The path the United States has taken to protect its semiconductor industry over the years

Latest update time：2019-11-23

Reads：

Source: The content comes from "Guotai Junan Securities ", thank you.

The technology industry represented by semiconductors is a key focus of the competition between China and the United States.

Long before the Sino-US trade friction became public, the United States launched the Congressional Semiconductor Core Meeting to study semiconductor industry policy. In 2016 and early 2017, the Congressional Research Service and the President's Science and Technology Advisory Committee successively released two reports proposing the need to curb the development of China's semiconductor industry, and coincidentally proposed policy recommendations to create a better environment for the development of the US semiconductor industry.

Looking back at history, the "visible hand", that is, the U.S. government, as the largest venture investor and strategic designer in the semiconductor industry, has provided long-term impetus and an excellent growth environment for the development of the semiconductor industry.

Sorting out the key factors in the development of the U.S. semiconductor industry from a historical perspective is very important for us to understand the independent development model of the semiconductor industry and the industrial and technological response strategies under the Sino-U.S. trade war.

Long before the Sino-US trade friction became public, the United States launched the Congressional Semiconductor Core Meeting to specifically study semiconductor industry policy. The United States is very sensitive to the fluctuations of core competitiveness in this field. Just like the previous threat from Japan, the United States believes that when its semiconductor industry is threatened by China, it should maintain its leading position through proactive government intervention. In 2016 and early 2017, the Congressional Research Service and the President's Science and Technology Advisory Committee successively released "Semiconductor manufacturing in America: industry trend, international competition and federal policy" and "Report to the President: Ensuring Long -Term US Leadership in Semiconductors".

In addition to proposing the need to curb the development of China's semiconductor industry, these two reports also put forward policy recommendations to create a better environment for the development of the US semiconductor industry:

1. Continue to invest in science and continue to attract talent from around the world;

2. The United States needs to increase investment in advanced technologies and must reverse the trend of stagnant growth in basic research and development funding in recent years;

3. Promote tax reform. The Research and Development Act passed by the United States in 1978 and 1982 stipulates that the profit portion of enterprises used for R&D is not taxed, and the S Amendment Act stipulates that knowledge-intensive companies can pay 1/3 less tax. The US government needs to give more tax incentives to companies engaged in high-tech R&D;

4. Increase investment in infrastructure to reduce the cost required for the semiconductor industry to achieve economies of scale and technological innovation.

In terms of foreign investment restrictions, it is planned to restrict China's investment in semiconductors, 5G and other fields in the United States. The U.S. Treasury Department intends to formulate these restrictions into a bill and expand the powers of the Committee on Foreign Investment in the United States (CFIUS) to set more stringent conditions for Chinese investment acquisitions. Among them, some proposals have been reflected in the "Foreign Investment Risk Assessment Modernization Act" proposed by the U.S. Congress. In the past two years, the U.S. government has vetoed a series of acquisitions of U.S. semiconductor companies by Chinese semiconductor companies (Table 1). In the semiconductor field, in the first quarter of 2018, CFIUS approved only one acquisition by a Chinese company - the acquisition of Akrion Systems, a U.S. semiconductor equipment manufacturer, by North Huachuang Microelectronics, and the transaction was worth only US$15 million.

ZTE paid a huge price to lift the ban, and China showed a more restrained strategy on this incident. After repeated mediation, the Trump administration's Department of Commerce agreed to lift the seven-year export ban on the condition that ZTE paid a total fine of US$1.7 billion, and the National Defense Authorization Act was not passed by Congress.

Sorting out the key elements in the development of the U.S. semiconductor industry from a historical perspective is crucial for us to understand the semiconductor industry's independent development model and the industry and technological response strategies under the Sino-U.S. trade war.

Having experienced the US-Japan conflict in the 1980s, the United States can be said to be repeating the same old tricks this time, using trade protection policies and foreign investment restrictions to suppress the development of China's high-tech fields, led by the semiconductor industry. my country's semiconductor industry started a short time ago and is still in its infancy. The United States has shown its targeting of my country's core industries in the trade war, pushing China to a stage where it must independently develop the integrated circuit and semiconductor industries. From the experience of the United States, we can see that government industrial policies play a vital role in industrial development. China should design semiconductor industry policies suitable for China's development based on the characteristics of the semiconductor industry, and strengthen China's semiconductor industry's competitive position in terms of funding, technology advancement, industrial structure and organization.

