By Vincent Nichols, Senior Investment Specialist, US and Global Thematic Equities
- Technological disruption, specifically in AI infrastructure, is fuelling significant market growth, as evidenced by record-breaking quarterly performances
- Eventually, there will likely be a meaningful deceleration in spending on computing; but for now, expenditure has strengthened and medium-term expectations have risen
- Investors’ attention is shifting from a primary focus on valuations toward the future trajectory of earnings
Technological innovations are disrupting virtually every part of the global economy – from consumption, financial transactions, manufacturing, energy and more.
And this environment continues to drive huge market momentum. Technology stocks experienced a boom in the second quarter of 2026, with artificial intelligence infrastructure investment serving as the primary engine.
Over the period, the Nasdaq’s Philadelphia Semiconductor Sector index delivered its best-ever quarter – up 88% – while the technology uplift led to the best quarter in six years for the S&P 500 and emerging markets were the strongest since 20091.
Memory and logic were central to the rally, supported by accelerating demand from data centres, AI servers, optical communications, power supply, and related infrastructure.
Regionally, the AI theme most clearly favoured semiconductor-heavy markets in Asia, particularly South Korea and Taiwan, while Europe lagged due to limited AI exposure.
Valuations vs. earnings
The question presently hanging over the technology sector is less about valuation and more about the future trajectory of earnings.
At some point there will likely be a meaningful deceleration in spending on computing and eventually an outright decline.
Predictions on how that unfolds are extremely difficult to forecast with a wide range of probable outcomes.
For now, however, spending on computing resources has only strengthened and future expectations have even increased.
For as long as this persists the various components of AI infrastructure should continue to experience impressive fundamental growth; but they are precariously dependent on this dynamic, which becomes increasingly difficult to improve upon.
Over time, as the infrastructure buildout slows and AI adoption increases, we expect the focus and momentum to shift towards cloud service providers and parts of the software sector.
Initially investors were highly focused on AI accelerators, both graphic processing units and custom-built application-specific integrated circuits as well as high bandwidth memory. But we have seen increasing awareness of the critical roles of optical, storage, and central processing units.
Cloud service providers play a critical role that we believe is underappreciated by investors who are focused on the capital requirements and not seeing the early positive signs of strong returns on invested capital. Infrastructure software, including software development tools, database platforms, and cybersecurity, also all play a critical role in AI.
Accelerating adoption
We do not believe the potential use cases for AI are overhyped. AI adoption is progressing orders of magnitude faster than in the dotcom era and the infrastructure being built today is being utilised immediately once deployed – and still fails to keep pace with demand.
It will take time for productivity gains due to AI adoption to appear in margin expansion and corporate profit growth though, and a period of some disillusionment is likely to occur at some point in this journey.
For example, some companies had incentivised workers to maximise their AI usage in their day-to-day work – known as token maxing – but upon seeing the associated costs, are being more pragmatic about their consumption of AI resources.
Fortunately, market optimism in this regard has mostly been concentrated in companies selling hardware to support exponential growth in demand for computing resources rather than potential beneficiaries of AI adoption.
The exception to this is of course frontier AI companies with very high cash burn rates albeit with revenues growing very fast.
Lofty valuations in this space have been almost exclusively a dilemma for private markets, but that is now changing with three enormous initial public offerings either planned or already having taken place this year (SpaceX, OpenAI and Anthropic), which together will add trillions of dollars to public markets.
Positive outlook
While the initial surge in AI-driven growth has been primarily captured by semiconductor-heavy markets and hardware infrastructure, we see the investment opportunity evolving.
Naturally, investors should remain mindful of the cyclical nature of hardware valuations and the eventual deceleration of infrastructure spending.
Ultimately the rapid deployment and use of AI infrastructure underscore a fundamental economic shift.
For investors, the key will be identifying the transition from the engine of AI infrastructure to the companies capable of converting this technological capacity into sustainable margin expansion and corporate profit growth.
[1] Source: Bloomberg
