does, then bookings have to start coming up.

Shakar Pramanick (ph): Right.

And booking kind of stayed at this 250 level or 240 level into the middle of next year.

Then there would be a problem.

Shakar Pramanick (ph): Then we have to bring the break-even down even further.

But it isn't so much if the bookings stay at 230, 240.

The question is, is the demand going to stay at 330.

If the demand is going to stay at 330, bookings have to come up.

We don't forecast bookings.

On the other hand, I look at them and I say -- we're not seeing anything that says there's a tremendous runaway of new orders.

We don't see evidence of that coming.

Nor do I see any evidence of significant disasters.

And at the same time, unit production is still up and increasing and the factories are pretty fully utilized.

So I think the -- we're betting on the 300 to the 330 range being what the demand level stays at.

Shakar Pramanick (ph): Also, are you seeing some of your new chips are taking longer test time, the tests (ph)?

Some do.

But I think it's more -- not because the -- I think it's testing longer because there's more new chips that are in the mix, and it takes a longer period of time for yields to get up and for people to get the whole process debugged.

I think on the one hand you have a unit test time (ph) going up but you have inventions (ph) that get higher ([inaudible) bring test sites (ph) down.

And I don't see a significant change in that.

The average test time, certainly.

Shakar Pramanick (ph): Thank you very much.

Your next question comes from Mehdi Hasingi (ph) of Downview (ph) Technology Group.

Mehdi Hasingi (ph): Going back to break even point, to what extent are you going to depend on new product introduction, particularly in the semi-test (ph) area to help you get that break even target.

And also I have a follow-up question.

There's two ways to think about to what extent you're depending on new product introduction.

One is, if you finish some new product work you're doing, you're able to free up resources, you can lower your break even by reducing your head count and not having to continue the investment in those programs.

As we get towards the middle of next year, we have some programs where we do expect they will be near completion and will free up people.

The other part of the question is, how much of those new products are you expecting in the revenue stream and -- the products we have fielded today are the products that will be part of that revenue stream as we get to break even.

Mehdi Hasingi (ph): And in regards to the backlog, could you clarify the estimate you made?

You said backlog is still at 540 and what was the percentage of shippable in the next six months?

71 percent.

Mehdi Hasingi (ph): And your gross margin guidance for 2003 was 25 to 28 percent?

George Chamillard (ph): For the fourth quarter, we provided guidance -- we said 37 percent, 350 break even model.

Mehdi Hasingi (ph): Okay.

Thank you.

Your next question comes from Vernon Esee (ph).

Vernon Esee (ph): Couple of questions, primarily on your end, Greg.

I wanted to go and revisit one point here on the break-even.

Can you help us understand sort of the gross margin results (ph) you're going to get going from -- reducing all these costs, how much of this is going to come from things you can't control, i.e. the reductions you took in overhead relative to leaning on your supply base and getting better pricing on that front?

I'm trying to figure out how much of that is sort of in Teradyne's bailiwick of cost control..

I think the vast majority is within our control.

We've gotten costs down on some material components, but it takes awhile for them to make their way through your P&L.

We're also going to source more low-cost regions (ph) and you heard the J 750 (ph) will be manufactured and shipped from Asia and that will lower our costs.

Much of it is in our control.

Also, some of our business in the electromechanical integration, we're going to take a hard look at some of that business that might be one, two, three percent profit business, low profit business that we may price that business away from us and keep the business that we can enjoy a greater return from.

So there's a bunch of things that we have on our list.

And I think virtually all of it is in our control.

Vernon Esee (ph): On the R&D side, on your sort of target, 20 percent at the 350 million break even, if we go back to '99, you were running at almost a 35 percent lower level on R&D spending.

I understand obviously you've grown.

There's other projects that are out there, but then on the other hand you've gotten out of some businesses since then.

I'm wondering why it's so high and if there's any solid explanation relative to the environment back then.

Well, I think first the smaller percentage was when sales were much more -- obviously a much larger number.

If the question is if we could get by on 10 or 15 percent, 10 or 12 percent engineering at that level, why can't we get by on 10 or 12 percent in engineering now, the answer is simple -- that's the lifeblood of the business.

If I was convinced it would be $350 million For the next decade I guess we'd figure out how to do that.

That's not what anyone expects.

And we're going to field products as aggressively as we can.

In fact, I'm more frustrated that I can't spend 25 percent on engineering rather than when (ph) I can't get in 15 at break even point because I think the engine that drives the whole business is that investment.

