AI is here to stay. Exceptional returns may not be.

Big Tech existed long before AI. However, the AI pivot has done wonders for their share prices.

The chart shows the growth of an initial investment of $10,000 in the Magnificent 7, comparing it to the performance of the S&P 500 index. 5 year view. Source: Portfolios Lab

The divergence with the S&P 500 began in early 2023, and continues to grow as AI exuberance reaches new heights, optimism inflates price premiums, and returns expand in response. Over the past 5 years the Magnificent 7¹ group has provided a return of over 3x that of the benchmark index

Moving to the present, the picture is less exceptional.

As above, but a 6 month horizon. Source: Portfolios Lab

What used to be an outperformance of magnitudes has become a mild elevation. This can be attributed to two things:

1. The Surge In Capital Expenditure

The first is the surge in capital expenditure. Big Tech’s projected spending for 2026 is eye-watering. With individual hyperscaler spending hitting $140-200 billion, the massive scale fuels concerns about both its sustainability and corporate profitability - especially given the challenge to revenues from the slow commercialisation of AI.

Whilst recent earnings partly neutralised some of the capex damage, this doesn’t detract from the fact that capex is rising, rather than falling - dragging down future profits.

Share prices have not been immune, as the top infrastructure spenders² lag behind the benchmark indexes they used to dominate.

However, it is important to note that these companies are essentially funding the infrastructure that the entire AI economy is going to benefit from. Economics teaches us that the private sector has no incentive to provide public goods, due to the free-rider problem. But in this case, the companies are left with no choice, as governments simply don’t have the money - nor the desire - to foot the bill.

The incentive to continue AI development far exceeds the short-term costs that come with its financing. But, projects with long-term potential at the sacrifice of short-term profit historically tend to weigh on stock performance.

2. A Maturing AI Division

The second idea is more nuanced: the AI divisions of these companies are maturing. And given that speculative excitement is the driver of record-breaking returns, the loss of this quality translates to a more muted market performance.

But, AI exuberance still exists. It has just diffused to a different sector.

Over the past two months, the Philadelphia Semiconductor Index, (which tracks 30 of the world’s biggest US-listed chip manufacturers), has gained over $5trn in market value.

In particular, Arm, AMD and Intel have done exceptionally well:

Source: Financial Times

Nvidia, one of the biggest beneficiaries of the AI boom, has lagged behind its chip making peers, in an exemplification of the maturation argument.

Despite beating revenue expectations and raising its quarterly dividend from $0.01 to $0.25 per share, the company’s stock price actually slipped upon announcement. This “apathetic” response signals a growing consensus that the company has transitioned from exponential growth into a “more mature tech stock”. Consequently, excitement has shifted to newer players with higher market cap growth potential.

This is one example in a broader story. The Magnificent 7 is no longer looking so magnificent in comparison to the return being granted by the chip sector.

But that is not necessarily a bad thing.

The Distinction: Structural vs Cyclical

As the technology itself develops, the AI industry that underpins it must evolve in tandem. We have reached a point of transition: to continue the boom, the infrastructure needs to be built to support it.

As discussed, hyperscaler capital expenditure is actively funding the data centres and physical equipment “needed to power the AI era this year.” The consequent demand for hardware is being met by chip firms - the very mechanism propelling the rally.

Here, the distinction between ‘structural anchors’ and ‘cyclical players’ emerges.

The market has deemed Big Tech as structural to the AI industry.³ They are the ones at the forefront of the technology’s development, and so will remain at the centre through its evolution. Thus, in this transition from speculative to foundational, the positioning of these companies aligns more with sustained, long-term growth; even if expenditure concerns may lead to temporary fluctuations.

On the other hand, “cyclical players” refers to the firms in AI-adjacent industries that stand to profit or lose from the developments being led by the structural component of the industry.

Take chip stocks; share prices have surged on the back of the infrastructure build-out. Alternatively, look to software stocks, which plunged when the coding prowess of Anthropic’s latest models were revealed. Since the trajectory of these adjacent firms are highly dependent on external drivers, their speculative potential aligns more with the volatile, short-term movements characteristic of extreme gain and loss.

Fundamentally, these cyclical industries remain vulnerable, since their value is tied directly to said external drivers.

What happens if demand for chips dries up? If a new form of hardware is required for the next stage of infrastructure? If chips become obsolete at a faster rate than they can be replaced?

This uncertainty only acts to reinforce the fervour.

Transition through Integration 

But, equally, it could be the case that there is sustained demand for the hardware produced by these chip firms; especially if they can adapt their products to meet the changing needs of the industry.

When this happens, the chip companies would have integrated into the AI industry, becoming structural anchors rather than just cyclical players.

The conclusion remains the same. They are unlikely to retain the speculative potential needed for outsized return. However, they will tend towards the steady, long-run growth akin to their Big Tech counterparts.

An expansion of the AI industry’s structural component will inevitably lead to new cyclical players. Whether they integrate and become structural, or sputter in the long run, depends on investor sentiment, the AI trajectory and the market environment their run takes place in.

Therefore, AI is here to stay. Exceptional returns may not be.

Footnotes

[1] - The Magnificent 7 is comprised of Apple, Amazon, Alphabet/Google, Meta, Microsoft, Nvidia, Tesla

[2] - The major “hyperscalers”: Amazon, Meta, Microsoft, Alphabet, Oracle

[3] - This structural component also consists of currently private companies like Open AI and Anthropic.