TARP fund math

Banking is a business of spreads. A bank borrows money from depositors at one rate and loans money to creditors at a higher rate. They collect the spread. The Treasury loans money at 8%. The Fed has set interest rates at near zero. How can a bank loan money to anyone but the most risky of creditors with that kind of spread? They can't really. They have to invest in risky assets. There is a disincentive to lend implicit in these TARP loans.

Moral hazard is in high gear, my friends.

NAND contract significantly higher

NAND flash contract prices soar by up to 40% in 2H January, says DRAMeXchange
Josephine Lien, Taipei; Jessie Shen, DIGITIMES [Friday 16 January 2009]

Contract quotes for mainstream 8Gb and 4Gb multi-level cell (MLC) NAND flash jumped by up to 40% and 33%, respectively, in the second half of January, latest quotes gathered by DRAMeXchange show. Meanwhile, contract prices for the 16Gb MLC NAND flash segment grew 2-8%.

The price surge has appeared as a result of suppliers' production cutbacks and lower 8-inch fab output, according to the research firm. In addition, demand for restocking inventory before the Lunar New Year will contribute to price stabilization for the near term.

Most contract prices of single-level cell (SLC) parts, whose demand has decreased, also rose slightly in the second half of January, DRAMeXchange said.

En-Min Jow, chairman of NAND flash controller IC designer Phison Electronics, said that a rise in NAND flash pricing is expected to continue throughout the first half of February. The price hike will start slowing down in March, as chip suppliers begin ramping up production on a restored supply/demand balance, according to Jow.

Mainstream SLC NAND flash contract prices, 2H Jan 09 (US$)
Item	High	Low	Average	High change	Low change
16Gb (2Gx8)	11.50	10.50	11.00	(4.17%)	(4.55%)
8Gb (1024Mx8)	5.00	4.00	4.50	2.04%	0.00%
4Gb (512Mx8)	3.50	2.50	2.98	16.67%	25.00%
Mainstream MLC NAND flash contract prices, 2H Jan 09 (US$)
Item	High	Low	Average	High change	Low change
16Gb (2Gx8)	2.70	2.20	2.46	8.00%	2.33%
8Gb (1024Mx8)	2.10	1.50	1.82	40.00%	15.38%
4Gb (512Mx8)	1.60	1.20	1.40	33.33%	14.29%

Source: DRAMeXchange, compiled by Digitimes, January 2009

The Skew: Contract and spot prices are generally in the same range now. Previously contract was significantly below spot. The good news is spot prices haven't given back their gains since jolting higher a couple of weeks ago. Production cutbacks should lead to higher NAND prices. Taiwan bailouts may quash a notional recovery as rejuvenated cash flow would likely lead to more production. Like Sandisk for IP royalty catalog though quarter is so bad I'm going to have to read about it through finger-covered eyes when they report.

LG Display (LPL)

Operating loss worse than their December 10^th re-guide down. Panel prices dove 23% in Q4 on 5% shipment growth. Co expects panel shipments down mid single digits for 1Q:09. Believes the "worst in terms of earnings" occurred in December, though prices will continue to decline in 1Q. Capacity utilization, presently at 60%, will rise to the mid 90% range in 1Q. Capex spending intentions down by 50%.

The Skew: Despite the uncertain spending environment and their inability to predict sustainability of a recovery, they'll be running production back up next quarter. Analysts are calling for a Q3 recovery. If LG is running higher production into uncertain demand, the company and/or analysts could easily be wrong about pricing recovery. No edge but seems risky. More interested in this company as a look-through to consumer spending and channel inventories.

Intel (INTC) still expensive, still rigging margins

See link

ThinkEquity calls it the kitchen sink quarter and expects 2H09 margins to be 50-60%.

