You are here

Amateur quants are suddenly everywhere in China


CHINA'S individual investors are catching quant fever.

Armed with cheap computing power and free online software, thousands of amateurs across the country are building trading algorithms that were once only available to market pros. JoinQuant, China's most popular do-it-yourself quant platform, now has more than 180,000 users, up four-fold since 2016.

The influx points to rising levels of sophistication among China's retail investors, who account for more than 80 per cent of the nation's stock trades and have long been branded as rumour-chasing speculators.

Even though most of the 1.7 million strategies developed by JoinQuant users have yet to demonstrate market-beating potential, a few are performing remarkably well. Ruan Zhe, a former Microsoft Corp engineer in Beijing, said his algorithms delivered a 70 per cent return in 2018, bucking a 25 per cent drop in the Shanghai Composite Index. He's up another 53 per cent this year through May.

"There are certainly pros within the retail masses, though they're still a very small percentage," said Yan Hong, director of the China Hedge Fund Research Centre in Shanghai. "Individual investors don't totally equal chives," he said, using a popular Chinese term for unsophisticated speculators.

Mr Ruan, 35, began his foray into algorithmic trading after stumbling across JoinQuant's website three years ago.

Started in 2015, the Baidu Inc-backed platform provides investors with data and tools to develop and test their trading ideas. Those who create promising algorithms can use their own cash to deploy them, sell them or license them to brokerages for a cut of trading commissions. They can also share them with other users for a monthly fee (JoinQuant takes a percentage for itself).

In rare cases, JoinQuant will fund users' trading ideas. It tapped Mr Ruan in December to manage 10 million yuan (S$1.98 million), which he deploys with four souped-up computers in his apartment in Beijing's Zhongguancun tech district. Ruan stands to earn anywhere from 20 per cent to 50 per cent of the portfolio's profits, depending on how well he performs.

For most DIY quants, the odds of success are low, said Mr Yan, who's also a professor at the Shanghai Advanced Institute of Finance.

"It's possible that some of the users on such platforms are extremely smart and able to quickly grasp some basic skills and even find strategies that can generate good returns in a certain period of time," he said. "But a long-term, sustainable investment process will still require the support of professional institutions."

That hasn't stopped individual investors from trying. Encouraged by a soaring stock market and signs that regulators will soon loosen regulations on algorithmic trading, new JoinQuant users signed up at a rate of about 500 per day earlier this year, with the pace slowing only slightly in recent weeks as Chinese shares pared their world-beating advance.

JoinQuant had a record 184,000 users as of May, up about 50 per cent from a year earlier, according to Li Xiao, co-founder of Beijing Xiaolongxia Technology Co, which operates the platform. That compares with about 230,000 at Quantopian Inc, a similar platform in the US, and 100,000 at Uqer, JoinQuant's main competitor in China. Almost all of JoinQuant's users are individual investors, while a small percentage work for money managers such as private funds, Mr Li said.

International heavyweights from BlackRock Inc to DE Shaw & Co and Bridgewater Associates LP have also bulked up their quant teams in China, as have local institutions including China Asset Management Co. They've done so despite a recent rough patch for quant funds in the US and other developed markets, where disappointing performance has fuelled investor withdrawals.

While the growing number of quants in China means there's more competition to find mispriced stocks, the market still offers plenty of alpha-generating opportunities, according to Li Yimei, the chief executive officer of China Asset.

"The A-share market is not yet a very efficient market, and therefore the room for enhancement above the index is quite ample," she said. "Fairly good alpha is indeed achievable." BLOOMBERG