Lately there hasn't been much good news in the communications semiconductor market. How bad has it been? Assuming a flat fourth quarter, total communications semiconductors will be down 28 percent in 2001 vs. 2000. Application-specific chips for wired communications will be down 35 percent. Application-specific chips for wireless will be down 23 percent. General-purpose chips will be down 27 percent. Will things get worse?
No one can predict the future. What analysts can do is provide scenarios based on given sets of assumptions.
Recently, I built a simple model that estimates 2002 communications semiconductor growth based on the two most important inputs. The first input is how much inventory the industry will have burned through by the end of 2001. The second is the growth (or decline) of communications equipment sales in 2002.
To model the impact of inventory, one must estimate how much inventory the industry started with at the beginning of 2001. Historically, semiconductors represent just less than 14 percent of total system sales.
But in 2000, the semiconductor market was nearly 16 percent of system sales. That left approximately $6.8 billion of communications semiconductor inventory overhang.
One can then model 2002 semiconductor growth by making assumptions about inventory burn and equipment growth. A fair estimate is that the industry has burned through 75 percent of the excess inventory. The 2002 equipment growth rate is harder to nail down. It includes carrier capital expenditure plus sales of communications equipment to businesses and consumers worldwide. A fair starting estimate is that those sales will decline 5 percent in 2002. Thus the communications semiconductor market would be essentially flat in 2002 (2.8 percent growth from $40.6 billion, to $41.7 billion). If that happens, we will need to see compound quarterly sequential growth of 13.3 percent starting in the first quarter of 2002.
Using similar assumptions, we can model best-case and worst-case scenarios. The best case is growth of 20 percent based on 100 percent inventory burn in 2001 and 3 percent equipment revenue growth in 2002. The worst case is a 20 percent decline based on 50 percent inventory burn and 20 percent decline in equipment sales. Even the worst-case scenario requires 2 percent compound quarterly sequential growth. So, if the first quarter is down, look out below.
Jeremey Donovan (jeremey.donovan@gartner.com) is a principal analyst at Gartner Dataquest.
RSS
RSS| 
Digital| 
Mobile
