2001 IC Industry at the Crossroads

		At a Glance
	Three major factors that typically cause, or are part of, IC industry downturns are global recessions, inventory burns and/or overcapacity. This article looks at lessons learned from past cycles and forecasts how the present cycle will evolve.

Many are familiar with the "pinwheel" cycle chart IC Insights first created and presented in early 1999 (Fig. 1). It has been fine-tuned since the original was released, with changes to both the years and descriptions of the cycle segments. This most recent version is specific to the IC market. The pinwheel shows the contribution of various factors to the ongoing cyclical nature of the IC industry.

Figure 2 shows a comparison of the six cycles that have occurred in the IC industry since 1971. The number inside each arrow is the deviation during that cycle's upturn or downturn from the 30-year historical average annual growth rate of 17%. Table 1 uses the 1993-1998 time period, Cycle 5, as an example. The 1993-1995 annual IC market growth rates deviated a total of 58 points on the positive side of 17% while the 1996-1998 annual IC market growth rates totaled 65 points on the negative side.

This article examines IC industry cycles from 1971 to the present. IC Insights refers to this 30-year period as the "modern" IC industry. Although data from 1958-1970 may confirm or refute the cycle trends presented here, we do not believe they are relevant to the post-2000 IC marketplace.

The cycles described are based upon worldwide IC industry revenue in U.S. dollars. IC market regional (e.g., United States, Japan, Europe and ROW) aspects with regard to industry cycles are not included. A brief overview of each cycle follows.

Cycle 1: The '1st oil shock' cycle

A huge increase in oil prices (i.e., the "oil shock") occurred in 1973 and was followed by a worldwide recession (defined as GDP growth of less than or equal to 2%) in 1975. As shown in Figure 2, 1973 was the last year of Cycle 1's three-year upturn. The IC industry downturn lasted four years, extending two full years after the global recession year of 1975.

1. The ongoing cyclical nature of the IC industry and how it is impacted by various factors.

Cycle 2: The '2nd oil shock' cycle

The second oil shock came in 1979, six years after the first. 1979 was the "middle" year of the three-year 1978-1980 upturn. A global recession occurred in 1982. This global recession was at the tail end of the relatively short two-year downturn of 1981-1982.

Cycle 3: The 'inventory build/burn' cycle

Including both the upturn and downturn portions, this entire cycle lasted only three years. In 1984, the IC industry witnessed the largest IC unit volume increase (50%) of any year in its modern history. Electronic system producers stocked warehouses full of ICs in 1984. In 1985, IC unit volume shipments plummeted 16%. This 16% decline is the largest annual percentage decline for IC unit volume in the modern history of the IC industry. In fact, 1985 is the only year since 1980 in which IC unit volumes decreased.

Table 1. IC Industry Cycle Example (Cycle 5)
Year	IC market growth (A)	Historical average (B)	Deviation from average (A-B)
1993	32%	17%	15
1994	34%	17%	17
1995	43%	17%	26
Total upturn Deviation points*	—	—	58
1996	-9%	17%	-26
1997	4%	17%	-13
1998	-9%	17%	-26
Total downturn Deviation points*	—	—	-65
(Source: IC Insights)			*Values used for Cycle 5

Cycle 4: The '3rd oil shock/ electronic system sales lull' cycle

Iraq invaded Kuwait in 1990 and the Persian Gulf War was fought in 1991. The invasion and subsequent war created the third oil shock since 1970. In 1991, the world also endured its third global recession since 1970.

2. A comparison of the six cycles that have occurred in the IC industry since 1971. The number inside each of the up and down arrows is the total deviation from the historical average annual growth rate of 17%.

It was during 1989-1992 that electronic system sales displayed lackluster market growth. Worldwide electronic system sales in 1989-1992 registered only a 5% average annual growth rate, compared with a long-term average of about 8%. This slow period for electronic system growth is reflected in the U.S. production of computer systems during this time. In 1989, U.S. production of computer equipment totaled $68B. In 1992, that figure was only $67B. A slow electronic systems market, especially computer equipment, has historically had an extremely negative impact on the IC industry.

Cycle 5: 'Big wave/double dip' cycle

Both the up and down deviations in the modern IC industry cycles were getting less pronounced until the "big wave" cycle of 1993-1998. In 1995, many believed that the IC industry was immune from a recurrence of its past cyclical behavior. In hindsight, it is very clear that this was not the case. It has been well documented how the overbuilding of capacity came crashing down in the form of weak IC pricing throughout the 1996-1998 time period.

In 1998, the IC industry appeared on the road to recovery until the Asian currency crisis came along. The Asian crisis helped cause the fourth global recession since 1970. Thus, the 1996-1998 IC industry downturn was caused at first by overcapacity and then extended by a global recession.