Vernon Esee (ph): Let me ask it like this.

You have a -- back then your dollar amount was running at a quarterly rate around about 45 million to 50 million and now you're targeting about a 70 million quarterly run rate.

That's 100 million annual difference in R&D over the last three years and I'm wondering what has substantially changed to justify that or quantify it.

One thing is the structure of the industry has changed as well.

For example, when the industry was all IDMs and you had highly vertical, make (ph) the way for test equipment, make the final test equipment, all in one factory, you developed a series of products that were aimed for that.

And it's a more complicated engineering support picture now that you have test houses (ph) who are driven by one set of characteristics and dynamics and IDMs who want something different.

As the outsourcing has occurred, complexity has gone up (inaudible).

That can change -- the rate of change in technology, in spite of the recession, hasn't slowed down.

And you need a new round of instrumentation -- you need two things in the system.

You need digital performance and you need instrumentation, particularly on SOC (ph) devices..

Every round of cycle, you need a broader and broader set of instrumentations that needs to be refreshed more frequently.

And those would be two structural changes I could think of.

Vernon Esee (ph): And so, going forward, do you think you may have more creative partnership arrangements with third parties on some of those issues, like instrumentation, what have you?

It's hard to tell because on the one hand you're going to say, what's the world seeing?

As people get to technology nodes that are now .13 (inaudible) and five levels (ph) and copper to connect (ph) and so forth and so on, you're finding that the cost to design a new device has gone way up.

For example, it's not inconceivable to have a radical (ph) set of mass (ph() that might cost a half million to a $1 million.

If you have to spin that a couple of times, the investment that people have to make to get the chip out may in fact slow down to the rate of change -- that's one hypothesis.

If that occurred you'd need less change in instrumentation.

On the other hand, because the new cost to get the new design out, as these designs take teams of a couple hundred engineering and cost a huge amount of money for the tooling, that either is going to slow down the rate of change or people are going to find some invention to change the formula to keep the high rate of change.

At this point no one knows.

If you ask me, do I think 10 years from now there might be a narrower set of instruments I could build an argument around it's going to be a narrow - yes, it's going to be narrower because only a few companies can tool the expensive, complicated devices.

On the other hand, I think there will be invention.

And right now it's sort of in the -- I don't know, what do you call it when it's fresh water, salt water coming together?

Vernon Esee (ph): Brackish.

Brackish.

And it's hard to tell what exactly it is.

But I know right now the investment you need to make is significant if you want to be positioned so the range of products that are coming out.

And that's why all the effort to get to fewer platforms and trying things like open architecture, et cetera.

That's why those are all popular thoughts right now.

Vernon Esee (ph): Just a last housekeeping question.

If you go over the Q3 order by business segment again.

I apologize, I missed that.

No problem.

We do it by business segment.

You wanted orders by business segment?

Vernon Esee (ph): I meant particularly through TCS test (ph) and what have you.

Semi-test (ph) was 52 percent of the total.

Connection systems was 20, assembly test 17 and the other tests I think was 10.

Vernon Esee (ph): Thank you very much.

Your next question comes from Timothy Arcuri (ph) of Deutsche Bank.

Timothy Arcuri (ph): Actually I have three quick ones.

First of all , can you maybe talk a little bit about what the optimum backlog level is?

Is there a point here -- we're kind of down to near four months of backlog, and my model shows that in Q1 maybe you might dip below four months of backlog.

Is there a point at which you start shipping out of backlog and you see a pop here in orders?

I think it's going back to the point I was trying to make, that I think that's close.

I don't know if it's fourth quarter or first.

But you obviously have to have a pause (ph) in bookings, if we're going to be able to maintain the shipment of the 330 million or so that we're at.

And that's the fundamental question.

Is the demand level going to stay at about the level of that now or go up or is it going to not stay there and (inaudible).

If it stays where it is and goes up, there has to be a pop in bookings, either in fourth quarter or first.

Timothy Arcuri (ph): Is there an optimum target you have for backlog, like in terms of months?

From an operations guy (ph), I'd like a couple years of backlog.

From the customer's point of view, he'd like to have about a day and a half.

You've got say, what's the efficient day in terms of running the factories and having reasonable control of efficiency in the factory.

And that certainly 13, 16 -- 16 weeks (ph).

Beyond that it's very nice.

Below that it's really tough (inaudible).

