Structure of product market & corporate strategy
Structure of product market & corporate strategy
Nihon Fukushi University
1. Market structure ＆ portfolio management
1-1. Cycle model approach
1-2. Product portfolio management
2. Semiconductor market and corporate strategy
2-1.Corporate strategy of Japanese manufactures
2-2.Analayses of market structure and product diffusion
Managerial approach of Product Portfolio Management :PPM (Abegglen, J.C.& Boston consulting group.1977)in decision-making of corporate strategy is suggestive to the dynamic analysis of market structure and product diffusion, especially in the market fluctuation and trend and positioning of competition. Prerequisite for PPM model requires two factors: pattern of product diffusion in the market, viz. PLC (Product Life Cycle)and the existence of experience curve. Product life cycle concerning market entry and market growth of products is observed remarkably in the semiconductor product market.
1.Market structure ＆ portfolio management
1-1. Cycle model approach
The feature of product life cycle model and experience curve appear in an under mentioned market trend of semiconductor products (SIC365-7),especially observed in the 1975-95 period.
(a) Existence of Learning curve (Henderson Rule)
Learning effect of technology causes sharp cost decrease for products sales in the market, which was discovered as experience rule by Boston
(b) Pai(π) Dollar & Bai(χ)1/2πDollar Rule
Market price of semiconductor unit, especially of memory component will fluctuate and then down to 3.14 (π)dollar or so in the first stage of market entry and secondly further down to about 1/2 (χ)dollar, the price trend of which was also the experience rule, discovered by the Bell
(c) Bai(χ)4 rule
Capacity of memory tips will increase about four times in three years with some tendency, especially in the product line of DRAM (Dynamic Random Access Memory) unit.
(d) IBM cycle (Silicon cycle)
Boom and depression of semiconductor product market has been raised in about four years period since 1974,the market trend of which appears owing to the adjustment of stock notably by IBM, relatively short period of depreciation with production line and timing gap of product switching (Egg Cycling).
Recently, this aforementioned trend is accelerated with expansion of the use in the private sector in which the factor of production cost counts a lot.
1-2. Product portfolio management
PPM model renders corporate strategy to manufacturers in deciding line and range of their products and also of their competitors’ in the future market, which will show the balance of competition in the market structure. ( matrix 1)
Positioning of competition/ rates of market concentration
Rate of market growth
Boy in question
2.Semiconductor market and corporate strategy
2-1.Corporate strategies of Japanese manufactures
Matrix chart :Positioning of semiconductor products applied for PPM model
Boy in question: PPRAM
Cash caw: Integrated circuit
Defeated dog: Discrete
2-2.Analayses of semiconductor market
According to WSTS, estimate of world wide whole sales of semiconductor product(1997) amounts to 138 B dollars( including logic:22B,MPU:49B and Memory:29B)and the sales of Japanese manufactures to 3717B yen. Remarkable sales increase is observed in the MPU sector.
Table: Sales and market share of semiconductor manufacturers
(specified for MPU use) 1994
Market share (%)
Vector of modulating/ Integration of assemble(→)
Vector of systemizing
/Integration of function(↑)
System integrated product
Industrial surveys (data source) show that more than half of the Japanese manufacturers point to suppliers’ technological capabilities as the most important factor for a successful design-in. These surveys can be interpreted to mean that the keys to a
Successful design-in is superior technology and accumulated experience. Three-fourths of the Japanese buyers expect foreign suppliers to bring these keys to their design-in activity. [See Chart]
They point to technology as the key to a successful design-in for analog, logic, and micro devices, whereas technology is regarded as less important for other products that are easier to design in. Good relationships with Japanese buyers, therefore, will be
Facilitated where foreign suppliers can offer continuous design-in support with superior technology.
Long-Term Cooperative Relationships
Semiconductor sales have grown dramatically over the past decade. At the same time, however, the expenditures required for new research and development and manufacturing have been soaring. These factors have fueled the explosion in international alliances that have taken place over the past several years. These alliances, which include the biggest names in the semiconductor industry, spread
The risks of new ventures and reduce the economic burden that such activities place on individual companies. [View Table 1] [View Table 2]
In addition to the forces driving international alliances, market forces are also spawning other forms of international cooperation. For example, the semiconductor industry faces a growing number of environmental problems.
A single company, or even the semiconductor industry of a single country cannot solve these problems. Today, environmental issues are being addressed on a global scale through
Similarly, standardization of the next generation of silicon wafer, along with the collection of international statistics, is two additional examples of areas being addressed through international cooperation.
