HONG KONG (Reuters) - Sometime in the third quarter of 2016, Blackrock’s (BLK.N) scientific active equity team, which manages $80 billion globally, began picking up increased signs of construction activity on the ground in China by using satellite imagery.
Using that as a starting point and adding freight traffic and other data to the picture, the U.S.-based analysts determined that the world’s second-biggest economy was on the verge of a cyclical rebound.
Their expectations were borne out when China reported last week that its economy grew a quicker than expected 6.9 percent in the first quarter, the fastest in six quarters.
Blackrock’s quantitative equities portfolios increased their exposure to China based on the data. The Shanghai Composite Index .SSEC gained about 11 percent between the end of September and early April, though it has lost some of those gains in recent days as regulators clamped down on speculative activity.
“We were picking up signals by simply looking at more metal on the ground and such unconventional data sources are very essential in an economy as large and diverse as China where data may not be very timely,” said San-Francisco-based Jeff Shen, co-head of the scientific active equity team.
The use of unconventional data sources to gather price-sensitive information about a company, or even an entire economy, before it is made public isn’t a new phenomenon, especially among the hedge fund community in developed markets.
But that behaviour has now spread across the world, and to many more mainstream mutual funds. Faced with fierce rivalry from low-priced exchange traded funds and desperate to show that they can outperform market benchmarks, the mutual funds are now increasingly seeking alternative kinds of information in Asia.
In the first two months of the year, actively managed equity mutual funds domiciled in Asia saw a net outflow of $1 billion compared to net inflows of $9.3 billion in all of 2016 and $40.3 billion in 2015.
In contrast, exchange traded funds saw net inflows of $15.16 billion in the first two months of this year compared to $39.1 billion in all of 2016 and $21.7 billion in 2015, according to Morningstar data.
That pressure is forcing fund managers to increasingly look at alternative sources of information about listed companies ranging from shipping logistics trends, and social media chatter, to payments data for unlisted private companies that supply parts to bigger listed companies.
Robert Ciemniak, founder of Real Estate Foresight, a China-focused real estate data analytics firm, routinely analyses headlines from a variety of publicly available blogs, local news feeds and government bulletins in local cities and provinces to understand the ground-level shifts in the Chinese property sector before it garners attention from mainstream media.
“Last year, online searches related to buying property in Nanjing spiked a few months before prices in the city jumped,” Ciemniak said referring to a property price surge of almost 40 percent in 2016 in the eastern Chinese city, with the authorities rolling out tightening curbs since August last year.
”In a funny way, China is quite transparent as long as you employ reliable filters and know where to look,” he said.
Only the top asset managers in the world can afford to hire top notch data scientists to mine mountains of data to hunt for investment ideas.
While a 40-member strong investment team at San-Francisco based Matthews Asia, which manages $26 billion in global assets, takes the conventional approach by conducting an average of 2,000 meetings on an annual basis in Asia, its investment strategists are looking at other data sources to supplement investment decisions.
“We look at various sources such as financial information released by multinationals with a big presence in China, tourist spending and visitor patterns by Chinese visitors abroad to get a grip on discretionary spending,,” said Sherwood Zhang, who manages the Matthews China Dividend Fund.
Some others rely on companies which aggregate data from a variety of unusual sources such as Toronto-based Quandl, though most of such companies are based in U.S and Europe for now.
Still, getting such information fully integrated into daily workflow of investors is not easy.
“The quality, capture rate, the representativeness of the data and accuracy of prediction are only some of the problems,” said Seth Fischer, founder of Hong Kong-based hedge fund Oasis Capital.
Some investors say that the increasing push for an edge from unconventional sources is inevitable. “Thirty years ago, the ability to rank 2,000 stocks on a price-to-earnings or a price-to-book ratio was considered pretty scientific but today the amount of information we process is larger than the U.S. Congress Library,” said Blackrock’s Shen.
“The challenge increasingly nowadays is to separate the signal from the noise.”
Edited by Martin Howell