The avalanche of plummeting shares of ShangTang (00020.HK) didn’t stop after the previous trading day when the company suffered a “one-day cut”, but on July 4, ShangTang’s shares continued to plummet, falling by more than 20 percent to close at HK$2.54, a drop of 18.85 percent. On the previous trading day (June 30), ShangTang’s share price had plunged 46.77% in one day, and the share price plunged more than 65% in two trading days.
The total market value of Shangtang Technology has shrunk to near HK$85 billion before, and its market value evaporated more than HK$100 billion in two days.
Unblocked day investors staged a “great escape”
On June 30, Shangtang ushered in the first unlocking day after the listing, that is, the pre-listing investors and cornerstone investors’ equity will be basically all unlocked, according to public information, involving more than 70% of the company’s total share capital.
Although early on June 30, Shangtang Technology issued an announcement that co-founder, chairman and CEO Xu Li, chief scientist Wang Xiaogang, executive director Xu Bing and several members of management voluntarily committed to extend their holdings of restricted Class B shares until December 29, 2022, with the share lock-up period extended to 12 months.
However, this still could not stop the collapse of the share price. on June 30, Shangtang’s share price plunged 46.77% to HK$3.13 per share. What is more noteworthy is that on that day, the turnover of Shangtang Technology was HK$6.198 billion, which is more than thirty times larger than the turnover of HK$158 million on the previous trading day, which undoubtedly means that the capital, especially the unbanned investors, is staging a “great escape”.
Shangtang Technology on December 30, 2021 “pressure line” to complete the IPO, market analysts pointed out that this may be forced by the cruel betting agreement with investors. If the IPO is not completed by the agreed time, the company will have to redeem its shares at the issue price plus annual compound interest and unpaid dividends.
But Shangtong Technology’s issue price was HK$3.85 per share at the time, and the share price once peaked at HK$9.7, but today its shares are trading near HK$2.50. The mood of major shareholders and institutional investors can be imagined.
In fact, as a star startup, Shangtang Technology’s shareholder list is luxurious. The cornerstone investors of Shangtang Technology are China Chengtong sponsored mixed-ownership reform fund, Guosheng Overseas Hong Kong, Shanghai Artificial Intelligence Industry Equity Investment Fund, SAIC Hong Kong, Shanghai Xuhui Capital, Guotai Junan Securities Investment, Hong Kong Science and Technology Park Venture Capital Fund, Sigma Ophthalmology, Taizhou Cultural Tourism and so on. And Softbank Capital, Ali, Chunghwa Capital, Silver Lake Capital, IDG, State Transfer Fund, and Dinghui also hold shares of Shangtang Technology.
However, retail investors seem to be uncomfortable with Shangtang Technology from the beginning. In the pre-IPO subscription phase, the subscription multiple of retail investors was less than 3 times, which is not very enthusiastic, especially when compared with the new economy companies that have landed in Hong Kong stocks in recent years, which is relatively dismal.
2021 loss is 3.6 times the revenue, “burn money for the market” will not work?
The most fundamental reason behind the flight of capital and the plunge of the stock price is still the loss, and it is a continuous huge loss.
In fact, the market and public opinion on Shang Tang Technology, the star unicorn company, have been divided into two views. On the one hand, the AI track is undoubtedly the “windfall” field that is favored by all parties. As one of the representative companies, and even known as the first of the “Four Little Dragons of AI” (Shangtang Technology, Kuangwei Technology, Itu Technology, and Cloud Technology are the most sought-after AI startups), Shangtang Technology is naturally a typical company. Especially as the first “AI first stock” in the capital market, Shangtang Technology is regarded by the market as an indicator. It is natural for a company to run in front of a popular track to be sought after by capital.
But on the other hand, the whole AI industry is facing a huge problem – not making money, and especially burning money, which is also very obvious in Shangtang Technology.
According to Shangtang Technology’s first annual report after going public, the company achieved revenue of 4.7 billion yuan in 2021, up 36.4% year-on-year from 3.446 billion yuan a year earlier. But the company’s total loss in 2021 reached 17.14 billion yuan, a loss of 3.6 times the revenue, which is apparently hard for investors to accept. One investor chimed in the comments section, “It’s really quite technical for a company to spend 17.1 billion yuan to only ‘buy’ back 4.7 billion yuan in revenue.”
And Shangtang Technology’s losses have continued for four years, as revealed in Shangtang Technology’s prospectus, from 2018 to 2021, the company’s net profit attributable to shareholders of the parent company (referred to as net profit) will be a loss of 3.428 billion yuan, 4.963 billion yuan, 12.158 billion yuan and 17.140 billion yuan, or a combined loss of 37.689 billion yuan.
