‘Big data’ is now being employed by the asset management community, but it remains a mystery to many.
It has been making waves around Asia, from traffic management in Singapore to peer-to-peer lending in China. Now asset managers are analysing how it can help them run portfolios and better service their clients.
Big data can be loosely defined as the use of mathematical algorithms to analyse large amounts of information in order to identify buying trends and economic activity.
BlackRock is one asset manager that has been investing in the technology, which may be fed into some of its Asian investment strategies in the future.
While the $4 trillion fund house does not currently incorporate big data in Asia, one of its senior executives noted that 20-30% of its funds for institutional and retail investors in Europe and the US were using it.
“What we are finding is there is an increasingly vast volume of information that is important for understanding security pricings and returns,” said London-based Richard Mathieson, head of strategy for EMEA at BlackRock’s scientific active equity unit.
“It doesn’t come in nice, pre-packaged databases of rows and columns and financial information and quantitative statistics. It is typically in textual form.”
BlackRock’s big data project includes analysing 15,000 corporate regulatory filings with the US Securities and Exchange Commission over the last 12 months, 5,000 transcripts of quarterly earnings calls, as well as 4,000 daily broker notes.
This is equivalent to two-and-a-half terabytes of data per day, something the average human would take too long to process for the information to be of any use. Algorithms are therefore used to scan the textual information, picking up on specific words to produce a rating score for listed companies ranging from +3 to -3.
Positive words such as “exciting”, “growth” and “opportunity” add to the score, while those with negative connotations such as “threat”, “deteriorate” and “competition” count against a company. It ends in a recommendation to overweight or underweight a stock. Biases are also filtered in based on BlackRock’s macroeconomic view of a market.
Other uses of big data employed by BlackRock include running analytical queries on internet search engines to gain a leading indicator on economic activity.
Specifically, the more search terms that Google users make within a certain timeframe – such as for cars, holidays or luxury goods – indicates they are doing research and are ready to spend. In contrast, words such as unemployment, benefit and welfare have the opposite connotation.
“Through use of these inquiries, we picked up on improving consumer sentiment in parts of Europe, especially Spain, and that moved our portfolio to overweight the Spanish economy, which was a very successful trade in 2013,” noted Mathieson.
However, big data is not being universally embraced and there is a healthy dose of scepticism amongst asset managers in Asia.
“We don’t do it and I am not aware of fund managers doing it,” said Dean Chisholm, head of operations at Invesco. “Our fund managers tend to be stock-pickers.” He argued there were inherent issues in the reliability and consistency of big data that make its value questionable, although that is not to say it won’t become valuable in the future.
“It’s a back-to-basics data problem, where it’s a bit unreliable and inconsistent,” said Chisholm.
Another regional chief operating officer at a large fund house in Asia also questioned the reliability of the data and, quite pertinently, the data source, as well as whether the information would be fully up-to-date once it reached the fund manager’s table.
“There is a question of whether the data can be accurate enough to make investable ideas,” said the COO, who preferred to remain anonymous. “That will stem from whether they are the relevant data, in the right format, sampled within a specific timeframe and sourced from reliable sources.”
Anecdotal evidence from Asia also seems to point towards the same conclusion, in part because of lack of understanding and lack of trust. Amundi, BlackRock, Deutsche Asset & Wealth Management and Invesco are not employing it currently, at least not in Asia.
Indeed, a 2013 survey on institutional investors carried out by State Street revealed that just 31% of respondents in Asia viewed data as their most important strategic priority. That was only marginally below Europe (37%) and North America (44%).
But some are putting greater emphasis on it. The same survey also revealed that 84% of Asian institutional investors were strongly upbeat about investing more in data over the next three years, again in line with peers in Europe (84%) and the US (90%).
Read the full feature on big data in the current (March) issue of AsianInvestor magazine.