The Barriers Behind the Semiconductor Investment Boom: "Capital" and "Talent"

The semiconductor industry differs from others in that, beyond capital driving forces, the accumulation of technology and product cycles are also insurmountable. "Simply throwing money at the problem is bound to lead to a scattered mess, with high costs for both the government and investors. Currently, some companies, after receiving support, do not earn money from the market but rely on subsidies and investments for profit. This means their pricing does not follow market principles, which harms the industry."

"The International Chess Game of Chips" Part Seven: The Flip Side of the Semiconductor Investment Boom

The "International Chess Game of Chips" series has reached its seventh installment. In the first six parts, we examined the international competitive landscape of the chip industry, as well as the experiences and current industrial dynamics of leading regions in semiconductor development, such as South Korea, Taiwan, Europe, the United States, and Japan. Additionally, we explored talent concerns in the semiconductor industry from a human resources perspective.

In this installment, we discuss capital and industrial investment funds. The semiconductor industry does not rely on abundant natural resources or convenient geographical locations. It is a highly technology- and capital-intensive industry, with "capital" and "talent" being the key drivers of its development.

Venture capital was a crucial catalyst for Silicon Valley's rise. Currently, China is leveraging policy support and substantial capital investment to accumulate technical expertise and talent, narrowing the gap with foreign counterparts. However, does the venture capital model that fueled Silicon Valley's early growth work in China today? Amid the ongoing semiconductor investment boom, we must also consider its potential long-term adverse effects on the industry. (By Li Yanxia)

On August 31, Allen Wu, Executive Chairman and CEO of Arm China, candidly stated at the JiWei Semiconductor Summit in Xiamen that from an investment perspective, the semiconductor industry was not highly regarded by investors in the past. Venture capital was not enthusiastic about the industry due to its inherent nature.

The characteristics of high investment requirements and long return cycles have long erected an "invisible barrier" for semiconductor industry investment. However, in recent years, China's semiconductor industry has experienced turbulent growth, with a new wave of investment fervor surging ahead. Three major forces—government funds represented by the National Integrated Circuit Industry Investment Fund (known as the "Big Fund") and local industrial investment funds, internal venture capital divisions of large tech companies, and private capital active in the semiconductor industry—have poured into the sector.

While capital is propelling the semiconductor industry onto a fast track, it has also raised concerns among domestic semiconductor professionals and investors about "excessive funding." Overfunding could lead to issues such as inflated project valuations, fragmentation of companies, and excessive homogeneous competition, thereby deviating from the core development model of semiconductor companies that rely on profits to continually invest in R&D.

Is the Silicon Valley Model Applicable to China Today?

When discussing the semiconductor industry, venture capital is an inseparable topic.

Huang Qing, Managing Director of Walden International, noted that the rise of Silicon Valley was closely tied to the growth of the semiconductor industry, and the startup phase of semiconductor companies heavily depended on venture capital.

George Doriot is often called the "father of venture capital." The American Research and Development Corporation (ARDC), which he founded in 1946, is considered the beginning of the non-family-style venture capital enterprise model. However, some scholars argue that the true "father of venture capital" is Arthur Rock, co-founder of Davis & Rock, established in the 1960s. Rock helped Fairchild Semiconductor become the first venture-backed company in Silicon Valley, later founding his own firm and playing key roles in the establishment of numerous Silicon Valley high-tech companies, including Teledyne, Intel, and Apple.

After signing an "agreement" on ten one-dollar bills with Rock and another banker, Bud Coyle, the "Traitorous Eight" of Fairchild submitted their resignations to William Shockley, the "father of the transistor," on September 18, 1957. This day was later named by The New York Times as one of the ten days that changed American history.

The "Traitorous Eight" pioneered the model of talent teams leaving to start ventures with venture capital support, which played a crucial role in Silicon Valley's rise.

On September 20, Chen Guanhua, a senior professor in the Department of Chemistry at the University of Hong Kong and co-founder of the tech startup platform Hong Kong X, explained that his team invested last year in a Silicon Valley startup focused on FPGAs. The two founders of the company came from Intel and Xilinx, respectively, the leading companies in the FPGA field.

"They developed a faster and more efficient new FPGA solution but were struggling at the time, nearly running out of funds," Chen said. His team of experts recognized the project's value and invested in the startup even before it generated revenue. The company has since collaborated with Samsung on 7nm chip development and attracted acquisition interest from Intel.

Sravan Kundojjala, Deputy Director of Strategy Analytics' Handset Component Technologies service, pointed out that the semiconductor industry has now reached maturity, with companies actively engaging in mergers and acquisitions to stabilize pricing and strengthen negotiating power with customers. Startups in the industry often focus on niche emerging technologies, playing a supplementary role to larger companies. "Compared to the internet and software industries, the semiconductor sector sees relatively few startups emerge."

