Technology StocksIt's Different This Time!

Previous 10 Next 10 
From: etchmeister6/18/2010 12:07:48 AM
   of 549
Perhaps one should figure out what TAM is for TSV going forward - that would be a meaningful contribution IMHO....

Share RecommendKeepReplyMark as Last ReadRead Replies (1)

To: Sam Citron who wrote (15)6/18/2010 12:34:33 AM
From: Cary Salsberg
   of 549
Hi Sam,

Thanks for posting. I am not troubled by the fact that you know me well. I am here to test my ideas. I can think of no more astute examiner. I will cut through the tech talk and get to the bottom line whenever possible.

I have followed SNDK for a number of years. I have a serious opinion. It is selling a commodity semiconductor which requires very high capital expenditures for its fabrication. There will be times, like now, when demand outstrips supply, prices are high, and they have a high EPS. I am not certain if the EPS is translating into high free cash flow, but you can check that.

I will correct your description of my stodginess. 3 of my 8, AMAT, KLAC, and NVLS go back to 1996. 4, ALTR, XLNX, LLTC, MXIM, go back to 1998. 1, ASML, goes back to 2001.

I have a question for you. If I am selling picks and shovels to miners or I am investing in the pick and shovel company, should I be able to tell which miners will strike it rich?

My wife has 3 best friends. One, friends for about 23 years, is married to an AAPL employee and I watched their stock go from 12 to 540 (split price 6 to 270). They struck it rich. I never wanted to buy AAPL and I still don't.

Part of the purpose of this thread will be to give my reasons why those 8 are "really the most interesting companies in the market for making money going forward."

A key issue going forward is your "I think will outperform, and only in their outperformance phase." We will need to quantify "think" and "outperform" to arrive at an expectation of performance." I like "think" to have a high probability and "outperform" a healthy premium to to market's performance. You might want to present the factors you believe support a high probability to outperform.

Share RecommendKeepReplyMark as Last ReadRead Replies (2)

From: KMcKlendin6/18/2010 12:41:59 AM
   of 549
With Regards to AMAT, it *is* different this time. They now have multiple large businesses all characterized by thin film nanotechnology.

1. Semi equip. Still the largest segment and the best margins
2. LCD equip. Extremely cyclical and currently on the upswing
3. Solar equip. Thin film (Sunfab) has been a flop, but crystalline silicon is booming. AMAT is the #1 solar equip supplier. Impressive for a company that entered the business 4 years ago.
4. LED equip. AMAT's LED tool has been impending for some time now.
5. Service, automation, software, abatement: For all the above

With the recent severe downturn, the relative proportions of the above businesses fluctuated substantially. And the diversification helped cushion the downturn. Now AMAT is benefiting proportionally less from the upturn, although AMAT's semi equip business is apparently robust.

IMHO AMAT now has a less cyclical business model and will likely outgrow a maturing semi industry.

Share RecommendKeepReplyMark as Last ReadRead Replies (1)

To: etchmeister who wrote (16)6/18/2010 12:42:00 AM
From: Cary Salsberg
   of 549
Hi Etch,

Total available market (TAM) for through silicon via (TSV).

The market is small now and growing slowly according to Rick Hill at a recent technology conference.

I think much of the future market is not known yet.

Share RecommendKeepReplyMark as Last Read

To: KMcKlendin who wrote (18)6/18/2010 1:51:23 AM
From: Cary Salsberg
   of 549

Thanks for posting.

I am interested in your statement, "a maturing semi industry." In a previous post, I attempted to characterize the aspects of the semi industry that were maturing and those that were not.

Please define what you mean by "a maturing semi industry."


Share RecommendKeepReplyMark as Last ReadRead Replies (1)

To: Cary Salsberg who wrote (20)6/18/2010 4:26:54 AM
From: KMcKlendin
   of 549
It's a good point. The "maturation" of the semi industry is a widely held assumption. And it is easy not to think critically about assumptions. In fact when I've been wrong about an investment, it has often been because my assumptions were wrong.

So I'll play the devil's advocate and argue for maturation:

1. The growth rate of semiconductor revenues appears to be declining with time. Even though new high growth technologies are always emerging, the percentage impact is less because of the size of the industry. Using the SIA three month moving average worldwide billings (1) I get an annual growth rate for the 1990s of 14.28% and the 2000s of 4.22%. Since 2000 was the top of the tech bubble I recalculated for 1985-1995 (19.81%) and 1995-2005 (7.60%). The numbers are very noisy but the overall pattern of a decreased annual growth rate appears to hold.

2. There has been industry consolidation with fewer and larger players among the top semiconductor producers (Fab owners) and the top semi equip companies. This is characteristic of a maturing industry.