I think maybe this will help, but I think there's a lot of history that says that the test (ph) business for years has run between three months and six months, depending on where you are in the cycle.

So there's one data point.

There's another point, though, I think from the companies and the business, the test (ph) companies, and it depends a great deal on whether you're trying to ship memory testers or logic testers which have less configurability available.

Each one is much more like the other one, and therefore you can run with shorter backlogs and configure in a shorter time than if you're building SOC (ph) testers, which have a huge range of configurability and are harder to configure in a shorter amount of time.

There's a bunch of variables but the data says it's run between three and six months.

Timothy Arcuri (ph): If we go much below four that's getting to the minimum levels.

I think if you go below a quarter the business becomes tougher is what I would say.

Timothy Arcuri (ph): I guess the second question, can you help me understand, you answer with a break even in TCS (ph).

Can you tell us what the break-even is in ATE (ph) and also PCB (ph)?

As of Q4, '02, semi (ph) would be 227.

Assembly test division would be about 64.

Timothy Arcuri (ph): Okay.

And the target, when you get to 350?

We're not going to provide the detailed numbers by segment.

But we have plans to work our way there.

It's safe (ph) to say that each division is going to lower their break even.

Timothy Arcuri (ph): Thanks.

I guess the last thing, maybe George, can you help me understand the TCS ( ph) business?

We've seen book-to-bill has been less than what I believed for the last four quarters.

I always thought that was a turns business and it didn't really operate much on a backlog basis.

Can you help me understand maybe is there kind of a backlog element to that business that I'm missing here?

Starting this quarter, their backlog by the way is about 175 million.

I think on one hand you would describe -- what you do is you get these huge orders and they really aren't orders until they're pulled.

And there's agreements -- we can pull them within - pick a number, 12 weeks, 14 weeks, 18 weeks.

That's a firm commitment you have.

But you have a very long horizon that you can do planning against, (inaudible) capacity in inventory.

And so what I get (ph) would be, we ought to be seeing that go up again because bookings have come down dramatically.

I expect we'll see that pull-up.

I don't believe we'll see -- because (ph) there's an order for eight weeks in the future, please deliver it without some way where people have the ability to have horizons and know what to plan sizing around and so forth and so on.

So I would expect to see that starting to come back.

I know on the other hand the contract manufacturing side of the business, the customers want to keep that as short as they can keep it, because they just don't have very strong commitments from their customers.

I think this issue on backlog, though, let me just shift back to that for the total issue, not just TCS (ph).

When we get to 13 weeks, there's a risk profile, but in fact there's a risk profile for our customers and for us.

Because if I have a 13 week delivery, the next question you have to say is, how many systems can I ship with 13 week delivery?

If you come in you want to buy a six-pack or two six-packs of catalysts (ph), right, 12 machines, which by the way is not -- would not be abnormal to the way this thing starts to go when it takes off and you're assuming I have 13 weeks.

I do have 13 week delivery for a small number of fiscals, but lead times go up dramatically.

Now, both our customers and the equipment providers understand that dynamic very well.

And both recognize the interdependency they have on that.

So I think at 13 weeks it's not just a high risk for the equipment provider, it's an increasingly high risk for the device manufacturer.

Timothy Arcuri (ph): Thanks.

Next question comes from Adam Wolfman (ph) of Lehman Brothers.

Adam Wolfman (ph): Adam Wolfman (ph) for Ed White (ph).

I was wondering if you could talk a little bit about -- seems comments from a lot of the subcons (ph) indicate that utilization rates are running in the 50% to 60% range, how does this mesh with your comment that overall utilization rates for your semiconductor test systems are running at higher levels.

We must have (ph) market share gain.

Adam Wolfman (ph): How do you expect the business will likely respond heading to what is a seasonally a slow first quarter from a unit standpoint?

I think it's very hard to tell.

We've seen pretty consistent -- I think with one month exception we've seen seven months of unit growth here.

And we wouldn't see why that would change in the first quarter based on the macro events that we're aware of.

So -- and if you look at kind of third party surveys, VLSI (ph) Research and others that are making projections of test utilization going forward, their projections have gone up every month for three months for near term forecasting, near term utilization increases.

So the data that we have available, although the future is always uncertain and somewhat murky, the data that's available doesn't show anything tapering off, and we're not (ph) discouraged about the prospect going forward.

It looks like it should continue.

Adam Wolfman (ph): One last question.

Can you give us some sense of how the outlook for Q4 breaks out by business?

We generally don't forecast by business.

