Category Archives: semiconductor manufacturing

INTC’s strong 2Q12

July 20, 2012 By in earnings conference calls, Industry Analysis, semiconductor manufacturing, Stock ideas, Technology, Wireless Tags: , , , , , , Leave a comment

strong 2Q results

After the equity market close in New York on Tuesday the 17th, INTC reported its 2Q12 earnings.  The company generated revenue of $13.5 billion during the three months ending in June.  That was up slightly from the $13.0 billion the company took in during the comparable period of 2011.  Earnings per share were $.57.  That was flat, year on year, but considerably ahead of the $.52 consensus of brokerage house analysts.

I wonder how real this positive earnings surprise actually was, although I haven’t been paying enough attention to the Wall Street consensus to be sure.  My sense is that the consensus was somewhat higher until rival chipmaker AMD reported very weak results a week or so ago. The knee-jerk reaction of analysts was to lower their INTC numbers.   As it turns out, INTC has been taking market share from AMD in the low end of the microprocessor market, where AMD tends to operate.

Despite weakness in afterhours trading right after the report, INTC responded to the earnings news on Wednesday by gaining close to 4%.  This, even though the company lowering revenue guidance for the rest of the year.

lower guidance for the second half

INTC now thinks its full-year 2012 revenue will be only 3%-5% higher than it reported for 2011.  That’s only about half the “high single digit” (read: 8%) sales gain it had anticipated three months ago.  The first half was up about 2%, yoy.

What’s changed?

The US and EU economies are remaining weak for longer than INTC had expected.  Emerging markets are slowing.  In response, as well as in anticipation of a buying slowdown before the debut of the Windows 8 operating system (slated for October 26), consumer PC builders are keeping their chip inventories very lean.

The lower revenues will probably clip about $.20-$.25 a share from the non-GAAP figure of $2.75 I had estimated INTC would earn for 2012 eight months ago.

profits won’t be hurt quite as much

The corporate business is strong.  The cloud is booming.  The traditional INTC consumer customers who are buying are tending to choose high-end PCs, which mean higher profit for INTC.  At the low end, INTC appears to be taking considerable market share from AMD.

On the cost side, the current lull is allowing INTC to satisfy demand and still close trailing-edge 32nm factories.  That way it can put that equipment into new state-of-the-art 22nm fabs, rather than buying more expensive new machines.  The company is also slowing down hiring.  All in all, cost savings could end up adding $.05 to second-half results.

2013 is much more important than 2012,

in my view.

2012 was always going to be a transition period.  It’s the last of several years of very heavy spending by INTC on R&D and cutting edge chip factories.  The investment is aimed at putting even greater distance in processing technique between INTC and its rivals, and at producing small, powerful, energy-stingy products that will gain INTC entry into the burgeoning market for smartphones and tablets.

Early results on both fronts are encouraging:

–the ramp-up of 22 nm is going faster than INTC had thought

–Ultrabooks are already doing ok.  140+ new designs are now in the pipeline, including 40+ with touch screens and 12 that will be tablet/laptop convertibles.  And entry level ultrabooks will be priced around $700 by fall.  So 2H12 Ultrabook performance stands to be considerably better.

–Lenovo, Lava and Orange have already launched INTC-driven smartphones.

Success in these areas is far more important for the long-term progress of INTC shares, I think, than whether 2H12 can live up to expectations formed when the we thought the economic problems of the developed world were capable of being solved by governments more quickly.

my bottom line

I think 2Q12 results and guidance underscore the idea that INTC is doing as well as can be expected in an unfavorable macro environment.  I don’t have any great insight into how successful INTC’s repositioning for mobile computing will be.  On the other hand, I think today’s stock price still assumes the worst.  More than that, I don’t think the price fully reflects the attractiveness of INTC’s corporate, server and cloud operations, considered by themselves.  So I continue to think that the stock is a very reasonable bet.

