Concerns about an AI bubble have become one of the most pressing topics in financial markets. With artificial intelligence driving unprecedented investment and innovation, many investors are questioning whether this surge is sustainable—or a sign of excessive commitment in AI.
Concerns about an AI bubble have become one of the most pressing topics in financial markets. With artificial intelligence driving unprecedented investment and innovation, many investors are questioning whether this surge is sustainable—or a sign of excessive commitment in AI. While history shows that bubbles are easiest to recognize after they burst, the scale and speed of current spending have raised questions about long-term returns and valuation risks.
One of the biggest concerns centers on the scale of investment in Artificial Intelligence by hyperscalers and large-cap technology companies. The chart below highlights the Cumulative Rolling 12-Month Capital Expenditure (CapEx) for the largest AI spenders—Microsoft (MSFT), Amazon (AMZN), Alphabet (GOOG/GOOGL), Oracle (ORCL), and Meta Platforms (META). Over the past 2-3 years, these firms have increased their annualized CapEx by two to three times altogether, underscoring the depth of their commitment to AI infrastructure.
This surge in depreciable spending has led to speculation that tech giants have attempted to artificially boost earnings through changes in accounting practices—specifically, by extending depreciation schedules on AI-related hardware. When a company lengthens the depreciation period for its assets, annual depreciation expense declines, which in turn inflates net income. Given that depreciation represents a significant portion of operating costs, such adjustments could create an upward bias in reported earnings—potentially adding over $150 billion in aggregate net income for AI-focused firms over the next two to three years.
Advocates for this approach argue that longer depreciation schedules are justified. Advances in engineering have improved the durability and useful life of GPUs and other hardware. Depreciation should reflect economic usefulness, not simply the cadence of purchases. If older GPUs and CPUs continue to deliver value through new use cases, extending their depreciation schedule becomes a rational move. In short, the argument is that these assets are lasting longer and providing sustained economic benefit, which supports a longer depreciation timeline.
Despite these accounting considerations, fears of an imminent AI bubble may be overstated. When looking at valuation trends across S&P 500 sectors provides important context. The chart below decomposes Price-to-Earnings (P/E) changes into Price Return and EPS Growth for all 11 GICS sectors. While several sectors have posted strong price gains this year, these increases have generally been supported by robust earnings growth. For example:
- Information Technology: Price Return rose 18.33%, while EPS grew 20.45%, resulting in a ~2.1% decline in P/E ratio.
- Communication Services: Price Return outpaced EPS growth slightly, lifting P/E by ~2.2%.
- S&P 500 overall: Aggregate P/E has dipped modestly, signaling stable valuations despite market gains.
The data indicates that while AI-related spending continues to accelerate, underlying fundamentals have not meaningfully changed recently. Valuations have not experienced significant inflation this year, and earnings growth has largely kept pace with price appreciation across most sectors. For long-term investors, this supports maintaining a disciplined, diversified approach to U.S. equities rather than reacting to short-term fears of an AI-driven bubble. A momentum-based strategy adds further value, as our tools are designed to detect emerging technical strength or weakness in AI-related areas, allowing investors to make decisions as trends emerge. For now, both the technology sector and domestic equities remain overweight positions within our DALI framework.
If you’re looking to gain exposure to AI-related companies, one practical approach is to use the new Security Screener (beta). Within the screener, select your investment universe and apply the “Held By” filter to identify stocks included in specific ETFs—helping you narrow down technically strong names. For AI-focused ideas, consider exploring holdings within the AIQ ETF as a starting point.
As implied by the name, the Global X Funds Artificial Intelligence & Technology ETF (AIQ) provides a global diversified exposure to companies involved in AI by tracking the “Artificial Intelligence and Big Data Index”. The fund caps large cap companies at 3%, providing a more balanced weighing mechanism towards mid-cap tech stocks. AIQ maintains a strong fund score of 4.86 and is up ~30% year-to-date. The fund sat on 4 consecutive buy signals before ultimately breaking a double bottom at $48.50 to give a sell signal. Last week, the fund reversed into Os against the market, but this short-term technical deterioration could provide an attractive entry point for those looking to gain a controlled and diversified exposure to AI. Initial support is at $46.50, with additional strong support at $43.50. Resistance can be seen at $53, its previous all-time high.