*LONG-TERM COOPERATIVE RELATIONSHIPS
The semiconductor industry is growing beyond national borders. Japanese semiconductor manufacturers have been forming long-term cooperative relationships with foreign suppliers in a variety of areas, including research and development, technology, manufacturing, and sales.
Several important new initiatives have been launched to promote
1.Research and development of new technology and products;
2.Manufacturing, including support for start up of mass production.
*THE GROWTH IN INTERNATIONAL COOPERATION
Cooperation between the U.S.
Industry has witnessed a proliferation of full-blown joint research and development and joint manufacturing projects among the industry’s leading semiconductor manufacturers. [View Table]
2.Present Condition of the Japanese Market Compared to World Market
3.Semiconductor Market Composition
4.The Changing Japanese Market
5.Japanese Semiconductor Market Size Changes
6.Forecast of Future Trends in Japanese Semiconductor Market
The production of electronics equipment in Japan
First, the Japanese household market for consumer products was already highly saturated by the outset of the 1990’s. Moreover, the Japanese economic recession suppressed private consumption for consumer products. Demand for consumer products has not increased since 1988.
Second, many Japanese electronic companies moved production facilities for conventional consumer products offshore over the past several years due to the increase in the value of the yen. For example, more than half of color television and videocassette recorder production has been shifted outside of Japan Japan
*PRESENT CONDITION OF THE JAPANESE MARKET COMPARED TO WORLD MARKET
Despite a weak economy, the Japanese semiconductor market rebounded in 1993 and again in 1994, after a significant downturn in 1992. The Japanese market grew to $29.4 billion in 1994. The U.S.
Semiconductors. The fast-growing Asian market also increased its world market share, as shown in the chart. [View Chart]
The shift in the relative positions of the Japanese and U.S. Japan Japan U.S.
*SEMICONDUCTOR MARKET COMPOSITION
The composition of the semiconductor market differs greatly between Japan United States Japan U.S.
This emphasis on products outside the consumer products sector, such as computers, for example, is one factor, which has led the U.S.
*THE CHANGING JAPANESE MARKET
Although Japan’s consumer sector is still relatively large, there are signs that the Japanese market is changing. Two major trends illustrate this observation.
First, as illustrated in the chart, the overall semiconductor demand ratio of major Japanese users is slowly decreasing in the consumer sector. [View Chart]
Second, the product mix is changing. Increases are taking place in demand for MOS LSI’s (micro, memory, and logic); with the greatest increase being seen in the memory segment. At the same time, demand for analog and discrete devices is decreasing. [View Chart] The consumer sector is experiencing similar trends.
This change in the product mix reflects the shift offshore of much of Japan’s production of conventional, more low-tech consumer products. Consumer product production left in Japan tends to concentrate on more high-tech products, including newly emerging products such as wide screen televisions and game applications. These types of high-tech products tend to require more MOS LSI’s rather than analog and discrete devices.
*JAPANESE SEMICONDUCTOR MARKET SIZE CHANGES
From the first half of FY 1993 to the first half of FY 1994, the Japanese semiconductor market increased by U.S. $1 billion. [View Chart]
In most of the product categories there were increases, but there were declines in consumer ICs reflecting structural changes in the consumer market.
*FORECAST OF FUTURE TRENDS IN JAPANESE SEMICONDUCTOR MARKET
Foreign semiconductor products with a very high share of their particular market segment are widely recognized by Japanese user companies. For these products, Japanese users have high expectations that the foreign suppliers of these products will be able to meet the users’ increased demand for these products in the future.
In certain other product categories, foreign products enjoy comparable recognition among Japanese user companies, despite the fact that their market share is less dominant than those described above. For these latter types of products, Japanese users have similar expectations that foreign manufacturers will be able to accommodate their supply requirements as demand increases in the coming years. [View Chart]
*JAPAN CHIP MARKET: FOREIGN SHARE UP
Tokyo - One year after the new U.S.-Japan trade agreement, foreign market share for computer chips in Japan rose to a record 32.6% during the first quarter of 1997, the U.S. Semiconductor Industry Association
“We are doing a better job of reaching customers in Japan because of the cooperative market access activities we established with Japan under the umbrella of the trade accord,” said George Sealise, president of the association, in a statement.
The semiconductor association said Japanese consumers are also buying more and more personal computers, a development that benefits U.S. and other foreign chip makers who design microprocessors, microperipherals and memory chips for computers. U.S. chip makers are the world leaders in designing chips.
