'We haven't solved the data problem', data overload meets limited infrastructure

Technological progress has simplified many aspects of life, including how retirement plans are designed and managed. For plan sponsors and participants, the growing role of data is enabling more informed decision-making and driving more tailored, participant-specific advice.

But therein lies the double-edged sword as one of the most challenging aspects plan sponsors regularly face is using effective data to manage plans. Cynthia Shaw-Pereira believes sponsors and asset owners are increasingly focused on extracting greater value from their data, but also must deal with constraints around resources, technology, and strategy.

There’s also a growing urgency among plan sponsors to improve how they consume and use information, driven in part by what she calls the “three Vs” of data: volume, velocity and the veracity of data. She underscores the core issue facing asset owners today isn’t access to data, rather it’s the ability to use it effectively. Many are “challenged significantly” because they don't have the talent, resources or the spend to manage or undergo a large transformation around data.

“We are at a stage right now where we haven’t solved the data problem,” said Shaw-Pereira, vice-president, head of asset owner segment at CIBC Mellon. “The volume of data has increased significantly, and the information is there. Across asset classes, there's been a lot of work to make sure that people have and can get to the information that's available. The challenge is, how do you consume it?”

That’s where she sees technology playing a pivotal role, helping firms streamline workflows and boost efficiency. She pointed to a clear distinction between large institutional plans like the Canadian Maple 8 and their smaller counterparts. While she noted the large plans are well-equipped with resources, technology and talent, the story is very different for mid-sized asset owners managing portfolios in the $5 to $20 billion range.

The main issue is ultimately scale. Because these smaller plans often can’t afford to build large internal tech infrastructures, many are turning to outsourcing models instead of building in-house capabilities.

Shaw-Pereira noted a trend toward relying on external partners who are already investing in the necessary tools and systems. That shift includes exploring solutions from fintech firms but even asset owners and plan sponsors aren’t always equipped to manage implementation on their own.

Consequently, there’s a growing demand for advisory support from service providers. While the goal is to help asset owners build a comprehensive, unified view of their portfolios, she acknowledged that most are still far from achieving that.

Chief to that, daily pricing and standardized data in public assets make them relatively easy to manage. But it’s the private assets that are presenting the real challenge right now as sponsors often have to manually retrieve data from investor portals, Shaw-Pereira explained.

They also need to understand how to use third-party data sources effectively. She underscored knowing how to integrate internal and external data is becoming a critical part of creating a comprehensive and accurate investment view.

“Third party providers are quite common within firms that we work with,” she said, highlighting Bloomberg, Bloomberg, FactSet and MSCI as other commonly used sources. “They are using other providers to supplement for asset servicing data, but they have that challenge of how that information marries with the information from the service provider and how do you actually pull that together?”

So how can plan sponsors solve the data dilemma? Shaw-Pereira highlighted while artificial intelligence is beginning to play a role in both data management and decision-making, many firms are still navigating how best to implement it.

“Everybody wants the ultimate use case of being able to use AI to get to insights,” she said, pointing to the ideal scenario where users could simply ask, ‘What’s your exposure to the US?’ and have an AI tool like ChatGPT or Copilot instantly search, aggregate, and deliver a clear answer.

While some asset owners are actively piloting AI solutions, she emphasized most are still working through key concerns, like getting comfortable around how they want to use it and adjusting to governance and guardrails around how the data is being used, she explained.

Instead, the starting point should simply be to figure out what dataset is a priority, said Shaw-Pereira, acknowledging that it also depends heavily on the makeup of the portfolio. Additionally, the data requirements vary widely based on asset class. For example, a plan with significant allocations to private equity or private credit will need different metrics than one focused on infrastructure or real estate.

As for the exact information or dataset that plan sponsors should focus on, she pointed to data in custody services, pension accounting and performance calculations. Yet, even consolidating these core datasets can be a struggle because each present data through a different lens. Shaw-Pereira also emphasized that many continue to rely heavily on Excel to make it all work. However, she warned that this approach isn’t scalable in the long-term.

“It’s not entirely sustainable to have everything based in Excel and to have that be the basis of how you’re going to run your technology and get access to your data,” she said.

Ultimately, she believes smarter tech adoption will be key to resolving long-term data challenges and enable plan sponsors to enhance their data strategies by outsourcing to partners with robust tech capabilities, using cloud platforms for integration, and adopting fintech tools that simplify data normalization. Peer collaboration and networking will also play a key role in identifying effective solutions, all with the goal of achieving a unified view across diverse asset classes.