Big data key to big alpha

The buy-side must harness ‘big data’ to generate alpha in equities investing by identifying trends ahead of the market, presenters at the inaugural Alpha Innovation Required summit concluded.

The buy-side must harness ‘big data’ to generate alpha in equities investing by identifying trends ahead of the market, presenters at the inaugural Alpha Innovation Required (AIR) summit concluded.

The concept that belies big data – using vast sets of information to distill actionable information – is not new to the buy-side, but the extent to which asset managers can use disparate sets of data will increasingly mark equities performance.

Bill Stephenson, head of global trading strategy for Franklin Templeton Investments, put together last week’s AIR Summit in Florida that brought together 20 companies offering innovative products to help the buy-side generate alpha, hewn from a shortlist of 220.

“There is simply a greater amount of data from a growing number of sources that impacts the investment process and asset managers who can best gather, manage, and contextualise this information will be able to make more informed decisions that drive alpha,” Stephenson told

One such firm, DISCERN, operates a platform aggregating sector-specific data on stocks for buy- and sell-side firms. Currently working across health, energy and real estate fields, the platform seeks to use the most granular data available to inform stock price forecasts.

Deep dive

Harry Blount, the firm’s founder and CEO, said such micro data included, for example, measurements from specific oil wells to adjust valuations of the owning energy firm. He said the key to buy-side use of data was a combination of machine and human curating, to block out the ‘noise’ and focus on relevant information.

“Seeing trends before they play out in the market by using larger and larger data sets will be the key to future alpha generation,” Blount said.

“There is too much time and resources spent on assembling and cleaning data, and not enough on analysing it,” he said. “The DISCERN platform taps into a vast set of data which can be correlated and modeled with other data sets to provide actionable information for buy-side investment decisions.”

Another offering, from solutions provider Kensho, seeks to save portfolio managers (PMs), quants and traders time by answering through a Google-like search engine macro questions addressing links between, for example, political events and stock prices. The search platform, named Warren, uses historical market data, which has been correlated to macro events.

“Warren is about productivity because it can respond in seconds to questions asked by quants, research analysts or PMs on a daily basis that have traditionally taken hours or entire days to answer,” said Adam Broun, head of strategy for Kensho.

Although Warren cannot yet deliver responses directly into algorithms, it instantly answers common assumptions PMs or traders may hold, such as improved retail sector performance during the holiday season, which Warren said has not occurred.

“The buy-side wants the vast amount of available data to be managed and packaged in a way that can be easily consumed,” Broun added.

Always connected

Social media will play an increasingly central role in this evolution of the buy-side’s use of data. Based on the growth of platforms like Twitter, Franklin Templeton’s Stephenson suggests information shared on social media will affect all elements of the investment process.

“Data from social media sources will likely play an increasing role in institutional investing, so asset managers should begin thinking about their social strategy in order to capture and take advantage of the exponential growth of financial related content,” he said.

“The initial key will likely be around the natural language processing of unstructured content within social media in order to automate interpretation and to integrate this sentiment as another variable in shorter-term decision making processes.”

Joe Gits, CEO of Social Market Analytics, a firm present at the AIR Summit that automatically monitors more than 50,000 profiles from Twitter and StockTwitz to provide sentiment information for specific equities, said social media was key, but only one segment, of the investment process.

“Social media is not [an asset manager’s trading] model, it’s only a factor in the model,” he said. “This is a statistical framework showing changes to stock prices.”