Artificial intelligence continues to drive stock market momentum this year, even on days when broader indexes falter. Investors are betting on the buildout of computing and power infrastructure.
BNN Bloomberg spoke with Doug Rogers, managing director and co-head of core/growth equity at Morgan Stanley, says the outperformance of AI-linked names depends on continued growth in data centre spending.
Key Takeaways
- Forecasts for 2027 data centre capital spending have doubled to nearly US$500 billion, underlining the scale of expectations built into AI stocks.
- Energy demand is surging, with Nvidia’s latest announcement implying power needs equal to New York City’s consumption.
- Efficiency is a priority, with data centres exploring lower-power chips such as custom ASICs to reduce electricity costs.
- Conventional application software firms, including Adobe and Constellation Software, may see growth pressures as AI alternatives gain traction.
- Stock opportunities extend beyond tech to areas like security software and infrastructure firms positioned to benefit from AI adoption.
Read the full transcript below:
ANDREW: Let’s stick with that theme of AI-related optimism. Our guest says outperformance by many of these AI names, including the chipmakers, has been fuelled by growth in data centre spending. We’re joined by Doug Rogers, managing director and co-head of core/growth equity at Morgan Stanley. Thanks very much for joining us.
DOUG: Thanks, Andrew.
ANDREW: Do you think the market is vulnerable in the sense that if this data centre spending slows, that could cast a shadow over these AI stocks?
DOUG: Yes, for sure. The growth of the market has certainly been driven by the capital spending that’s being reflected at the data centres. If you step back and look at the beginning of 2024, the estimates for 2027 data centre capital spending by the big hyperscalers was just over US$200 billion. Now, the estimates for 2027 have more than doubled, approaching US$500 billion in anticipated data centre spending. So there’s a lot of expectations baked into these stocks.
I think that growth is there, maybe best evidenced by yesterday’s Nvidia announcement about its investment in OpenAI. There’s some circularity to that — Nvidia investing in OpenAI, which invests in Oracle, which then buys Nvidia chips. But it’s a clear indication the demand is there and the spending will continue.
ANDREW: What about the energy names? I’m thinking of Constellation or Vistra. Without commenting on those stocks specifically, are the electric generators set for more upside?
DOUG: I think so. If you look at the Nvidia announcement yesterday, I believe it was for 10 gigawatts of power. That’s about the power requirement of New York City. That’s a tremendous amount of power generation that will be required and delivered. So I think some stocks that play in that area could certainly see upside in the future as well.
ANDREW: It’s interesting that an investment in computing power is expressed in terms of power consumption. I haven’t heard that metric used so directly before.
DOUG: It is. It’s a big driver of expense in data centres, which is why there’s a lot of emphasis on lowering chip power consumption. Some data centres are considering lower power chips. I’m drawn to a company like Avago, which plays in custom ASICs. It’s a more streamlined, function-driven chip that results in lower power consumption. That’s a big focus for data centres given the power requirements needed by all the chips.
ANDREW: There are concerns among investors in Adobe or, in this country, Constellation Software, about application software companies. Conventional players could be challenged by AI. Do you think there’s anything to that?
DOUG: I think there is. The first thing likely to be displaced is a lot of lower-end applications used by big enterprises. If not disrupted directly, there may be cheaper alternatives, so growth could slow for some of these companies.
That said, companies like Adobe and Constellation are uniquely positioned. Constellation focuses on a broad array of vertical software with very specific needs. Adobe meets certain requirements, including managing a large digital library with copyright issues for content production. Those are barriers that will be difficult to overcome with AI. Still, AI could delay spending on some of these providers or at least limit growth before AI matures further.
ANDREW: You have an idea for us — Okta. They’re a security or access control company with an AI angle. What’s the case for Okta?
DOUG: As you mentioned, Okta is a pure-play security vendor and a leader in identity and access management, or IAM. They control who is able to access apps or networks. The company had some missteps — including a security breach and a salesforce execution issue — but has done a good job recovering, regaining trust and strengthening its products. They now have a new go-to-market strategy. It’s a bit of a self-help story.
The AI angle is important. Right now, IAM is mainly used for people — making sure whoever accesses a network is who they say they are. But there will be many more machine-to-machine interactions, and those machines will vastly outnumber people. Any time a machine tries to access inventory or customer data, you’ll need to verify that source. Okta is well positioned to lead in that area, and we think that market will grow meaningfully in the coming years.
ANDREW: I guess we may be heading toward a world where wars are fought with drones and cybersecurity means AI agents battling each other. Doug, thank you very much indeed. Doug Rogers, managing director and co-head of core/growth equity at Morgan Stanley.
---
This BNN Bloomberg summary and transcript of the Sept. 23, 2025 interview with Doug Rogers are published with the assistance of AI. Original research, interview questions and added context was created by BNN Bloomberg journalists. An editor also reviewed this material before it was published to ensure its accuracy and adherence with BNN Bloomberg editorial policies and standards.
