The past few years, commentators have confidently predicted the “AI bubble” is about to burst and every year, those forecasts prove wrong. The reason is simple: bubbles are defined by speculation on worthless assets. Prices rise only because investors believe someone else will pay more later. Regardless of intrinsic value.
The AI expansion is nothing like that. It’s built on something bubbles never have:
- Hard earnings
- Meaningful infrastructure investment
- productivity gains
- structural demand.
Chart 1 & 2 – S&P 500 market returns during the dot com bubble and the last two years
Source: Bloomberg. Data as of December 2025.
The key question for 2026 isn’t “will AI crash?”, it’s “are high valuations already pricing in all the future growth?” The answer is: not even close.
- Valuations have risen, but earnings have risen faster
This is the misunderstood part. Since 2022, the major AI leaders’ earnings have outgrown their share prices.
- Revenue revisions for Microsoft, Amazon, and Alphabet are all up since 2023
- Nvidia's earnings revisions have increased +900% in 24 months.
- Data-centre suppliers like Broadcom, Marvell, and Super Micro have seen EPS forecasts double over two years.
Investors aren’t paying more for the same earnings; the earnings base has exploded.
Chart 3 & 4 –EPS and stock price growth for Nvidia, Meta, Amazon and Alphabet
Source: Bloomberg. Data as of December 2025.
- 2026 will be the biggest AI capex year yet. The market isn’t fully pricing it in
Despite all the recent headlines, 2026 is shaping up to be the year where AI spending broadens:
- Telecommunications begin full scale network AI deployments
- Banks, insurers and healthcare move from pilots to production
- Infrastructure spending accelerates to support data centre demand
- APAC & Europe add the largest pipeline of new cloud regions ever announced
- Enterprise agent-based AI systems reach commercial maturity
Table 1 – AI adoption across industries shows many businesses are still in early stages
Analysts expect another double digit global AI capex increase in 2026, following record spending in the last 2 years. Yet valuation models still assume AI spend decelerates from 2027 onward.
Chart 5 - Capex actual and forecast 2023–2027 across Magnificent 7 stocks
Source: Bloomberg. Data as of December 2025.
- Productivity gains are only just beginning to show up in data
This is the biggest piece not priced in. In 2024–25, AI adoption was mostly experimental. In 2026, organisations will begin capturing real, trackable productivity benefits:
- Software development productivity gains of 20–40%
- Customer service automation reducing cost-to-serve 30–50%
- Hospitals deploying AI documentation + imaging workflows
- Manufacturers adopting AI agents for routing, maintenance, and logistics
Productivity gains expand corporate margins, yet these gains are barely reflected in long term earnings models. If AI becomes a margin tailwind, today’s EPS forecasts are still too conservative.
Table 2 - Enterprise AI adoption shows the productivity benefits across business units
- Structural bottlenecks prevent the one thing that pops bubbles: oversupply
Dotcom crashed because supply massively overshot demand. AI cannot overshoot as the system is constrained by:
- Limited high-bandwidth memory (HBM)
- Power grid bottlenecks
- Data centre zoning/land scarcity
- Transformer and substation backlogs
These structural bottlenecks protect profitability. They keep pricing power high, reduce competitive entry and prevent a 2000 style overbuild.
Chart 6 - Global data vacancy rates fall
Positioning for 2026: The investable opportunity
To capture both the platform leaders and the infrastructure powering AI:
- ETFS Magnificent 7 Plus ETF (HUGE AU): Exposure to mega cap beneficiaries such as Microsoft, Amazon and Alphabet driving cloud, AI software, and capital light AI economics.
- ETFS US Technology ETF (WWWW AU): Broad exposure across the full technology ecosystem; semiconductors, networking, compute, memory, and data centre suppliers.
AI bubble forecasts fail because they ignore fundamentals. But 2026 is shaping up to be the most fundamental driven year yet. On a static snapshot, valuations look high. But once you factor in what lies ahead, they’re far from bubble territory.
Disclaimer:
The issuer of units in ETFS Magnificent 7+ ETF (HUGE) (ARSN: 685 356 183) and ETFS US Technology ETF (WWWW) (ARSN: 685 355 971) is the responsible entity of the Fund, being ETF Shares Management Limited (ABN 77 680 639 963, AFSL: 562 766). The product disclosure statement (PDS) for the Fund contains all of the details of the offer of units in the Fund. Copies of the PDS are available from ETF Shares Management Limited or at www.etfshares.com.au. In respect of each retail product, ETFS has prepared a target market determination (TMD) which describes the type of customers who the relevant retail product is likely to be appropriate for. The TMD also specifies distribution conditions and restrictions that will help ensure the relevant product is likely to reach customers in the target market. Each TMD is available at www.etfshares.com.au
The information provided in this document is general in nature only and does not take into account your personal objectives, financial situation or needs. Before acting on any information in this email, you should consider the appropriateness of the of the information having regards to your objectives, financial situation or needs and consider seeking independent financial, legal, tax and other relevant advice.
Investment in any product issued by ETFS are subject to investment risk, including possible delays in repayment and loss of income and principal invested. The value or return of an investment will fluctuate and an investor may lose some or all of their investment. Past performance is not a reliable indicator of future performance