1. In the infancy of the industry, the U.S. government was the largest venture investor in the semiconductor industry

Looking back at history, the "visible hand", that is, the U.S. government as the largest venture investor and strategic designer in the development of the semiconductor industry, its intervention in the semiconductor industry has truly provided it with long-term development momentum and an excellent growth environment.

Since the 1950s, the U.S. semiconductor industry has gone through various stages of industry start-up, development, and globalization, and still maintains its position as the world's leading industry. Looking at its development history, we can find that the U.S. government is not just a fixer in the "market failure" process of industry evolution. In different periods of the development of the semiconductor industry, the U.S. government has actively intervened and made strategic adjustments from multiple angles, including funding, technology promotion, industrial structure, and organizational transformation.

On the one hand, as the largest venture investor in the semiconductor industry, the U.S. government has driven private capital R&D and investment through direct procurement expenditures and semiconductor R&D funding. On the other hand, the U.S. government has encouraged the formation of a competitive industrial structure during the period of rapid industrial growth. More importantly, in the face of international competition, it has actively intervened in industrial restructuring and used trade policies to suppress foreign competitors.

1. Government procurement provides “demand pull” for the U.S. semiconductor industry in its infancy

After World War II, semiconductors became a top priority in the U.S. industrial development system. After World War II, facing the Soviet Union's military and technological competition, all sectors of the U.S. believed that the U.S.'s technological advantage and even "national security" were threatened. In 1957, the Soviet Union launched the first artificial satellite, which shocked the U.S. government and the public. The development of cutting-edge weapon systems and aerospace technology has made it necessary to miniaturize and lighten the electronic components used in military and aerospace equipment.

In the initial development stage in the 1950s, private capital lacked the motivation to invest heavily in the semiconductor industry, and the U.S. government invested in the semiconductor industry at almost any cost.

The R&D costs and risks of semiconductor solid-state electronic devices such as transistors, microprocessor chips, and dynamic random access memory are extremely high, with long investment cycles and fast technological updates. Famous semiconductor R&D institutions and companies, such as AT&T's Bell Labs and Texas Instruments, received a large number of procurement contracts from the US military during this period. Take Texas Instruments as an example. In 1954, the silicon junction transistor it developed was favored by the US military and purchased on a large scale for use in radar and missile systems. Its epoch-making invention, the integrated circuit, received a large number of orders from the military, laying the foundation for its subsequent development.

From 1955 to 1970, the U.S. government accounted for more than 40% of all semiconductor equipment shipments. U.S. semiconductor equipment shipments increased from almost zero to about $1.4 billion. Before 1962, the government's semiconductor equipment procurement accounted for more than 40%, and the highest reached about 50%. In the mid-to-late 1960s, with the widespread opening of civilian and commercial applications, the semiconductor industry entered a period of rapid growth. Government procurement was no longer the main source of market demand, and the U.S. government's procurement of semiconductor equipment remained stable at more than 20%.

In the early stages of integrated circuit development, the expansion of the market size and the creation of demand were almost entirely dependent on government procurement support.

After the advent of integrated circuits, private enterprises were hesitant due to the high risks. For example, in 1960, IBM finally decided to abandon the use of monolithic integrated circuits as the core components of computers. On the contrary, the US government chose integrated circuits as the technical standard among various options. Integrated circuit components are widely used in the navigation computer of the Apollo spacecraft and the guidance system of the improved Minuteman missile. In the early 1960s, the integrated circuit products required for the US "Apollo" moon landing program accounted for the vast majority of the total sales of such products.

At a stage when private investment is insufficient, the funds provided by large-scale government procurement can be used as "living water" for corporate research and development.

