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.

Strategies & Market Trends : Nasdaq 100 Analysis -- Ignore unavailable to you. Want to Upgrade?

To: Robert Scott who wrote (11)	12/2/2001 8:55:02 AM
From: Steve Lee	Read Replies (3) \| Respond to of 238

I thought your recent newsletter discussing semiconductor valuation was a very well thought out piece of analysis. Two important points you made were: 1) P/S ratios are at or above historical highs 2) Semi sales are 4% up from the '98 downturn but SOX is 84% higher I think the competitive situation needs to be addressed more. Semi sales may be slightly higher but how does that relate to the number of companies competing for that total revs figure? Take mobile phones for example. Not long ago you bought either a Motorola, Ericcson or Nokia. Now you can also choose from Siemens, Samsung, Panasonic, Sony etc. All big names looking for a piece of a pie that is not growing as fast as it used to when there were only 3 players. Same goes for PC manufacturers. The number of top tier names such as Dell, HP etc has probably not changed much in recent years, but there are plenty of quality vendors that are well known enough in their own regions such that buyers don't have to resort to beige-box-builders. The result of competition is of course lower cost and lower margins. That makes for lower profits. Now semi companies, especially large ones, cannot grow as fast as they did in 98 when the PC and mobile phone boom was in full swing due to the dual effects of less growth and lower margins. You mentioned that PCs and servers are renewed on a 3 year cycle. I wouldn't argue with that as a general statement. Some customers renew on a four year cycle and even a bit longer for servers. Some renew more often. The net effect is that the sales come spread out over any 3 year period. This spreading out was not the case in 1999. At that time companies were panic buying PC and server equipment, especially server equipment. There were three factors at play causing all this technology refresh demand: 1) The intro of the Euro currency. 2) The Y2k panic 3) The dot com panic I use the word "panic" because fear was driving IT buyers who were scared of a disaster through having unworking systems, or systems that couldn't cope with the new demand for internet services. Often, new systems were run in parallel with old systems in advance of a switchover. I own a business providing consultants to the IT industry, specialising in installation and refresh of datacentres. Much of the demand in late 98 and 99 was coming from board level decisions, rather than from technical staff that had identified a real need. Much of the buying was misguided in my opinion, but company officers were not spending their own personal money and were more scared of "missing the boat". Any one of those three factors in a compressed upgrade cycle was enough to create a demand boom. We got all three factors at once. Then the semi companies decided they needed to build new fabs. Remember the Pentium shortage when Coppermine was introduced? Now we have a semi industry that has the capacity to produce 3 years worth of demand in 2 years or less. That is why we have low fab utilisation. Some new fabs were put on hold, but are suspended projects, ready to be restarted if demand returns - that puts a floor on any high prices due to future shortages. I guess I could summarise this post by saying that the current downturn is due to overcapacity, rather than lack of demand. Of course, demand is lacking but that is because noone needs to buy additional capacity. Demand will further worsen because the lack of buying reduces profits that flow through the economy, stifling further demand. Thus economic stimulus will not work as well as it would in a demand led recession. As demand increases, the overcapacity will result in no shortages, keeping competition high and profits low. Finally, we need to understand that Intel is a special case company, as is Microsoft. These two companies have monopolies and as such are partially protected from having to drop prices during times of low demand. Sure, there is some price elasticity with their products, but not much due to lack of alternatives. Intel has competitors but has been through a process of change recently, which has diverted from their chip producing capabilities. The changes were the intro of the P4, with its larger die, the intro of the Itanium, and the intro and proliferation of the Rambus and DDR memory that accompany the new P4. These changes have now had many early problems ironed out, and can start showing results for Intel. I have been saying for about 6 months that Intel's fortunes may turn upwards and give a false hint of a recovery. That would be a nice shorting opportunity. Of course, I could go into the effects on capacity of 300mm and reducing design rule geometries, but this post is already long enough. I will also not go into the failure of broadband proliferation to reestablish performance, rather than bandwidth as the bottleneck in total system design.