"Companies like Intel, Samsung, Qualcomm, Google, and MediaTek have their own internal venture capital divisions," Kundojjala said. "They identify and invest in next-generation technologies, and many semiconductor companies have been acquired by these large firms."

However, Feng Jinfeng, a senior integrated circuit industry expert currently working at Shanghai Beyond Moore Fund, believes that small and medium-sized enterprises are always the most innovative, a conclusion that applies universally, including the semiconductor industry. "Therefore, in today's environment of semiconductor industrial investment funds often amounting to tens or even hundreds of billions, the 'venture capital' model remains crucial for China's semiconductor industry development," he told reporters. "Of course, this model has also evolved in China."

Feng pointed out that, on one hand, domestic industrial parks offer substantial incentives for investment, with semiconductor startups often enjoying benefits such as rent exemptions and settlement subsidies. On the other hand, these parks typically have their own investment platforms willing to invest in semiconductor companies that locate there. In his view, these are characteristics absent during Silicon Valley's rise and pose challenges to traditional venture capital.

"Outstanding domestic semiconductor venture capital institutions in China often possess expertise in one or several areas, such as professional knowledge, legal proficiency, or financial acumen for IPOs," he said. In the context of China's current push to develop the integrated circuit industry, venture capital remains an irreplaceable model.

Is There Excessive Funding in the "Trendy" Industrial Investment?

Yin Jiayin, Executive Partner of Beijing Hailin Investment Co., Ltd., noted that compared to industries like the internet, semiconductor investment requires large capital and has long return cycles. Additionally, foreign giants have established high barriers in this field, creating many uncertainties for latecomers seeking success. "Every investor hopes to achieve ideal returns within a relatively short period. In this context, it's understandable that most funds prefer short-term, quick-return projects."

After the state designated semiconductors as a strategic pillar industry, with the introduction of a series of supportive policies and the establishment of central and local official industrial funds, the development speed of China's semiconductor industry has accelerated, and the likelihood of enterprise success has significantly increased. "Any far-sighted capital would not easily pass up such an opportunity," Yin said.

Many industry insiders consider 2014 as the "first year" of rapid development for China's integrated circuit industry. That year, integrated circuits were listed as an emerging industry in the government work report for the first time. Additionally, the "National Integrated Circuit Industry Development Promotion Outline" was released, and the "Big Fund" was established. The fund operates under a corporate structure, functioning in a manner similar to venture capital.

Around 2016, local integrated circuit industry funds began to emerge. For example, the Shanghai Integrated Circuit Industry Fund was established in 2016 with a target scale of 50 billion yuan. Guangdong also announced the establishment of an integrated circuit industry investment fund in June of that year.

However, multiple semiconductor industry investors told reporters that the semiconductor industry investment, currently in a "trendy" phase, is experiencing "excessive funding."

Su Renhong, Founding Partner and CEO of Hoosea Capital, pointed out that many participants in the current wave are not funds specialized in semiconductor industry investment, which has indeed driven up project valuations. While there are still many investment opportunities in the semiconductor industry, some are hidden beneath the bubble, increasing investment difficulty.

"Most critically, if the industry becomes unhealthy, it won't generate profits, and companies cannot sustain their development," Su said. "Semiconductors are not the internet; companies must be profitable to support growth, and short-term capital returns are unlikely." Furthermore, the semiconductor market has limited capacity. Ideally, each category or niche market should be orderly, with only a few dedicated and deep-rooted companies. "It shouldn't be a rush where everyone jumps in, ultimately leading to no one making profits and the industry being ruined."

When asked about challenges constraining industry development at the JiWei Semiconductor Summit, Li Min, Chairman of Rockchip,直言不讳地 stated he feared the industry might become "overheated" like "a collective adrenaline rush." Zhang Fan, Chairman of Goodix Technology, said at the same event that the current investment fervor in China's semiconductors has led to inflated valuations for almost all projects. "Receiving such high financing or valuations at the outset may not be beneficial for a company's future development."

However, Feng Jinfeng believes that the current perception of excessive industrial funding might be an illusion. "In reality, capital is still quite tight, especially private capital," he said. "The state has recently imposed strict restrictions on bank and insurance funds investing in industrial funds. In the coming period, fundraising difficulties may become a common phenomenon."

Lin Jianhong, Research Manager at TrendForce's拓墣产业研究院, noted that Samsung, Intel, and TSMC have each had annual capital expenditures接近 $10 billion over the past three years. If China's lag in memory chips or foundry technology behind leading companies is estimated at five years, the total capital required to support a leading company in these two fields could reach $200 billion. "Moreover, industrial development is not limited to just two companies."

Kundojjala also stated that as semiconductor manufacturing processes advance, chip design is becoming more complex and challenging, requiring massive investments. For example, Qualcomm spent $27 billion on R&D between 2013 and 2017, compared to $15 billion between 2008 and 2012. NVIDIA's R&D expenditures increased 2.5 times between 2012 and 2018.