From the mid-1950s to the early 1960s, at least 70% to 80% of semiconductor companies' R&D funds came from government procurement contracts. According to statistics from the U.S. Department of Commerce in 1960, the average unit price paid by the U.S. military for purchasing electronic components and equipment in the late 1950s was about twice that of private users. For example, Fairchild Semiconductor was in a liquidity crisis in developing new planar transistor technology. The Air Force's "Minuteman Missile Program" signed a $1.5 million silicon material transistor supply contract with it, which greatly alleviated its tight R&D funding situation.

The industrial demand created by the government allows semiconductor companies to quickly strengthen the learning effect of the learning curve. The industry transforms exogenous resources into endogenous dynamic comparative advantages, giving industrial development sustained momentum.

Government procurement sets higher performance standards for equipment than the general level and requires that the requirements be met in a short period of time, which accelerates the pace of technological updates. In other words, government support allows companies to quickly move down the learning curve, achieve production scale effects, and reduce product costs. By reducing costs, products have the prospect of large-scale commercial use, embarking on a spiral path of industrial development with self-research and development, market expansion, and growth.

2. The government’s determination to invest in the semiconductor industry and introduce private venture capital

As the industry gradually develops and the market scale expands, more and more venture capital and companies' own funds are invested in research and development, promoting the technological upgrading of semiconductor and integrated circuit products.

The government's large-scale procurement of semiconductor and integrated circuit products and the guidance of government funds are equivalent to sending a clear signal, which makes enterprises and capital believe that the government is determined to develop the industry. Take the research and development funds for communication equipment as an example. During the booming period of the US semiconductor industry from 1965 to 1979, the proportion of research and development funds for communication equipment from outside the government increased year by year (Table 2). From the Small Business Investment Company Act of 1958 to the Credit Guarantee Act of 1993, the United States has enacted more than a dozen related laws. These laws have greatly promoted the development of venture capital and promoted the technological innovation and industrialization of semiconductors.

2. During the growth stage of the industry, government R&D funding supports technological progress

The semiconductor industry gradually entered its growth stage in the 1970s. The proportion of direct procurement by the US government gradually decreased, but it began to pay more attention to direct support for corporate R&D projects.

Compared with the infant stage, the government's direct intervention in the market has decreased. This does not mean that the government's capital investment in the semiconductor and integrated circuit industries has stopped. At this stage, the expansion of market demand and the open competition of enterprises have made it crucial for the survival of enterprises to win the competition through continuous research and development and technological progress.

According to statistics from the NSF, the amount of research funding received by the information technology (electronic engineering and computer) field from the US government has long been at the forefront of all industries. It is second only to biomedicine and is on par with the field of physics, which has long been heavily funded by the US government. After 2000, it has gradually surpassed it. Looking at the funding sources for the information technology field, in addition to the relatively high investment by NASA in the early 1970s, the research funding provided by the Department of Defense is the main source.

3. The industry is growing rapidly, and the US government relaxes regulations and encourages competition

The US semiconductor industry has entered a growth phase, while Japanese and European semiconductor manufacturers are temporarily catching up and have not yet formed sufficient competitiveness. Under such circumstances, the US government's direct intervention in the semiconductor industry has been relatively weak, mainly removing institutional barriers to full competition among companies.

1. The government actively encourages the "Second Source" strategy to accelerate the diffusion of cutting-edge technology to the market.

The US government is worried that centralized procurement of semiconductor and integrated circuit products will lead to dependence on a single supplier. Therefore, the government actively provides purchase orders to many new manufacturers. The government also actively recommends its suppliers to open up "backup supply sources". Most of the contracting companies in the government plan have signed contracts with multiple component suppliers. This has enabled cutting-edge technology to spread widely and quickly throughout the industry.

2. The implementation of antitrust laws prohibits industry monopolies, eliminates barriers for new manufacturers to enter the semiconductor industry, and stimulates the formation of a dynamic industrial structure.

In the late 1940s, Bell Labs, which was affiliated with AT&T, achieved significant research results in solid-state electronic devices. AT&T tried to maintain its monopoly in the market by selling transistors and patent certificates at high prices. In 1949, the U.S. Department of Justice launched an antitrust lawsuit, and the final result was that in 1956 AT&T was required to restrict its business activities in electronic communication services and equipment, and to sell licenses to new entrants seeking to obtain AT&T patents at a "reasonable" normal price or allow other manufacturers to use AT&T's patents in the form of patent exchanges.