Cycle 6: 'Highly anticipated' cycle

The 1999 and 2000 upturn years represented the early stage of the sixth IC industry cycle and were boosted by very strong worldwide GDP results. In 2000, worldwide GDP growth registered a 4.8% surge — one of the strongest increases encountered over the past 30 years and significantly above the historical average of 3.4%.

Of all the IC industry cycles since 1970, the downturn associated with Cycle 6 is by far the most anticipated.It was one or two quarters into the current upturn (in late 1999) when IC Insights was first asked about the next downturn. Whenever IC Insights was asked about the upturn portion of the cycle, it was in the context of, "When is the upturn ending?"

IC Insights believes that the relatively long lead-time anticipation of upturns and downturns will reshape future IC industry cycles. Most IC suppliers are more sensitive to the IC industry cycles we are now describing. This new-found sensitivity, coupled with other infrastructure factors, is expected to cause more rapid IC producer capital spending adjustments (both up and down) in response to market changes. This, in turn, is likely to help continue to moderate the magnitude of future IC industry cycles.

Excluding the "big wave" of Cycle 5, the magnitude of IC industry cycle highs and lows has lessened over the past 30 years. Using IC Insights' 35% increase for 2000, Cycle 6's upturn will register 20 points (+2 in 1999 at 19% and +18 in 2000 at 35%). Assuming a continuation of the cycles, IC Insights expects Cycle 6's downturn to register about -18 points. This 18-point deviation is likely to occur over a two-year period with 2001 and 2002 IC industry growth being 10% or less each year.

Could another "big wave" cycle similar to Cycle 5 occur again? IC Insights believes it can. Although the IC industry trend is toward less deviation from the average, given the right set of circumstances, an unexpected Tsunami cannot be ruled out.

IC industry cycle summary

What can be learned from the past 5.5 IC industry cycles? Table 2 shows IC Insights' observations from examining the past 30 years of IC industry history. It is interesting to note that the very "neat" depictions of IC industry cycles using dollar volumes do not work when using IC unit volumes.

Table 2. IC Industry Cycle Observations

• Four out of the last five downturns were accompanied by a global recession. Global recession years include: 1975, 1982, 1991, 1998.

• There are three triggers for an IC industry downturn: 1) Global recession 2) Overcapacity 3) Inventory burn

• IC industry deviations come close to balancing within one cycle.

• In general, IC industry cycle deviations are becoming less pronounced.

• Most upturns last about three years.

• Downturns last one to four years, about whatever it takes to offset the upturn.

(Source: IC Insights)

To help understand where the IC industry is headed, we must determine its current position in the cycle. In 2000, almost all the major indicators IC Insights tracks (worldwide GDP, electronic system sales, IC industry capital spending, etc.) were at extremely high levels compared with their historical averages.

When viewing such data, IC Insights believes that only one of two conclusions is logical: 1) Historical averages are too low and future results can stay above these averages for an extended time period; or 2) These averages are realistic and for each extended period spent above the average there lurks a subsequent period of time where growth will be below the average, at or near the same order of magnitude.

IC Insights believes in conclusion number 2.

2001 summary

The downturn portion of an IC industry cycle is usually triggered by global economic recession, IC industry overcapacity or inventory corrections. In 2001, IC Insights believes that the IC industry will be affected by all three.

Global GDP is forecast to slow from 4.8% growth in 2000 to 3.5% or less in 2001. Although not a worldwide recession, the reduction in growth will negatively impact electronic system sales.

Worldwide MOS IC wafer capacity is likely to increase 24% in 2000, which is greater than the 21% increase registered in the "boom year" of 1995. The main cause of this capacity surge was a 79% increase in capital spending by IC suppliers in 2000. IC Insights forecasts that worldwide IC capital spending will increase by <10% in 2001.

Many electronics producers accumulated excess IC inventory throughout 2000. The inventory adjustment period is expected to last through the first half of 2001. Inventory corrections and overcapacity usually lead to IC pricing weakness.

Although a cyclical IC industry slowdown is anticipated to begin in 2001, IC Insights believes that it will be short-lived. After modest growth in 2001 and 2002, we expect the market to increase 20% or more beginning in 2003.

In a perfect world, annual global GDP growth would be 3.4%, electronic equipment sales would increase 8% each year, and every IC producer would spend 21% of sales on capital each year. In this scenario, the IC industry would grow exactly 17% every year. As we all know, the world is a long way from being perfect and the IC industry is still likely to exhibit its cyclical pattern in the future. •

Bill McClean is president of IC Insights Inc. He began his market research career in the IC industry in 1980 and founded IC Insights in 1997. During his 20 years of tracking the IC industry, he has specialized in market and technology trend forecasting. McClean has a B.S. in marketing and an associate's degree in aviation from the University of Illinois.