GAM Preps for a Different Quant Future

GAM Systematic, a division of Swiss asset manager GAM, is investing in research and development related to artificial intelligence and big data to ensure it is at the center of what it thinks will result in a restructuring of the industry. 

GAM Systematic grew out of the firm’s acquisition last year of multistrategy systematic manager Cantab Capital Partners, which has $4.1 billion in institutional assets. The quant shop was part of the first wave of quantitative investing, replacing discretionary managers with computer algorithims to make investment decisions. The firm says AI and big data will represent the next wave. 

“Implementing algorithms that quantitatively sift through large amounts of data was a revolution. AI and big data is the next step change, or ‘scientization,’ of investing,” says Matthew Killeya, head of research for GAM Systematic. He adds that clients — seeing the potential of self-driving cars and programs like Google’s AlphaGo, which beat the world champion player of the Asian board game Go last year — are asking whether these types of advances can be applied to investments.

[II Deep Dive: Younger, Cheaper, Smaller, Smarter: Meet the New Breed of Quants]

Killeya says one area GAM is focused on is building algorithms to make sense of text — including news articles, speeches or reports — and then creating sentiment indices. Whether daily news or online market chatter is positive or negative toward a security could be a valuable data point, he says. Although there are multiple third-party vendors that analyze text, GAM insists on building its own proprietary systems, in part because of the nuances in interpreting this type of information. 

“With a third party, we could make more progress more quickly, but there would be less transparency into how the numbers are derived,” he says. “It’s important to understand what your systems are doing, rather than buying a stream of numbers,” adds Tom Howat, GAM Systematic’s chief technology officer.

GAM is also evaluating how to use weather data, including the specific amount of sunlight or rainfall in a given location, which would affect agricultural markets or securities like Kansas wheat futures contracts. Another data point GAM is looking at is the location of ships. Publicly available information shows, for example, a ship’s cargo, capacity, schedule, and other data. Killeya says delays or difficulties that a ship may encounter can affect the price of commodities. 

“We can invest in coal, for instance, and transportation is a critical link in the price of coal. That could give us an information edge and enable us to have a better-informed portfolio,” says Killeya. 

New data sets are turning up daily, and GAM can only seriously evaluate a small fraction of them. “There are many people knocking on our door with everything from aerial photos of gas tanks in China to hotel room bookings,” says Killeya. “All these things are fascinating, but we don’t have infinite resources.” 

Anthony Lawler, co-head of GAM Systematic, says the firm has used machine learning — albeit in a less advanced form than what is now available — for ten years to perform tasks such as scrubbing data and helping with trade execution. Although Lawler thinks machine learning will have a big impact on GAM and the industry, he says it’s important not to get carried away with the hype around these technologies. “Think of it as another avenue of normal research; not something that is totally game changing,” he says. “At least yet.”