Wafer area relevance?, Chapter 2

September 6, 2006

View from Silicon Valley- "Wafer area" relevance, Chapter 2

In response to our August 19 missive, "Does wafer growth impact sales?" (http://www.viewfromsiliconvalley.com/id248.html) a reader who studies such things for a living asked our further opinion about the impact of solar power wafer demand and 200mm-to-300mm fab conversions on our findings. We addressed the fab conversion aspect in some detail yesterday (September 5) with, "'Wafer area' relevance?, Chapter 1" (http://www.viewfromsiliconvalley.com/id259.html).

While we were busy researching the solar power issue, the same reader explained he shared our August 19 missive with a Paul McWilliams.

Mr. McWilliams agreed to let his comments be published, "If he wants to quote me it's ok in full context and with full credit (including the fact that I caught his error and that I'm editor of nextinning.com). Beyond that I'll view it as a violation of my copyright." You will quickly notice he feels it's reasonable to assert copyright over his critique of our copyrighted material. As I understand the law, when material is received unsolicited, he is no position to demand anything.

Even so, in the spirit of granting him his full voice and, frankly, hoping to never have to deal with anybody else's lawyers over the content of our site, we decided to publish in full his treatise criticizing ours (including all 573 words and 13 paragraphs of copyright and legal-disclaimer verbiage).

Correcting the referenced error invalidated a couple of our intermediate points but, contrary to Mr. McWilliams' sarcastic insults below, did nothing to change our basic conclusion. To wit,

"SEMI's y-o-y wafer area growth in 2Q06 was +22.4% while SIA's y-o-y IC dollars sales increased only +9.4%. This is a warning sign for semiconductor profits."

As long-time readers have come to expect, this is not the end of the story. We will venture a response to Mr. McWilliams' "insights" in a future chapter.

Meanwhile, here you go:

* * * * * *
>> Member Q&As - Evaluating Semi Data & Sustaining the Current Rally

Q) Here is an interesting article that I'd like to get your take on in your newsletter.

A) Thank you for alerting me to yet another pundit who can't seem to get it right. The contention of the article you linked from the service, "View from Silicon Valley," is that the growth rate in wafer shipments is significantly outpacing the growth in semiconductor shipments. What makes this particularly interesting is that the author is suggesting what we are seeing today mirrors what we saw in 2001. Based on this flawed thinking, the author makes the following conclusion:

"Conclusion: Clearly, wafer area grwoth [growth] does NOT drive semiconductor sales growth. (We think the impact on profits will soon become clear.) If asked to bet which way semiconductor stocks were likely to go from here, we definitely would have an opinion."

The first step anyone should take when evaluating the opinions of others is to review the data upon which the opinion is based. This is why I almost always "show my work" and provide members with detailed data I'm using to make my case, particularly in our quarterly State of Tech reports.

In this case, the author is using data from SEMI (Semiconductor Manufacturing Industry Association) and contrasting that with data provided by the SIA (Semiconductor Industry Association), which gets its data from the WSTS (World Semiconductor Trade Statistics). SEMI specializes in tracking data for fabrication equipment and also publishes quarterly data for millions of square inches of wafer area shipped (semiconductors start life as wafers). The WSTS collects monthly revenue and unit volume data from member semiconductor companies and then augments this collected data with estimates for non-member companies. The SIA then compiles and distributes this data.

In the first table used by the author of the aforementioned article, he lists the data for his bearish conclusion. In this table he makes the case that wafer area shipments increased in 2001 while sales fell. This, he says, is why profits fell so dramatically in 2001. However, when we look at the numbers, we can very quickly see the author has made a serious mathematical error and then used the error as the cornerstone of his thesis. Rather than show that wafer area shipments declined in 2001 by 15.4%, he shows that they grew by 15.5%. Maybe if someone pointed out this flaw in his data he would respond with the famous Emily Litella line, "never mind." However, there was no one there to stop this misguided train of logic, so like Emily, he careened aimlessly towards his flawed conclusion.

