The explosive growth in private satellites, social media, and the Internet of Things has delivered a wealth of new alternative data sets that portfolio managers and traders can use to capture additional alpha.
“Some of the larger quant trading firms have been doing this for a long time,” said Valerie Bogard, an analyst at industry research firm Tabb Group and author of a recent research note on the subject. “But now, we are seeing growth in more long-only firms and more traditional asset managers are starting to explore this area more.”
What is the definition of alternative data? Bogard describes alternative data as what it is not.
Valerie Bogard, Tabb Group
“It’s not the structured market data the one would expect,” she said. “It would be social media data, data coming from Twitter as well as image data from satellites, sensors, location data. It is a wide variety of data sets.”
For Point72 Asset Mangement, the firm focuses on deeply understanding the fundamental drivers of the companies in which it invests, according to Matthew Granade, the chief market intelligence officer at Point72.
“Alternative data is instrumental to our process in that it gives us a great lens – often a high fidelity, high-frequency look — into the detailed functioning of the companies we are trying to understand,” he said. “It literally can be used to paint a picture of exactly how the company is functioning.”
For example, through the analysis of credit card data, it is possible to know a tremendous amount about how the business functions, according to Granade. “Other things we focus on are geo-location, satellite imagery and information on app downloads.”
Point72 looks to invest more into alternative data sets and has established Point72 Ventures, an early stage private equity firm that invests in some emerging alternative data providers.
“This will arm the firm with the most powerful technologies available and set us apart from others in the industry,” he said. “We want to be the Google of finance.”
The challenge of incorporating alternative data into trading and investment strategies is evaluating the new data’s overall worth, noted Bogard.
“You have to evaluate the new data source, ensure the data is clean and actionable, and that you can monetize it,” she said. “You also have to look at who else is using the data, how much alpha you gain compared to the time and energy spent analyzing the data.”
Point72’s process “begins with the companies we want to trade and a deep understanding of the fundamental drivers of those companies,” he said. “We then broadly map types of data and data sources to those drivers. Every vendor must then undergo a detailed compliance review to ensure the data is something we can use.”
Point72 also strips out the personally identifiable information from the alternative data that it uses.
“We’re not interested in Adam Smith drinking Starbucks at 9:15 AM on a Monday,” said Granade. “We’re more interested in the aggregate-level data. We then get into the hard work of understanding what the data is, what it’s telling us, how it fits with other sources, and how it maps to the fundamental company drivers we are trying to understand.”
Although Bogard classifies many of the alternative data vendors as startups, with less that two years of historical data, she foresees the vendors establishing distribution relationships with established data providers.
“It may be akin to the social media analytics firms that partnered with larger data providers to distribute their data more widely,” she said. “It wouldn’t surprise me if a similar trajectory occurred for some of these other players.”
More on Alternative Data:
