Featured Image

Data Overhaul Paid off This Year, Says AIA's Konyn

Data Science and AI

By Jame DiBiasio  |  October 20, 2020

This article originally appeared on DigFinGroup.com.

Large asset owners such as insurance companies are devouring data like never beforewhich requires a firm-wide strategy to manage.

“Data needs total integration,” said Mark Konyn, Group Chief Investment Officer at AIA, the Hong Kong-based insurer. AIA operates across 18 markets with group assets under management standing at $284 billion as of end 2019.

The firm has spent the past four years investing in the infrastructure and governance around how it sources, verifies, stores, organizes, and utilizes data.

That work paid off earlier this year when markets were roiling under the impact of the COVID-19 pandemic. Fund managers had to make important trades in very short periods of time, which meant they needed the means of analyzing the right kind of data and having it move efficiently through the organization to act on the investment team’s insights.

Flexible Foundations

Getting to this point is a monumental task that in AIA’s case has led it to completely overhaul how it operates. Konyn explains that traditionally financial institutions relied on proprietary systems or vendor solutions that were designed bespoke. The idea was that vendors should adapt their products so a customer didn’t have to change their operations to use it. This led to expensive, multi-year implementationsand a huge overhead.

Although this approach could give big firms a powerful system, it means they’d be left behind whenever there was an innovation in the marketplace or a new set of regulations for which to adhere.

“When you’re the only user of your system, you only find out about problems when they arise,” Konyn said. “Bespoke systems become an albatross.”

The software-as-a-service model has turned this on its head. Now firms are framing their systems around vendor platforms so that data can flow through the organization seamlessly.

Todd Hartmann, Senior Vice President of Strategy within the Content and Technology Solutions group at data vendor FactSet, says the company has responded to this trend by moving from an IT focus to a business focus, providing enterprise solutions to chief data officers beyond just providing data to various siloed departments.

The Rise of Alternative Data

The influx of alternative datasets has made this even more important. Vendors now spend a lot of time on symbology: putting consistent, uniform identifiers at the level of the underlying security, like ID building blocks. This helps ensure investors receiving data from an array of sources can readily integrate it into their systems.

“Investors spend a lot of time wrangling with the challenge of connecting new datasets,” he said.

Konyn says making sure data is clean is one of the biggest challenges to an organization. Consistency is another; for example, data for evaluating a stock isn’t the same as that used to price a derivative on it. Even something simple as a stock price is actually complicated depending on factors such as whether the price was struck midday or at a market’s close.

“Downstream analytical tools need consistent management of data,” Konyn said. “It’s coming from different vendors, who use different methodologies.”

Bryan Lenker, Vice President, Director of FactSet’s Content and Technology Solutions, says vendors spend effort on “concordance,” ensuring data fits with existing master-security models—a task that has become more important with the rise of alternative data sets.

Konyn says asset owners must ultimately put in place their governance models to get the most out of data, which includes new standards to evaluate vendors as well as upskilling the firm’s people. AIA now has a data-management office function within its investment team.

“The use of data requires a design, knowing what’s required, and investing in governance, management, and systems,” Konyn said. Doing so allows organizations to know where to find what they need and pull things together quickly.

It also allows investors to put controls on data, which means building in safeguards so that AI-driven investment decisions don’t breach risk budgets or portfolio mandates.

“We have vast pools of liabilities in different locations, matching different product lines,” Konyn said. “If we tried to do this manually, we could not implement investment decisions at scale.”

Jame DiBiasio

Founder and Editor, DigFin Group

Mr. Jame DiBiasio is an award-winning financial journalist and author based in Hong Kong. He launched DigFin in March 2017, serves as co-chair of the FinTech Association of Hong Kong's WealthTech Committee, and is the founding editor of AsianInvestor magazine (2000). Prior, he served as editorial director at Haymarket Financial Media (2013-2016) and was the first-ever journalist in Asia to be granted the “Outstanding Contribution to Institutional Journalism” by the State Street Institutional Press Awards. Mr. DiBiasio earned a BA in International Relations and Affairs from American University and a master's degree in International Relations and Affairs from Johns Hopkins University.

Comments

The information contained in this article is not investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.