Tejek,
First, let me say that (although it may seem obvious), that there is no longer a monolith called "the semiconductor industry" (save that there are a bunch of companies that sell very small slivers of silicon for much more than the cost of materials). Two-three decades ago, there was an industry dominated by "broad line semiconductor companies" that all sold the same thing (or a sub-set of the same thing) to the same customers. What affected, say, TXN, affected MOT, and NSM, and even little INTC. Now, companies are far more specialized, and what affects INTC may or may not have bearing on TXN.
That said, there are always two issues: inventories, and end-demand. In general, end-demand (broadly defined), has always grown, but at quite variable rates. In the long, flat period of 1989-1991, when the industry kept lead times under control, end demand slumped, and semiconductor sales flattened- but no worse than that. That was a "River Platte" recession. By contrast, in 1973-1975, and 1983-1985, when lead times spiraled out of control, and consequently semiconductor customers built inventories, demand collapsed- but then, rebounded, as customers had been buying less than they were using.
The run-up in lead times late last year and earlier this worried me. OTOH, I had felt the overshoot would last longer. My GUESS is that this downturn will be brief- maybe over by early next year- but not a lot of fun. A lot will depend on how bad end-demand really is.
Stuart |
