An investment frenzy triggered by a chip

In September 2020, Li Ping, a former investment manager of the National Big Fund, appeared again in the office of an external foundry of SMIC. This factory has an 8-inch wafer production line, mainly engaged in sensor foundry. The major shareholder is the local government, "which is equivalent to the local government providing capital and SMIC providing technology."

Three months after Li Ping's last visit, the company's valuation has increased by 50%. The CFO of the foundry, who is also responsible for financing, is too busy to check the accounts, and sometimes meets with four investors a day. This is completely different from his previous pace at the livestock breeding company.

Perhaps because he had met too many investors, the CFO could talk about the company's valuation and production scale more professionally. There were no longer dozens of business cards scattered on his desk. Instead, there were four neatly stacked business cards, about 15 cm thick, with an estimated number of thousands, prepared for visiting investors.

Chip manufacturing is an extremely difficult business. In the past five years, Samsung and TSMC have invested more than 10 billion US dollars in manufacturing capacity each year. SMIC, which has been losing money for 20 years, is still far away from them. At present, SMIC's 14nm chips have begun mass production, but TSMC has already installed 5nm chips in iPhone 12 and started to test 2nm technology.

However, when it went public in July, SMIC still became the largest IPO in the A-share market in the past nine years, with an oversubscription of 566 times. Its affiliated companies also attracted much attention.

A chip listed company exchange traded fund (ETF) managed by China Asset Management was listed in February this year. In more than half a year, its market value has ranked 15th in the country (excluding money market funds), only behind some traditionally large index or bond funds.

In 2015, more than 100 listed companies sold houses across industries, and now there are jokes about real estate bosses making chips. According to Tianyancha data, in the third quarter of 2020 alone, 13,000 companies expanded their business scope to integrated circuits in their industrial and commercial registration information.

In fact, companies no longer need to hype the chip concept themselves. One of the questions that recently went public, Golden Dragon Fish, received from investors was, "Does your company have a semiconductor business?"

The craze for chip investment has been transmitted from the secondary market back to the primary market, which in turn has driven startups to accelerate their listing.

At a chip investment forum in 2019, an investor lamented that 10 years ago, people investing in chips could sit at two tables, but 5 years ago there was only one table left, "Now it can fit a football field, there must be thousands of people."

In the past, funds that mainly invested in consumer Internet and other fields basically had investors who only looked at chips. According to data from Zero2IPO Private Equity, from January to October 2020, China's VC/PE invested in 345 semiconductor projects, 10% less than the same period last year. However, the scale of financing increased significantly, reaching 71.13 billion yuan in the first 10 months, 2.5 times that of the same period last year.

With the influx of hot money, the number of entrepreneurs in the chip industry has increased significantly. Chen Datong, founder of chip design company Spreadtrum, once mentioned that when Spreadtrum was first established in the early 2000s, there were only a few chip design companies in China. After 2005, the number increased to 500 to 600, and now there are more than 2,000, while Silicon Valley has never had more than 100.

Chips were not originally considered a good business. Chips are too expensive, and compared with the consumer industry, which has a market size of trillions of dollars, the ceiling for chip companies selling to enterprises is also lower. Bai Zongyi, partner of Glory Capital, believes that there is a bubble in the Science and Technology Innovation Board, and it is very challenging to become a chip listed company with a market value of more than 10 billion US dollars in China.

However, the Sino-US trade conflict has gradually pushed chips to the forefront, and the responsibility of China's chips has become increasingly greater. Chips are China's largest import product, far exceeding crude oil. In 2019, China imported $305.5 billion in chips and $238.7 billion in crude oil. Only four companies, Huawei, Lenovo, BBK, and Xiaomi, purchased more than 20% of the chips. According to Guosen Securities, if all semiconductor imports are localized, China's total GDP will increase by 3.2%.

Under the trend of domestic substitution and independent control, the investment boom triggered by a chip is also spreading.

The turning point of the chip investment boom occurred on May 16, 2019, when Huawei was included in the Entity List by the U.S. Department of Commerce and was unable to sign new orders with U.S. chip companies. Turning to domestic chip suppliers became a top priority for Huawei and other companies.

"This is a choice without choice," said Cheng Miaoqi, partner of Qingsong Fund. According to Gartner data, Huawei's chip procurement expenditure in 2018 exceeded US$21 billion, making it the world's third largest chip buyer.

