Key takeaways:
- Power availability, permitting challenges, and debt market capacity to absorb AI-related paper could constrain the pace of the AI infrastructure buildout in the years ahead.
- As AI- and hyperscaler-related issuance grows, debt market capacity could force the buildout to draw on alternative sources of capital including equity markets and potentially even government funding.
- A more measured buildout could result in a less pronounced investment upcycle and therefore a less severe down cycle – should one occur. In the interim, the value of existing infrastructure, grid connections, and power capacity, is likely to increase.
IMPORTANT INFORMATION
Artificial intelligence (“AI”) focused companies, including those that develop or utilize AI technologies, may face rapid product obsolescence, intense competition, and increased regulatory scrutiny. These companies often rely heavily on intellectual property, invest significantly in research and development, and depend on maintaining and growing consumer demand. Their securities may be more volatile than those of companies offering more established technologies and may be affected by risks tied to the use of AI in business operations, including legal liability or reputational harm.
Fixed income securities are subject to interest rate, inflation, credit and default risk. The bond market is volatile. As interest rates rise, bond prices usually fall, and vice versa. The return of principal is not guaranteed, and prices may decline if an issuer fails to make timely payments or its credit strength weakens.
Sources and definitions
1 Source: Morgan Stanley Research, as of Q3 2026.
2 Source: Hyperscaler issuance data as per Janus Henderson estimates, as of Q3 2026.
Capital investment cycle: The purchasing of fixed assets by a company for the purpose of supporting day-to-day operations.
Hyperscaler: a large-scale cloud service provider that operates global networks of data centers, managing servers to deliver on-demand computing, storage, and networking.
View transcript
James Donahue: Here at Janus Henderson, we see three key bottlenecks to this AI build-out. We view this through the three Ps. Number one: Power. Number two: Permitting. And number three, and perhaps most overlooked: Paper.
We continue to believe that power is a key gating factor to this build-out. When you look at some estimates on the sell side, some folks are calling for 36 gigawatts (GW) of incremental power added in 2028.1 The U.S. grid can generally only add 20 to 25. Some of that gap can be perhaps addressed by behind-the-meter solutions, such as energy cells or bring your own generator, so to speak, but not all of it. And of those specific behind-the-meter solutions, they come with their own litany of regulatory issues – which brings us to permitting.
You can see that while these projects may ultimately go through, they’re being hit with a series of delays vis-a-vis permitting and folks, particularly the constituents that have to live next to this technology, not being super thrilled about the technology being within a five-to-ten-mile radius of them. One example is in southern New Mexico in Doña Ana County. There’s a 2.1 GW project, known as Project Jupiter, that came under a series of project and permitting delays – notably with respect to pipelines that were surrounding the project or leading into the project. And as a result, delays are sort of inherent to this build-out as well.
Lastly, paper. While power and permitting are fairly well discussed in most market circles, we think paper is a less known aspect to this build-out. Ultimately, our view is that the broader market just expects the debt market to absorb this paper. We think it may be a little more nuanced than that. When you look at next year’s estimates, there are some folks on the street that are calling for an excess of $400 billion in hyperscaler-related debt issuance. This year, we’ll probably land at around $275 billion.2 So that’s a big step up in a market that’s already fairly saturated with this paper. That also doesn’t account for some ancillary uses of capital or sources of capital as well, namely GPU- (graphic processing unit) related paper. Fixed income investors may reach a threshold of the amount of paper that they can own, vis-a-vis AI and hyperscalar debt. Ultimately, that may mean that other uses or sources of capital may need to emerge. Namely the equity market or perhaps even governments.
When you look at all of these different bottlenecks, the build out is actually a lot more complicated and challenging than previous cycles. There are a lot more hoops to jump through. In other words, the builders here have to be disciplined not because they want to, but because they have to. This could beget a less steep build out, which in our view is a good thing. A less steep build out and a less steep upcycle could ultimately beget a less steep downcycle, should that occur.
Lastly, we see a lot more value in the existing players as well. If it’s harder to add one net new GW, then your existing one GW sees more demand. Ultimately, this could benefit power shell providers and people who already have a connection to the grid, along with the GPUs that are ultimately slotted into those associated data centers as well.
Related insights
These are the views of the author at the time of publication and may differ from the views of other individuals/teams at Janus Henderson Investors. References made to individual securities do not constitute a recommendation to buy, sell or hold any security, investment strategy or market sector, and should not be assumed to be profitable. Janus Henderson Investors, its affiliated advisor, or its employees, may have a position in the securities mentioned.
Past performance does not predict future returns. The value of an investment and the income from it can fall as well as rise and you may not get back the amount originally invested.
The information in this article does not qualify as an investment recommendation.
There is no guarantee that past trends will continue, or forecasts will be realised.
Marketing Communication.