The Skew: Yes, 1Q:09 is intended to be the kitchen sink quarter. The trajectory of a recovery remains uncertain. The company is going to bomb pricing. The only thing keeping $30 Atom chips from the larger form factor notebook market is Intel's restrictive sales policies – they're impeding the price curve to try to keep pricing of Centrino 2 from tanking. Intel estimates are going to be about 50 cents for 2009 now – what do you want to pay for a company making 50 cents that hasn't grown in the last 4 years? Now what do you pay for that company if sales drop 30%? Is the answer really 28 * 2009 estimates? It shouldn't be. I'd say even at 20* you're overpaying and that's down 35% from here. Definitely not a buyer of Intel.

NPD December video game results

NPD released December video game sales information. Software grew 15% y/y, better than analyst expectations of roughly 5-10%. Activision (ATVI) remained the number one publisher despite Guitar Hero band kits costing more than consoles in some cases – they were up 7% y/y overall with Guitar Hero up 14% y/y on a higher ASP. Electronic Arts (ERTS) was up 11% y/y. THQ Interactive (THQI) was down 23%. Wii titles dominated the top selling list, though Nintendo software sales +37% y/y may be a bit below street estimates of +50%, according to Deutsche Bank. Nintendo (NTDOY) Wii dominated platform sales with 2.1 mil units, up 59% y/y but up only 5% m/m – Nintendo essentially doubled shipment rates in the last couple of months of 2008 – market share grew substantially to 46%. Xbox sold 1.4 mil units, up 14% y/y and Sony (SNE) PS3 sold 700k units, down 10% y/y.

The Skew: I'm surprised Guitar Hero's ASP held up as well as it did. I've been thinking tough year to year comps would make keeping the franchise going difficult. Layering new functionality and drums and microphones into the product allowed them to keep pricing strong. Despite the franchise's age, 42% of Guitar Hero sales were band kits. Nintendo may be showing some signs of slowing – industry hardware units sales accelerated in December yet Nintendo grew only 5% m/m. Activision should work toward ~$10.30 near-term.

 

Intel (INTC) 4Q:2008 results

Intel reports in-line revenues, gross margins at the low end of their 55 +/- 2% forecast. We knew that, they preannounced twice.

Guidance is a little fuzzier than usual. They're planning for revenues around 7 billion. Putting that into typical Intel speak, that would be $6.7 - $7.3 billion, below the consensus of ~7.3 billion. Really this is a down 15% guide down, which is at the upper end of semiconductor industry guidance of down 15-25%. The lowest number I heard was $6.8 billion. The stand-out number from the guidance is gross margins in the low 40s. That's off the charts bad and frankly suspicious – one would think they could beat that number pretty easily. Clearly Intel is planning some price cuts.

Typically Intel bottoms out on large inventory charges which enable them to inflate gross margins as they sell previously written off inventory. I guess they're saving the charge for next quarter when they're planning for gross margins to bottom. Speaking of inventory, they're showing absolute inventory up 10% q/q on a revenue decline of roughly 20%. The lower gross margin guidance suggests they'll be throttling back manufacturing over the course of the quarter… and possibly implies a big inventory writedown. That's what they should do.

Intel's capex came in at $5.2 billion for the full year and guidance of flat to slightly down is better than the street's expectation of down 20%. Capital equipment companies exposed to Intel include: FORM, NVLS, ASML, AMAT, LRCX, COHU. There are probably more but those come to mind. That group is starved for positive news and Intel has one of the largest capital spending budgets in the world. Of course, my initial reaction to that number is it has to come down. They're spending as if revenues didn't just drop 30% for them in 6 months. Uh guys, it did.

Eye-balling this "full year spending" number they've never given before, 10.4 – 10.6 billion looks like its slightly higher than analyst models so there's some operating margin pressure.

Intel has the potential to be a reversal event – historically many bottoms and tops are set by Intel's quarterly report. That said, I would prefer to see the stock down off these numbers. We're getting close to $12 which is where I think picking away at a long position makes sense. Unfortunately, it doesn't look like we're going to get near-term capitulation in the name or the sector, which would make things a hell of a lot easier. It's never easy, is it.

