Budapest, Hungary Semiconductor market analysts are still under estimating how good a year 2006 could turn out to be according to Malcolm Penn, chairman and CEO of Future Horizons. Speaking at his company’s Annual International Electronics Forum, Penn was still confident that his forecast made in January of a growth in dollar sales of semiconductors topping 20% this year was realistic while unit sales would grow 12%. Other major analyst predictions are coming out showing an average growth of 8% with only one willing at present to break the 10% barrier.
Penn is bullish that a combination of an improvement of 4.9% in world GDP which the IMF upgraded from 4.3% last week, a rise in unit demand and a shortage of capacity, pushing prices up, will provide the catalyst for a good year. “All the evidence is strong and everything is in place – if it is not a strong year then the semiconductor industry does not deserve a strong year,” says Penn.
No-one believes that the chip industry can bounce back says Penn and there is too much of the 'Ill see believe it when I see' attitude but by the time it becomes evident it will be too late too take advantage.
Penn say the recovery in the unit demand started last May while a recovery in the average selling price started in the third quarter of 2005. Last year is seen as a market correction not a recession with the second half coming on strong after a first half depressed by inventory purges. “Markets normally bounce back fast from a correction,” according to Penn. While average selling prices (ASPs) have bottomed out, they take a time to recover, anything between 4 to 8 quarters after unit demand has improved.
Reacting to criticisms of his use of the IMF GDP forecast as the economic basis for his analysis Penn says that only twice in the last ten years have they been significantly off with their figures – in 1997 which was affected by the SE Asian financial problems and in 2000 when they was a stock market drop post 9/11. “All we can see is the IMF is not good at forecasting world disasters, in all their another forecasts they are correct or slightly pessimistic if anything, never optimistic.”
According to the IMF growth in the major industrialised geographies is stable but growing in the emerging economies including central and eastern Europe, Korea, Taiwan, Singapore and Brazil. “The new economies are driving the World GDP and where the GDP grows … so goes the chip market,” so Penn says the time is to “go east” to improve sales.
The tightness in wafer capacity looks like continuing as capacity is not being added as fast as demand. While unit demand has risen 4.2% per quarter from Q1 2003 to Q4 2005, fab capacity has risen only 2.7% in the same period. The underlying capacity utilisation trends are good and as Penn stresses “having fab capacity does not guarantee new sales, but if you can't make wafers you can't get sales.”
One trend that Penn is keen to point out is that process technology advances do not significantly increase capacity, it simply enables increased complexity. And it is now too late to order new equipment too boost this year's production – if it is not already on the way to the fabs it will not have effect on the coming shortages so says Penn “unless unit demand collapses, 2006 will be a year of shortages and allocation.” This undersupply imbalance is structural not temporary and while the back end is 'easier' to fix no one is willing to make the move, the front end takes longer – 6 to 9 months minimum – and its going to get worse before it can even get better.
Penn believes that the industry has nowhere near run out of steam, it is expanding in both geographies and new products, and nor is it anywhere near to 'maturing' only some sectors and companies are maturing not the industry as a whole. But there is a note of caution to companies to check their plans. “The third digital wave will change the status quo with business being 'one deal at a time' but if your markets are maturing, you're playing in the wrong space.”