Probably even more ironic than the author's math error is the idea that wafer area and semiconductor revenues should be evaluated in a vacuum. As I've frequently documented, the best proxy for semiconductor demand is unit volume, not revenue. This topic, along with the fact that we need to focus on IC statistics versus all semiconductor data, that I covered in great detail in my 2005 report covering semiconductor cycles.

Had the author gotten his math right and known to focus on only IC unit volumes, which exclude discrete semiconductors, opto and sensors, his article probably would have not seen the light of day. Why? Well, IC unit volume changes have tracked very closely with wafer area shipments during the entire timeframe (2000 to 2006) covered by the author's study. As a matter of fact, IC unit volume shipments increased by 25.3% and that compares so closely with the reported unit volume increase of 25.4% that it is clearly a non-factor. As you can see in the following chart that compares SEMI wafer area data to SIA IC unit volume, this correlation is pretty darn consistent.

Please note that this chart uses two "y" axes and, therefore, two scales. Due to this and several other variables, we should not be terribly concerned about modest changes in the separation of the lines. In addition to the normal perturbations these variables cause, we also have to consider the fact that solar cell production, which uses silicon wafers just like ICs, has gone up considerably during the last 24 months. As a matter of fact, I've expressed concerns this year about solar cells absorbing so many wafers that it might impact availability and raise prices for semiconductor manufacturers. So far, this has not happened, but we're still keeping an eye peeled for any evidence of shortages or measurably higher prices.

While it's interesting to contrast the data from SEMI, I think it's probably more meaningful to run the same study with data from SICAS (Semiconductor International Capacity and Statistics). Unfortunately, SICAS data lags even that of SEMI. However, the data is considerably more detailed. Not only does SICAS provide actual wafer start data, but it also includes capacity and utilization data. And, in addition to providing these two additional views, SICAS also breaks down the data into details such as processing technology and wafer size, but for our current purposes, we'll stick with the aggregate data.

In the following chart we've included both wafer start and capacity data. Both data sets are based on the peak week for the quarter and are, therefore, not directly comparable with SEMI data. Be that as it may, even though absolute data points for SEMI and SICAS do not match, there is still a strong relative correlation. In other words, both short- and long-term trends between the data sets match very well.

Beyond again seeing a very strong correlation between the trends of IC unit volumes and wafer areas shipped, we can also now see what really happens during semiconductor down-cycles. During these down-cycles, there is generally a very noticeable separation between both wafer starts and IC unit volumes, which tend to track each other during both up and down-cycles, and available capacity. As common sense would suggest, when there is significant excess capacity and utilization drops below 85%, there is a strong likelihood we are having or that there will be at least a modest downturn in the semiconductor cycle. As you can see from the above chart, we are currently at a stage of high relative capacity utilization.

To put a cap on the validity of the long term correlation between wafer starts, wafer area and unit volumes, we'll look at two more charts. The first compares the longer-term study we can do using just SICAS data to SIA unit volume data. The second compares the shorter-term data we have for SICAS, SEMI and the SIA.

As you can see from these charts, the correlation between wafer starts, wafer area, and IC unit volumes correlates very well during both up and down cycles. Therefore, it is my opinion that using the lagging data on actual starts or wafer area shipments is not a meaningful method to predict what may happen in the future. However, as I noted above, keeping an eye on capacity utilization, which we publish in each of our quarterly State of Tech reports, is an important factor, albeit the importance is minimized by the fact that the SICAS data lags SIA data by over a quarter.

Bottom Line: While none of this data can be realistically evaluated in a vacuum and there are usually many exceptions and caveats to consider, I believe we are currently at a reasonably healthy aggregate balance of capacity (supply) and demand. There are many things that could affect this balance both negatively and positively and for the most part, these things are pretty hard to predict. However, as it stands today, I think the odds favor the continuation of a healthy semiconductor market and, if anything, a capacity shortage in DRAMs that will likely lead to more favorable pricing and profitably for companies such as Micron (MU) and Qimonda (QI), formerly the memory division of Infineon (IFX), during the next six to nine months.

