2003 Economic Forecast

Advanced Forecasting Inc. David F. Crume Director

Following our projections last year for a slow recovery, total IC revenues in 2002 were certainly hampered by the weak economic factors impacting underlying demand, resulting in a dismal annual growth over 2001.
 
As our quantitative-based Recovery Index had indicated, the IC industry emerged from its minimum point at the end of 2001. While the economic factors affecting the sluggish recovery are now improving, the timing of the next significant industry upturn was first discussed in our February 2002 report. Our long-term model indicated major improvement in underlying demand to take place mid-2003. We have maintained this forecast, which was in contrast to other analysts, who projected for 2002 total IC revenues in the range of 14-20%.

Note: There are close similarities between the 1986 and the 2002 recovery periods. Both recovery periods shared a non-symmetric recovery relative to their previously deep decline. However, the long and slow recovery caused industry observers in 1986 to underestimate the growth for 1987. Actually, 1987 was a great year, having experienced at least 25% growth, as predicted by AFI's model.

Due to our unique quantitative methodology, which avoids human opinion and extrapolations as input, the measurements of economic factors allow us to project what the IC industry will experience over the next 19 months. Currently, we do not detect another plunge in underlying demand during 2003.

For the many executives who can't see the drivers that will fuel the upswing, we identify the following: The tendency of human beings to extrapolate darkens their current vision, discounting drivers that during 2000 would have been hailed as the next "must have" gadget.

Indeed, some drivers of previous IC booms were not known well in advance. And numerous start-ups as well as mature firms are developing future drivers that will become viable when underlying demand strengthens. In sum, underlying demand will drive the drivers, not vice versa.

Gartner Dataquest Dean Freeman Principal Analyst

After surviving two consecutive years of a dismal business climate, the industry is looking ahead to 2003 for what should be improving conditions. Corporate America is beginning to return to profitability and with it tech spending is finally seeing some positive growth with a 4.3% burst in Q3 of 2002. The U.S. and world economies are gradually improving. In 2003, the world and U.S. GDP are expected to grow ~2.6%, which — while not a strong year — allows for some encouragement.

Electronic equipment should grow at 6%, which translates into 12% growth for semiconductors in 2003 after a flat 2002. Capital expenditures for the semiconductor business are expected to be in the 14% range with spending on wafer fab equipment at 15%. Packaging and assembly historically lead the way out of a downturn, and this year is no exception with growth in 2003 expected to be in the 24% range.

While all of these numbers sound encouraging for the year it is expected that the industry will continue to struggle in the first half of 2003 as the economy gains momentum. The second half of 2003 should see stronger growth as the economy continues to improve.

Technology purchases will be dominant in the first half of the year as factories prepare for the introduction of 90 nm. 300 mm purchases will lag until the second half when at that time the increasing demand for capacity will jump-start the 300 mm fabs that are currently running on idle.

Asia-Pacific will lead the growth spurt as foundries and China gear up to meet demand. Europe and the United States will follow as the industry regains its legs and gets ready for another steep climb. Although 15% growth for the year is tame by semiconductor equipment standards, growth in 2003 will set the platform for 2004 and beyond.

IC Insights Bill McClean President

In 2003, the global economy is forecast to grow 3.1%. Although this growth rate is a half percentage point better than in 2002, it is still below the 30-year average annual increase in worldwide GDP of 3.4%.

For the electronic systems industry, the 2001-2002 time period can be legitimately labeled an "industrial depression." After the first-ever recorded decline in 2001 (-14%), worldwide electronic system sales declined another 6% in 2002.

For 2003, there appears to be no "killer" application in the electronic systems industry to spur very high growth in the IC industry. However, in 2003, IC Insights forecasts that a 5% increase in electronic system sales will be enough to drive IC market growth of 15%.

On a cautionary note, the "below average" semiconductor industry capital spending level of 2002, which is expected to carry over into 2003, coupled with an already high MOS capacity utilization rate for leading-edge devices, leaves the IC industry with no fab capacity "safety margin" in 2003. Thus, any surge in IC demand (a distinct possibility in the second half of 2003) is likely to cause spot shortages of advanced ICs and a spike in IC average selling prices.

In-Stat/MDR Steve Cullen Director and Principal Analyst

Worldwide semiconductor revenue in 2002 was about even with 2001's figure, give or take a percentage point or two, but the good news is that the year ended on a much more positive note than it began.

Third-quarter revenue was 21% above the low point of the downturn, which occurred in the final quarter of 2001. Complete fourth-quarter data was not available when this was written, but we believe the slowing of the growth that occurred in the fall of 2002 was a temporary phase in the recovery and that further increases in demand along with higher ASPs will make 2003 a better year.

There are still some bleak spots in the markets, to be sure. The communications and computing infrastructure at the core of the network, which was severely overbuilt during 1999-2000, still has capacity that has yet to be fully utilized. Telecom capital spending is not expected to resume growth until 2004 and major segments of the WAN equipment market will have no growth in 2003. Growth in the LAN and wireless areas, however, has resumed and In-Stat/MDR's outlook for these products is for still stronger growth in 2003.

The long-awaited growth in the business PC market is expected to kick in by mid-2003 as the economic recovery continues and business confidence builds. Recovery in these major markets, combined with continued growth in the smaller but healthier consumer-driven automotive and consumer electronics sectors, is expected to raise 2003 semiconductor revenues to the $164B level, about an 18% increase over 2002.

