Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Semi-Equips - Buy when BLOOD is running in the streets! -- Ignore unavailable to you. Want to Upgrade?

To: Cary Salsberg who wrote (7992)	7/17/1999 8:37:00 PM
From: Investor2	Respond to of 10921

Excellent analysis. You may want to consider the following two thoughts. 1. Your analysis seems to assume that the current up-cycle will extend 5 years from the prior peak. If the current up-cycle peaks before 2001, your analysis may be too optimistic. If the cycle peaks after 2001, your projections could be too low. 2. The market P/E has gone up significantly since 1996. If the market P/E stays this high until 2001, your assumed P/E of 20 is probably too low. Of course, if the market P/E drops to a point lower than it was in 1996, your assumed P/E of 20 may be too high. Best wishes, I2

To: Cary Salsberg who wrote (7992)	7/17/1999 9:18:00 PM
From: Gottfried	Respond to of 10921

Cary, thanks for doing all that work. To succeed with this analysis requires you to guess correctly several times. And, you have to rely on unreliable earnings estimates more than a year out. I recently expanded an AMAT price vs SEMI orders chart back to 1995. [I don't have SEMI orders data prior to that]. It shows that AMAT price peaks a few months before orders do. [Orders are plotted on the date they are ANNOUNCED on the SEMI web site]. With this method you only have to guess ahead of time WHEN orders will peak. Then sell a couple of months before that. I don't know how to guess that, unless I assume the cycle LENGTH will repeat. I'm throwing this in, in case someone has a smarter idea than mine. geocities.com Gottfried

To: Cary Salsberg who wrote (7992)	7/17/1999 9:58:00 PM
From: Ian@SI	Read Replies (2) \| Respond to of 10921

Cary, A very interesting piece of work. I agree with Gottfried. By the time we have empirical evidence that earnings have peaked, we'll be 5 months beyond peak prices. And in this sector that could be devastating. Similarly, a lot of money could be taken off the table if we wait until we've confirmed a peak in orders. And orders sometimes zig zag along the way up, so that's a difficult indicator to use to predict what's going to happen. Let's not forget that it takes at least 3 months for the final number to get published going through a cycle of Preliminary then Revised then Final in successive months. I've never religiously studied how the revisions affect the original number, so it could be 4 or 5 months before one really knows that a peak in orders was put in place. It's always so much more obvious after the cycle has completed and we have the advantage of being able to look at the graphs to know precisely when we should have sold to have maximized profits. Same applies to using PSR ratios. You never know when the Sales growth is going to accelerate or regress. My thinking hasn't progressed much from a couple years ago when this thread first started talking about this subject. I think some symmetry might be in order. You bought in thirds. Perhaps, a "dollar cost averaging" selling plan is in order for these cyclicals. Scenario 1. Up leg will last at least another 4-6 years; and will include migration to 0.18µ, Cu, 0.13µ, 300mm. Action: Reserve 1/3 of holdings as a core portfolio to be held as long as one has a positive long term outlook for the sector. Divide remaining 2/3 into 5 equal pieces and sell 1 piece (about 12% of the sector portfolio) each year. If an aggressive investor, use the proceeds to upgrade portfolio quality on sector pullbacks of more than 25%. Scenario 2. Up leg will take a 1 1/2 to 3 year pause after 0.18µ and Cu; before proceeding down the path to 300mm. Sell 2/3 of the portfolio between now and yearend in 5 equal pieces 1 month apart; or take the money and run now. ++++++++++++++++ Personally, I think that the chip demand growth resulting from the Communications revolution, Internet, etc makes Scenario 1 more likely. But there are those that think that there will be a pause before 300mm because of 300mm. i.e. No one wants to be known for building the last 200mm fab. And Siemens-Mot (Infineon) has already built the first 300mm one. [Begin Blatant MTSN plug] I think this is a crock. MTSN has already sold CVD 200mm/300mm bridge tool (to Samsung I believe) in quantity. Fab managers can and will buy bridge tools to handle their technology and capacity requirements prior to 300mm thus reducing their costs of migrating to 300mm. Several vendors announced bridge tool intents at Semicon West. MTSN is the only equipment maker to sell a bunch of them. [end blatant MTSN plug] :-) I accept that regardless of which strategy I deploy, money will be left on the table. I think that one has to decide what would be an acceptable after tax profit; take it, and hopefully still have some money in play to take advantage of further upside; and sufficient cash to take advantage of the next time the Street thinks no equipment maker will ever again sell another tool. We're always going to have less information than we need to make an optimal buy / sell decision. but isn't that what makes a market? As usual, I welcome your insights; and rebuttal. ;-) Best regards, Ian.

To: Cary Salsberg who wrote (7992)	7/18/1999 1:28:00 PM
From: semi_infinite	Read Replies (2) \| Respond to of 10921

Cary, Thanks were taking on this difficult subject. Looking at the combined chart over the last 5 years: techstocks.com The stock prices peaked about a year before the earnings peak, and a few weeks after the Semicon West show in 1995. Therefore, I would think that the sell-analyses would have to be based on forward looking PE and other factors such as chip demand and prices (and this may be reflected in the bookings and BTB). I recall that Soundview's analyst, Rick Whittington (sp?), turned negative at about the right time. Many people disagreed. I also remember Merill Lynch increasing Micron's price target to about $200 when it hit $90 back in 1995. Now, there is an idea, we will just use ML as contrarian indicator :-). Back on topic -- where can one get detailed historic data on earnings expectations?