Profile: Joy Capital's Erhai Liu

In Chinese kungfu, attacking with a stick or sword is considered straightforward while using nunchaku or a whip relies on a mastery of indirect power. In early-stage technology investment, the entrepreneur is the equivalent of the stick: direct, single-minded, sometimes authoritarian. Venture capitalists are the whip: exponents of indirect power or influence, prone to self-harm if clumsy in their approach.

If results are anything to go by, Erhai Liu is adept with the whip, having nurtured a string of start-ups to full bloom. He made his mark with Legend Capital, backing the likes of China Auto Rental, Bitauto, and iDreamSky Technology, which generated returns of 18x, 13x and 11x, respectively. Having struck out on his own with Joy Capital, Liu has accumulated $1.5 billion in assets under management within the space of four years. The first two of Joy’s three US dollar-denominated funds rank among the top 5% of VC funds globally for their vintage.

During this journey, one key learning has been reinforced time and again: an VC can only wield the whip if he first wins the trust of the entrepreneur. “If an entrepreneur doesn’t trust you, you will not be able to exert any influence,” he explains. “To gain that trust, you need to be professional, do your research, and give concrete advice. And if you give your word, keep to it.”

Liu’s early interest in technology led to a somewhat predictable outcome: bachelor’s and master’s degrees in engineering and communication. His first job in 1994 was with state-owned Jitong Communications, where he initiated the VoIP phone services department, enabling calls to be placed through internet networks rather than traditional switchboards. It came to account for 70% of Jitong’s revenue.

In 2000, Liu joined foreign communication services provider Clarent as deputy general manager and head for marketing for China. From there, he was recruited by a customer, China Tietong Telecom, to head up its internet business. Liu juggled these duties with MBA courses at Peking University and that is where he met Nengguang Wang, co-founder of Legend Capital. This opened up a new career path as a VC.

Valued counsel

Liu soon came to understand, as an investment director at Legend Capital, that his job was to help companies make the decisions on key development nodes, not to get involved in day-to-day operations. Other lessons came thick and fast – as did opportunities to apply the learnings. 

For example, Liu led Legend’s early-stage investment in China Auto Rental (CAR) and was then called upon in 2013 to advise on a curious request from Hertz. The US-headquartered auto rental giant was coming as a strategic investor with a 20% stake and wanted a “first look” right that would give it first option on any shares sold by other investors. Liu told CAR to reject it out of hand.

This put CAR in a difficult negotiating position. Moreover, the management team didn’t really understand why it was so important to put up a fight. But Zhengyao Lu, CAR’s CEO, trusted Liu – Legend Capital first backed the business in 2008 – and heeded the advice. Years later he reaped the rewards. CAR and Lu wanted to repurchases a large volume of shares and were able to do so without any third-party interference.

“Strategic partners want this kind of [first look] condition to facilitate raising their shareholding in companies,” Liu says. “Not only could this affect how quickly VC investors achieve liquidity, but the companies themselves could be blocked from repurchasing their own shares and potentially end up losing control.”

Another seemingly innocuous deal condition that can come to haunt entrepreneurs are veto rights. In this regard, Liu proved his value to apartment rental platform Danke Apartment as recently as last year when Tiger Global Management and Ant Financial were poised to lead a $500 million Series C round.

Liu, who took Joy into Danke’s extended Series A two years earlier, suggested the company take back the veto right because it ceded too much control over decision making. Management was unable to close the deal with no veto right given to the new investors, so Liu proposed making every participant in the Series A through C rounds give up their veto when the Series D arrived. A majority of investors could block new initiatives, but not an individual acting alone. Danke raised a $190 million Series D last October and went public in the US last week.

Chinese bike-sharing start-up Ofo learned the power of the veto the hard way. Between 2015 and 2018, the company raised $2.2 billion across eight rounds of funding as it battled Mobike for market share. Ofo then collapsed in late 2018, unable to refund the deposits of some 10 million customers. Pony Ma, chairman of Tencent Holdings, blamed the veto – it had allowed a relatively broad group of investors to block various exit avenues. A merger with Mobike and acquisitions by Didi Chuxing and Alibaba Group were reportedly rejected.

