Active Managers Believe Data Analytics Is Key to the Future — But Many Struggle with the Technology and Know-How

Fundamental active asset managers are rushing to incorporate data science into their investment processes and are increasing their spending accordingly. But, across the industry, the state of managers’ systems and technology needed to analyze and interpret the data are far from being streamlined and sophisticated.  

According to a Northern Trust survey of 300 asset managers released Tuesday, 98 percent of respondents said they are currently using, planning to adopt, or are interested in incorporating data science tools into their firms’ investment strategy in the next one to two years. Managers use data sets to inform their decisions on risks and opportunities for portfolio adjustments and to automatically execute trades based on signals and trends. They also increasingly use data analytics to inform and guide a firm’s investment thesis and evaluate performance. 

“There is a trend in the industry of moving away from an analog investment process to one that takes more advantage of data and technology,” Marc Mallett, Northern Trust’s head of strategy for asset servicing, told Institutional Investor. 

In just the past year, managers and banks have made significant investments in data science capabilities, so they can incorporate new sources of alternative information into investment strategies. In June 2020, Citigroup tapped Muir Paterson to lead a newly-formed strategic advisory group as part of a data science initiative for its banking, capital markets, and advisory teams, and at Wellington Management, a firm with over $1 trillion in assets under management, data scientists work alongside portfolio management teams. 

In the survey, the majority of respondents said they used environmental, social, and corporate governance data and factor data, such as that from MSCI, Wolfe, and Axioma, in their investment processes. Once firms acquire the data they need, asset managers analyze the data to help make decisions and generate alpha, or returns beyond what’s delivered by passive strategies.

Firms are increasingly using data science to measure the skills of their portfolio managers and better understand their investment decision-making process. Still, 48 percent of survey respondents said their organizations use a qualitative measurement to gauge the investment skill level of their team. From Mallett’s perspective, managers need to use data to have an accurate representation of their teams’ abilities. 

“I think it’s incumbent upon managers to really understand where they are getting their best ideas from, whether that’s certain individuals within the firm or certain processes,” Mallett said. “And having a clear understanding of that requires more than just qualitative or anecdotal evidence.” 

According to Mallett, the nature and scope of data has changed throughout the decades: “Today, everybody has access to more information than they can possibly analyze,” he said. “So what we're seeing now is managers asking themselves the question: how do we best use the information that we have?” 

For asset managers, the central challenge of data science is harnessing the data and synthesizing them for key insights. In the survey, 57 percent of respondents said their firms needed a centralized platform for consolidating investment data, and 83 percent of respondents said they expected the investment capital allocated to investment data sourcing or alignment to stay the same as or increase from the 2020 numbers in the next two years. 

“We’re seeing an increase in the number of data sources that firms are consuming,” Mallett said. 

In fact, 66 percent of survey respondents said their firms leverage around six to eight sources of investment data. In order to integrate the data into their firms’ process, 52 percent of respondents said they use spreadsheets. According to Mallett, firms will integrate various new data aggregation platforms with commonplace tools, like Excel.

But the report pointed out the pitfalls. “The dangers here are obvious. When people use spreadsheets, they tend to develop their own links, formulae, charts and even macros, creating a single person dependency.”