The U.S.-Japan trade agreement is designed to ensure continued progress in opening the Japanese market. The industry association said foreign market share was 14.6% five years ago.Meanwhile, Japan’s producers of chip-making equipment said that in
order to fill increased demand from U.S. and Taiwanese chip manufacturers, they plan to expand production.Orders for chip-making gear were down on a year-to-year basis from last summer to January, as Japanese and South Korean makers reduced
capital investment due to the collapse of the market for 16-megabit DRAMs.
Japan’s largest producer of steppers, Nikon Corp., said this week it will spend about 8 billion yen, or about $67 million, to increase capacity at one of the company’s plants to 500 systems from 200 per year.Canon Inc. will also increase stepper capacity by 20% at one of the firm’s plants by year end.
*GLOBAL INCREASE IN CHIP SALES REPORTED IN JULY
The Semiconductor Industry Association announced a 13% increase in worldwide sales of computer chips in July 1997. Sales rose to $11.6 billion compared to $10.2 billion in sales recorded for the same month last year.
The United States, after months of strong shipments, saw the largest jump in sales up nearly 20% to $3.8 billion. Notable increases were also seen in Asian-Pacific and European markets.
The North American semiconductor equipment industry posted a book-to-bill ratio of 1.11 for August. This ratio, determined over a 3 month time span, suggests a rising demand in the industry. Electronics exports represent 22% of all US exports.
*SEMICONDUCTOR INDUSTRY UPDATE
GLOBAL SEMICONDUCTOR SALES ROSE 4.5 % IN DECEMBER
Wall Street Journal
February 12, 1998
Worldwide semiconductor sales rose 4.5% in December to $11.48 billion, ringing up a total of $137.2 billion for 1997, according to the Semiconductor Industry Association.
The 1997 total represents a 4% increase over the 1996 total of $131.97 billion and is slightly below an October forecast by the trade group of 5.5% growth for the year. Some of the disparity stems from depressed prices in DRAM memory chips and recent economic uncertainty in Asia, the group indicated in a press release.The 1997 sales total is the second highest on record, after 1995 peak sales of $144 billion, the SIA said.Dataquest, a market-research firm also based in San Jose, said that the Asian crisis would restrict growth in worldwide semiconductor sales to around 7% in 1998, down from a previously forecast 17%. Its new estimate of the 1998 global chip market was revised to $160 billion, down from $175 billion.
*GLOBAL CHIP SALES IN ‘97 UP 5.5% TO $150 BILLION
Investor’s Business DailyJanuary 6, 1998
Researcher Dataquest says Intel remained the leader, with semiconductor sales up 19% to $21 billion. NEC held the No. 2 spot at $10.6 billion; followed by Motorola at $8 billion. The currency crisis in Asia exacerbated low memory-chip pricing. Sales at Hitachi, the No. 4 chip supplier in 1996 fell 20% as it dropped to No. 6. Toshiba held No.
5 though sales fell 7%. No. 7 Samsung’s also fell 7%. [View Chart]
January 20, 1998
Leading Japanese semiconductor manufacturers are scaling back production and capital-spending plans because of a drop in the price of computer chips and the economic slowdown in Japan. In a highly unusual move, Hitachi will cut production by 20 percent at eight domestic plants in an attempt to stem losses.
l EIAJ News Releases．EIAJ Semiconductor Policy Papers．
EIAJ-SIA Agreements on Semiconductors．Commentary on Semiconductor Issues
Semiconductor Studies Facts & Figures
A DRI/McGraw-Hill Study
THE GLOBALIZATION OF THE SEMICONDUCTOR INDUSTRY
TABLE OF CONTENTS
2.Summary of Key Findings
3.Discussion of Findings In-Detail
trade and sales of semiconductors are a function of demand, and the demand end-markets for semiconductors vary considerably across regions. single measure of the international semiconductor market tells the whole story. Globalization is reality in the international semiconductor industry,
With a significant shift of production and assembly by U.S. and Japanese Firms outside of their respective countries. international alliances among semiconductor companies have become the norm. The largest portion of these alliances are between Japanese and
5.Download relevant sections of the full text of DRI’s study that apply to the above keypoints.
The last decade saw the semiconductor industry go through a period of remarkable and far-reaching change. This vitally important, $100 billion industry is shifting from a vertically integrated, monolithic, single nation-based business to a horizontally integrated, cooperative industry based on multi-national partnerships and alliances. According to World Semiconductor Trade Statistics (WSTS), in 1995, the world semiconductor industry generated sales of approximately $144 billion. The market is growing rapidly, at about 15% annually. Sales in 1996 are expected to total some $154 billion, and to reach $270 billion by the year 2000. This growth in semiconductor sales and trade is the result of growth in demand for existing products, and the development of new technologies, applications, products and markets. This study of the international semiconductor industry explores the outlook for the industry by closely examining its current trends and prospects for the future. In the study, we look at the key factors affecting the trade, growth, and competitiveness of the industry. We also closely examine developments at the micro, or business level. As the basis for the study
we draw upon published information from outside research organizations, such as the World Semiconductor Trade Statistics, Dataquest, Elsevier’s, and INSEAD, as well as DRI/McGraw-Hill sources.