No green lights from me here one way or the other. It's a lousy number. I still need to see a big inventory write-off and the stock at $12 to think about buying it.

Tech Roundup – January 15, 2009

 

PC Industry (various symbols)

From WSJ:

 

PC shipments world-wide dropped 0.4% in the fourth quarter from a year

earlier, the first year-to-year drop in six years, according to Interactive Data

Corp. Rival market-research firm Gartner offered a slightly more upbeat estimate

Wednesday, reporting that fourth-quarter shipments world-wide rose 1.1%.

PC sales in the U.S. were particularly weak, with shipments falling 3.5%,

according to IDC. Gartner said U.S. shipments dropped 10.1%.

The poor showing represents a "pretty dramatic drop" in the market compared

with the year-earlier quarter, when world-wide sales grew 15.5%, said IDC

analyst Loren Loverde. He said sales of netbooks -- laptop PCs that sell for

less than $500 -- were the market's single bright spot.

About five million of the small PCs were shipped in the fourth quarter, he

said, bringing the 2008 total to about 10 million. At the same time last year,

that market was "basically nonexistent," he added.

But netbooks are typically priced from $300 to $500 and carry low profit

margins, contributing to what Gartner described as a "record decline" in PC

revenue.

Hewlett-Packard Co. remained the world's largest PC maker, increasing

shipments 3.1% as its global market share grew to 19.6% from 19% last year,

according to IDC.

But rival Dell Inc., the No. 2 PC maker saw its world-wide shipments drop

6.3%, IDC said.

The Skew: This all makes sense, right? Desktops are getting slammed. It was only last quarter that notebooks replaced desktops as the majority of new sales. Netbooks in the $300-500 price point are the fastest growing segment and showing significant unit growth at the expense of higher priced notebooks. Underlying trends like this coupled with a severe contraction in consumer spending has led to one of the weakest quarters on record for the industry. Intel guided down twice – no one should be shocked here.

Within the numbers, Acer is picking up a lot of unit share due to their successful Aspire One line of netbooks, Dell is losing significant share. As Dell added retail channels late last year, this is particularly ominous. HP held share. Apple gained share.

It is extremely likely that the March quarter, without the significant holiday spending catalyst, will be catastrophically bad for the entire industry from a historical perspective. Business conditions continue to deteriorate worldwide and consumer spending is still extremely reserved. I continue to think Dell is a short. On a relative basis, HPQ will outperform because they aggressively manage costs and because printer ink is the fattest margin product in technology which subsidizes marginal hardware profitability elsewhere.

 

*

 

Xilinx (XLNX)

Reports 458 mil revs, beating 440 mil consensus. Gross margins exceeded street estimates by half a point. Guide for decline of 15-25% in revs – 344-390 mil versus consensus of 410 mil. Gross margins are expected to be 61-63% versus street at 63.3.

The Skew: Consumer and auto down 12% q/q, data processing down 20% q/q, communications down 3% q/q, industrial flat due to defense up 20% q/q. Numbers are going to $.70ish for 09. Chart looks broken and dangling. No near-term catalysts. Doesn't sound good, does it? Some will pick on Altera (ALTR), who reports on January 27^th. I suggest you ignore people telling you to trade Altera off Xilinx. It's a totally derivative idea. In fact, if there's a big move in Altera one way or the other I suggest you fade it because it's a bunch of cattle driving the move.

*

Apple Computer (AAPL)

Steve Jobs is taking a leave of absence until June.

The Skew: Steve says he recently became aware that his health problems are more complicated than he had thought. No comment about the quarter from the company but they've been pretty good about keeping Steve's health issues and the company's health issues separate. No edge here, I'm still not a doctor. I stumbled upon a commencement speech Steve gave at Stanford after being diagnosed with cancer. It won't make you any money but it might make you reflect a little.

 

*

More coming pre-opening. Check back.

