Dogpile

I am borrowing a title from a recent INFRASTRUCTURE newsletter for this article. It really fits the scenario we are seeing today. A few months ago I wrote about the possibility of a "dogpile" in the semiconductor industry. Sure enough, we are seeing chipmakers dive on top of each other to build new fabs and install new manufacturing capacity. Not that this was an unexpected event or anything...

There are good reasons why this is happening. During the past three years, capital equipment spending ratios collapsed to levels not seen since the early '90s. In fact, equipment purchases as a percentage of total semiconductor revenues approached the single-digit level in the third quarter. For assembly and packaging companies, the percentage of chip revenues allocated to new equipment fell, on a monthly basis, below 1%. We hit new lows in several equipment categories this year. All this, right in the face of chip unit volumes tracking toward record levels.

Per the usual script, chipmakers have found themselves under-invested. This always seems to happen during a period where seeds of strong recovery are poking up through the ground. Aside from a few large players, namely one large West Coast-based manufacturer of microprocessors, most chipmakers are scrambling to add new capacity.

What I find really positive about the outlook is the growth in new applications and end markets. Just wandering about the electronic goods shops will tell you that semiconductor content is broadening. What is really striking is the amount of chip content used by new display technologies. Yes, the PC and the cell phone are still the two biggest drivers, but there is much more to come.

Another positive is the potential for broadening strength in end demand. The corporate spending front has yet to kick into high gear and device pricing, though stronger in recent months, is still far below the levels recorded in previous cycles. If these two gain some traction next year, we will see the dogs stacked a mile high.

Now that the dogpile is underway, I will admit I am concerned about this turning into a one-year phenomenon. Yes, I am aware of the prognostications by the forecasters. Of late, several have revised and raised their industry tallies for 2003 and 2004. As for me, I have been talking about an increase in activity for a while. I like to use Jim Morgan's approach: Prepare for the upturn during the downturn. Prepare for the downturn during the upturn.

I probably do not have to tell you that stocks in the sector have been partying like it was 1999. The only way you could have lost money was to fight the tape. Will this continue in 2004? I think we are going to see a period of gain digestion during the coming months. I believe traders will lean toward profit booking in late 2003 and early 2004. I suspect that some will voice concern about post-holiday electronics sales and that will spook some of the weaker hands. I am also keeping fresh in my mind the fact that chip equipment stocks, during the first quarter of 2000, peaked almost a full year ahead of the peak in shipments. Don't forget that Wall Street wants to be way, way ahead of the curve.

Looking longer term, I can make a pretty good case where dips in the stocks are buying opportunities. This fits with my positive view of the digital revolution. I am not suggesting that one should rush out and buy every dip. But applying capital to the companies that are profiting from the upturn is certainly an appropriate strategy to pursue.