As we approach the semicap mothership’s C1Q earnings call, I like looking back at previews to see who is getting the datapoints right. And I came across this gem.

MS’ Atif Malik makes the case that chip equipment utilization will fall in Q2 due to Japan on weaker demand and lacking wafer supplies, and as a result “2Q earnings are at risk for the group.”

Really? Lower equipment utilization is the reason semicap will miss Q2? Let’s quickly examine how semicap companies make money.

They sell machines.

Wow, that was easy. The bulk of sales/earnings don’t come from running machines or supplying materials. They sell machines. How does decreasing utilization in one quarter mean semicap will miss?

Perhaps what he meant to say was that utilization is part of the equation used to determine future demand. If there is an expectation that utilization will come down long enough to impact orders and/or sales of equipment then manufacturers will say “Meh, why do I need to buy more equipment?”

Attie goes on to say he expects utilization to pick up again in 2H11.

Surely there must be something else in his note to explain these first two lines on his cover page. At this point he looks like an incompetent boob with two nipples.

We get to Pg 2 and he is kind enough to supply us with a chart of industry utilization vs semicap stocks. While there are a lot of knobs to turn between “utilization” and “stock performance,” this is a pretty common type of chart because in the past folks knew manufacturers would order a steady stream of machines if utilization was above 80%. Everyone became smarter since then (specific manufacturers became more industry independent, sell-siders developed better insight into company-specific order decisions…) and in fact TSMC now says 95% is the new 80%. But we all still like looking at this chart.

Huh. Look at this. Everything seems reasonable until June 2007. If this chart were cropped from June 2007, a newbie would correctly say “you’re a crackpot, there is no correlation between semicap stocks and utilization.” What you can say is that semicap stocks sort of lead chip utilization, implying that there is some other indicator out there defining what moves semicap stocks. It also means measuring chip utilization as an indicator of semicap performance is a little silly.

Hey Attie, stick to datapoints on order flow. Semicap is no longer sexy and you aren’t going to make it sexy by pretending it is related to near term volatility in the chip market.

Tweet