 

INTC and ASML (III): the structure of the agreement

July 13, 2012 By in Industry Analysis, semiconductor manufacturing, Stock ideas, Technology Tags: , , , , , , Leave a comment

six months in the making

ASML says it began negotiations six months ago with its largest customers about  their contributing to R&D and making commitments to purchase next-generation machines.  As far as I’m aware, no one has yet said who initiated the talks.  It wouldn’t surprise me if INTC got the ball rolling.  The salient points would be that:

–INTC needs next-generation equipment in four years, not six;

–it would contribute a significant amount of money to the required R&D effort; and

–it was going to partner with someone, but would prefer that the partner be ASML.

Because every supplier studies his customers extremely carefully, it’s equally possible that ASML anticipated INTC and made the first move, giving the same rationale.

the major provisions of the agreement

1.  A group of customers will provide up to $1.7 billion in R&D money to ASML.  The funds will be spent on next-generation lithography and on migration from machines that use 12″ silicon wafers to 18″.  Part of the money will be treated as an advance on purchase of next-generation equipment and will be returned through a lower price on machines bought.  Part won’t.  We don’t know the proportions.

By doing this, ASML reduces the amount it will have to put up itself of the several billion dollars it will cost to develop the new equipment.  At the same time, it locks in customers for its next-generation equipment.  ASML scientists will doubtless get excellent access to its partners’ R&D staffs, as well.

INTC, which has agreed to put up 60% of this money, estimates it will get next-generation machines two years earlier than it would otherwise.

ASML is still talking with TSMC and Samsung–and possibly with others.  In my view, it’s not a sure thing, however, that any others see the same pressing need INTC does.

2.  ASML intends to sell up to a 25% ownership interest in itself to the R&D partners, at a fixed price of €39.91 per share.  The share purchase will presumably be proportionate to the R&D commitment.

INTC has agreed to buy 60% of this stock, or about $3 billion worth.

The stock that customers purchase will be restricted by being placed in a special trust.  Under normal circumstances, the stock will have no voting rights and will not be transferable to another party.  So the customers will truly be passive investors.

We’ll know more clearly why the equity sale is necessary when the trust documents are available.

My take is that this is a provision insisted on by INTC, not ASML.   I see it as protection against the possibility that current management might lose control of the company, either to an activist investor group or to a bid from a semiconductor equipment conglomerate.  The worry would be that the R&D partners might lose access to the fruits of next-generation research, or that a new owner might slow down the pace of progress or shut down the next-generation project altogether.

3.  ASML won’t keep the money it gets from the stock sale.  It will distribute it instead to shareholders (ex the customers).

Why?

ASML’s market cap would be about $25 billion after the share sale, if it did nothing–and if the stock continued to trade at yesterday’s price.  The company would have net working capital, after repayment of all debt, of around $8 billion.  Most of that would be cash.  I think the temptation to a prospective acquirer of a large acquisition-funding cash pile would be too great to pass up.  And ASML doesn’t need the money.

4. ASML will conduct a “synthetic buyback” of shares from existing shareholders in the same amount it sells to customers.

Here’s where things get a little weird.  But ASML has done this before, in 2007.

I think I understand what’s going on, but don’t count on it.

With shareholder approval, ASML will

–use accounting legerdemain to transform the proposed payment to shareholders into a return of capital, meaning (I think–but don’t bet the farm on my read) it won’t be subject to tax

–make the distribution

–proportionately reduce each shareowner’s holding (ex the customers) through a reverse split, so that the number of outstanding shares will be the same as before the sale to customers.

The result of all this will be to transfer ownership of up to 25% of the company away from existing shareholders to the customer group, without triggering any tax effect.  So existing shareholders will be able to lighten up, near the peak of this semiconductor equipment cycle, and at a price about 4x the low of late 2008–all without having to pay tax.  Who’s going to complain about that?

what to watch for

I think this is a really good deal for both INTC and ASML.  I’ll be interested to see if TSMC or Samsung sign up as well.

