Welcome to Thoughts on the Market. I’m Ed Stanley, Morgan Stanley’s Global Head of Thematic Research. Today I'll discuss how understanding AI’s rate of change can generate alpha in the year of AI agents.
It’s Tuesday, the 14th of January, at 2 PM in London.
Even if you haven't been using artificial intelligence in your work or home life yet – you’ll doubtless have heard about its capabilities by now. Tasked, for example, with drafting an elevator pitch for a 100-page report; it's a tedious task at the best of times. But using an AI model not only does it become a breeze, but these models can also generate you a podcast – if you so wish – through which to disseminate it, and almost in any language conceivable. But now imagine the algorithm begins thinking through multi-stage processes itself – planning, executing – to generate that 100-page report itself, in the first place. That … is an example of Agentic AI.
As the name implies, this next phase of AI development is where software programs gain agency, transitioning from reactive chatbots that we’ve been using into proactive task fulfillment agents. And this transition is happening now.
Over the past 36 months, we’ve gone from reliable output that can displace or supplement 5-second or 5-minute tasks, such as translation or quick summaries, to models that are providing reliable output for 15-minute tasks, 1-hour tasks – like the ones that I just mentioned. And each time the skeptics have claimed that model improvements are slowing down, and thus call into question the returns on hundreds of billions of dollars that have been spent on AI infrastructure, the AI research labs manage to take another leap forward, surprising even seasoned analysts.
That’s why we think this is such an important trend for 2025. AI Adopter companies that can leverage these agents will start to pull ahead of their peers. And as a result, tracking AI’s evolution in the materiality of companies’ investment cases, we think, has never been more important.
Since our first AI Adopter survey in January 2024 to our latest just published in January 2025, we've seen profound shifts in the thousands of stocks that we cover globally. This ongoing transformation not only underscores that AI’s diffusion is advancing rapidly, but that we’re still very much in its early innings.
To understand the breakneck speed of the AI evolution through the lens of its impact on the stock markets, we need to wrap our heads around the concept of “rate of change.” We just published the third iteration of our AI mapping survey of 3,700 global stocks under coverage. And it reveals that 585 of those stocks had their AI exposure or materiality to investment case changed by our analysts – and that is just versus 6 months ago. And it impacts around $14 trillion of global market cap.
And this rate of change in AI isn't just a buzzword; it's a tangible metric driving outperformance. So, if we look back in the second half of last year, 2024, stocks where our analysts previously increased both AI exposure and materiality in our last survey – went on to outperform broader equity markets by over 20 per cent in the second half of 2024. If we apply the same logic looking forward, where do we think most outperformance is going to come from? It’s in those same stocks where our analysts have just upgraded the exposure and materiality to the investment case.
Beyond this simple screen for AI outperformers we think there are three other key conclusions from our latest survey. The first is AI Enabler stocks with Rising Materiality, within which we believe that Semiconductors, which have outperformed well, might soon pass the baton to the Software layer in terms of equity market dominance. Second, Adopters with Pricing Power. These are companies that adopt AI early and use it to expand their margins but sustainably, without having to give it back to their customers. And the third is Financial stocks, in particular, where AI Rate of Change has been the fastest of any sector in our global coverage – in terms of the efficiency gains that we think it can manifest for the share prices.
So all in all, 2025 promises a slew of significant developments in AI, and, of course, we’ll be here to bring you all of the updates.
Thank you for listening. If you enjoy the show, please leave a review wherever you listen to your podcasts and share Thoughts on the Market with a friend or a colleague today.
Thoughts on the Market is also available on Apple Podcasts, Spotify, Google Podcasts and other major podcast platforms.
For more Ideas, visit www.morganstanley.com/ideas