3. Fundamental metrics of valuation (P/E and P/S ratios) have declined and most players now offer a significant dividend. This is also characteristic of a mature industry.

4. Semi equip capital intensity has also declined, as you noted.

I personally believe that some maturation of the industry, as defined by declining total revenue growth rates, has indeed occurred. However, the bursting of two large bubbles in the 2000s may lead the market to overestimate the maturation in the industry, and undervalue the companies. I think this may be your underlying point. And certainly any subset of the semi industry, such as your PLD companies, may show growth significantly above the overall trend line.


Share RecommendKeepReplyMark as Last ReadRead Replies (1)

To: KMcKlendin who wrote (21)6/18/2010 1:06:10 PM
From: Cary Salsberg
   of 549
A mature industry is one that has penetrated its available market to the point where industry growth rates slow to the rate of growth of the available market and the available market growth is similar to macroeconomic growth rates such as the GDP or population growth.

The semiconductor available market keeps expanding, so the industry is not mature. In this context, I will examine your points, which make it seem that the industry has matured.

Revenue is the product of unit volume and price. The semiconductor industry has posted continuously declining prices. This makes price elasticity a very important factor in the semiconductor industry, much more important than for other industries.

Falling semiconductor prices are normally attributable to the "shrinks" at each node along the Moore's Law road map. In the last decade, the move to 12 inch wafers (300mm) also contributed significantly to the decline in unit prices. This combination lowered prices at a faster than normal rate for the industry.

The theory is that lower prices will lead to sufficiently higher volumes and revenue growth will be achieved. Revenue growth was achieved this past decade, but it slowed. There is always a lag between lowering prices and getting the higher volumes due to the lower prices. Semiconductors are not consumed directly. Lower prices must spur innovation and the products that result must spur unit volume growth. That will, by its nature, produce lumpy lags.

If prices fall at a faster rate than normal, volumes lag more and revenue growth rates slow. It is entirely possible that a combination of falling prices returning to a more normal rate and innovation, necessary for higher unit volumes, continuing to be stimulated by the lower prices and playing catch up, will produce a rising industry growth rate. Anecdotal evidence points to more innovation in more markets.

There is consolidation in the industry. I believe it is the result of the maturing of the Moore's Law road map, not declining growth rates in the industry's markets. As the industry moves along the road map, the product improves dramatically, but production becomes very much more difficult and expensive. The process weeds out the weaker players and rewards the stronger. The industry has fewer players, but they are the winners. The combination of growing markets and industry winners should be very attractive to investors.

I discussed reasons for declining industry capital intensity at length. I will add my belief that it was not caused by the industry maturing and it will stop declining and possibly reverse to modestly higher levels. It will not approach historical peak rates.

During this past decade, P/E and P/S have been buffeted by the DOT COM boom and bust, the options expense, the back dating option fiasco, and, most recently, the "Great Recession." Add to this is the irony that everyone knows the industry is cyclical and hardly anyone has taken to the trouble to move beyond the numbers and really examine the dynamics of recent cycles.

The dividend is a big plus. Many of these companies have business models which support high levels of R & D, yet produced plenty of free cash flow which is returned to investors as dividends and stock buy backs.

Share RecommendKeepReplyMark as Last ReadRead Replies (1)

To: Cary Salsberg who wrote (17)6/18/2010 2:01:42 PM
From: Sam Citron
1 Recommendation   of 549

NAND flash may be a commodity in the broad homogeneity sense of the term where I can comfortably plug any brand of flash memory into my camera, whether it be Kingston, Transcend or SanDisk, and know that it will work and the performance will be quite similar. But it is no standard commodity like durum wheat where there are so many producers that near perfect competition reigns to the extent that no individual producer can have a meaningful impact on prices. In fact, due to the high barrier to entry imposed by the cost of a fab ($5B or so), there are very few producers and Samsung, Toshiba and SanDisk account for over 80% of the entire market. So if NAND is a commodity, let's recognize it as an oligopolistic commodity where the behavior of individual firms matters (e.g., controlling capacity utilization to balance supply with available demand) and the likelihood of collusive behavior to maximize profits is quite high. As a speculator, I tend to like such markets for the regularity of their cyclical behavior.

Unlike its primary competition who are large Asian based conglomerates, SanDisk is practically a pure play on flash. It thus has a commitment and a focus that is extraordinary because it lives or dies mainly based on this one product category -- like if you had all your retirement assets in just one stock instead of a portfolio of assets.

As one of the lowest cost producers of NAND, SNDK seems to be in an attractive market position, especially as demand rises as flash seems to be gaining market share against other types of memory such as hard disk drives. Notice how SNDK has been outperforming STX and WDC since the iPad has been introduced about 3 months ago. I would expect such outperformance to continue as flash-enabled tablets steal market share from HDD-enabled netbooks.