They might not.  They could do nothing.  Or they could cut a similar deal with another lithography company.  Yes, the others would be Japanese, who are notoriously difficult to deal with and who are part of a legal and cultural environment that is not friendly to foreigners.  But the Netherlands is no walk in the park for an outsider, either.

How the others act will give us a good read on how much of a lead, if any, ASML has in EUV or 18″.

For what it’s worth, I think ASML is a great company, but I haven’t owned it for years and have no intention of buying it now.  I do continue to own INTC, though.

 

 

INTC and ASML (II)

July 12, 2012 By in Industry Analysis, semiconductor manufacturing, Stock ideas, Technology, Wireless Tags: , , , , , , , , Leave a comment

recap

INTC has agreed to pay $3 billion to acquire a 15% ownership interest in ASML.  In addition, it is pledging to make a $1 billion contribution to ASML’s R&D efforts to develop next-generation semiconductor lithography tools.

Yesterday, I wrote about the deal in general and about what ASML gets from it.  Tomorrow, I’ll write about the specifics of the arrangement, as far as they’re known.  Today’s topic is why INTC is interested.

a little history:  evolutionary threats to INTC

A semiconductor factory costs, say, $3 billion to build and outfit.  It will spew out maybe $8 billion worth of annual output.

Who has that kind of money to spend?   …INTC and Samsung, and basically no one else.

More important, who has a high enough level of sales to use the plant effectively, once it’s up?    …the same two, INTC and Samsung.

On the surface, this huge capital investment requirement represents a significant barrier to entry into the semiconductor-making business.

It did once.  No longer, however.

The industry evolved.

Specifically:

–foundries like TSMC have started up.  They own semiconductor plants but don’t manufacture products they design.  They specialize in making chips under contract from third-parties.  The fact the foundries aren’t direct competitors makes semiconductor design firms much more willing to trust their intellectual property to them.

–an intellectual property company, ARMH, began to develop and sell standardized chip designs to others, plus toolkits to customize them.

This laid the groundwork for lots of little semiconductor design firms to spring up, who create innovative enhancements to the basic ARMH design and have the chips made through foundries.

ARMH-based chips don’t have as much computing power as INTC offerings.  But they’ve smaller and more flexible.  They generate less heat and use less power.  In other words, they’re perfect for use in the smartphone/tablet business they’ve enabled.

Over the years ARMH chips have also been gaining in computing power, to the point where they’re now beginning to be designed into PCs and low-end servers.

the INTC response

Just as ARMH chips have been adding computing power so they can enter traditional INTC markets, INTC has been working to make its chips smaller, cooler and less electric power-hungry.  That way it plans to become a presence in the burgeoning mobile market now dominated by ARMH.

the Cloud is a plus

News on the mobile front isn’t all bad.  The storage and computational limitations of mobile devices means they have to be backed up by very big and very agile servers in the “Cloud.”  Virtually all Cloud servers run on INTC chips.

INTC thinks it has already pulled even with ARMH…

..in terms of small size, low heat and low electricity use for the chips it intends to be used in mobile devices.  It’s done this both through chip design and by very aggressively adopting new semiconductor manufacturing technology.  Most industry observers think INTC has at least a 1-2 year advantage in process techniques over foundries like TSMC.  We’ll be seeing the first INTC-based smartphones and tablets late this year or early next.

where to from here?

This is where the ASML deal comes in.

Let’s assume that INTC presently has a two-year lead in process technology over other chip makers and that it anticipates needing next-generation manufacturing equipment four years from now.  Will that equipment be available?

Maybe not.

If I were ASML and I thought I’d make some sales of next-generation equipment in 2017, but that no one else would be ready to buy until 2019-2020, I’d be in no rush to complete my research and development in the shortest possible time.  I’d go to a lot of extra expense to do that, but I wouldn’t be able to really cash in until a couple of years later.  And of course, I’d have the additional risk that rivals would reverse-engineer my machines and have better ones available when the the big market ultimately developed.