SUMMARY OF KEY FINDINGS
semiconductor trade and sales are a function of demand, and the demand end-markets for semiconductors vary considerably across regions. How well certain national industries or specific businesses perform in a particular market depends in large measure on the specific demand for semiconductors within that market. This is true for virtually all products, and is especially true for semiconductors. Indeed, it is
somewhat misleading to talk of a monolithic semiconductor industry, since the industry is composed of many different sectors, from low-end, simple commodity semiconductors, to high-end application specific semiconductors. Each sector has its own demand and business characteristics.For most world regions, semiconductor trade and sales are driven by the demand for computer applications, which include microprocessors and
memories. Only in Japan has the consumer products end-use segment consistently rivaled or led computers in consumption of semiconductor products. But recent trends have the Japanese market moving away from consumer-oriented analog products, towards digital electronics and computer products. Different demand structures in the world are an important factor in the current sales and exports markets, as well as
their future growth.single measure of the international semiconductor market tells the
whole story.There has been much discussion among trade and policy experts about shares or sales percentages of one industry or country in another country’s market. One of the many difficulties in using such a simple market share analysis is that it can distort what is really happening in the semiconductor industry.There are a number of ways to look at the international marketplace, including: global sales, exports by country of origin, and exports by capital affiliation. Each paint a different, but incomplete picture of the marketplace. These simple sets of data are further complicated by the complex and dispersed organizations of production within each company, and the growing inter-relationships between firms of all nationalities at different stages of production. globalization is reality in the international semiconductor industry, with a significant shift of production and assembly by U.S. and Japanese
firms moving outside of their respective countries.An examination of trade by ownership and by country of origin, together with an examination of the trade flows of semiconductor wafers, parts, and finished semiconductors, evidences the significant shift of investment and production of semiconductors to countries outside the U.S. and Japan, to other locations in Asia.In fact, U.S. firms have been more aggressive than other semiconductor companies in moving production facilities offshore in search of lower labor and facility costs, mainly at the finishing stages of production.international alliances among semiconductor companies have become the norm—a requirement for survival and growth in today’s competitive international marketplace. The largest portion of these alliances are between Japanese and U.S. firms.The semiconductor industry is one of the most integrated and international industries in the world. The number of alliances between companies in this industry has increased significantly throughout the 1980s and the early 1990s.
Regardless of national policies, this trend will likely continue due to:
1.increased development costs and the need to share risk;
2.advances in technology and the need to swap technological and production expertise;
3.the need to respond to new applications and consumer demand; and
4.the need to take advantage of production economies, as well as marketing and distribution networks.
This trend is led by Japanese and U.S. companies favoring alliances with
each other more than with any other economic region.the trend towards continued globalization of the semiconductor industry is most likely irreversible.The technology, investment, and marketing demands of the semiconductor industry will require continued cooperation among individual businesses and industries both in Japan and the United States, as well as from the emerging industries in Southeast Asia, and Europe. Moreover, the continued dynamic evolution and rapid expansion of the international semiconductor industry will provide new products for consumers and new
market opportunities for manufacturers across all geographic regions and product lines for years to come.
*DISCUSSION OF FINDINGS IN-DETAIL
1. Trade and sales of semiconductors are a function of demand, and the demand end-markets for semiconductors vary considerably across regions. Any examination of the globalization of the semiconductor industry must begin with a study of semiconductors and what they are used for.Not surprisingly, the end-use markets for semiconductors vary considerably across regions. This is because the demand for various types of semiconductors differs dramatically from one country or region to the next. This, in turn, is reflected in the uneven degrees of penetration of semiconductors in given applications in different regions of the world. It also points to the fact that the potential for future growth of a number of market segments is still very large. The Japanese end-use market stands out as somewhat unique, with a strong consumer goods segment which has traditionally challenged computer sales. In other regions of the world, it is the computer market which is the largest user of semiconductors, producing approximately 62% of sales in the US, 41% in Europe, and 52% in the Asia/Pacific region. Recent trends, however, have the Japanese market moving away from consumer- oriented analog products, towards digital electronics and computer products. This variation of semiconductor end-uses helps to explain the differences in market penetration of some companies in different geographic regions.