On the same day a year later, the US Department of Commerce banned Huawei from using US technology and software to design chips, and the chip fever reached its peak.

On the other hand, as market sentiment is high, the market value ceiling of chip companies has risen sharply. After the first generation of relatively large chip companies, such as Spreadtrum and GigaDevice, went public, their market value remained at 30 billion yuan for a long time. Today, the market value of chip companies on the Science and Technology Innovation Board is much higher. Even at a 50% discount, most of them are still above 30 billion yuan.

Zheng Weihe, chairman of the RMB fund Tongchuang Weiye, also used the word "dream" to describe the Science and Technology Innovation Board. For example, the price of SMIC's Hong Kong stocks was less than HK$20 most of the time this year, while the A-shares reached a high of RMB 95.

"The greater the responsibility, the higher the market value," said a former semiconductor investor. According to incomplete statistics, as of the end of the third quarter of this year, among the 28 semiconductor companies listed on the Science and Technology Innovation Board, excluding two loss-making companies (with no P/E ratio), the average P/E ratio was 145 times. Cheng Miaoqi, partner of Qingsong Fund, believes that the normal P/E ratio should be around 40 times, and the high P/E ratio has overdrawn the company's growth space too early.

A former Hua Xia Fund manager also said, "Absolute valuation is no longer applicable in the secondary market." They use relative valuation methods and combine their relative positions in the industrial chain for comparison. For example, they compare chip manufacturing leader SMIC with packaging and testing leader Changdian Technology to estimate their value.

Investors and entrepreneurs have also smelled the opportunity of Chinese chips. Once chip companies enter the supply chain of Chinese consumer electronics companies, especially Huawei, they will be close to listing on the Science and Technology Innovation Board.

A typical example of domestic substitution is SiRuiPu, an analog chip supplier invested by Walden International in 2014.

SiRuiPu was founded in 2008. Its annual revenue from main business was only 113 million yuan in 2018, and its valuation after fundraising the following year was only 900 million yuan.

But in 2019, an undisclosed major customer brought 170 million yuan in revenue to SiRuiPu, increasing the company's revenue to 300 million yuan. This quickly met the listing standards of the Science and Technology Innovation Board. Analysts believe that the major customer is Huawei: SiRuiPu was certified as a qualified supplier by Huawei at the end of 2017 and received investment from Huawei Habo in 2019.

In September this year, SiRuiPu was listed on the Science and Technology Innovation Board, and its stock price soared 77% on the first day. Its current market value is 27.7 billion yuan.

Even though there was no talk of domestic substitution, chips began to rekindle investors' interest five years ago.

Globally, chips are gaining attention along with the popularity of artificial intelligence. Zhang Fei, a partner of 5Y Capital who invested in Horizon Robotics in 2016, told LatePost that he expected artificial intelligence to be a huge paradigm shift that could create opportunities worth more than a trillion dollars, which could be 3-5 times the market value of the IT industry.

2015 was the first year of AI investment. Google, Amazon, Facebook and other companies also entered the field of AI chips. Alibaba, Tencent and Baidu also paid attention to this field.

During the same period, a number of chip companies emerged, such as Horizon Robotics, Pony.ai, Cambrian, Bitmain, and DeePhi Technology. Since 2017, artificial intelligence algorithm companies such as Unisound and Yitu Technology have also begun to design chips to build a moat.

At that time, the chip industry also underwent tremendous changes. According to IC Insight data, the global semiconductor M&A amount reached $103.3 billion in 2015, and the total transaction amount in 2016 exceeded $98.5 billion, almost the sum of 2010-2014. An investor said that this is a way to expand categories and scale. Chip companies have high technical barriers and extremely segmented scenarios. Even if they have technical strength, it is difficult to grow on their own.

Chinese chip companies that landed on NASDAQ in the 3G era also experienced a wave of privatization and mergers and acquisitions. In 2013, Tsinghua Unigroup acquired Spreadtrum Communications for a total price of US$1.8 billion; in 2014, Tsinghua Unigroup completed the acquisition of RDA Microelectronics for US$907 million.

This has also driven a group of veterans to start a second business. In 2015, RDA founder Dai Baojia founded Aojie Technology, and veteran employee Zhang Liang founded Hengxuan Technology, both of which received investment from Alibaba and other companies.