Disclosure: At the time of this publication, out of the stocks discussed herein, Paul McWilliams held no positions.

Posted by Paul McWilliams - 8/18/06 8:55 AM EST
-------------------------------

Indie Research
254 Witherspoon
Princeton, NJ 08540

Except for making one printed copy of this report or any other materials, files or documents available from, accessible through or published by Indie Research LLC. for your personal use (or downloading for the same limited purpose), none of these said materials, files and/or documents may be reproduced, republished, rebroadcast or otherwise re-distributed without the prior written consent of Indie Research LLC.

LEGAL TERMS/DISCLAIMER:

Indie Research, LLC, Vizue, LLC and Paul McWilliams are not registered investment advisors, broker/dealers, or research analysts/organizations.

Readers are advised that the commentary and reports on the nextinning.com website and in its newsletters are issued solely for information purposes and should not to be construed as an offer to sell or the solicitation of an offer to buy any security. The opinions and analyses included herein are based from sources believed to be reliable and written in good faith, but no representation or warranty, expressed or implied is made as to their accuracy, completeness or correctness.

YOU SHOULD VERIFY ALL CLAIMS AND DO YOUR OWN RESEARCH BEFORE INVESTING IN ANY SECURITIES MENTIONED IN THE NEXTINNING NEWSLETTER OR WEBSITE. INVESTING IN SECURITIES IS SPECULATIVE AND CARRIES A HIGH DEGREE OF RISK. YOU MAY LOSE PART OR ALL OF YOUR PRINCIPAL INVESTMENT.

Readers are urged to consult with their own independent financial advisors with respect to any investment. All information contained in this report should be independently verified with the companies mentioned. Neither Indie Research, LLC, nor its officers, directors, partners, contributors or employees/consultants, or Vizue, LLC nor its officers, directors, partners, contributors or employees/consultants accept any liability whatsoever for any direct or consequential loss arising from any use of information on this website or any use of information in its newsletters.

Indie Research, LLC, and its officers, directors, partners, contributors or employees/consultants and Vizue, LLC and Paul McWilliams may on occasion hold positions in the securities mentioned in the nextinning.com website and in its newsletters. These positions may change at anytime. Please note that Indie Research, Vizue or Paul McWilliams do not receive compensation of any kind for mentioning a company on the nextinning.com website and in its newsletters.

Our newsletter is limited to the dissemination of general information pertaining to investments. Accordingly, the newsletter should not be construed by any existing or prospective subscriber as the rendering of personalized investment advice for compensation. We are neither attorneys nor accountants and no portion of the web site content should be interpreted as legal, accounting or tax advice.

Past performance may not be indicative of future results. Therefore, no prospective or existing subscriber should assume that future performance of any specific investment or investment strategy, including the investments or investment strategies made reference to directly or indirectly in our current or previous newsletters, will be profitable or equal any historical performance levels.

Please remember that different types of investments involve varying degrees of risk, and there can be no assurance that any specific investment will be suitable for a prospective or existing subscriber's investment portfolio. Historical performance results for investment indices or categories generally do not reflect the deduction of transaction or custodial charges, the deduction of an investment management fee, nor the impact of taxes, the incurrence of which would have the effect of decreasing historical performance results.

USE OF THIS SITE SIGNIFIES YOUR AGREEMENT TO OUR TERMS OF USE AGREEMENT

* * * * *

We hope you enjoyed this piece written by one of our critics.

* * * * * *

Opinions expressed are not necessarily those of View from Silicon Valley. The opinions are subject to change, are not guaranteed and should not be considered recommendations to buy or sell any security.

The above is not intended as advice to buy, sell or hold any stock, bond, real estate nor any other financial product or service. Buy and sell at your own risk (just like we do.)
* * * * *