"On average, a semiconductor company needs to invest about 20% of its revenue in R&D," Kundojjala said. "Compared to other industries, the semiconductor sector's demand for capital is at a very high level."

Focusing on Industry Chain Investment

Regarding concerns about potentially inflated semiconductor project valuations, Su Renhong emphasized that the semiconductor industry differs from others; beyond capital驱动, technological accumulation and product cycles are also insurmountable. "If people approach it with an overly浮躁的心态, it could actually harm the industry."

Su believes that to avoid this situation, local governments with genuine long-term commitment, capability, and suitability for semiconductor development need to collaborate with professional semiconductor investment funds. "Simply throwing money at the problem is bound to lead to a scattered mess, with high costs for both the government and investors," he said. "Currently, some companies, after receiving support, do not earn money from the market but rely on subsidies and investments for profit. This means their pricing does not follow market principles, which harms the industry."

In his view, some companies that should have been淘汰 by the market receive support from不够专业、理性的 investors and local governments eager to develop the semiconductor industry, leading to an excess of similar companies.

"Market-oriented solutions should be applied where possible," Su said. "For China's semiconductor industry development, many high-end projects require government support, including in equipment, materials, and cutting-edge design areas. These are the non-market-driven segments where government support should primarily focus."

"Market-driven segments are not without need for government support, but such support should incorporate market experience," he added. "We should not invest in companies that generate no economic效益 and may harm the entire industry."

Chen Guanhua also pointed out that while China's venture capital industry has been "booming" in recent years, the vast majority focuses on business models. "Business models yield faster returns; once the model is set, capital is injected, and a unicorn emerges in a few years, but it's essentially built with capital," he said. "In truly high-tech industries, this is unrealistic."

In his view, Walden International is a good example of investing in high-tech industries today. "Walden has invested in many semiconductor companies, over 50% of which later went public. Of course, the return周期 can be long, sometimes up to 10 years."

Su Renhong previously worked at Walden International before leaving to establish Hoosea Capital. "The investment landscape in this industry has been consistent over the past decade; those who succeeded were funds专注 in this industry chain," he said. The semiconductor industry chain is extremely complex, involving numerous segments, with constant changes in talent and upstream/downstream supply. Unless one specializes in and has deep understanding of this chain, it is difficult to make sound investments.

"It requires investors to深耕 in this field, with profound industry understanding. Regardless of the external environment, we will follow the industry's development规律 and invest in the semiconductor industry according to our own strategy," Su said.

Three Capital Forces Complement Each Other to Form Synergy

Feng Jinfeng believes that in the context of China's push to develop the integrated circuit industry, government-led funds, tech company investment funds, and private capital have been attempting collaboration. For example, a typical overseas semiconductor M&A case often involves private capital identifying and connecting with the target, then leveraging government fund resources for the acquisition, and eventually integrating it into a large tech company at an opportune time.

"None of the three parties can replace the others; each has obvious advantages and shortcomings," Feng said. "Private capital is the most agile, government funds have the most financial power, and tech company investment divisions have the strongest底气."

In Feng's view, large tech companies' external investments often focus on upstream/downstream integration, creating entry opportunities into上游; others aim for strategic expansion of their core business.

Lin Jianhong also noted that while tech company investment funds also primarily invest in startups, unlike venture capital, their primary goal is often aligned with upstream/downstream companies related to the parent company's business. Therefore, some countries restrict the持股比例 of tech companies in their invested firms. "Additionally, when an invested company grows to a certain scale, venture capital tends to sell its stake, but tech companies might acquire it, as seen with Xilinx's acquisition of Deephi Tech."

Government industrial investment funds prioritize the invested entity's contribution to regional industrial enhancement, including industrial scale, follow-up settlement of upstream/downstream enterprises, and tax revenue. "Government industrial investment funds focus on the long-term economic development potential of the country, region, or city," Feng said. "This is also the main difference between government-led industrial investment funds and market-oriented investment funds."

Lin Jianhong believes that precisely because of their focus on upstream/downstream industrial development, government funds often invest in existing domestic leaders in specific fields, and their investment horizons are typically longer than those of venture capital.

Private capital is purely profit-driven. Feng Jinfeng thinks private capital may lack the "boldness" of government industrial investment funds, which emphasize social benefits, and also lack the inherent upstream/downstream resources of large tech company investment divisions. However, they possess a core competency often missing in both: professional expertise.

"Especially market-oriented private integrated circuit funds, whose core teams have decades of rich industry experience, hold immense value for semiconductor startups in avoiding pitfalls and achieving rapid catch-up," Feng said.

Lin Jianhong also noted that the three types of investment firms have different positioning and objectives, creating a complementary relationship.