So I don’t think ASML has any natural incentive to speed up its R&D to suit INTC’s schedule.  Therefore, without the ASML deal, INTC risks seeing its semiconductor manufacturing advantage frittered away as it is forced to wait for new machinery to be perfected.  That’s why $4 billion may be a reasonable price for INTC to pay to eliminate this risk.

funding will come from offshore

INTC intends to use cash it has permanently invested outside the US to pay for the ASML deal.  Were that money repatriated to the US, it would be subject to federal income tax of about $1.4 billion.   So finding a potentially high-return use of the funds is in itself another plus.

That’s it for today.  Tomorrow, the deal structure.

INTC and ASML

July 11, 2012 By in Industry Analysis, semiconductor manufacturing, Stock ideas, Technology Tags: , , , Leave a comment

broad outlines of the deal

Yesterday INTC and ASML announced an agreement under which INTC will pay ASML a bit more than $4 billion over the next five years.  The money, which will come entirely from balances INTC has on deposit outside the US, comes in two piles:

–$3 billion+ will go to acquire a 15% ownership interest in ASML, and

–$1 billion+ will be a contribution to ASML’s R&D efforts.  Part of this will be purely a research expense; part will be a prepayment on the future delivery of the tools that research will enable.

ASML’s goal is to raise a total of $6.7 billion through such deals.  It’s now in talks with its other two big customers, TSMC and Samsung.  INTC’s commitment brings ASML 60% of the way to its goal.

three posts

I’m going to write about this deal in three posts.  Today, I’ll look at it from ASML’s perspective and tomorrow from INTC’s.  On Friday, I’ll write about the deal itself in detail, including some of its more eccentric aspects.

Here goes…

from ASML’s perspective

who ASML is

ASML is a Netherlands-based maker of semiconductor production equipment.  Its specialty is lithography machines, devices that place the semiconductor design onto the silicon that will ultimately become the semiconductor chip.

ASML has two traditional rivals, Canon and Nikon.

increasing capital intensity

The increasingly capital-intensive nature of semiconductor manufacturing has several implications for ASML:

–R&D budgets are getting bigger.  ASML will spend at least several billion dollars developing the next generation of machines.

–the machines themselves are getting more expensive.  Top-of-the-line lithography machines can cost $70 million apiece today.  Next generation ones will probably cost $100 million+.

–as a result of the big capital requirements, ASML has only three really important customers.  They are:  INTC, TSMC and Samsung.

today’s inflection point

INTC is currently using lithography machines that can “write” the lines of a chip design onto silicon as close to one another as 22 nanometers.  TSMC is at 28nm.  INTC is planning to be at 14nm next year.  As a general rule, the smaller the line spacing the faster the chip, the less power it uses and the less heat it throws off.  So smaller is (very) good.

The semiconductor equipment industry is starting to work on next-generation lithography machines, using an Extreme Ultra Violet (EUV) process that will allow lines to be drawn within 10nm or less of one another.  The industry expectation is that they will be available in about six or seven years.

At the same time, semiconductor equipment makers are also starting work on the less expensive task of increasing the maximum size of the silicon disks they can handle from the current 12″ to 18″.  That has the potential of doubling the number of chips that one disk will yield, and dropping unit costs by 30%-40%.

Q:  bet the farm or not?    A:  not.

ASML is in a boom or bust business.  It’s currently in boom.  At the end of 2011, the company had net cash of over $3 billion.  The current stock price is 4x its level during the bust days of late 2008, when ASML was losing money.