INFRASTRUCTURE Carl Johnson President and Co-Founder

As we enter 2003 I confess to being very concerned about another year of "nano-sized" profitability in the semiconductor and semiconductor equipment industry. Chipmakers and equipment companies are still scrambling to find the correct balance in their businesses. Unfortunately, the excessive behavior of the past still haunts the sector.

It goes without saying that the economic climate holds the key. To me it seems very clear that the economies of the world still have severe dislocations to work through. That process will continue to hamstring corporate IT and consumer spending patterns. I say this even though we have witnessed a nice rebound in unit volumes. What we really need to see is strength in semiconductor ASPs.

There is optimism that high-speed, wireless Internet access will usher in a new round of strong demand. Frankly, this smacks too much of the past bubble and I am reluctant to get overly optimistic with the possibilities.

Many people in the industry agree that we have reached an inflection point. That point is often compared to inflections that took place in the automobile and aircraft industry. Perhaps the semiconductor industry is maturing and we are moving much closer to an industrial manufacturing model? If that is the case, businesses have to adjust expectations for profitability and growth to thrive during this transition. More adjustments ahead? Most definitely.

InsideChips.com Steve Szirom President

After compiling the preliminary numbers for the 2003 forecasts from the industry's analysts, it was apparent that there was quite a bit of positioning among the analysts in the last few months to come up with numbers for 2003.

The forecasts tend to cluster in the 12-20% growth range for next year. The two semiconductor associations, SIA and WSTS, are holding firm to 19.8% and 16.6% forecasts, respectively. SIA represents the North American view and WSTS is more representative of the world view.

Future Horizons, which stuck its neck out last year with a -5.5% forecast for 2002, is one of the optimists for next year with a 26.4% growth forecast for the industry. Semico, which foresaw 19.6% growth for 2002 at the end of last year, is bullish again and prognosticating a 30% growth next year. Other 2003 growth predictions announced in the final quarter of 2002 include VLSI Research, 19.8%; Gartner Dataquest, 12.1%; In-Stat/MDR, 18.1%; iSuppli, 15.0%; and IC Insights, 15.0%.

The worldwide digital cellular handset market will be one of the modest drivers creating growth as users shift from second-generation (2G) to 2.5-generation (2.5G) cellular. 2.5G is also the beginning of camera-enabled handsets. These feature-rich phones will provide a positive boost to an otherwise lackluster rebound. Revenues from the PC market will be essentially flat as modest sales gains are offset by strong competitive price pressures.

Even with inventories at OEMs at historically low levels, market demand from the enterprise PC and IT buyer has to pick up. We believe that this demand cycle will not start until the second half of 2003.

iSuppli Corp. Gary Grandbois Principal Analyst

As semiconductor recoveries go, 2002 was a non-event. Pockets of growth were overshadowed by regional and application market declines, resulting in virtually no growth in the semiconductor industry for the year. However, 2003 will bring a return to the more consistent, monolithic growth that typically characterizes a recovery.

Electronic equipment sales will exhibit low growth during the first half of 2003. But sales will swell in the second half, resulting in revenue growth of 7% for the year. This growth will be the consequence of concurrent increases in all major electronic equipment categories — an event not seen in two years. A strong surge in wireless handset sales will combine with the steady growth in the automotive and consumer markets, the reawakening of the computer business and the thawing of the networking segment after a disastrous two-year slide. This equipment growth will be mirrored in the electronic components markets.

Unfortunately, semiconductor growth will be virtually flat at 1.5% during the first half of the year because sagging average selling prices for DRAMs will drag on the revenue. However, the 7% growth in electronic equipment will drive an increase in unit volume and semiconductor revenue growth will blossom to a 10% rate in the second half of 2003, bringing the year to better than 11% growth.

Additional semiconductor fab capacity will become available just as the market slows in 2005, resulting in semiconductor revenue growth dropping into the low single digits.

Reed Electronics Group Daryl Delano Director of Economics

A review of sales and order data compiled by SEMI covering the world as a whole was unusually encouraging for the month of September 2002. Since this global data summary lags by a month the book-to-bill information compiled for the North American market data, we don't yet know how sales trended during October. We can only hope that data for the final quarter of 2002 confirms the tentative signs that September represented an inflection point for trends in semiconductor equipment sales and orders.

The value of worldwide equipment billings during September 2002 was 19.8% higher than during the same month a year earlier, although it should be acknowledged that September 2001 billings (particularly to North America) were at very low levels in the immediate aftermath of the 9/11 terrorist attacks. Nevertheless, the underlying trend does appear poised to turn more positive (or, at least, less negative) in the months immediately ahead.

With 2002 numbers so disastrously bad, of course, any sign of good news is welcome. During the first nine months of 2002, global semiconductor equipment sales were worth ~$14.42B — about 40% less than the $23.94B worth of equipment sold worldwide during the first three-quarters of 2001.

There's little question that the vast majority of semiconductor capital equipment purchases during 2002 were driven by the desire to upgrade technology, not to increase production capacity. Although the average capacity utilization rate among U.S. semiconductor manufacturers had moved up from 63.1% at the beginning of 2002 to 67.8% during October, it remained extremely low by any standard of historical comparison. Utilization rates of 90% or above were the monthly norm from 1998-2000. And even the industry's more modest 35-year-average capacity utilization rate of 79.6% far exceeds the current dismally low rate.

Nevertheless, as more firms become convinced of the sustainability of a global economic recovery during the first half of 2003, pressure should slowly but surely build throughout the world for investment in new more-efficient, cost-reducing chipmaking equipment and technologies. So we remain reasonably confident that the global market will bottom out during the early months of this year, and that we'll start to see some sequential gains (from the new — very depressed — base of activity, of course) by the early summer of 2003.