In 2018, Zhao Lidong, who worked in chip design and development at AMD and was once the vice president of Tsinghua Unigroup, founded Suiyuan Technology. Since May 2018, Tencent has increased its investment three times in a row. This is also the chip project with the fastest valuation growth among those invested by ZhenFund. "We met on the same day and signed on the same day," said Yin Le of ZhenFund. The valuation was 200-300 million yuan, and now it has reached 5 billion yuan.

The AI ​​boom has passed, and asset targets have gone through a round of reshuffle. Cheng Miaoqi, partner of Qingsong Fund, mentioned that the number of investments has dropped significantly since 2019.

But a new wave has arrived. In the eyes of investors, the future will be an era of 5G, the Internet of Things, cloud computing, and autonomous driving.

In 2017, Intel acquired the Israeli self-driving chip company Mobileye for $15.3 billion. At the same time, the market value of Tesla, Mobileye's largest customer, was only $40 billion. Two years later, Mobileye became Intel's fastest-growing business line, with a year-on-year growth of 26% and revenue exceeding $1 billion.

In China, automotive chip companies such as SemiDrive Technology have received investments from institutions such as Sequoia Capital and Matrix Partners.

There are benchmarks in each sub-segment. BiRen Technology, founded by Zhang Wen, the former president of SenseTime in 2019, has become the "next Nvidia" in the eyes of investors. Within two months, BiRen Technology announced that its A round of financing exceeded 1 billion, and on August 18, it completed its B round of financing of 2 billion.

BiRen Technology “did not even have a PPT at first,” an investor familiar with the situation recalled to LatePost. But this did not prevent it from attracting the attention of hundreds of investment institutions. Not long ago, Nvidia’s market value exceeded Intel’s for the first time.

Silicon Valley, which was founded in the 1950s with military orders, has many veteran venture capital firms with experience in chip investment. But in China, after learning from the mistakes of investment, only a few investors have reached the harvest period.

The first batch of chip investors were represented by the earliest venture capital firms to enter China, such as IDG, Walden International, and Legend Capital. The second batch was mainly composed of established US dollar funds that entered China from Silicon Valley in 2005, such as Northern Light Venture Capital. After the launch of the GEM in 2009, RMB funds also became more active in the chip sector. But overall, few people invested.

Early successes basically have experience in related industries. IDG partner Li Xiaojun is a semiconductor major. He worked in chip design companies such as Marvell Technology and seized chip design projects such as RDA and VeriSilicon. Marvell Technology founder Dai Weili and VeriSilicon founder Dai Weimin are siblings. Walden International founder Chen Liwu is still the global CEO of EDA company Cadence. In 1998, he invested in chip design company Xintao Technology. Northern Lights Venture Capital founder Deng Feng was a counselor for the 1985 class of the Radio Department of Tsinghua University. Zhao Weiguo, chairman of Tsinghua Unigroup, and others were his students.

But accumulation does not guarantee success. After the ZTE incident in 2018, Zhu Xiaohu, a partner of Jinshajiang Venture Capital, said at an investment forum, "In fact, we are not unwilling to invest in chips. We have invested in several chip companies before and lost all our money. We have also contributed to China's technological innovation." Wu Xinjun, a former partner of Jinshajiang Venture Capital and a former employee of Bell Labs, led investments in projects such as Jingneng Optoelectronics and Jingbaoli, but they failed to successfully go public or be acquired.

Zhou Zhixiong, who also majored in semiconductors and was a former partner of Softbank SAIF Partners and now the founder of Triumph Ventures, also mentioned that "my return in this industry is not the worst, but it is the second to last." He invested in chip design company Alchip in two rounds from 2004 to 2006. Alchip has not raised funds or IPOed since then.

Chip investment cycles are long and uncertain. Sequoia invested in GalaxyCore in 2006, and GalaxyCore applied for listing on the Science and Technology Innovation Board this year, which took a long 14 years. In 2004, Lightspeed Capital invested in chip equipment company AMEC, which was listed on the Science and Technology Innovation Board 15 years later. US dollar funds generally have an investment cycle of ten years.

Zheng Weihe, chairman of Tongchuang Weiye, mentioned in a speech reviewing the 20 years of China's venture capital industry that the most profitable case in the investment industry was SIG's investment in Toutiao, and if nothing unexpected happens, the return should be 100 billion yuan. However, there are many IPOs of RMB funds, and it is very difficult to return 1 billion yuan from a project.