Nortel exposed

A quick Nortel supplier table:

% of sales
to NT

AIRV 93.0%
BKHM 18.0%
AUDC 17.0%
AVNX 13.0%
AMCC 7.8%
TTIL 6.3%
RSYS 5.6%
FLEX 3.5%

Courtesy: Connexiti.

Clues from 5N Plus conference call

Last quarter, First Solar (FSLR) was 75% of sales of 14 million (Canadian). This quarter, First Solar (FSLR) was 78% of sales of 18 million. So sales to FSLR were up 3 mil, or roughly 22%. First Solar consensus estimate for this quarter is up 17.5% Q/Q. Backlog at 5N Plus was 54 mil, which is where it was last quarter, though they referred to it as "up 19 mil", which I don't understand. The notes I have say it was 54 mil last quarter too. Backlog for 12 months has the same mix as the rest of the business -- if backlog is flat q/q, it indicates the moving average of First Solar's orders isn't changing, which it should be as they've got hypergrowth and its expected to continue.

They see no need to add new capacity having completed an addition recently. They referred to having adequate capacity for "growth, if any", which struck me as a warning. As they didn't give specific guidance for next quarter, there's no silver platter here of information. Margins moderated somewhat. They're slowing capacity additions. They're seeing some softness and starting to wonder if there will be growth.

First Solar's revenue ramp for next year is modeled to be pretty strong (100% y/y growth in Q1, levels out at 40% y/y in Q4) which doesn't seem consistent with this suppliers' tone. 5N Plus is a very small player and a very very small percentage of FSLR's sales and not necessarily directly indicative of a trend at their largest customer. But it might be.

Optisolar cuts half of workforce

This is what I mean about non-functional credit markets and the potential to derail solar projects.

-- California solar company slashes half its staff --
LOS ANGELES, Jan 13 (Reuters) - Solar power company OptiSolar
Inc has cut nearly 300 jobs, or half its work force, and halted
construction of a manufacturing plant because it could not secure
the funding it needed to expand, a spokesman said on Tuesday.
The privately held company, which made headlines last year
when it secured a deal to build a 550-megawatt solar power plant
in Central California for utility PG&E Corp , said its
biggest contract would not be affected by the move because
construction was not scheduled to begin until 2010.
OptiSolar spokesman Alan Bernheimer said the company had
eliminated 185 positions at its Hayward, California headquarters
and an additional 105 jobs in Sacramento, the location of a
planned manufacturing plant.
"It's on hold until we are able to attract financial support,"
Bernheimer said, adding that the company was applying for loan
guarantees from the U.S. Department of Energy.
The financial crisis and tight credit markets have put
expansion of the high-flying renewable energy industry on hold in
recent weeks, though many companies are encouraged by
President-elect Barack Obama's pledge to double alternative energy
production in three years as part of a plan to stimulate economic
activity.
"We'd like to take the Obama administration at their word,"
Bernheimer said. "We're sort of the poster child for what they're
talking about -- renewable energy, green jobs and U.S.-based
high-tech manufacturing."
Once it is completed, the Sacramento plant will be capable of
producing more than 600 MW of OptiSolar's thin film solar panels a
year and is expected to employ about 1,000 people, Bernheimer
said.
Until that plant is able to begin production, OptiSolar is
relying on a smaller manufacturing facility in Hayward to fulfill
its current commitments to customers.
OptiSolar so-called thin film solar panels are made from
amorphous silicon and are cheaper to produce than traditional
crystalline silicon panels.
Its deal with PG&E, announced in August, comes as California
utilities are scrambling to meet the state's mandated goal of
producing 20 percent of their electricity from renewable sources
by 2010.
(Reporting by Nichola Groom; Editing by Phil Berlowitz)
((nichola.groom@thomsonreuters.com; +1-213-955-6755; Reuters
Messaging: nichola.groom.reuters.com@reuters.net))
Keywords: OPTISOLAR JOBS/