3. During the rapid growth period of the US semiconductor industry, corporate spinoffs prevailed in the relaxed and competitive environment created by the government.

According to statistics, in 1971, when the number of spin-offs peaked, the ratio of spin-offs from existing companies to new companies in Silicon Valley reached 4:1. The most representative of these was the eight people who left Shockley Laboratory and were dramatically called the Fairchild "Traitorous Eight". They founded Fairchild Semiconductor and discovered "Moore's Law" for the development of integrated circuits.

In the early 1970s, the U.S. semiconductor industry formed a highly competitive industrial structure. From government intervention to deregulation, this is actually a continuous process. In fact, the formation of the highly competitive U.S. semiconductor industry is itself a direct consequence of the U.S. government's previous policies and strategic encouragement of a highly competitive industrial structure during the high-growth stage.

(1) The proportion of delivery volume of the top 4 and 8 semiconductor manufacturers in the total industry delivery volume has dropped sharply. According to data from the U.S. Department of Commerce, in 1965, the delivery volume of the top four and eight semiconductor manufacturers in the United States accounted for 69% and 91% of the total industry delivery volume; in 1972, this proportion dropped to 53% and 67%.

(2) From 1966 to 1971, the ratio of the semiconductor industry’s net income to total sales was lower than the average level of the manufacturing industry as a whole every year. From 1966 to 1971, the output of semiconductor components increased significantly, and the number of new manufacturers also reached a peak. However, the ratio of the semiconductor industry’s net income to total sales fell from 5.3% in 1966 to 2.7% in 1971. In turn, the decline in average profits prompted manufacturers to improve R&D efficiency.

(3) Technology and market leaders are rapidly iterating and updating, and the probability of retention is decreasing. For example, of the top ten manufacturers in 1955, only three were still on the list 20 years later. And by calculating the probability of retention every ten years, it can be found that the probability of retention is lower in the growth stage of an industry with intensified competition than in the start-up stage.

The 1970s was the golden age of the US semiconductor industry, with US semiconductor manufacturers firmly occupying the world's leading position in terms of shipments and market share.

4. Strengthen industry integration and adopt protective industry policies in the Japan-US trade war

Since the 1980s, American semiconductors have been gradually surpassed by Japan. Behind this is the opening of the internationalization process of semiconductor production and the transformation of industrial technology. With the discovery of "Moore's Law", the integration of semiconductor components has become increasingly higher, and the demand for backward integration and vertical integration by semiconductor manufacturers has become increasingly strong. Japanese companies are large vertically integrated companies, and in an environment where antitrust laws are flexibly applied, American companies have gradually fallen into a disadvantage in the competition. With the penetration of Japanese companies and the relocation of some semiconductor product production by American companies overseas, the shipment volume of American semiconductors has declined from about 60% of the world and has been surpassed by Japan.

The highly competitive landscape of the U.S. semiconductor industry is not enough to cope with the development trend of the semiconductor industry. The U.S. government has embarked on industry integration and adopted protective industrial policies. The highly competitive landscape in the United States has caused a sharp rise in the cost of technology development and product production in the era of large-scale integration. Throughout the 1970s, the design cost of integrated circuits increased by more than five times. The relatively low average profit level of the semiconductor industry has led to limited internal funds for manufacturers and an unfavorable external financing environment. This dilemma has greatly limited the "backward integration" capabilities of semiconductor manufacturers.

Faced with the threat of Japan's semiconductor industry, the US government has regained its "visible hand", adjusting the semiconductor industry structure internally and working to suppress the Japanese government's influence on the international competitiveness of its semiconductor industry externally. President Reagan declared that "the health and vitality of the US semiconductor industry is vital to the future competitiveness of the United States. We cannot allow this field to be harmed by unfair trade."

First, calls for relaxing antitrust laws in the United States have reached a climax, with some believing that the U.S. government must examine corporate competition behavior from the perspective of world market competition.

