Over the past 50 years, technology has been dominated by mega-themes, from the dawn of the commercial internet to the rise of cloud computing, mobile, machine learning and now artificial intelligence (AI). With any theme of this magnitude, the question inevitably arises of whether it is over- or under-hyped – and whether the valuations of companies tied to those themes are justified.
Recent volatility in AI-related equities sharpens this enquiry. To gain perspective, investors must understand the magnitude of the transformation and how it set to touch most corporate sectors, impacting revenue growth, productivity and, in many cases, sending business models toward obsolescence. AI is no longer just a question for equities investors, as typically conservatively funded mega-cap tech companies are tapping debt capital markets at record levels.
Perhaps the most important thing to consider when weighing valuations for today’s mega-cap tech stocks is that it’s still early days in the AI investment cycle. The technological shifts that made AI possible occurred over 25 to 30 years, and those advancements were critical in laying the groundwork for the progress seen in the past three years.
Dispelling the myths of an AI bubble
There are a couple of key areas about AI investing that dispel concerns around an AI bubble. First, the reason AI has been one of the main drivers for the market is because investors are recognising how profound this technology shift is – not only in terms of the revenue opportunity created by the winners but also the meaningful productivity and margin lift expected as AI proliferates into the broader economy.
Second, the focus tends to be on the digital manifestation of AI, which could have a significant impact on many sectors. But the physical manifestation of AI could be equally profound, with progress being made in areas such as autonomous driving and robotics.
The tech stocks of the early 2000s looked expensive at the time but ultimately grew earnings much faster than anticipated. The same is likely to occur with AI as companies compound earnings and free cash flow materially faster than investors expect. And as the AI buildout cascades from infrastructure companies (enablers) to well-positioned software providers (enhancers) – and ultimately into the broader economy – that is the point where forward-looking companies (end users) could reap significant economic benefits.
Funding the AI boom
With the bullishness around AI comes historic CapEx budgets. Many AI hyperscalers have deviated from their script and have aggressively tapped debt capital markets, increasing leverage in the process. This shift has increasingly made the AI investment theme just as much of a fixed income play as an equities play.
New issuance varies widely across subsectors. Hyperscalers are tapping into the investment-grade (IG) credit markets for AI financing. This could put pressure on currently tight spreads in IG as that debt is absorbed. The high-yield market appears better positioned, as direct AI debt issuance is under 1 per cent of Bloomberg US Corporate High Yield Index as of February 2026.
Additionally, there is room for growth in AI-related issuance in securitised credit. Commercial mortgage-backed securities (CMBS) are a major focus given the funding required to build new data centres. Furthermore, shorter time horizons associated with projects funded by CMBS may provide greater visibility in issuers’ ability to meet their obligations. On that front, data centre growth is likely to persist. Some industry estimates suggest that meeting AI-related demand could require power equivalent to roughly three New York Cities by 2030.
As some AI hyperscalers seek to own data centres, private deals could increase. Ultimately, companies will pursue the lowest cost of capital and look to diversify funding sources. The massive CapEx programmes of investment-grade hyperscalers stand to benefit companies across the fixed income spectrum as spending flows through the AI data centre supply chain.
CapEx of the big three hyperscalers (Microsoft, Amazon, Google)
Capex trends suggest data centre growth is here to stay.
Source: Dell’Oro, JP Morgan estimates. CSP capex by: Microsoft, Google, Amazon.
From rising tides to the imperative of selectivity
From both an equity and debt perspective, one of the biggest concerns around AI is obsolescence, especially since it has advanced so quickly. While AI applications such as ChatGPT may feel as though they became ubiquitous overnight, it is important to recall that the foundation for these technologies were laid over decades. Likewise, the work of continuing to build out the AI training infrastructure are likely to continue for many years. And the time horizon for end users to not only implement but reap benefits from AI is longer still. Meanwhile, the inference stage of this technology appears poised for rapid growth as the shift towards a more agentic model accelerates.
The rising tide will not lift all boats forever. Looking across the spectrum, some AI infrastructure companies are emerging leaders and continue to differentiate themselves; based on projected cash flows, their valuations do not appear stretched given their role in driving the technology transformation. Likewise, there are companies in the enhancer space that have developed competencies over many years and that are navigating the AI infrastructure more efficiently than their competitors.
At the same time, many areas will face significant disruption. For example, AI stands as a threat to software as a service (SaaS) companies, although those that embrace AI not as a feature but as the foundation of their operating models could emerge as winners. It’s also worth noting that some tech stocks are trading at 50 times revenues due largely to the “AI halo” effect, even though they have limited revenues or products and services that may not even be deployed until 2030.
AI disruption will not be linear. While the direction of travel over the next decade is clear, what happens over a year or two is not. Technological shifts of this magnitude are a fertile environment for active management. Selectivity is likely to be key to identifying the next wave of winners and capitalising on the broad spectrum of opportunities that arise as the AI transformation continues in the years ahead.
By Denny Fish, Portfolio Manager, Research Analyst, Janus Henderson Investors
John Kerschner, CFA, Global Head of Securitised Products, Portfolio Manager, Janus Henderson Investors
John Lloyd, Global Head of Multi-Sector Credit, Portfolio Manager, Janus Henderson Investors
