Chinese artificial intelligence companies are over-valued. As market liquidity conditions continue to tighten, these AI companies are likely to see their valuations drop by 30% to 50%, according to Wayne Shiong, a partner at Chinese venture investment firm China Growth Capital.
"We expect companies to lower their valuation expectations for the second half of the year. People start to fear whether they can get funding in the private market now (because of tighter liquidity). Everyone is trying to list in the public market before liquidity dries out. A lot of really expensive companies should see their valuation corrected before IPO," Shiong told China Money Network in an interview last week in Beijing.
China’s AI industry is already massive and the country has the highest valued AI company in the world, Shenzhen-based humanoid robot maker Ubtech Robotics currently worth US$5 billion. China also has the most number of AI companies worth more than US$1 billion or more, or commonly referred to as "unicorns", than any other countries in the world. There are at least over 10 AI unicorns in China, according to China Money Network’s own proprietary data.
The Chinese AI market size stood at RMB23.74 billion (US$3.45 billion) in 2017, representing a 67% increase compared with 2016, according to an AI report by Tsinghua University released in July. Also last year, China’s AI firms attracted US$27.71 billion funding, accounting for 70% of the total fundraising in the industry globally, according to the same report.
That has led many investors to question if the Chinese AI market is over-heated. But to suggest AI companies will see their valuations drop by 30% to 50% – as Shiong reckons – still puts him among the more pessimistic, or sober, group of investors.
Founded in 2006, China Growth Capital is a early-stage venture capital firm funding seed to series B rounds in fintech, enterprise tech and Internet consumer sectors. It has RMB6 billion (US$873 million) asset under management across its RMB and U.S. dollar funds.
Shiong sat down with China Money Network last week to share his insights on the AI industry in China. Below is an edited Q&A.
Q: You mainly invest in frontier technology. What new trends have you seen this year?
A: For deep tech, this year has been pretty fragmented. Blockchain took a lot of attention in the first and second quarter. But now people are looking at general AI, or AI for everyone type of opportunities . Right now, there is a gap between infrastructure software technology and AI application. We are still yet to see a lot of live AI projects.
Q: When do you think AI applications will be more mature?
A: We have to see whether autonomous driving will be successful. Autonomous driving is a very independent AI-driven sector so that will be a huge indicator. We will see whether Waymo will be more meaningful commercially, by either get external funding or real businesses being rolled out. And a lot of AI start-ups, like Element AI in Canada, have raised a lot of capital.
But it is easier to generate revenue for AI companies in China than in the U.S., partially because some of the markets such as public video surveillance are very Chinese market centric.
Q: How would you compare the Chinese and American AI markets?
A: Application. If you look at real applications, in the US, companies like Google and Facebook are using AI to enhance their revenues. In China, we have huge sectors like surveillance and IoT. Because big buyers of these services are provincial government or government departments that are trying to improve public security system or law enforcement. Government spending creates a huge test bed for AI in China.
We don’t have enough original academic resources in China, however. Also, we don’t have enough visionary companies. But if you look at the U.S., everyone is trying to get ahead. Not just Google, not just Amazon. We need large companies to lead the way and to build up infrastructure. For example, we don’t see Baidu being aggressive in hardware.
Q: Among Chinese tech giants, the BAT, who is the most visionary, in your opinion?
A: Alibaba is probably the most visionary company. Because it started with a very humble e-commerce business, which required little advanced technology. Now they are moving toward semiconductor as they acquired chipmaker C-Sky. They are also aggressively building research institutions.
Tencent is pretty happy with their established empire. They don’t touch hardware. Their cloud business is lagging behind others.
As a leader, you have to invest in the future. The challenge for large companies is whether they are able to invest in the unknown. If you look at Baidu’s portfolio, it invests in anything that could enhance its quarterly earnings in the next two to three quarters. Tencent, has a similar philosophy but longer range, probably six to eight quarters. Alibaba is trying to build something new.
Q: Many AI chip companies have received funding, or existing AI companies have announced plans to make their own chips. Will they succeed in helping China to become more self-sufficient in chips?
A: My suggestion for these companies is to sell quickly to be part of a bigger company, or become an application company with strong hardware expertise. Don’t define yourself as an AI chip company, it is too late. Instead, position yourself as an application company with strong tech background.
And raise money based on your application, revenue and growth trajectory instead of saying you will be the next Nvdia. It is simply not realistic for a start-up to achieve that.
Q: Are Chinese AI companies over-valued now?
A:We expect companies to lower their valuation expectations for the second half of the year. People start to fear whether they can get funding in the private market now (because of tighter liquidity). Everyone is trying to list in the public market before liquidity dries out. A lot of really expensive companies should see their valuation corrected before IPO.
Companies need to raise money but they need to lower their expectations on valuation. Especially for those earlier growth and series B companies, we are in a wait-and-see mode. We don’t want to close the deals within a month. For early stage founders, it is usually their first time to raise capital and their expectations are very high, but that should change.
Q: How much do you see valuation being corrected?
A: I think 30% to 50% (drop from current level).
Similar things happen in the fintech sector. During the recent P2P crackdown, lots of companies went bankrupt. There were lots of VC-backed companies. When VCs got into those companies, they used price-earning or price-to-sales multiples, which was not right to value those asset-based companies. They should just use multiples on book value and qualities of the assets. But a lot of VCs didn’t have experience in this sector and helped drive valuation to very high levels. That will see huge corrections too.
Q: What factors do you consider when investing in an AI company?
A: We need strong academic proof. We look at the numbers of papers published and citations and whether the CTO is a top-notch scientist. Then their commercial capability, whether they could think as a business operator, not a professor. And lastly, whether their solutions or approaches are unique.
Q: You recently invested in AI2B, an AI solutions provider. Can you explain your investment thesis?
A: They have a top team who are extremely commercial and capable. They target verticals such as retail and travel industry. They launched a partnership with shopping mall operator K11. AI2B will optimize the internal operation of K11 to improve the shopping mall operator’s return on investment. On the user experience side, they are working on AI applications and UI to improve user interactions at the mall.
For travel sector, they are trying to integrate facial recognition systems at high way toll, high-speed train, airport, parks and other points of interests. We see these applications having great potential.