Chart 1 illustrates the major categories of semiconductors used in the major world markets. Six product groups are distinguished: microprocessors, logic, memory, bipolar digital, analog and discrete. (View Chart 1)
Chart 2 illustrates the six key semiconductor end-use categories: consumer goods, computers, communications, industrial uses, automobile and a blanket category for other uses (for example, transportation). (View Chart 2)
The high-end, computer-based segments are the fastest growing segments in all regions, including Japan. The continued evolution of end-use demand markets will drive, in large measure, the market penetration and share growth of semiconductor suppliers across regions.
2. No single measure of the international semiconductor market tells the whole story.
There are different ways of measuring size and shares of semiconductor markets, none of which is perfect. One measure of total market size is total world exports for semiconductors; $105 billion in 1995. This measure, however, excludes the sales of semiconductors within the market in which they are produced. For example, exports would not include U.S. made semiconductor sales in the U.S. or Malaysian made semiconductors sold to other companies in Malaysia.
Another commonly used measure is the total value of the industry’s sales, which was $144 billion in 1995, according to WSTS. The sales-based measure, however, does not take into account the captive market for semiconductors, and is thus also short of providing an accurate picture of the total market value.None of these measures is thus appropriate by itself to understand the complexity of the world semiconductor industry or assess the relative weight of different countries or regions in this market.
3. Globalization is reality in the international semiconductor industry, with a significant shift of production and assembly by U.S. and Japanese firms outside of their respective countries.
The industry has indeed undergone a major globalization pattern over the past 10-15 years, which has been driven by technological change, competitive pressures and changes in demand (both in terms of product mix and geography). This globalization trend, which is reflected in the continued long-term growth in foreign investment and changes in international sourcing, and in the extensive formation of alliances and other international links between firms and countries, has now taken such proportions that the process is practically irreversible. Very different conclusions can thus be drawn based on whether the relative importance of individual countries or regions in the world
semiconductor market is measured based on company ownership, or based on export shares. Any assessment of the relative importance of individual countries or regions in the total world semiconductor market is further complicated by the complex organizations of production within each company and the growing inter-relationships between firms of all nationalities at the different stages of production. In past few years,
the inter-relationships between firms have indeed grown to such an extent that it becomes totally irrelevant to use a single measure to assess market share or the contribution of different countries/regions to the global market.An examination of all of the data points to this significant movement of production and assembly across world regions. Charts 3A and 3B illustrate this by comparing the percentage of ownership shares with world trade share in the world market by region from 1993 to 1995. They
show that Asia-Pacific exports have been growing in importance, while in North America, the gap between ownership of items traded and trade figures has broadened. In the period examined, for example, US and Japanese-based firms each had approximately 40 percent of total world sales.In the beginning of the industry’s development, market shares measures based on trade or ownership shares were similar. The increased globalization of the industry and manufacturers’ global sourcing and
relocation strategies, however, have created a growing gap between these two indicators. (View Chart 3A; View Chart 3B)
In 1994, Asian exports represented 36% of world semiconductor exports. European and Japanese exports followed in a close tie for second place, by each representing above 20% of the world semiconductor trade in 1994. The U.S. was fourth with 17.2% of the world trade that year. However, these figures only give a partial view of what is happening. For example, Europe’s large share of world exports mainly reflects the importance of intra-European Union trade. The same holds true for Asia, as this region’s exports to other Asian partners represent 36% of its integrated circuit exports, 47% of its discrete exports, and 65% of its parts exports. These figures increase to above
50% when Asia is expanded to include Japan. What is most surprising, however, is the comparatively low U.S. share of world exports, which, at about 16%, is barely half its market share based on company ownership, the latter being estimated at 32% in 1994. This suggests that a very high share of U.S. production is taking place outside the U.S. - mainly in Asia.
Table 1 compares the market shares of different world regions based on ownership and on export market shares respectively.