We do it in total and within any business you can have ups and downs and we generally are able to estimate it in total, but by division, we just don't do that as a practice.

Adam Wolfman (ph): I think last quarter you mentioned that you expected double digit sequential growth in the second half of the year in semiconductor test, and your hope was that the other businesses might grow around five percent sequentially.

We did say sometime back that connection systems would turn first and semi-test (ph) would be the next engine of growth and that occurred as well.

When we look into the fourth quarter, we said sales would be above flat, and I'd say the divisions will be pretty much close to where they were in the fourth quarter.

I don't see significant changes.

But it's possible semi could be down a little bit.

And we might get a little bit of help from the other test segment.

But you'll not see significant changes from what you will see in Q3.

I think, Adam (ph), your quotation of what we have said in the past was not based on the last call, it was based on the call before that, when we were experiencing significant growth -- 25 percent -- in the second quarter over the first quarter and expecting that to continue as of the April call.

The July call we clearly signaled that the growth was slowing and it was slowing in semi test and it was slowing in connections.

And I think that changed the tone of our outlook for the year substantially as of the last call.

Adam Wolfman (ph): Thanks very much.

Next question comes from Steven Pelayal (ph) of Morgan Stanley.

Steven Pelayal (ph): Just a little follow-up there.

I'm trying to reconcile here -- we're getting requests for accelerations from the subcontractors, we have full tester utilization rates, (inaudible) subcontractors, but yet I think your bookings are down in Taiwan and Southeast Asia, and we're hearing wafer (ph) starts, double-digit declines there and utilization rates.

The last question suggested you're 50%, 60%.

I guess I'm trying to figure out what's really going to happen here, maybe in the next one to two quarters in light of the backdrop and the seasonal impacts we can see.

I'm not sure we know, Steve (ph).

We aren't claiming we do.

There is a dichotomy here which is in some ways multifaceted.

You've got pretty persistent unit growth for the customers back to 90 percent of the 2000 peak, up from 85 percent of the peak in the last call.

You've also got slight growth by comparison in their sales month-to-month.

So they're under pressure not to order unless they absolutely need it.

So if they can delay, believe me, they're going to delay in this kind of environment.

The second issue is the subcontractors, clearly weakened, as we said in the last call, that they would.

They did weaken in the third quarter, and on the margin, they're not getting encouraging signs from their customers about the end markets related to back to school first and then Christmas.

And so they're nervous.

If they can hold off ordering again they're absolutely going to do that.

On the other hand, the utilization information that we gathered, and it is from a couple dozen different IBM locations and 30 different subcontract locations, persistently point out our utilization is extremely high and is running about 90 percent at both IDMs (ph) and subcontractors.

So they don't have a lot of margin for error here.

They have to be very cautious about projecting their capacity utilization needs for tests or they're going to get even more squeezed.

It's an uncomfortable environment for them, but we seem to have substantially higher utilization than others, at least in the data we look at.

And any slight increase in demand has to result in more test orders.

And if units keeps growing that has to result in more test orders.

So I agree with the conflicts, but I mean customers don't have an easy road here either.

They're balancing a lot of things that on the margin, if they can delay ordering, they'll do that.

But they're under a lot of pressure to order to meet their capacity and growing capacity needs.

I don't have any better answer than that unfortunately.

Steven Pelayal (ph): I hear you loud and clear.

I'd like to see George's two years of backlog, too.

In the mix of bookings in the fourth quarter, do you expect much of a change there?

Still ATE (ph) to be roughly to be roughly 50%, 55% or so, and then maybe I guess within ATE (ph) it was mixed back over IDMs (ph) over subcons (ph)?

Probably, yeah.

I think that's the best scenario we would have.

If you look at IDMs (ph) on semi-booking (ph), Q1, Q2, Q3, IDMs (ph) have been 70%, 72.5%, 75%.

So the biggest portion of the orders are coming from IDMs (ph), which somehow surprises me a little.

I don't quite understand that, because - but I also find that encouraging, because for the IDM (ph) to be making the cap ex risk (inaudible), he (ph) must know something I don't know.

And consequently, at the same time, the fabs (ph) and subcons (ph) were down slightly.

Steven Pelayal (ph): George, do you know that ratio show at the peak, what the subcons represented as a percentage?

I assume it was a majority.

In the peak bookings quarter at $1 billion, the subcons were 35 percent, up semi-test (ph).

