APG: How to balance man-machine using AI | Alternatives

APG Asset Management is a strong believer in 'man-machine' balance, a concept the fund first introduced in 2017 when it began experimenting with artificial intelligence (AI). Director Peter Strikwada said. $615 billion pension investor.

“While we use machines to increase intelligence, we will never fully automate decision-making or perform it without human oversight. All kinds of decisions in our investment portfolio People are always involved in decisions,” Strikwerda said. Asian investors.

Although human decision-making remains paramount, AI-driven modules are now being leveraged within the fund’s investment strategy.

“We have something called ‘real-time trading analytics,’ which advises traders on the best strategy to make a particular trade. For example, if you want to invest $50 million in company X. AI-driven The module provides guidance on what trades to do: at specific times and in the right venue,” he said.

APG’s best-known AI model is called “Samuel,” a digital portfolio manager that gives the fund’s alternative investment team instant access to information and analysis.

“This is a big problem because information in private markets is not as organized as it is in capital markets, so by using generative AI and language models, even people with no programming experience can use prompts to access the right data. We can do that,” Strikwerda said.

The company's next step is to create predictive AI models. This can help determine which alternative assets offer the greatest return on investment.

“That's what we're experimenting with right now,” he said.

think big, act small

Since beginning the experiment, APG has taken a structured approach to assessing the potential value of AI across the value chain. Part of that process includes keeping the heatmap updated regularly.

“This is essentially a top-down driven approach to pinpointing the areas where we think AI, particularly generative AI, will have the greatest impact,” Strikwerda said.

“We choose a few areas where the impact is the highest and the risk assessment is acceptable. And we typically think big and act small. That means, for example, we start with an experiment, or our colleagues experiment with it. We start by creating an infrastructure that allows us to collect evidence points and determine whether they make sense.”

For highly supervised pension investors like APG, managing the balance of opportunity and risk when it comes to AI is critical.

“I'm proud that our board of directors has taken such a positive stance on AI adoption,” Strikwerda said. “We are working with regulators to address dilemmas in risk management and strive to be at the forefront of implementing AI in investments.”

Augmented intelligence

Asset manager T. Rowe Price has also been building data science, machine learning and predictive analytics capabilities since 2017 to help investors, according to Jordan Binalb, director of the firm's New York City Technology Development Center. It is said that they are doing so.

“Through access to alternative data, natural language processing solutions, and predictive models, we have been able to provide data-driven insights to investment decision makers,” said Vinarub. Asian investors.

Like APG, the asset manager is adopting an “intelligent augmentation” strategy rather than automating decision-making with AI.

“We bring the power of data and insight to human decision makers in existing processes,” said Vinarub. “This approach has allowed us to evolve our capabilities in a thoughtful way that leverages the power of AI to move human decision makers to higher-value tasks within the investment process.”

As generative artificial intelligence (GAI) continues to evolve rapidly, he said continued collaboration and evaluation of new products is essential to understanding how best to incorporate this technology into business processes.

“GAI can be thought of as supporting the three Cs: consumption, characterization, and creation. As the amount of information inundating research analysts increases, GAI rapidly consumes textual content and It helps by finding signals among the noise and making it easier to create new content in notes, emails, and presentations,” Vinarub elaborated.

These features promise to improve the productivity of not only a company's overall investment process but also its business processes.

“Companies that navigate these changes and learn how to integrate these capabilities into their businesses will be in the best position to evolve and grow,” Binalb said.