Jean-Philippe Dauvin, corporate economic and market research manager at SGS-Thomson Microelectronics NV, predicted particularly strong European growth in the semiconductor industry while Japan continues its slide in a presentation entitled Semiconductor Market and Industry Outlook: Growth and Changes. However growth in the industry has always been and remains cyclical, although fluctuations in demand have become […]
Jean-Philippe Dauvin, corporate economic and market research manager at SGS-Thomson Microelectronics NV, predicted particularly strong European growth in the semiconductor industry while Japan continues its slide in a presentation entitled Semiconductor Market and Industry Outlook: Growth and Changes. However growth in the industry has always been and remains cyclical, although fluctuations in demand have become less pronounced. 1993 was the high point of the seventh cycle and, consequently, slower growth is expected in 1994, with demand increasing by 25% over 1993 levels, but by just 8% in 1995. Dauvin expects the current cycle to bottom out in 1995, thereafter seeing stronger growth. In 1993, the world semiconductor market was estimated at $77,000m and of this the US held 32%, Japan 31%, Europe 19% and Asia 18%. In terms of product families, microprocessors, microcontrollers and peripherals represented 25%, memories 28%, dedicated integrated circuits 18% and discrete components and optoelectronics 15%. Over the past 30 years, demand for semiconductors has grown at an average of 15.5% per year, and Dauvin has an annual growth forecast of 13.5% between now and the end of the century, to take the market from $50,000m in 1990 to $200,000m in the year 2000. In semiconductors, unlike other sectors, the consolidation progressively increases as the industry reaches maturity and as the cost of developing and manufacturing new products rises.
As a result, in the year 2000, the world’s top 10 semiconductors will probably meet 60% of total demand against 55% today. Semiconductor technologies are crucial to the progress of the world electronics industry, through constant innovation and cost reductions, and their importance is increasing. For although semiconductors may account for only 10% of the total cost of electronic equipment, the performance and development of the equipment is totally dependent on that 10%. Central to the stability and growth in the profitability of the semiconductor market are fabrication capacities. And for Dauvin, the most important event for the last 13 years has been the excess supply and capacity overhang created by Japanese semiconductor suppliers, driving down profitability as Japanese chip makers overinvested in production capacity. However the Japanese, forced by recession, now appear to be retrenching and their capital expenditure has dropped from a peak of 28% in 1985 to 12% in 1994, as percentages of revenues. This shrinkage will affect the profitability of semiconductor manufacturers very positively as there now exists excess demand coupled with an undercapacity. In terms of macro-economic trends, the US should enjoy growth of around 3% a year until 1995, fuelled by domestic demand, boosted by NAFTA and by a structural return to high productivity. In Europe, Dauvin believes that after five years of continuous decline, economies reached the bottom of the cycle in 1993, and 1994 should ‘ring the bell’ of a new five year cycle upwards, with a peak in 1997-98.
By David Johnson
In 1995, the Japanese economy will hit the bottom of a structural six year decline, however relative short-term growth thereafter will not alter Japan’s fundamental problems. By contrast, the Asia Pacific Rim market is fuelled by a boiling domestic demand, still boosted by China, comparable to Japan in the 1950s, and the region has entered a long period of prosperity. For the electronic industry, Dauvin is sure that the bad times are over and that the industry is now entering a completely new period of growth. The equipment industry returned to growth in 1993, propped up by personal computer sales, where the US brand enjoys an 80% domination, and the emerging applications such as radio telephone and new markets. In 1994 and 1995, the overall European recovery should add significant growth to US and Asian growth, balancing the slowing of personal computer demand. However he expects the Japanese electronic industry to remain in the doldrums until 1996 at the earliest. It will not enjoy the fruits of the persona
l computer revolution, as Japan remains a mainframe-centric country, suffering high production costs, and is presently enduring, for the Japanese, an unacceptably high level of unemployment of around 6% in real terms, which serves further to starve domestic demand. To Dauvin’s mind the electronic industry in Japan has missed all the new technologies and is not in the driving forces of electronic development. According to Bank of Japan figures, the growth of production of equipment in the electronic industry began to decelerate in 1989 and started shrinking in mid-1991, slumping to a 20% contraction in 1993. Its forecast for the future sees a deceleration in the rate of reduction, with a 10% fall in 1994 and a 5% fall in 1995. The future for Europe is much rosier, says Dauvin. It is the dominant power in the telecommunications field, particularly switching, and this is where rapid global growth is occuring. The explosion of cellular telephony has surprised all analysts, and Dauvin has been forced to multiply his prediction for the numbers of cellular phones produced in Europe by a factor of five for the year 1998, to 30m units. This is good news for the semiconductor market, for whereas a corded phone requires around $5 of semiconductors, a cordless phone requires $20, and a cellular phone $100. Similarly, Europe is foremost in car production and the semiconductor market will enjoy the increased penetration of automotive electronics. From $35 per car in 1990, to $100 in 1994, the value of semiconductors will rise to around $160 by 1997. For example, the new Ford Motor Co Mondeo has five microcontrollers and around $100 of semiconductors, whereas the car it replaced, the Sierra, had only $5 or one microcontroller. By 1998, Dauvin expects 80% of cars to have an engine management system, averaging $20 of silicon and 50% of cars will have automatic braking systems as standard, around $25. Whereas the Japanese market will grow by 24% to $29,500m between 1993 and 1995, Europe will see 37% growth to $20,000m in the same period. However in global market share Japan will fall to 28% in 1995 from 39% in 1990, while Europe has been steady at 19% in a market whose value has more than doubled.
Europe has continued to grow in the specialist telecommunications area and is also now seeing the return of manufacturing transplants, particularly to the UK and Eire. Forecasts about the future of the semiconductor market are fairly consistent and most analysts believe there will be above 20% growth in 1994, positive but lower growth in 1995, with a slight pick-up in 1996. Although the market will continue to be cyclical, Dauvin believes we will not have to undergo the same boom-bust of the 1983-84 cycle. He has revised his November growth rates from 10% in 1994 and 5% in 1995, to 22% and 8% repectively. Comparing the 1983-85 and 1993-85 cycles, in the 1983-85 cycle homogeneous and strong world economic conditions meant that all regions moved up and down together. There was low penetration of chips, hugely over-stuffed inventories of equipment and semiconductors, and manufacturing overcapacity. In the present cycle he sees heterogeneous economic conditions with Europe and Japan moving independently. There is a much higher level of penetration in terms of the value of semiconductors: 12% of the cost of the product against 6% 10 years ago. Inventories are under control and and there are some shortages. More and more silicon is used in electronic equipment, representing higher percentages of the total cost: the 80286 represented 18% of the cost of the computer, the 80486 is 28%, all chips are 50% of the value. By 2000, he sees the world semiconductor reaching at least $200,000m, more likely $250,000m.