Steven Pelayal (ph): So none of them (ph) were really the majority (ph).

Last question was, when you look at AT (ph) representing approximately half of the bookings and revenues, is there a large service component in there?

I'm curious, how much does that represent?

It's generally less than 10 percent of the revenues -- the service revenue that's separately ...

Steven Pelayal (ph): And at today's levels, that sounds about right?

It's around 10 today - that neighborhood.

We don't break it out in the 10-K because it's not over 10%, but I know it's under that amount.

Your next question comes from Jim Covello (ph) of Goldman Sachs.

Jim Covello (ph): Thanks so much.

I have a quick question on the competitive environment, the impact you're seeing from the competitive environment on pricing, market share, you touched how your share is going up, but I wonder if you could provide us some granularity there.

And at what point is there going to be enough pressure in this industry to force some consolidation?

Thanks so much.

I think the competitive environment is always tough in downturns, and in that regard I don't think we see a dramatically worse or better situation in this downturn than we've seen in others.

I think pricing is difficult, but it's not the extreme situation and the great deterioration that we've seen in the printed circuit business in TCS.

So it's competitive, but it's not disastrous.

I think we're feeling pretty good about our position in the marketplace at this point in the non-memory area.

There's not much going on in memory at all for us, at least in a respect for others in the DRAM space.

There may be in Flash but we're not in that segment anymore.

So in non-memory we're feeling pretty good at this point.

On consolidation, I think I'll let George try to answer that..

Consolidation is tough in the industry, because the fact is that everybody has an installed base and they have products that overlap and the time it takes and the cost that you incur shifting a customer from the acquired company's product to your product, you're just as well off beating them in the market, winning the business, and converting them that way.

I don't see a big consolidation issue coming.

I think we'll find out that somehow everyone lives through this, and the number of players doesn't change very much on the outside.

Who has share may change dramatically.

And I hope that's us.

I feel good about market share in the past quarter and where we're at now.

But for the year it's too early to say who's gained share or not.

I think probably we're holding our own over the year.

But I see right now with the emphasis on high volume, low mix parts, smart cards and less complex logic devices and less complex mixed signal and image processing, that's playing to an area where we have a great portfolio.

So that's helping us right now.

So maybe we'll gain share over the year.

Maybe we won't.

But I don't think there's going to be a lot of consolidation.

Your next question comes from Glen Yeung (ph) of Salomon Smith Barney.

Glen Yeung (ph): If you look at your view that you're probably going to hang around the $330 million number for the next couple of quarters, when you talk about how much of that is real end market demand versus potentially share gain for your class, all your segments?

I think too close to call in this kind of environment.

I don't think we have enough examples to point to to say that we've got X million from share gain in this kind of market.

The problem at this stage of the year, you don't really have good estimates of the total market size.

Everybody is guessing.

So we kind of go back and compare our shipment information, augmented shipments against everybody else's that's public.

But that's only a wet finger in the wind kind of estimate, it's not precise.

I don't feel we can make an estimate on that at this point.

We feel good on the analyses we can do.

But it's very early to tell, as George said.

Glen Yeung (ph): There's no changes in terms of other types of mixes - I think you've talked about mix of your divisional businesses.

Geographically, any changes in mix that you're expecting to see, even at the same levels of revenue?

I think the biggest mix change that we talked about is kind of the high volume segment of the market, whether it's emmon censors (ph) or Logic or mixed signal analog stuff with the flux (ph) system.

And we're benefiting from that.

This kind of big portfolio that we have of products and semi tests and also businesses in the company helps us a lot on this kind of environment.

So no matter where the option is, we tend to be a beneficiary, but the biggest trend that we said is toward the really high volume portions of the market and we're benefiting from that.

Glen Yeung (ph): Thanks.

Your next question comes from Greg Kanisny (ph) of U.S. Bank.

Greg Kanisny (ph): Somewhat of a follow-up to Glen's (ph) question.

What are the end applications you think we should be tracking here in terms of what could drive tester (ph) demand for next year?

Sounds like it's a little more moving towards kind of consumer electronics, maybe wireless, or if you could add any color to where you think that's kind of headed for you guys.

(inaudible) you're going to say who buys semiconductors, and who uses them, and who is the end customer for those products.

And of course it breaks into the three high level - it's consumers, it's enterprises, whether for communications or computers or whatever.

And it's the government.

In the areas (ph) that have been called (ph) to debt have been the communication/EIT segment.

I think if there's an upside to this thing, it has to come from that segment getting stronger.

