How APG used AI to develop a 'digital portfolio manager'
APG Asset Management, one of Europe's largest pension fund managers, has emerged as an early adopter of artificial intelligence (AI) in the investment management industry, particularly within private markets, according to Genio van der Schaft, chief operating officer of APG Investments Asia.
"We started innovation initiatives around 15 years ago, which gradually evolved into digitalisation efforts, with AI being part of that," van der Schaft told AsianInvestor’s COO Forum in Hong Kong on September 24.
With approximately $670 billion in assets under management, including a nearly $50 billion allocation to Asian markets through its Hong Kong base, the Dutch pension fund aims to remain at the forefront of integrating AI into both its workforce and investment processes.
"Our focus areas are automation, efficiency, analytics and visualisation, machine learning, and workforce skills development."
DIGITAL PORTFOLIO MANAGER
One of APG's most notable AI-driven innovations is "Samuel", a project designed to make information and analytics in the typically opaque private markets instantly accessible to the fund’s alternatives investment team.
APG Asset Management
"We embedded a team of data specialists within our private markets team. This team created Samuel, which started in 2017 as a 'digital butler' providing data and analytics. It has since evolved into what we call a 'digital portfolio manager'," van der Schaft explained.
"Samuel captures data and has an embedded language model so you can query it. It now provides not just insights but also recommendations. Our portfolio managers work with Samuel, scrutinising its recommendations to develop solutions together."
The AI tool's applications are diverse and impactful.
"We've also combined this with GPS data to verify underwriting assumptions - for example, looking at traffic patterns for shopping malls or toll roads," he said.
Also read: APG: How it strikes the man-machine balance with AI
Perhaps most importantly, Samuel also functions as a repository of institutional knowledge.
“Samuel captures the collective knowledge of the entire team, building our institutional memory so knowledge isn't lost if people leave."
The development of Samuel aligns with APG's strong focus on private markets, which include areas like real estate, infrastructure, and private equity.
“These markets aren't readily accessible, and we are able to generate returns by capturing things others can't."
AI-EMBRACING CULTURE
Van der Schaft's approach to fostering a culture that embraces AI and digitalisation at APG is multifaceted.
"As COOs, we're primarily facilitators. The core business is investing and trading - that's where the value comes from. However, there needs to be supporting machinery, and we as COOs can facilitate the underlying basics like IT infrastructure and data management," he said.
To embed AI into the company culture, van der Schaft highlighted several initiatives.
"We've set up what we call 'Digi tribes' across all locations. In Hong Kong, we have a data scientist who works with representatives from each team. Globally, we have tribes focused on tools like Power BI or Python."
"The key with these tribes is what I call the three E's: explore, exchange, and enjoy. This approach helps make the technology less threatening and brings the latest innovations to our teams."
APG also organises various activities like lunch-and-learns, Digi tribe meetings, hackathons, and workshops.
Last year in Hong Kong, the fund held a Digi competition where teams could propose ideas, with the winning idea guaranteed to be implemented with expert help.
Van der Schaft also underscored the importance of talent management in APG’s team.
"For workforce and skills, when hiring new talent, we look for those with digitalisation skills. We also work with existing staff to empower and inspire them to use these technologies, helping them see it as an opportunity rather than a threat as our industry evolves."
FUTURE OF AI
Looking ahead, van der Schaft envisions numerous opportunities for AI in investment management.
"We're working on a private ChatGPT-like environment, as we have proprietary data we don't want shared externally. We're also focusing on giving our people 'data at their fingertips' and developing self-service models so they can work more independently without always needing expert help," he said.
The technology is being utilised across a diverse range of applications.
"We're applying AI to responsible investing, such as in engagement and proxy voting. In some cases, we're using AI for seemingly simple tasks, like helping our responsible investing teams visualise risks when signing off on deals."
However, van der Schaft emphasised that AI’s true value lies in enhancing the investment process.
"We're using tools like Samuel to find and source new deals, shorten due diligence cycles, and add real value to our investment process. That's where we're really trying to push the frontier."
When it comes to measuring the success of AI initiatives, van der Schaft is clear.
"The most important measure is our investment return and the ideas we can generate as investors. While process automation and efficiency are important, the real business value for us is related to investments."