The inherent requirement of backward integration and vertical integration is to increase the scale of the company. The US government gradually relaxed the restrictions of antitrust laws after the 1980s. The important regulations of the US antitrust policy, the "Merger Guidelines" and the "Horizontal Merger Guidelines", were revised in 1982, 1984 and 2010. In a relatively relaxed antitrust atmosphere, the number of mergers and acquisitions in the US semiconductor industry has risen rapidly since the 1980s. For example, from 1980 to 1985, the number of mergers and acquisitions with a value of more than one million US dollars initiated by US semiconductor companies averaged about 6 per year, while from 1986 to 1990, it rose to about 18, and from 1991 to 1996, it averaged 34.

Second, American computer and semiconductor companies formed research and development alliances to expand the overall scale of scientific research and improve the efficiency of resource utilization.

In 1982, 11 chip manufacturers and computer companies in the United States raised $4 million to form the Semiconductor Research Corporation (SRC), which is dedicated to basic research in semiconductor technology. Just four years after its establishment, SRC had 36 members, and its budget soared from $4 million to $18 million. It participated in funding more than half of the research projects related to silicon materials and semiconductors in American universities. In 1987, the US government allocated $100 million to guide more than 10 semiconductor companies to form the Semiconductor Manufacturing Technology Strategic Alliance to promote development assistance, R&D cooperation, and standardization of unified technical standards among companies.

Third, the government will increase funding for semiconductor research and development, stabilize industry investment, and respond to Japan's challenges.

In the 1980s, due to the challenge of Japanese semiconductor companies, the market share of American semiconductor companies once lost its position as the world's top market. The government continuously provided R&D funds in the late 1980s. From 1986 to 1994, the proportion of semiconductor electronic equipment R&D funds from government funding increased from about 7% to 13%, providing an important guarantee for the revival of the US semiconductor industry.

Fourth, in order to suppress the influence of the Japanese government’s semiconductor industry policy, the US government has begun to adopt a highly protective trade policy.

In 1985, the Semiconductor Industry Association of the United States filed an anti-dumping lawsuit against Japanese semiconductor manufacturers under Section 301 of the Trade Act of 1974. The US government began to force the Japanese government to negotiate with the semiconductor industry. In September 1986, the result of the negotiations, the US-Japan Semiconductor Trade Agreement, officially came into effect. The main contents are: (1) Japan must further open its domestic semiconductor market and allow foreign manufacturers to enter; (2) it must not dump products in the US semiconductor market.

In 1987, the conflict between the United States and Japan escalated. Claiming that Japan did not fully comply with the US-Japan Semiconductor Trade Agreement, the US government announced that it would impose tariffs on Japanese products imported into the United States in accordance with Section 301. In order to avoid a greater conflict, Japan responded by encouraging more US semiconductor products to be exported to the United States in accordance with the agreement and raising the prices of Japanese semiconductor products sold abroad. The United States also established the International Semiconductor Cooperation Center (ISCC) and the User's Committee of Foreign Semiconductors in 1987 and 1988 respectively to monitor whether the US-Japan Semiconductor Trade Agreement was being followed.

Due to the price increase and the continued appreciation of the yen after the Plaza Accord, the international competitiveness of Japanese semiconductor companies began to decline. By 1991, the market share of Japanese semiconductor manufacturers was taken back to the top by the United States.

*Disclaimer: This article is originally written by the author. The content of the article is the author's personal opinion. Semiconductor Industry Observer reprints it only to convey a different point of view. It does not mean that Semiconductor Industry Observer agrees or supports this point of view. If you have any objections, please contact Semiconductor Industry Observer.

Today is the 2137th issue of content shared by "Semiconductor Industry Observer" for you, welcome to follow.

Latest articles about

■SiC giant, rebirth: how to predict the future?

■Apple chips may hit Qualcomm hard

■Chip cost per car: soaring to $1,000

■TSMC 2nm, important information

■Huang Renxun's latest views

■The risks of this type of chips that are promising have increased significantly!

■NPU, how to see it?

■Storage giants are abandoning DDR 4

■Intel, why?

■Nvidia will definitely be disrupted