I think the reason -- consumers generally, devices that consumers consume generally fit on the high mix, low volume, lower cost.

That's not pure, it's just where the majority is.

That's why I think we're seeing a steady shift of our testers going to Asia from North America.

For the year 2001, North American shipments were 44 percent.

The last four quarters they're 34 percent.

And I think that is a -- that reinforces to some extent the dependency on the low-volume high-mix devices.

And I think we'll see that continue for a while.

Greg Kanisny (ph): Is it safe to say that consumer electronics is a bigger driver overall for your business than it was say a year or two ago?

It certainly wasn't the peak.

I'm not sure about a year ago.

Greg Kanisny (ph): And then just want to ask you a question about equity in the business and the restructurings that you have planned going forward and the cash that could be coming out for severance and other uses, and also the cash burn.

When you look at what you expect to happen here over the next few quarters, how much equity do you think that will take out of the business?

Is that 100 million or a few hundred million?

Could you just give us maybe a rough idea of what you think might happen there?

That's not something that I can forecast with accuracy, but I can tell you that the actions this past quarter first in terms of cash, the severances will cost us about $12 million.

The special items we detailed give us 12 million going out and we also received a gain of seven million in cash.

So specialized (ph) this quarter was a net five million.

As you go forward, some of the facility consolidations, as an example, may require some small amount of cash in terms of fitting (ph) up a building.

I don't think it's going to be significant.

If you look at the material costs now, we're going to save money there by lowering our break even, so there's no cash cost.

And some of the people actions (ph) will require some cash but we don't think the bulk of it by any stretch is going to come from the people, and our attrition is going to get us a long way there, as well as about 50 people a month.

So when you look at the cash, I don't see any issues on that front.

In terms of stockholders equity, sure there could be certain non-cash charges that reduce stockholders equity.

But that's not something that I think -- that I worry too much about in terms of an accounting entry that wipes out an asset if it's the right accounting and there's no cash impact.

Greg Kanisny (ph): Would you expect to have further writedowns of either intangibles or PP&E type assets here the next couple of quarters?

I don't know of any.

So there's none that I have in mind but it's certainly possible in this environment that there could be impairments in the future.

Greg Kanisny (ph): Okay.

Great.

Thanks.

David, we're a little over our time.

We'll take one last question.

Your final question comes from Kevin Vasily (ph) of Thomas Weisel Partners.

Kevin Vasily (ph): Two quick questions on bookings.

Well, actually one in bookings and one in cost savings.

On the bookings side, what was the percentage of bookings from ATE (ph) in the June quarter?

Is assembly test division (ph) you're asking about?

Greg Kanisny (ph): ATE (ph).

Semi test or all of test?

Greg Kanisny (ph): Semi test.

Semi test bookings were 53 percent or $123 million in the second quarter.

Greg Kanisny (ph): And then one question on the cost restructuring you guys are doing.

You made mention of a $12 million number for -- as a result of some of the moves you've made.

Is that for the full year or does that take effect fully in Q4?

GB (?): That's fully in Q4.

Greg Kanisny (ph): That's something that annualized is in the neighborhood of $40 million to $50 million.

Greg Beecher (?): Correct.

And let me go back to the question that was asked earlier, in terms of charges that could occur in the future.

The one item that was disclosed in our 10-Q and 10-K for some period of time that has accounting consequences is our deferred tax assets, and I think you'll see more technology companies disclose that there's a possibility that the deferred tax assets might require a valuation allowance.

And basically, quickly, the accounting rules require you to look more at the past results and weigh that more heavily than the recent past, by the way, versus your forecast for a longer-term history, and therefore it's certainly possible we could provide a valuation allowance or a write-down to deferred (ph) tax asset in the foreseeable future.

That doesn't change the fact that we have these NOLs or tax attributes that are still available.

It's just a journal entry.

George Chamillard (?): If that helps answer your question about what could happen to stockholders equity - that's one I think could affect it in the future.

Greg Kanisny (ph): OK.

Thank you.

Ladies and gentlemen, we have reached the end of the allotted time for questions and answers.

We will now go back to Mr. Tom Newman (ph) for any closing remarks.

Tom Newman (ph): Thanks, David (ph).

We'd just like to thank everyone for their interest in the company and for their participation, and we'll look forward to talking to you all next quarter.

This concludes today's Teradyne third quarter earnings release conference call.

Thank you for participating.

You may now disconnect.