A Comparison of 1994 World Semiconductor
Trade Shares Region by Ownership by Country of Origin of Exports Differential
North America 32% 18% 14%
Europe 20% 22% -2%
Japan 29% 22% 7%
Asia/Pacific 19% 38% -19%
The two regions showing the greatest difference are, in fact, North America and Asia/Pacific. To a large extent, these are mirror images of the same phenomenon: if we assume that the data on nationality of ownership are correct, then Asia/Pacific companies supplying semiconductors account for just 19% of the more than 38% of finished semiconductors sold worldwide. The differential, about 19%, can be compared to that of North America, where ownership share represents 32% of world trade and exports only 18%.Indeed, based on the country of origin and corporate nationality data, the story being told here is that the U.S. firms have been more aggressive than other semiconductor companies in moving production facilities offshore in search of lower labor and facility costs, mainly at the finishing stages of production. While this contributes to job creation and economic growth in the offshore region, it may not
necessarily affect the job creation or economic growth prospects of the country where the owner companies are located. For example, U.S. exports of semiconductor wafers and parts (estimated at $6 billion in 1994, or 43% of the U.S. total semiconductor products exports) still represent a sizable share of the world exports of parts. Moreover, 78% of these U.S. exports are destined to facilities in Southeast Asia and the NIEs. In
comparison, Japan’s exports of wafers and parts are smaller ($2.8 billion in 1994, or 20% of the country’s total semiconductor products exports), with 67% of these exports were destined to this region in 1994.
A Comparison of U.S. & Japanese Semiconductor Exports
Trade Flows United States (1994) U.S. $ billions Japan (1994) U.S. $ billions
Exports10.818.4 Net Parts4.73.8
Imports23.75.9 Parts Share of Exports43%20%
Source: WSTS, DRI/McGraw-Hill
Table 2 highlights the relative importance of parts exports for the US and Japan. Together, these two countries represent the majority of parts exports (48% and 34% respectively). Again, an examination of all of the data sources suggests that the U.S. and Japan are exporting parts to be used in offshore labor-intensive manufacturing (some of which may be owned by companies from those countries) and provides a partial
explanation for the ownership/trade data differences shown in Table 1.
4. International alliances among semiconductor companies have become the norm. The largest portion of these alliances are between Japanese and U.S. firms.Alliances are being formed at all stages of the product development process, from R&D and conception to production and distribution. The majority of these are pre-competitive in nature, i.e. formed at the product development stage. Table 3 lists the top 10 companies ranked by activity in strategic alliances. It should be noted that five of the 10 were positioned in the top 10 of the most active in U.S. patent awards in technology in 1994, reinforcing the importance of alliances for technological advance. This data also reveals that U.S. companies have led in the formation of these strategic alliances with partners in different geographic regions.
Top 10 Companies Most Active in Strategic Alliances
Alliances Semiconductor Revenue
1993/1994 ($M)Corporate Revenue
IBM Corporation (1,298 Patents) (US)312,510/2,885 62.71/63.5
AT&T (US)241,110/1,307 67.15/72.44
Hewlett-Packard (US)17469/51220. 31/24.99
Siemens (Germany) 171,501/2,075DM 81.6/DM84.7
Motorola (838 Patents) (US)155,957/7,23716.96/20.78
Toshiba (983 Patents) (Japan)135,727/7,52743.04/NA
Matsushita (788 Patents) (Japan)132,344/2,88761.56/NA
Texas Instruments (US)124,083/5,2808.52/9.90
Samsung (Korea) 103,044/4,893 8.68/12.5
Hitachi (999 Patents) (Japan)85,015/6,48568.78/NANA = Not available
Source: Dataquest (January 1995)
The increase in global alliances can be attributed to a few key factors. These factors include a move toward specialization, the growing cost of R&D, changes in the geographic distribution of production linked to the different rates of growth of demand across regions, and to the companies’ search for lower cost production bases. Additionally, semiconductor companies have looked for technological complements in areas where they had limited expertise, and have sought to gain access to partners’ specific strengths in resources, such as products, engineering, technology, finance, customer base, sales forces, and/or distribution. Movement toward specialization has been a key factor leading to industry cooperation. Different regions have become increasingly specialized in the production of different types of semiconductor devices, and in different stages of the production process. Chart 4 illustrates the reasons given
for semiconductor industry alliances in the 1990-1992 time frame. For example, U.S. companies lead in production of microcomponents (microprocessors and microcontrollers) and ASICs (Application Specific Integrated Circuits) in the logic family of products, while Japanese companies are strong in memory devices such as SRAM and DRAM (Static and Random Access Memory), and devices geared to consumer-oriented products. (View Chart 4)
This specialization has lead to an interesting breakdown of reasons given for alliances. A survey of Japanese companies, illustrated in Chart 5, indicates that over 60% of alliances forged by Japanese companies in emerging Asia were for the purposes of production cooperation, while approximately 60% of alliances between Japan-U.S. and