However, several RMB funds have achieved single returns of more than 1 billion yuan in the chip field, which was not optimistic in the past.

On April 28, 2018, shortly after the U.S. Department of Commerce sanctioned ZTE, Cornerstone Capital, Yuanhe Puhua, Wuyuefeng Capital and others invested in Huawei's camera supplier OmniVision Technologies. In 2019, OmniVision Technologies was successfully incorporated into the listed company Will Semiconductor, and Will Semiconductor's stock price soared. On July 10, Will Semiconductor's market value exceeded 200 billion, while it was 20 billion in 2019.

This transaction created high returns for two local RMB funds. Tongchuang Weiye, which invested 40 million yuan in Will Semiconductor in 2011 and 2014, exited in 2019 with a return of 900 million yuan. Cornerstone Capital, which invested over 100 million yuan in OmniVision Technologies, has a floating profit of 1.5 billion yuan in two years.

An investor told LatePost that an institution invested in chip equipment company AMEC in 2018 and has now made more than ten times the profit. On July 22, 2019, the day of listing on the Science and Technology Innovation Board, AMEC's ​​market value exceeded 60 billion yuan, and then exceeded 100 billion yuan.

According to the latest market value, iFlytek's angel round investment in Cambrian in 2016 may have made as much profit as its annual net profit. In early November, Cambrian's market value once exceeded 75 billion yuan, and iFlytek held 1.19% of the shares, worth nearly 900 million yuan. In 2019, iFlytek's annual net profit was only 788 million yuan.

In a 2019 report, the U.S. Department of Defense stated that the concentration of chip production in East Asia is a "single point-of-failure" for the U.S. technology industry. If there is a problem with the chip supply in the region, it will jeopardize the entire U.S. technology industry, including the military industry.

On the other hand, this single weak link is the same for China, or even more urgent. After Huawei was sanctioned by the United States, the problem of Chinese companies' incomplete coverage in the chip industry chain was put on the table.

In the field of chip design, Huawei HiSilicon has squeezed into the top ten in the world. However, as the United States stopped licensing EDA chip design software and required TSMC to cut off supply, Huawei phones could not get 7-nanometer and 5-nanometer process chips, making it difficult to achieve the same performance as Apple and Samsung.

At present, all fields of the chip industry chain, including materials and design in the upstream, equipment and manufacturing in the midstream, and packaging and testing in the downstream, are very popular. However, most capital is still invested in asset-light chip design companies like HiSilicon.

According to Yunxiu Capital statistics, as of August 12, 2020, nearly 71% of chip-related investments went to chip design companies, and only 15.3% went to companies in the chip materials and equipment fields - 1.3 percentage points higher than the whole of 2019.

At the same time, the investment targets chosen by investment institutions are becoming safer. In the first half of this year, the proportion of investment after round C increased significantly, from 8% in 2017 to 21.9%.

Another change is that it is becoming increasingly difficult for financial investors to compete with industrial capital for investment targets. Industrial capital has great advantages, not only can it bring funds to the invested companies, but also orders.

In 2011, Walden International received support from the National Science and Technology Investment Group and Shanghai Science and Technology Venture Capital Group under the National Development and Reform Commission to establish the China Semiconductor Industry Fund Shanghai Huaxin with a scale of 500 million. Its LPs also include semiconductor companies such as TSMC, ARM, and Fujitsu.

Industrial funds dominated by consumer electronics companies also have more advantages than pure financial investments.

Lingming Photonics originally tried laser radar chips, which were not favored by most investors at first. This year, Lingming Photonics launched 3D sensing distance measurement sensors for mobile phones, which attracted more institutional attention, but the founder was no longer so easy to make an appointment with, said a PE investor involved in the project. In October, Lingming Photonics accepted a stake in Xiaomi Industrial Fund.

Even bigger is Huawei. In April 2019, Huawei established Habour Investment with a registered capital of 700 million yuan. In January 2020, its registered capital had reached 1.7 billion yuan.

"In the real world, our evaluation of things swings between two extremes, 'very good' and 'not so good.' But in the investment world, the two extremes of the emotional pendulum are 'good to perfection' and 'hopelessly bad,'" Howard Marks, founder of Oaktree Capital, summarized in "The Cycle."

In just a few years, Chinese investment institutions' sentiment towards chips has shifted between two extremes.