Today's Buzzword: Overheated

The most widely used term to describe the current semiconductor cycle is "overheated." The cycle wasn't described this way at the beginning of the year. At that time, there was the "danger of overheating." Turns out, those warnings were right on target.

During past upcycles we have seen valuations of semiconductor and semiconductor equipment companies drop below the average value of the market. The measurement I am talking about is calculated by looking at the forward price earnings (PE) ratio of stocks in the chip sector and comparing it to the forward PE ratio of the S&P 500. It's rather simple really — companies do not make much money during downturns, so PE ratios soar. During upturns, when companies are bringing earnings to the bottom line, PE ratios contract. This is the nature of cyclical stocks.

With this volatility comes the impact of anticipatory pricing. It is unlikely that we will see chip and chip equipment stocks trade at a level equal to the market PE multiple when we are at the top of the cycle because, at the top, earnings have nowhere to go but down. The inverse happens at the bottom of the cycle.

Looking at the recent history, I see where stocks in the capital equipment sector traded below the value of the market during the latter months of 1998 and several times during the summer months of 1999. This is what we are facing today. Visibility in the sector is very good and the earnings outlook is positive, but the potential for a downturn in 2005 is weighing heavily on valuations.

As you know, Wall Street always wants to be positioned far ahead of these cycles. In fact, the rally in the stocks, from October 2002 to the early months of this year, pretty much priced in the current recovery. Share prices have weakened because, just like in October 2002, investors want to get out far ahead of the turn in the cycle. Yes, external elements have influenced the direction of share prices, but the lion's share of the selling has been caused by statements saying, "The end is near."

In the near term, all is not lost. There were several rallies during the summer months of 1999, and valuations actually grew until the first quarter of 2000. Personally, I believe we are going to see another rally phase after this summer lull passes. External factors could come into play, but it is clear that we are beginning a run on the 300 mm order and shipment front. I believe we are set for a rally.

Longer term, capacity additions will be much more incremental. The efficiencies of 300 mm will likely put pressure on device average selling prices, and there are a number of technology hurdles to overcome.

It is clear that the industry is getting much more efficient at adding capacity. I mean capacity in a very broad sense — 200 mm and everything above 0.18 µm are included in this judgment. This is where many of the new, Southeast Asian players are focused. The goal is to commoditize, proliferate and consumerize well-understood manufacturing techniques. I've seen some numbers that suggest the industry is about 50% better at adding manufacturing capacity than it was during 1998-2000 cycle.

What will happen to the industry in the long run? I think we are finally going to see the Great Sort. The sort will not get in full swing until the downturn starts to come into view. It will drive companies in several directions. Some will band together. Many will diversify into other industries. Some will get completely out of the business. Last but not least, many will develop large-scale service models that are attached to the semiconductor industry run rate. The latter category is something I am monitoring carefully because I believe there is a lot of opportunity in this arena. I would anticipate hearing a lot about "fab services" during SEMICON West.