Japan-Europe were for development (R&D and technology exchange) reasons. U.S. and Japanese alliances focused almost equally on technology
exchange, R&D, and production. (View Chart 5)
This pattern is further confirmed when looking at the reasons Japanese companies listed for partner selection. Chart 6 shows that the majority of U.S. and European partners were chosen for their superior technology and/or products, while Asian partners were chosen for their production capabilities and/or to help share the financial burden of the production facility investment. (View Chart 6)
Chart 7 reveals the global breakdown of alliances in the semiconductor industry, with 84% of alliances between companies of different nationalities. (View Chart 7)A growing share of Japanese and U.S. alliances are international in scope, i.e. with companies from different countries. Japanese companies have more alliances with U.S. companies than with any other economic region, and U.S. companies have more alliances with Japanese companies than any other region. This partly but not solely reflects the comparative size and number of semiconductor producers within these economic regions. Based on this survey, 69% of all the alliances between Japanese and foreign companies were between Japan-U.S., 15% Japan-Europe and 13% Japan-Emerging Asia. The number of alliances with emerging Asia is, however, set to grow, as the industry in this region develops further. Indeed, whereas 25% of all alliances reported in the INSEAD
database involved companies from the rest of the world - mainly emerging Asia, the Japanese survey reports that over the last 2 years, 43% of the alliances made by Japanese companies were with companies from other Asian countries.The cost of semiconductor R&D has grown dramatically.Cooperation and the pooling of financial resources are essential to defray the huge and permanently growing capital and technological development costs for the leaders of the semiconductor industry. For example, in the 1970’s, total R&D and capital expenditures did not exceed 30 percent of sales. By 1992, this figure had climbed to 40 percent, divided equally between R&D and capital expense. Table 4 shows the growing cost of R&D in the development of DRAM since 1985.
Development of DRAMCommercial year
Cost of R&D ($ million)
256 Kbit 1985110
1 Mbit 1988250
16 Mbit 199485064,
256 Mbit 199510001
Source: SCRL Report, DRI/McGraw-Hill, Company Reports
The changing and increasing demand for semiconductors is also affecting the formation of alliances.Another reason alliances will continue to develop in future years is that in order to maintain market share companies must permanently adjust to changing customer’s needs and identify new applications for products. As the main markets for conventional consumer electronic products, such as stereos and VCRs, have become saturated, consumer electronic companies have seen their traditional customer base shift attention and money to information technology, equipment and multimedia. With the development of the multimedia/information sector, products become ever more sophisticated. Table 5 is the forecast for selected semiconductor product sales through to 1998.
Worldwide Sales and Forecast for Selected Semiconductor Products
(U.S. $ million)
1993 1994 1995 1996 1997 1998
MOS Microprocessors 8,539.710,995.514,278.6
MOS Microcomponents, Total19,071.523,820.1
MOS Memory, Total21,266.932,450.353,457.953,640.0
55,093.363,384.3Source: WSTS April, 1996
In addition to supplier-supplier cooperative ventures, the industry is now witnessing a proliferation of global design-in projects between suppliers and users of different nationalities. Design-ins, or the practice of designing semiconductor devices for specific commercial products, will become increasingly common as the technology becomes more
cost effective.The growth of applications in the information technology sphere will increase the number of uses and demand for semiconductors and, as such, offers a potential new range of partnerships and design-in projects.Competitive market pressures will also force companies to focus on core competencies and buy-in other services.Another indication that the world market for semiconductors is becoming increasingly integrated is that even the huge, once vertically integrated Japanese semiconductor firms are shifting paradigms. Instead of doing everything from building processing equipment to manufacturing, packaging and shipping chips themselves, they are finding it can be more cost-efficient to buy equipment and services from outside suppliers. Japanese supplier companies are increasingly crossing traditional lines to sell to (and buy from) former competitors. Outsourcing decisions for operational functions are increasingly based on cost-savings as opposed to proprietary concerns. This process has already gone so far that changing regulatory market access conditions will likely not reverse this pattern towards internationalization of production and outsourcing.The fastest growth has been in High-value products in Asia and Japan.
The fastest growth sector in the industry is in high-end, high-value products. Simple semiconductor products such as discretes will continue to decline in importance relative to the growth rates of integrated circuit exports. By 2010, integrated circuits will represent a full 77.6% of the semiconductor trade, up from about 74% in 1995 (as seen below in Table 7). This growth is expected to occur at the expense of parts exports, which will represent less than 10% of the semiconductor trade by 2010.
