A discussion around gaining an edge through data during TradeTech Europe 2023 led asset managers to agree they would rather it be delivered in a ‘raw and messy’ format than structured, while simultaneously showing optimism around future tools and technologies. 
During the liveliest parts of the debate, speakers were asked how they prefer their data – raw or perfectly cleansed? To which two of the buy-side panellists selected the former.
“We would rather get raw data as close to the source as possible, and if there’s cleaning then we want to do it ourselves, that’s mostly because we want to understand any transformations that have been done to it,” said George Patterson, chief investment officer, PGIM Quantitative Solutions.
“The risk is always that something has been adjusted, or someone has applied a filter that makes it less useful. So our view is: always give us the raw data, and we’ll try to understand as much as possible what is in there.”
This was echoed by Gareth Shepherd, co-head, Voya machine intelligence and portfolio manager, Voya Investment Management, “It’s the old Keynes quote: ‘we’d rather be roughly right than precisely wrong’. Raw and messy data is what we would prefer to work with.”
Shepherd also hailed the use of artificial intelligence for data and highlighted how tools are enabling much smaller asset managers to pre-process the aforementioned raw data.
“To put number on it I know five years ago I would need 10 data scientists doing some expert system coding to pre-process raw financial data, but now I know I can plug into up and coming AI platforms that can do it almost automatically,” he added.
“The trend is in that direction so buyers of data want the raw and messy – why would you care about structured data and whether ESG is in certain fields when AI just looks through the noise?”
With the theme of the panel ‘which providers should you partner with to ensure best-in-class data that drives real-time market insights for your multi-asset trading desk?’ the speakers acknowledged they need to stay on top of the latest set of tools and offerings from data providers due to the incredible pace of change occurring.
“I remember having to beg and borrow, and now you get pestered by people having an innovation day,” said Patterson. “The problem is that most data is correlated or there’s nothing new in it.
“We’re at the beginning at the interesting period of time, the types of data that is getting collected – types of data from cell phones, censors in clothing, world will be very different in the next 10 years than we’ve seen in the past 20.”