The company thinks that, in addition to the cash flow it expects to generate, it will need to spend close to another $2 billion to develop EUV and 18″ products.  Should it bet the farm and take on the expense alone–possibly having to do an equity offering to raise funds?   …and without any assurance that customers would buy the new machines?   Or should it try to get the three big semiconductor producers to help fund development and agree in advance to purchase  ASML’s next-generation output, rather than that of a Japanese rival?

It has chosen the more conservative course.  INTC is the first to sign up.

cashing out existing holders

Raising the needed R&D money from customers rather than raising it from existing holders through a rights issue is one thing.  But ASML is going a step farther.  It’s proposing to use a “synthetic buyback” to distribute to existing shareholders the entire $5 billion it gets from issuing new stock, while reducing their equity interest from 100% of the company to 75%.  It’s kind of like a rights issue in reverse.

Presumably, major Netherlands holders have already given their okay (telling you something about what they think of today’s ASML stock quote).

More on this on Friday.

TSMC’s 28 nanometer problems: significance

May 2, 2012 By in Industry Analysis, semiconductor manufacturing, Stock ideas, Technology Tags: , , , , , , , , 1 Comment

Rumors have been swirling for some time in tech circles about difficulties the Taiwanese foundry, Taiwan Semiconductor Manufacturing Company (TSMC), is having in bringing its latest cutting-edge chip fabrication lines into full production.  The stories were confirmed when QCOM warned in its latest earnings conference call that over the next quarter or two it would be unable to supply customers with all the most advanced chips they wanted. (Interestingly, in its quarterly earnings call, AAPL said it would be unaffected because it isn’t using 28 nm chips.)

Why is this important?

background

1.  For many semiconductor chips, the history of their manufacture is one of constant attempts to make more complex and faster speed, but also smaller, less power-hungry and cooler output.  One of the main ways of accomplishing all but the first of these goals has been to shrink the spacing between the lines of the chip patterns written onto silicon.

2.  A nanometer is a billionth of a meter.   The 28 nanometer spacing that TSMC is having trouble with is, therefore, a distance between lines of 28 billionths of an inch.

3.  About twenty years ago, the foundry–or third-party manufacturing–business began to come into prominence, as several positive factors for that industry converged.  The increasing complexity of semiconductor “fabs” meant that it cost $3 billion to build one.  Even worse, a fab churned out $7+ billion in output, far beyond the sales of all but the largest companies.  At the same time, a generation of ambitious chip designers wanted to break away from stodgier established firms and develop chip designs on their own. Many focused on customizing templates provided by ARM Holdings (ARMH).

4.  The unquestioned leader in the foundry arena is TSMC.

a paradigm shift in the offing?

There are two big integrated semiconductor designer/fabricators left–INTC and Samsung.  Neither is having fabrication problems.  INTC is beginning to produce 22nm chips in volume, and promises 14 nm for 2013.  In addition, it is using a new production technique that it calls “3-D,” that gets an unusually large benefit from its current linewidth shrink.  Most important, in my view, is that the company seems increasingly concerned with providing customers with products they want, rather than just the latest engineering tour de force.

Samsung already provides foundry services to others–it builds many AAPL chips, for example.  And INTC’s mammoth capital spending campaign of 2011-12 has analysts asking–and the company denying–that it intends to offer similar foundry services in the future.

my thoughts

ARMH, which has–with justification–been an immense market outperformer as the one-stop-shopping way to play the mobile device chips that the design firms/foundry model has been churning out.  But the stock (at 55x historic eps) is down about 20% over the past year, a time when INTC shares (12x) is up by 25%.  Over the same period, Samsung Electronics (5930.KS) (15x) is up 50%.

Yes, the issue with ARMH may just be the high PE multiple.  And, yes, Samsung isn’t just chips.  It’s a force in smartphones and dominant in TVs.  And it trades in a market that marches to its own drummer.  But I think the market is saying that the old integrated model has more going for it than the consensus appreciates.  I also think the market is right.

TSMC’s fabrication difficulties may be the trigger that gets a wider group of investors to focus on the change.