Leveraged free cash flow is $566M and profit margins are in the 20-25% range.

It would be interesting to hear why you have never wanted to buy Apple given your interest in innovation and technology.

As for the factors that support a high probability of outperformance, in my view, in the short-term nothing compares to recent past performance -- in other words, Newton's first law of motion -- "a stock in motion..." Citron's second law of stock motion is that stock outperformance or underperformance tends to continue until it goes to an extreme, at which time it begins to revert toward the mean, but it rarely stops there.


Share RecommendKeepReplyMark as Last ReadRead Replies (2)

To: Cary Salsberg who wrote (17)6/18/2010 2:28:33 PM
From: Sam
   of 549
re: Sandisk and "It's different this time" for NAND

Free cash flow is easy enough to look up:

Cash Flow pS 4.72 $
Free Cash Flow pS 3.73 $
It will, IMHO, get better as the year goes on.

As why it is different, I suggest you read this piece on Sandisk:

Here is an excerpt that quotes Eli Harari, the Chair of Sandisk, during the Q&A at their February Investors' Day Conference and a graph from the presentation that Harari gave. While reading the excerpt, recall that currently, leading edge NAND vendors are mostly at 3x, are only now beginning to ramp at 2x, and no one knows yet how to ramp at 1x.

Here is Eli’s math from Investor Day:

“Q: OK very good. And how many new fabs would need to be built then to reach your estimated demand in the out years [2013]? At one point a few years back you were talking about 6 new fabs.

Eli: OK, and don’t hold me to it. To do that math.

Lets say that 1x generation product is between lets say 2.5 and 3 TB per wafer, approximately that- lets say 3 TB per wafer. You have to be really really good to get 3 TB per wafer, and probably use 2 to 3 bits per cell.

That means 30,000 PB in 2013, an incremental 30,000 PB would require another 10 million wafers which is 5 mega-fabs, assuming 1x @ pretty good yields @ 3 bits per cell.

This is the point I was making earlier- that the projected growth in demand requires very substantial new capacity.

The existing capacity is already generating 10,000 PBs- 11,000. So lets say it can go to 30,000 PB. If you want to build another 30,000 you need to have another 5 or so fabs- mega-fabs.

Q: On an average of $8B per fab?

Eli: Roughly.

Q: Roughly, plus maintenance capex. That’s a big investment- a big risk.

Eli: Yes. Somebody told me that a nuclear aircraft carrier cost about $5B. That gives you an idea how big $8 billion is, but as I said, it’s spread over a three year period- investment wise. In any kind of new capacity it would be very likely with Toshiba so our number would be less than that number.

Q: I hope so.

Eli: Me too. [laughter]”
So far, Toshiba/Sandisk and Samsung have announced new fabs to be built over the next 12 months, and gradually brought up. Micron/Intel (IMFT) has made some noises, but as far as I know, hasn't definitively announced plans for a new fab--we will likely find out more at MU's upcoming CC sometime in the next 2 weeks or so. Hynix has announced an investment total of around $3.5b, but hasn't broken that out as to what will be NAND and what will be DRAM. They are stronger and a more competitive number 2 in DRAM, and it would make sense for them to defend that position rather than try to go up against 3 much stronger and better capitalized competitors in NAND. The other players? There are no others to speak of--Numonyx and Powerchip have a grand total of about 3% of the NAND market (see Message 26550757 ). I think I can guarantee that there will not be any new players entering the market. Why? First, because of the complexity and cost of NAND manufacturing, and the long lead times of building a fab and bringing up production, even for companies that are experienced in it. Second, because many of the production processes that are currently used are patent protected. Third, because no one currently projects that manufacturing NAND will be possible after the 1x node. Instead, people have been working on post-NAND solutions. That is a story for another post, but I'll just say here that none of these post-NAND solutions have been commercialized yet (or are even close), and that Sandisk has its own post-NAND 3D process that it hopes to bring on line sometime in the 2013-15 time frame.

Share RecommendKeepReplyMark as Last ReadRead Replies (1)

To: Sam Citron who wrote (23)6/18/2010 2:36:20 PM
From: Sam
   of 549
Sam, I hadn't read your post before penning mine, but they seem to complement each other pretty well (although our fcf numbers differ, but nevermind that--the exact number isn't as important for this thread as the trend and the fact that it strong).

Share RecommendKeepReplyMark as Last Read
Previous 10 Next 10 

Copyright © 1995-2018 Knight Sac Media. All rights reserved.Stock quotes are delayed at least 15 minutes - See Terms of Use.