AI and the Price of an Easy Answer

AI may lower the cost of asking questions but raises the stakes for answering them.

Whether you’re an asset owner fine-tuning your portfolio, a bank or hedge fund creating systematic investment strategies, or a wealth manager assembling investments for clients, AI can help you sift through ideas, speed research and operate more efficiently.

But while no one can afford to ignore AI, no one fully trusts it either. Investors worry about the consequences of acting on an AI-generated answer that proves unreliable.

“Investors worry about the consequences of acting on an AI-generated answer that proves unreliable.”

This concern highlights a tension at the heart of AI’s arrival in global investing: You can’t be wrong — and you can’t stand still. Unless you anchor AI in data that can be traced to its source and validated, it can produce answers whose origins are unclear or, worse yet, that are simply wrong.

Our work as a provider of data, models, analytics and indexes gives us a window into how firms are using AI to lower costs, identify risks and opportunities, and assess performance. Here are three things we’re seeing.

With AI, the history and provenance of data become sacrosanct. For example, investors use indexes to understand markets in the present, but their ability to benchmark portfolios, test investment ideas and analyze risk depends on a historical record that has been built and maintained over decades.

While clients can access our index data through the AI tools they already use, I am struck by how many choose to access it directly via our MSCI ONE app. Portfolio managers, analysts and compliance teams say they believe they are more likely to get a correct answer within our own environment. (To be sure, clients choose: We offer the same validated data through the MSCI Connector.)

AI is also making it practical to tailor investment strategies on a much larger scale. Consider a sovereign wealth fund or public pension plan creating a custom strategy to mitigate risk or pursue outperformance, or a bank building a customized payoff profile of a financial instrument. With AI, they can explore, refine and manage customized strategies with greater efficiency and flexibility.

As AI expands the range of strategies investors can consider, we anticipate they will be able to describe their objective in everyday language and explore different ways to turn it into an index, testing and refining the strategy along the way.

While AI may hasten convergence among broad-market strategies — there are, after all, only so many indexes if you’re tracking global markets — the technology holds promise for investors who aim to see emerging risks and opportunities more clearly. AI can open access to fundamental data, such as geospatial intelligence and economic exposures, that can reveal risks and opportunities traditional financial data may miss.

Asset owners and managers are already engaged with AI but are grappling with what role AI should play in their investment processes. The most quantitatively sophisticated firms such as banks or hedge funds are already using it to accelerate research, strategy development and time to market, while other institutions remain earlier in the journey.

We know first-hand how AI is changing the way asset owners and managers work with their index provider. Among clients accessing our data through their AI, index portfolio managers are using advance estimates of benchmark changes; risk managers are monitoring multiple benchmarks for performance, return rankings, concentration and top holdings; and chief investment officers are combining alternative datasets with index information to uncover insights that constituent weights alone cannot provide.

Firms that pull ahead won’t necessarily adopt AI fastest; they’ll rebuild the investment process around data they can defend. In a world where anyone can ask a question in seconds, the advantage goes to those who can trust the answer.