Capital may be patient, but it is still fundamentally profit-seeking. The capital “Battle Royale” faced by Shangtang Technology is not an AI company’s predicament, but will likely turn into a turning point that will affect the financing ability and valuation level of the whole industry.
According to the reporter, now the AI track companies want to get financing and high valuation is not easy, because most investors no longer believe in the logic of “burn money for the market, a doctor is high technology, high technology can be profitable ……”.
AI is hot, but why is it “not profitable”?
“Independent AI algorithm companies like Shang Tang, to achieve profitability requires at least two of the most basic conditions: first, there are enough application scenarios to achieve commercial real estate; second, the algorithm and other technical strength is really excellent, to achieve no one I have.” A senior cloud computing vendor Zhang An (a pseudonym), who did not want to be named, told China Economic Weekly that many domestic AI companies are customers of this vendor.
Although Shangtang Technology in the listing prospectus “self-introduction” as: “an empowerment of a hundred industries, industry-leading artificial intelligence software company”, is to meet the two conditions of profitability, but the ideal and reality is obviously a gap.
Zhang An analysis said, first of all, “thousands of industries need AI” this prediction is not a problem, but the problem is that now can have a more mature commercial realization of the industry scene is still very small, mainly in the Internet industry.
“But large and medium-sized Internet companies almost all do their own AI business, after all, this involves the core of the company’s business, and is likely to be the core competitiveness, generally only relatively low-end, labor-intensive algorithm business will choose to outsource third parties. And the prospects for AI applications in government, finance, automotive, manufacturing, medical and other fields are good, but it is far from a realistic demand for scale, and once the demand explodes, there is a high probability that these companies will also choose to do it themselves, the logic is similar to the Internet companies.” Zhang An said.
Therefore, Zhang An believes that AI in the end to make money or not to make money to look at two levels, one is the third party independent AI company to make money or not; the second is the AI business to make money or not. “If the Internet company’s AI business is taken out separately, the volume does not know how many times larger than the ‘AI four dragons’, and it will certainly be profitable as well.” Zhang An said.
So, another condition for independent AI companies to be profitable is to have a technical barrier so that customers “can’t do without you”. “But frankly speaking, these so-called ‘AI four dragons’ are far from reaching this condition.” Zhang An said.
According to Zhang An, one data can actually reflect the real technical strength of an AI company sideways, and that is the salary cost of personnel. The core competitiveness of an algorithm company is talent, and good talent must be expensive, but the salary level of the ‘AI Four Dragons’, including Shangtang Technology and others, is actually not high in the industry.
According to the prospectus disclosure of Shangtang Technology, as of June 30, 2021, Shangtang Technology has more than 5,000 employees, including 40 professors and 250 doctors, and the total compensation of employees in the first half of 2021 is 2.493 billion yuan. According to this calculation, the average salary expense of Shangtang Technology is only 560,000 yuan, and the per capita salary after removing the share payment is even only 390,000 yuan.
And this is only the average figure. The prospectus also disclosed that for the six months ended June 30, 2021, co-founder, chairman and CEO Xu Li’s salary totaled 512 million yuan, chief scientist Wang Xiaogang’s salary totaled 375 million yuan, and executive director Xu Bing’s salary totaled 305 million yuan, with the combined salary of the three executives reaching 1.192 billion yuan, accounting for about 48 percent of the total compensation of all employees.
However, there is a considerable gap between the compensation cost disclosed in the company’s financial report and the actual monthly salary of the employees in hand. Because the company’s compensation cost includes housing allowance, free shuttle bus, holiday gifts, free canteen, reunion expenses, training expenses in addition to the employees’ broad pre-tax salary ……
The reporter also consulted a senior finance director, she said, in the case of an employee whose monthly salary in hand is 10,000 yuan, the actual salary cost paid by the company for it should be at least around 13,000 yuan. And according to the “Artificial Intelligence Industry Report” released by the Lajia Recruitment Data Institute, among the popular positions in the AI industry, architects have the highest salaries, with an average monthly salary able to reach 40,870 yuan; driver development is 30,750 yuan; algorithm engineers are 28,775 yuan.
Zhang An introduced the AI industry is an industrial chain, in which there are high-tech, high value-added links, but also labor-intensive, low value-added links, “sugar cane is not which section is sweet, so you can not say in general, as long as the AI company is high-tech, high returns.” He said.
But Zhang An is still very optimistic about the future prospects of the AI industry, except that today’s star companies, whether they end up as “pioneers” or “martyrs”, the answer is difficult to predict, but there will definitely be great companies in this field.
(Zhang An is a pseudonym in this article at the request of the interviewer)