Distribution of world exports by type
Shares (total in millions)
1989 1995 2000 2005 2010
Total(in million $’s)54,480140,136251,950434,771725,839
The fastest growing supplier and consumer of semiconductors is not the U.S. or Japan, but Asia. By 2010, it will represent above 40% of semiconductor exports, except in the parts category. In Asia, the newly-industrialized countries (NICs) are expected to export 25% of the semiconductors world wide, with the Southeast Asian countries handling
another 18%. Asian export growth will be especially fast in its exports of integrated circuits.Asia will also show the biggest growth in demand for semiconductors. Based on normal developments of trade following the changing patterns in demand and competitiveness, 54% of the North American and 61% of Japanese exports of integrated circuits will go to Asia in 2010. Of Asian imports in 2010, almost 47% will be expected to originate within that region.
The worldwide semiconductor industry is made of many diverse components, which are designed and manufactured, assembled and distributed through an international network of companies and countries. Characterization of the semiconductor industry is complicated by this diversity, as well as by other business strategy considerations. No single measure of sales or production ownership of the market can effectively translate into an accurate understanding of the industry. However, from our study we can draw the following conclusions: there has been and will continue to be strong growth in the production, sales and trade in semiconductors well into the next century. That growth is a result of the continuing increase in demand for existing products, and the development of new technologies, applications, products and markets. personal computers, with 48 percent of world sales, currently represent the largest single end-use of semiconductors. As such, it may be said with good reason that as the personal computer business goes, so goes the semiconductor industry. Computers represent by far and away the largest end-use segment in every regional sector of the world, except Japan. Only in Japan has the consumer products end-use segment consistently rivaled or led computers in consumption of semiconductor products. However, Japan is currently experiencing a shift in demand from analog-based consumer electronics to digital information technology applications. there are three primary categories of internationally traded products: integrated circuits, discrete devices and parts (including semiconductor wafers). Based on an examination of the trade flows of these three categories, U.S. based exports in the parts and wafer category is very large (2/3 the size of U.S. integrated circuit exports), the primary destinations being Taiwan, Malaysia and Singapore. There has also been a shift to Asia/Pacific of the manufacturing and export of products that rely heavily on semiconductors. The shift in production and assembly of semiconductors is paralleled by a shift in the production and assembly of products that use semiconductors. with U.S. companies’ leadership in sales of microcomponents, and Japanese companies having similar strengths in memory, an implicit dependence or specialization seems to have developed between the two on
the manufacturing and technology of those families of semiconductor products. That specialization is helped along by the high cost of entry into new technological markets with the price of building leading edge fabrication facilities, exceeding $1 billion per fabrication plant.the international semiconductor market is dominated by the strategic
behavior of large companies, rapid technological change and the shortening life-cycles of products. In the past five years, business alliances in the semiconductor industry have been used increasingly by companies to advance their competitive advantage. Semiconductor companies look for technological complements in areas where they have
limited expertise, and seek to gain access to partners’ specific strengths in resources, technology, or logistics. These alliances allow the partners to obtain reciprocal benefits from the combined use of complementary assets.strategic concerns, such as control of labor costs, avoidance of trade barriers, currency fluctuations, and maturing of certain industry segments, have sent phases of manufacturing to sites overseas from the U.S. and Japan. U.S. companies appear to be the most active in moving final production and assembly off-shore, particularly to the Asia/Pacific region.semiconductor industry manufacturer-supplier relationships in general are reacting to competitive and financial pressures. The Japanese model of cooperative business groups is evolving as it spreads to other areas of the world. Cost-control imperatives and global market realities are also working to open domestic alliances in Japan to outside companies.specialization is expected to continue for the immediate future, with U.S. companies capitalizing on strengths in high-end products such as microprocessors, microcontrollers and logic ICs. Japanese company strengths in SRAMs, DRAMs, and research into x-ray lithography
techniques continues to bode well for them in miniaturization, ultra thin packaging, and emerging consumer-oriented technologies like “smart” cards. South Korean spending to produce DRAM manufacturing capacity may spin off into technological gains, but at the moment, they are too far behind and too focused on memory products to challenge the overall lead of companies in the U.S. and Japan.The trend towards globalization of the semiconductor industry is most likely irreversible. The technology, investment, and marketing demands of the semiconductor industry will require continued cooperation among individual businesses and industries both in Japan and the United States, as well as from the emerging industries in Southeast Asia, and Europe. Moreover, the continued dynamic evolution and rapid expansion of the international semiconductor industry will provide new products for consumers and new market opportunities for manufacturers across all geographic regions and product lines for years to come.
If you would like to view relevant sections of the full text of DRI’s study that apply to the above keypoints, you can download them to your local hard drive by selecting one of the following:
Download DOS text file
Download Microsoft Word windows file
You can reach us directly at firstname.lastname@example.org
 EIAJ Semiconductor Policy Papers EIAJ-SIA Agreements on Semiconductors
Commentary on Semiconductor Issues