Allocators Struggle to Predict Which Managers Will Outperform. IOR Is Teaming Up With Xponance to Fix That.

Investment Office Resources will offer clients access to Xponance’s fintech platform Aapryl, which uses machine learning to evaluate how investment strategies are likely to perform in different market conditions.

art_xponance_0809.jpg

Xponance is headquartered in Philadelphia, Pennsylvania. (Matthew Hatcher/Bloomberg)

Past performance doesn’t predict future results — but by using machine learning and factor analysis, Xponance says its fintech platform can get close.

Now, the platform will be more widely available to allocators through a partnership with Investment Office Resources, the outsourcing firm founded by ex-Mercy Health CIO Anthony Waskiewicz.

IOR, which provides nondiscretionary resourcing to existing investment offices, will provide its clients with access to Xponance’s proprietary fintech platform, Aapryl, which uses artificial intelligence to predict performance for investment strategies. In return, IOR will offer insights and product feedback to Aapryl.

“A lot of institutional allocators are looking for ways to build portfolios that don’t look like anyone else’s,” Waskiewicz said by phone. “That’s driving decision-making right now, but you still need some discipline around some of the selection process.”

Aapryl aggregates performance of more than 20,000 separately managed accounts, mutual funds, and ETFs in both public equities and fixed income, sourcing the data from Informa Financial Technology.

Using a factor-based analysis, the platform assesses past performance of funds. Aapryl then runs the investments through an algorithm in order to score them based on suitability for an investor’s portfolio.

Xponance launched Aapryl as internal technology more than 15 years ago. The asset management firm began offering outsiders, including managers and outsourced chief investment officer firms, access to the platform in 2017.

“Allocators rely on peer group rankings and ratios, which are useful in characterizing past performance,” said David Andrade, general manager of the platform. “They are limited in their ability to be predictive.”

Aapryl solves for this by using machine learning to predict how an investment strategy could perform in different economic environments.

“A lot of institutions talk about manager diligence through the lens of trying to examine what’s the skill and the luck component of a manager’s results,” Waskiewicz said. “It’s not that easy to get to the final assessment.”

Aapryl’s factor analysis, he said, can help sus out whether a manager is achieving strong results because of the market’s outperformance — or because they’re truly generating alpha.

Allocators can use the platform to assess investments on a manager-by-manager basis, or by using an overall portfolio.

“For us, it gets at more than just: ‘Is this a good manager?’” Waskiewicz said. It’s also, he adds “how we fit it into a portfolio.”

Related