Mobike was luckier as Meituan-Dianping bought the business for an equity valuation of $2.7 billion in 2018. Joy, which was the sole investor in the company’s Series A in 2015 and re-upped in the next three rounds, exited with a 10x return.

Liu offered Mobike sage advice in two areas. First, when the company was raising its Series C in 2016 and bike-sharing was all the rage, management wanted to align with as many as investors as possible but avoid diluting their own shares too much. Liu suggested structuring the deal as equity plus debt, with $3 going into a convertible bond (CB) for every $5 in equity. The ratio was not arrived at by accident. “If you suggest half equity and half CB, investors will be annoyed,” says Liu.

Second, at the height of the battle with Ofo when cities were being flooded with bikes, Mobike was at a cost disadvantage. Its machines were of higher quality and had GPS tracking devices built into the electronic locks. The start-up could have capitulated and joined the race to the bottom, but it opted to compromise. The GPS devices remained and the electronc locks were switched from self-charging to solar-powered.

“We wanted to be friends with time,” Liu explains. “GPS helped us track our bikes and repair them if necessary. Without GPS, we would have lost their locations, which might have meant replacing entire bikes. Over time, our competitive advantage became clear.” In the end, Ofo replaced its mechanical locks with electronic locks and introduced GPS.

Seeing the future

Bike-sharing was arguably a controversial investment choice. Some VCs scoffed at the idea in Mobike and Ofo’s early days, then looked on bewildered as fundraising activity and valuations went into overdrive. Liu doesn’t mind going against the grain, provided he understands the essence of the business. However, there are some recurring themes, notably the new infrastructure.

“When a new technology emerges, it generally has several typical applications. For the internet era, it was web, email or FTP. For the AI era, it’s voice recognition, facial recognition, autonomous driving. But that’s just the first stage. In the second stage, new technology becomes the new infrastructure, leading the transformation of all industries,” he explains. “Changes to the planet caused the extinction of dinosaurs, right? We need identify new species that can adapt to the new environment.”

Luckin Coffee is a new species. The company is essentially an asset-light version of Starbucks – an emphasis on takeout allows for smaller stores, a reliance on digital ordering and payment removes the need for cashiers. Luckin claims that its tech-enabled model reduces customer acquisition costs because it tracks customer activity and devises precisely targeted promotions, while better inventory management boosts supply chain efficiency.

The company has plenty of critics who question the sustainability of the business model – rapid store expansion at enormous cost – and whether technology can be as transformative as claimed. Still, this didn’t prevent Luckin completing a US IPO last May less than 18 months after its formal launch. The board considered waiting longer, but Liu was among those advocating for an early listing.

“When Starbucks went public in 2002, its value was $300 million, and now it’s worth 110 billion. Who has earned the added value? The public market. When Uber listed last year at a valuation of $69 billion, the private sector earned the money. The public market won’t be happy, right? That’s why we see many companies adjusting their IPO prices downwards, which will affect some late-stage private market investors. Why should we stay private if this would damage some investors’ interests?” says Liu.

He believes that, following the WeWork fiasco, the phenomenon of ever larger private rounds for unprofitable companies might be ending. In the meantime, Joy is sitting on a big paper gain from Luckin. It took part in two funding rounds, the second at a valuation of $2.2 billion. Luckin now has a market capitalization of $12.6 billion, which means Joy’s stake is worth $850 million – more than the combined size of its first two funds.

Joy has raised capital aggressively, closing its third VC fund at $391 million last year and securing $327 million more for a debut growth vehicle. Liu claims to be comfortable with his 20-strong team, noting that venture is not a scale business. He also recognizes that success is built on reputation. Indeed, some of the early relationships forged at Legend are still paying out. Luckin was established by the former COO of CAR and Lu was an early backer and now serves as chairman. Joy had no problem getting access. Similarly, Bitauto founder Bin Li helped create Mobike.

Ask to identify his biggest advantage, Liu says he respects ordinary people and takes them seriously.  For example, when a friend recommended Nanami Shiono’s “Stories of the Romans,” he worked through all 15 volumes. When a start-up calls for advice, it can expect that kind of attention to detail.  

• Tweet  
• Facebook  
• LinkedIn  
• Google plus  
• Save this article  
• Send to  

• Topics
• Greater China
• GPs
• Early-stage
• Technology
• China
• Joy Capital